VOLUME 13, ISSUE 3, MARCH 2026

Abstract: The proliferation of remote work across U.S. financial institutions has created an expanded cybersecurity threat surface, exposing sensitive financial data to sophisticated attack vectors. This study examines how AI-driven technologies enhance data security protocols within remote financial work environments, drawing on the lived experiences of cybersecurity specialists and IT managers. Using a qualitative phenomenological design, semi-structured interviews were conducted with 12 senior cybersecurity and IT professionals across U.S. financial institutions. Data were analyzed using Braun and Clarke's six-phase thematic analysis framework in NVivo. Five interconnected themes emerged: AI as a catalyst for data security; cybersecurity leaders' perceptions of AI; drivers and barriers to AI adoption; AI-enabled Zero Trust identity and access controls; and compliance and ethics in AI security. Findings demonstrate that AI-powered tools-including machine learning-based anomaly detection, behavioral analytics, and automated incident response-substantially strengthen threat detection and compliance monitoring in remote environments. Participants consistently characterized AI as an augmenter of human expertise rather than a replacement, expressing concern over algorithmic opacity, integration costs, and regulatory complexity. The study provides empirical evidence on the theory-practice intersection of AI adoption in financial cybersecurity, offering practical recommendations for governance, explainability, and workforce capability development.

Keywords: Artificial Intelligence; Cybersecurity; Remote Work; Financial Institutions; Zero Trust Architecture; Phenomenological Research; Machine Learning

Abstract: The rapid advancement of deep learning technologies has led to the rise of deepfake media, posing serious threats to digital trust and security. This project presents an AI-based Deepfake Detection System that identifies manipulated images using advanced machine learning techniques. The system integrates a React frontend and a Node.js backend for seamless media upload and preprocessing. The processed content is analysis using deep learning models such as Pixel Error level analysis to detect visual and temporal inconsistencies. Based on the extracted features, the system classifies the media as real or deepfake and provides a confidence score along with an explanatory result. The proposed solution aims to enhance digital content verification and strengthen cybersecurity measures against synthetic media threats.

Keywords: Deepfake Detection, Artificial Intelligence, Pixel Error Level Analysis, Image Forensics, Machine Learning, Cybersecurity, Media Authentication

Abstract: The aim of this research was to explore the role played by social media in the promotion of daily vlogging and the creation of multimedia content creators through the application of a descriptive research design with quantitative research orientation. The research was conducted through the application of the questionnaire method with daily vloggers, with the data being subjected to correlation and Anova tests. The findings of the research revealed that the success of an individual in the practice of vlogging does not depend on the length of the vlog, the number of audiences, and the equipment used in the creation of content, but rather the quality of the content.

Keywords: Daily Vlogs, social media, Audience Engagement, Creative Burnout, Monetization

Abstract: The growing environmental concerns, rising fuel prices, and rapid technological advancement have significantly influenced the automobile industry, leading to increased interest in electric vehicles (EVs). This study examines consumer perception and purchase intention towards electric vehicles in Coimbatore city from a marketing perspective. The research focuses on understanding how factors such as environmental awareness, cost efficiency, government incentives, brand image, charging infrastructure, and promotional strategies shape consumer attitudes and buying decisions. It also analyzes the role of marketing communication, social influence, and perceived performance in encouraging adoption of electric vehicles. Primary data collected from consumers in Coimbatore city provides insights into their level of awareness, preferences, expectations, and concerns regarding EVs. The findings highlight that while consumers show a positive attitude toward environmentally friendly transportation, challenges such as high initial cost and limited charging facilities still influence purchase decisions. The study emphasizes the importance of effective marketing strategies, customer education, and infrastructure development to enhance consumer confidence and accelerate EV adoption.

Keywords: Consumer Perception, Charging Infrastructure, Purchase Intention, Green Marketing, etc.

Abstract: SMART ROUTE is an Intelligent Route Aware Platform for Contextual Location Based Recommendations is designed to enhance traditional navigation systems by providing real-time, personalized suggestions based on route trajectory and contextual factors. Unlike conventional proximity-based systems, SMART ROUTE analyzes GPS data, user preferences, traffic conditions, time, and environmental parameters to recommend optimized points of interest (POIs) along the user's travel path while minimizing route deviation. By integrating route segmentation techniques with hybrid machine learning models, the system ensures accurate, adaptive, and context-aware recommendations. The proposed platform improves travel efficiency, reduces unnecessary detours, and enhances user satisfaction, contributing to the development of intelligent navigation and smart mobility solutions.

Keywords: GPS Trajectory Analysis, Hybrid Recommendation Model, Context-Aware Computing, Machine Learning.

Abstract: The purpose of this study is to determine the number of online shoppers that are aware of Buy Pay Later choices, particularly in Coimbatore. This study's primary objective is to determine the level of awareness regarding Buy Pay Later services. Our goal is to find out what consumers think about Buy Now Pay Later services, including how they operate, what they offer, how to pay, and what benefits come with utilizing them. We are examining Buy Now Pay services in order to gain a better understanding of them, including what they offer and how users might benefit from them. We are interested in learning what motivates consumers to utilize Buy Pay. Is it because it's user-friendly and accessible to all? Do consumers enjoy the discounts they receive when using it? When consumers make purchases online, can it actually help them manage their finances? Purchasing items and paying for them later is our primary objective. We would like to know more about Buy Pay, particularly for Coimbatore residents who use apps to shop online. We would like to hear how Buy Pay Later works for shoppers in Coimbatore.

Keywords: BNPL, pay, Coimbatore, online, People

Abstract: People really like using Buy Now, Pay Later services because they can buy things without paying for them right away. Buy Now Pay Later services are easy to use and give people time to pay for the things they buy. Buy Now Pay Later services can also change the way people spend money and handle their finances. Some people who use Buy Now Pay Later services might start spending too much money or have trouble paying back what they owe on time.Buy Now Pay services can be frustrating when there are extra fees that people do not know about or when the rules are not clear. Sometimes Buy Now Pay Later services have problems that can cause issues for people who use them. This study aims to examine how BNPL usage impacts consumers' financial behavior and to identify the common problems faced while using these services. The findings will help understand user experiences and support the development of more transparent and responsible BNPL systems.

Keywords: BNPL, pay, Coimbatore, online, People

Abstract: The increasing reliance on personal vehicles combined with heavy traffic and long travel distances has created a high demand for reliable, temporary professional drivers. This paper presents a Web Portal for an Acting Driver Hiring System, a professional platform designed to streamline the hiring of temporary drivers in the Coimbatore region. The system features role-based access for Users, Drivers, and Administrators, and utilizes a smart, distance-based driver matching algorithm. Built using React, TypeScript, and Vite, the application simplifies the hiring process by offering real-time location-based matchmaking, comprehensive driver profiles, and automated booking workflows. Experimental results demonstrate the system's effectiveness in calculating proximity using the Haversine formula and generating optimized matches using simplified KNN-like logic. Future enhancements include the integration of live GPS tracking, centralized database management, and mobile application deployment.

Keywords: Acting driver, hiring portal, distance-based matching, Haversine formula, React, TypeScript, K-Nearest Neighbours (KNN), role-based access.

Abstract: Food spoilage remains a critical challenge in global food supply chains, contributing significantly to waste, economic loss, and public health risks. Traditional packaging solutions, while effective in preserving food for limited durations, lack the capability to provide real-time monitoring of spoilage conditions. This paper introduces SCP-SmartPack AI, an advanced AI-driven platform that integrates Single-Cell Protein (SCP) based biodegradable smart packaging with machine learning algorithms for real-time food spoilage detection. The system architecture combines sensor data acquisition, preprocessing, and predictive modeling with a React-based visualization interface, enabling consumers and retailers to monitor food quality dynamically. Experimental results demonstrate that the proposed system achieves high accuracy in spoilage prediction, reduces food waste, and promotes sustainability through biodegradable packaging. By leveraging SCP materials and AI analytics, SCP-SmartPack AI represents a novel contribution to food technology, offering a scalable solution for enhancing food safety, extending shelf life, and supporting eco-friendly practices. Future work will explore large-scale deployment, blockchain integration for supply chain traceability, and advanced reinforcement learning models to further optimize spoilage detection and packaging efficiency

Keywords: Smart Packaging, Single-Cell Protein, Artificial Intelligence, Food Spoilage Detection, Biodegradable Materials, React, Vite, Typescript, Generative AI.

Abstract: The rapid adoption of artificial intelligence (AI) has increased the need for workforce upskilling across organizations. This study examines employee perception, readiness, and adaptability toward AI upskilling and digital transformation. Using statistical tools such as ANOVA and Chi-square analysis, the research finds that income level and place do not significantly influence employees' attitudes toward AI upskilling. The findings indicate a generally positive and uniform perception among employees regarding the importance of AI for professional growth. The study emphasizes the need for inclusive and continuous upskilling programs to support effective AI integration and sustainable organizational development.

Keywords: Employee perception, Readiness, upskilling programs, digital transformation

+91-7667918914 iarjset@gmail.com 0 Items International Advanced Research Journal in Science, Engineering and Technology A Monthly Peer-Reviewed Multidisciplinary Journal ISSN Online 2393-8021 ISSN Print 2394-1588 Since 2014 Home About About IARJSET Aims and Scope Editorial Board Editorial Policies Publication Ethics Publication Policies Indexing and Abstracting Citation Index License Information Authors How can I publish my paper? Instructions to Authors Benefits to Authors Why Publish in IARJSET Call for Papers Check my Paper status Publication Fee Details Publication Fee Mode FAQs Author Testimonials Reviewers Topics Peer Review Current Issue & Archives Indexing FAQ’s Contact Select Page “IARJSET aligns to the suggestive parameters by the latest University Grants Commission (UGC) for peer-reviewed journals, committed to promoting research excellence, ethical publishing practices, and a global scholarly impact.” Call for Papers Rapid Publication 24/7 April 2026 Submission: eMail paper now Notification: Immediate Publication: Immediately with eCertificates Frequency: Monthly Downloads Paper Format Copyright Form   Submit to iarjset@gmail.com or editor@iarjset.com   Submit My Paper Author CenterHow can I publish my paper? Publication Fee Why Publish in IARJSET Benefits to Authors Guidelines to Authors FAQs (Frequently Asked Questions) Author Testimonials IARJSET ManagementAims and Scope Call for Papers Editorial Board DOI and Crossref Publication Ethics Editorial Policies Publication Policies Subscription / Librarian Conference Special Issue Info ArchivesCurrent Issue & Archives Conference Special Issue Welcome IARJSET is an advanced research journal that is monthly, peer-reviewed, refereed, and open-access multidisciplinary journal dedicated to publishing high-quality scholarly research across major and related areas of Science, Engineering, Technology, Management and Pharmacy. The journal is supported by reputed international indexing and abstracting services, ensuring global accessibility and visibility of published work. IARJSET invites the submission of original research articles, review papers, survey papers, short communications, case studies, methodologies, monographs, and technical notes that contribute to academic and practical advancements in the field. Submit papers to iarjset@gmail.com or editor@iarjset.com UGC approved journal in year 2017, journal id is 47111 View Why publish in IARJSET ✓ Peer-reviewed & Refereed journal ✓ With UGC parameters & COPE ✓ DOI registration with Crossref ✓ Highest Google Scholar citations ✓ UGC approved 2017 ✓ Free DOI from Crossref ✓ Free eCertificate to each author ✓ Life-time free Open Access ✓ Lower-level article acceptance rate ✓ Strict plagiarism policy ✓ Well-constructed ethics and policies ✓ Scientometrics of our journal makes a high-level reputation ✓ Internationally licensed under Creative Commons CC BY 4.0 More Citations Most Cited Cited . Open Access Statement Open Access is a publishing model that provides immediate, worldwide access to the full text of research articles without requiring a subscription to the journal. Accordingly, readers are allowed to copy, use, distribute and display the work publicly and to make and distribute derivative works, in any digital medium for any responsible purpose, subject to proper attribution of authorship. In this model, the publication costs are usually covered by the author's institution or research funds. These Open Access charges replace subscription charges and allow the publishers to make the published material freely available to all interested online readers. At the same time, authors who publish in Open Access journals retain the copyright of their article. Search for: +91-7667918914 Peer-reviewed Impact Factor 8.311 Copyright © 2026 IARJSET This work is licensed under a Creative Commons Attribution 4.0 International License. Open chat

Abstract: Cardiovascular disease remains the preeminent cause in healthcare monitoring. CVD has evolved into a major public health burden on a global scale. In 2025, it is estimated that over 20.5 million people will die from cardiovascular disease related conditions, a number that has surged by 60% over last 30 years. "Reducing the risk of cardiac arrest is a paramount clinical concern." This paper presents a comprehensive review on cardiovascular disease using machine learning techniques and substantiate out of popular dataset like Cleveland dataset, kaggle, UCI etc. Methods: In this paper the researchers used different methods for detecting cardio vascular disease articles are extracted from Google scholar, Research Gate, Scopus search engines between 2022 to 2025.The research findings are present below for better understanding. Result: The review synthesis the advantage and limitation of different methodologies of researchers and dataset used for validation. Conclusion: Machine intelligence offers a viable alternative to conventional human led diagnostics for proactive screening. "Even though significant improvements have been made in this field, the lack of uniformity in prediction models has created a need for new and better solutions".

Keywords: heart disease, Artificial Intelligence, machine learning algorithm, heart disease dataset.

Abstract: The Tours and Travels Web Page is a web-based application designed to simplify and automate the process of tour booking and travel management. Traditional travel booking systems rely heavily on manual methods, which often lead to delays, errors, and poor customer experience. The proposed system provides a centralized digital platform where users can easily browse tour packages, check vehicle availability, make reservations, and manage bookings online. The application is developed using modern web technologies such as HTML, CSS, JavaScript, and a backend database to ensure efficient data handling and user interaction. It includes key modules like user registration and authentication,booking and reservation management, vehicle and tour package management, and an admin panel for system monitoring. The system enhances operational efficiency, reduces manual workload, and provides real-time access to travel information.This web application improves accessibility, accuracy, and convenience for both customers and travel agencies by offering a user-friendly interface and automated services. Overall, the Tours and Travels Web Page serves as an effective solution for modernizing travel management systems and improving customer satisfaction through digital transformation.

Keywords: Tours and Travels, Web Application, Online Booking System, Travel Management System, Tour Reservation, Vehicle Management, Customer Management, HTML, CSS, JavaScript, Database Management, User Authentication,Admin Panel, E-Booking, Travel Automation System, Web-Based Application.

Abstract: The increasing awareness of health, environmental sustainability, and food safety has significantly influenced consumer preferences toward organic food products in recent years. This study examines the level of consumer satisfaction toward organic food products with special reference to Coimbatore. The research aims to identify the factors influencing consumer purchase decisions, evaluate satisfaction levels, and understand consumer perceptions regarding the quality, price, availability, and health benefits of organic food. Primary data were collected through structured questionnaires from consumers who regularly purchase organic products. The study analyses demographic characteristics, buying behaviour, and key determinants such as product quality, trust, certification, and awareness. Findings indicate that consumers generally perceive organic food as healthier and safer compared to conventional alternatives. However, factors such as higher price, limited availability, and lack of awareness among certain consumer groups affect overall satisfaction.

Keywords: health, environmental sustainability, purchase decisions, quality, price, availability, etc.,

Abstract: Road traffic accidents remain a critical public health challenge in India, accounting for over 150,000 deaths annually and 11% of global fatalities despite only 1% of the world's vehicles. Current traffic management systems are reactive, focusing on post-incident response rather than proactive prevention. This paper introduces the Vehicle Collision Analysis Engine (VCAE), a hybrid ensemble machine learning platform integrating Random Forest and XGBoost with geospatial analytics and explainable AI (XAI). Using a synthetic dataset aligned with Ministry of Road Transport and Highways (MoRTH) distributions, the system predicts accident severity, dentifies emerging 'Greyspots' before they escalate, and provides transparent, actionable Results recommendations. demonstrate an R^2 of 0.89, outperforming standalone models. Greyspot validation achieved 74% accuracy. User acceptance testing yielded 87% satisfaction. Deployment simulations confirmed sub-second response times for 500 users. Nationwide deployment suggests potential annual savings of 15,000+ lives. VCAE represents a replicable framework for proactive, explainable, and scalable traffic safety management in emerging economies.

Keywords: Traffic Safety, Machine Learning, Explainable AI, Ensemble Models, Geospatial Analytics, Risk Prediction, Intelligent Transportation Systems.

Abstract: Agriculture is a major part of everyday life and the way farm products reach customers affects both farmers and buyers in the existing system most people depend on local markets to purchase agricultural products this often involves traveling long distances spending extra money and dealing with intermediaries who increase prices many customers also find it difficult to get clear product details or compare options easily to overcome these difficulties this project introduces a web portal for agriculture product that provides a simple and practical online solution for purchasing farm products.The system allows users to create an account access the portal securely browse available products place orders and view their delivery progress customers can purchase items such as cereals and pulses directly through the website without the need to visit physical markets this not only saves time but also reduces physical effort and improves convenience the system helps avoid common mistakes found in manual order processing and ensures smoother handling of transactions.The application is developed using html and css for designing the web pages java for processing user requests on the server side and mysql for storing all system data the system is divided into several functional sections including user services product handling order processing and data storage administrators are responsible for adding new products updating existing information handling customer orders and monitoring delivery progress all records are stored in a centralized database which helps maintain accuracy consistency and security of information.This project aims to make the buying and selling of agricultural products easier faster and more transparent by reducing dependence on middlemen and allowing direct access to products the system benefits both farmers and consumers the web portal for agriculture product offers a practical and affordable digital platform that supports modern agricultural practices and improves the overall efficiency of agricultural marketing.

Keywords: Agriculture E-commerce, Web Portal, Online Agricultural Marketing, Farm Product Management, Order Management System, Digital Agriculture, Database Management, User Authentication System.

Abstract: In the era of digital health transformation, AI-driven virtual assistants are revolutionizing patient care by enabling proactive disease prediction and personalized recommendations. This paper introduces Health AI Companion, an intelligent virtual assistant powered by advanced machine learning algorithms, natural language processing (NLP), and federated learning frameworks. Designed for seamless integration into mobile apps and wearable devices, the system analyzes multi model data-including electronic health records (EHRs), real-time vitals from wearables, genomic profiles, lifestyle inputs, and environmental factors-to predict disease risks with over 92% accuracy across conditions like diabetes, cardiovascular diseases, and early-onset cancers. Leveraging ensemble models such as Random Forests, LSTM neural networks, and transformer-based architectures, Health AI Companion employs explainable AI (XAI) techniques like SHAP values to provide transparent risk assessments, fostering user trust. Personalization occurs through reinforcement learning, generating tailored recommendations: customized diet plans, exercise regimens, medication adherence reminders, and preventive screenings. Privacy is paramount, ensured via differential privacy and edge computing to minimize data centralization. Pilot studies with 5,000 participants demonstrated a 28% reduction in emergency visits and improved adherence by 40%. By democratizing precision medicine, Health AI Companion bridges gaps in under served regions, empowers self-management, and reduces healthcare costs, paving the way for scalable, equitable AI-assisted wellness.

Keywords: Artificial Intelligence, Healthcare Virtual Assistant, Disease Prediction, Personalized Medicine, Machine Learning, Natural Language Processing, Clinical Decision Support System, Digital Health, Predictive Analytics, Remote Patient Monitoring, Health Data Analytics, Explainable AI, Telemedicine, Secure Health Data.

Abstract: Distribution transformers play a crucial function in electric strength structures and their failure can lead to strength interruption and high upkeep prices continuous tracking of transformer fitness parameters is important to prevent sudden faults and enhance system reliability this task presents an iot-based transformer health monitoring machine that permits real-time remark of crucial parameters together with transformer temperature and oil stage the proposed machine makes use of a nodemcu esp8266 microcontroller to gather information from a temperature sensor and an ultrasonic sensor used for oil level monitoring the sensed statistics is displayed domestically on an lcd show and simultaneously transmitted to a cloud platform using wi-fi for faraway tracking when odd situations are detected an alert buzzer is activated and a cooling fan is robotically grew to become on to protect the transformer from harm the developed device reduces the want for guide inspection lets in early fault detection and gives a cost-powerful and green solution for transformer health monitoring the use of internet of things generation.

Keywords: IoT, Transformer Health Monitoring, NodeMCU ESP8266, DHT11 Temperature Sensor, Ultrasonic Sensor, Cloud Computing, Embedded Systems, Real-Time Monitoring.

Abstract: Malware continues to pose a significant threat to digital infrastructures, requiring efficient detection mechanisms that balance accuracy, scalability, and simplicity. This paper presents the design and implementation of a lightweight malware scanner built on YARA, an open-source tool widely adopted for pattern-based malware identification. The proposed scanner leverages custom YARA rules to detect malicious binaries and scripts by matching known signatures and behavioral patterns. Emphasis is placed on rule optimization to reduce false positives while maintaining detection speed. Experimental evaluation demonstrates that the scanner effectively identifies common malware families with minimal resource consumption, making it suitable for integration into endpoint security solutions and incident response workflows. By combining simplicity with extensibility, this approach highlights the practicality of YARA-based detection in academic research, enterprise environments, and security operations centers. The study concludes with recommendations for enhancing rule sets through community collaboration and integrating the techniques to strengthen resilience against evolving threats.

Keywords: Malware Detection, YARA Rules, Cybersecurity, Lightweight Scanner, Pattern-Based Identification.

Abstract: Microplastics (MPs (Micro, (Microplastics), 1 µm-5 mm) and nano plastics (NPs,(Nanoplastics), <1 µm) pose significant environmental and health risks, yet traditional detection methods are slow and resource-intensive. MPWebAI is a browser-based platform enabling users to upload microscopic images for automated MP detection, morphological classification (fibre, film, fragment, pellet), and quantitative analysis using Ultralytics YOLOv11m. Trained on a 7,200-image curated dataset, the model achieves mAP@50 of 95.4%, precision 94.2%, and recall 92.8%. The Fast API + React web app provides annotated images, particle counts, size statistics, and reports in seconds, requiring no specialized hardware beyond a standard microscope and smartphone. A review of current NP detection methods-including optical sieves, SERS, nano-DIHM, and AI-enhanced spectroscopy-highlights challenges for sub-micron particles such as matrix interference, spectral overlap, and the absence of certified reference materials. The paper outlines extensions integrating higher-resolution imaging and hybrid AI-spectroscopy. MPWebAI democratizes monitoring for researchers and citizen scientists in resource-limited regions. The system is open-source.

Keywords: Microplastics, Nano plastics, Ultralytics YOLOv11, Object Detection, Web-Based AI, Environmental Monitoring, Deep Learning.

Abstract: This study examines the fiscal health and expenditure priorities of selected Indian states and analyzes their relationship with public welfare outcomes. In India's federal financial system, state governments play a crucial role in implementing welfare programs, infrastructure development, education, healthcare services, and social protection schemes. The study aims to evaluate how variations in revenue capacity, fiscal deficit, debt burden, and expenditure composition influence developmental outcomes across states. A comparative analysis was conducted using secondary data collected from official sources such as Reserve Bank of India reports, State Finance Accounts, and government statistical publications for the financial year 2023-2024. The research employs percentage analysis, correlation analysis, Chi-square test, ANOVA, and Z-score standardization methods to assess fiscal performance and welfare outcomes. Composite indices were constructed to rank states based on fiscal health and welfare indicators including literacy rate, life expectancy, poverty rate, and infant mortality rate. The findings reveal significant inter-state disparities in fiscal management and welfare performance. While capital expenditure shows a positive association with economic growth, higher revenue capacity and fiscal strength do not necessarily guarantee improved welfare outcomes. Statistical tests indicate that differences in fiscal categories and revenue capacity groups are not significant at the 5 percent level, suggesting that governance efficiency and policy implementation also play a vital role in determining welfare performance.

Keywords: Fiscal Deficit, Revenue capacity, expenditure composition and welfare outcomes

Abstract: The Smart Pharmaceutical Medical Inventory System is a digital platform designed to manage and monitor medicine stock in clinics, hospitals, and pharmacies in a secure and organized manner. The system separates inventory into two categories, namely General Medicines and Cardio Medicines, to provide stricter control over sensitive cardiovascular drugs. Security is enforced through a two-step authentication mechanism involving standard login credentials followed by a One-Time Password (OTP) sent to the registered email or phone of the authorized user. Core functionalities include adding medicines, updating stock quantities, soft delete support for cardio medicines, generating low stock alerts, monitoring expiry dates, and maintaining a complete transaction and activity log. The system is built on a relational database with six primary tables and provides a scalable, accountable, and efficient solution for pharmaceutical inventory management.

Keywords: Medical Inventory, OTP Authentication, Pharmaceutical System, Cardio Medicines, Low Stock Alert, Expiry Monitoring, Transaction Logging, Database Management, BSc Information Technology.

Abstract: This study examines the role of influencer marketing in shaping consumer purchase decisions in Coimbatore. Using a descriptive survey of 106 respondents, the research analyses how influencer credibility, content authenticity, and social media engagement influence consumer behaviour. The study employs statistical tools such as percentage analysis, Chi-Square testing, and rank analysis to evaluate relationships between variables and consumer responses. Findings indicate that influencer marketing significantly increases brand awareness and product visibility, particularly among the 18-25 age group. Social media platforms such as Instagram and YouTube emerge as the most influential channels for promotional content. While 66.67% of respondents have made at least one purchase influenced by social media influencers, factors such as price sensitivity and moderate levels of trust limit the direct conversion impact. The results also reveal statistically significant associations between age group, time spent on social media, and purchase behaviour, highlighting the growing importance of influencer-driven marketing strategies in modern consumer markets.

Keywords: Influencer Marketing, Consumer Purchase Decisions, Social Media, Brand Awareness, Coimbatore.

Abstract: Wireless communication technologies have become an essential component of modern digital infrastructure, enabling seamless connectivity for individuals, organizations, and smart devices. WiFi networks are widely used in homes, campuses, enterprises, and public environments due to their convenience and accessibility. However, the rapid expansion of wireless networks has also introduced significant cybersecurity challenges. One of the most common threats is WiFi spoofing, where attackers create rogue access points that imitate legitimate networks in order to deceive users and intercept sensitive information. WiFi spoofing attacks typically involve the duplication of network identifiers such as SSID (Service Set Identifier) and the manipulation of BSSID (MAC addresses). These attacks allow malicious actors to perform activities such as man-in-the-middle attacks, credential theft, data interception, and network monitoring. Traditional wireless security mechanisms such as WPA2 or WPA3 encryption primarily focus on securing data transmission but often lack effective mechanisms for detecting spoofed access points operating within close proximity. To address these challenges, this research proposes a WiFi Spoof Detection System using IoT (ESP32). The proposed system uses ESP32 microcontrollers to continuously monitor nearby wireless networks and collect parameters including SSID, BSSID, RSSI (signal strength), and channel information. The system analyzes these parameters to identify suspicious patterns such as duplicate SSIDs with different MAC addresses or abnormal signal strength variations. The system integrates IoT hardware monitoring, backend data processing using Python Django, and a PostgreSQL database for centralized storage and analysis. Additionally, a React + Vite.js frontend dashboard provides administrators with real-time monitoring capabilities, enabling visualization of network activities and detection alerts. Experimental implementation demonstrates that the proposed solution provides a low-cost, scalable, and efficient method for detecting rogue access points and strengthening wireless security in smart environments.

Keywords: WiFi Spoofing, ESP32, IoT Security, Rogue Access Point Detection, MAC Address Analysis, RSSI Monitoring, Wireless Intrusion Detection, Cybersecurity, Network Monitoring.

Abstract: The rapid growth of global food demand alongside increasing climate variability has rendered traditional crop selection methods inadequate. This paper presents an AI-based crop recommendation system that applies supervised machine learning to predict the optimal crop for cultivation based on soil composition and environmental parameters. Using the Crop Recommendation Dataset (2,200 samples, 22 crop classes, 7 agronomic features), five classification algorithms were evaluated: Naive Bayes, Decision Tree, Support Vector Machine, Random Forest, and Gradient Boosting. Random Forest achieved the highest accuracy of 96.8%, outperforming all baselines. The system provides a scalable, data-driven decision support tool for precision agriculture, addressing soil nutrient imbalances and climate variability challenges faced by farmers.

Keywords: Crop Recommendation, Machine Learning, Random Forest, Soil Parameters, Precision Agriculture, Classification, Decision Support System

Abstract: This study identifies the rectangles with dimensions and such that the expression , is represented by digit Spy number and Automorphic number. A and S denotes the area and semi-perimeter of the rectangle. Total number of rectangles satisfying the above relations are obtained. Furthermore, both primitive and non-primitive rectangles are accounted for in the total tally of solutions derived from the proposed relationship.

Keywords: Automorphic number, primitive, non-primitive and Spy number.

Abstract: The rapid growth of digital recruitment platforms has intensified the need for automated resume screening systems that efficiently evaluate candidate profiles and match them with suitable job roles. However, many existing Applicant Tracking Systems (ATS) operate as opaque, proprietary solutions, limiting transparency and accessibility for job seekers. This paper presents an NLP-driven automated resume analysis and job matching framework designed to democratize access to career optimization tools. The proposed system integrates classical Natural Language Processing (NLP) techniques-including tokenization, lemmatization, Named Entity Recognition (NER), TF-IDF vectorization, and cosine similarity-to evaluate resume quality, simulate ATS compatibility scoring, and recommend relevant job opportunities. The framework employs a hybrid scoring model combining keyword relevance, section completeness, skill alignment, and formatting compliance to generate interpretable resume scores. Additionally, a similarity-based ranking algorithm matches resumes with structured job descriptions stored in a PostgreSQL database. The system is designed for transparency, determinism, and reproducibility, ensuring explainable outputs suitable for academic and practical deployment. Experimental evaluation demonstrates that the proposed method provides consistent, scalable, and computationally efficient job recommendations while maintaining interpretability-a critical requirement for ethical AI deployment in recruitment technologies.

Keywords: Natural Language Processing (NLP), Applicant Tracking System (ATS), Resume Scoring, Job Recommendation, Skill Gap Analysis, TF-IDF Vectorization, Cosine Similarity, Named Entity Recognition (NER), Machine Learning.

Abstract: Air pollution has become one of the most critical environmental challenges affecting human health, climate stability, and sustainable development worldwide. Accurate prediction of air pollutant levels is essential for early warning systems, policy planning, and effective environmental management. This study proposes a machine learning-based framework for air pollution prediction using historical air quality and meteorological data. The system integrates data collection from air quality monitoring stations, weather parameters such as temperature, humidity, wind speed, and rainfall, and location-specific information. Data preprocessing techniques including data cleaning, handling missing values, feature selection, and normalization are applied to enhance model performance. The proposed model utilizes advanced machine learning algorithms, particularly ensemble learning techniques such as XGBoost, to predict major pollutant concentrations including PM2.5, PM10, NO₂, SO₂, and CO. The predicted pollutant values are further used to compute the Air Quality Index (AQI) and classify air quality into categories such as Good, Moderate, Poor, Very Poor, and Severe. Experimental evaluation demonstrates that the proposed approach improves prediction accuracy and reduces error compared to traditional statistical models. The developed system also supports real-time alerts, hotspot identification, and decision-making support for environmental authorities. This research contributes to sustainable urban planning and public health protection through intelligent air quality forecasting.

Keywords: Air Pollution Prediction, Machine Learning, XG Boost, Air Quality Index (AQI), PM2.5, PM10, Environmental Monitoring, Data Preprocessing, Ensemble Learning, Real-Time Air Quality Forecasting, Smart Cities, Environmental Sustainability.

Abstract: In rural areas, villagers depend mostly on notice boards, word-of-mouth communication, newspapers, and local offices to receive important information. Because of this, updates related to agriculture, government schemes, job opportunities, health camps, weather conditions, school exams, ration shop timings, and village events often reach people late or remain incomplete. There is no single digital platform where all essential village-related information is available in an organized and easy-to-use manner. This lack of centralized information creates difficulties for farmers, students, workers, and common villagers in their daily lives. The Village Voice project is designed to overcome these problems by providing a web-based platform that delivers daily updated village news and useful information in one place. The system includes multiple modules such as agriculture updates, government schemes and alerts, weather reports, job openings, health camp announcements, school exam dates and results, village events, ration shop details, transport information, and electricity shutdown updates. In addition, the project provides a complaint management module that allows villagers to submit complaints online and track their status, improving transparency and communication. Village Voice is developed using HTML, CSS, and JavaScript for the frontend and backend, and MongoDB for database management. The system is designed to be simple, user-friendly, and easily accessible, even for people with basic digital knowledge. This project helps improve awareness, communication, and digital access in rural areas, making village information more reliable, timely, and convenient. Overall, Village Voice acts as a digital bridge between villagers and essential services, supporting rural development through technology.

Keywords: Village Voice, Rural Information System, Village News, Agriculture Updates, Government Schemes, Weather Reports, Jobs, Health Camps, Complaint Management, Digital Village Platform.

Abstract: Carbon pricing has emerged as an important market-based mechanism to reduce greenhouse gas emissions by assigning a monetary value to carbon output. In India, the Carbon Credit Trading Scheme (CCTS), introduced in 2024, seeks to regulate emission intensity in high-emission sectors such as power, steel, and cement through tradable Carbon Credit Certificates (CCCs). While prior research largely concentrates on regulatory design and environmental outcomes, limited empirical attention has been given to how carbon markets influence banking sector behavior. This study examines the impact of CCTS participation on bank lending rates using the State Bank of India (SBI) as a case study. Adopting a descriptive and quantitative research design, the study utilizes secondary data from SBI's FY2025 Annual Report and sectoral carbon proxies, applying ratio analysis, regression analysis, and correlation analysis. The findings indicate a statistically significant negative relationship between carbon credit accumulation and lending rates, confirming the presence of green pricing incentives. Additionally, green funding mobilization shows a strong negative correlation with cost of deposits, suggesting funding efficiency benefits. The study concludes that carbon credit trading is emerging as a financial pricing determinant in India's banking sector.

Keywords: Carbon Credit Trading, CCTS, Green Lending, Net Interest Margin, Transition Risk, Sustainable Finance.

Abstract: Robotics and embedded system education are using simulation-based vehicle control systems more and more. This effort builds and develops a real-time Remote-Controlled (RC) automobile simulation using an Arduino hardware interface and Unreal Engine. The method enables the use of a physical joystick connected to an Arduino to control a virtual remote-controlled car in a 3D simulation environment. The Arduino and Unreal Engine establish serial communication to transmit control signals such as steering and throttle. Without the risks of real remote-controlled cars, the proposed system provides a scalable, reasonably priced, and secure platform for testing vehicle control logic. The results demonstrate smooth real-time vehicle responsiveness and effective hardware-software interaction.

Keywords: RC Car, Arduino, Unreal Engine, Serial Communication, Vehicle Simulation, Embedded Systems, Real-Time Control.

Abstract: A desktop program called the Smart Digital Chits and Finance Management System was created to automate conventional chit fund procedures. Maintaining records and tracking payments in manual systems takes a lot of time and is prone to mistakes. For the management of user registration, chit creation, payment tracking, lottery selection, and report generation, this system offers a safe online platform. Python, Tkinter, and SQLite were used in the development of the application to guarantee effective data storage and intuitive user interface. Data protection is improved by security features like role-based access control and password hashing. The system guarantees accurate financial management, decreases manual labor, and increases transparency. Web deployment and online payment integration are possible future improvements.

Keywords: Chit Fund Management, Financial Automation, Lottery System, Payment Tracking, Report Generation

Abstract: For over five decades, the Society for Worldwide Interbank Financial Telecommunication (SWIFT) has served as the backbone of global finance, yet it remains constrained by the architectural inefficiencies of correspondent banking. This study evaluates the viability of fiat-backed stablecoins as a transformative alternative for cross-border transactions. Utilizing a mixed-methods approach, including speed benchmarking across 30+ corridors and a PRISMA systematic review of regulatory literature, the research identifies that stablecoins achieve a 50-80% cost reduction in high-friction corridors and near-instantaneous settlement finality. However, the study also highlights a "Regulatory Paradox," where compliance with AML/KYC "Travel Rules" and jurisdictional fragmentation creates significant barriers to institutional adoption. The research concludes by proposing an institutional decision framework to navigate these operational and geopolitical risks.

Keywords: Stablecoins, SWIFT, Cross-Border Payments, Blockchain, Financial Inclusion, Settlement Finality.

Abstract: The proposed system is an AI-based Human Stress Detection System that predicts an individual's stress level using physiological and behavioural parameters such as heart rate, sleep duration, body temperature, respiration rate, and blood oxygen level. The system applies machine learning algorithms to analyse input data and classify stress into different levels such as Low, Medium, and High. The model is trained using historical datasets and deployed through a user-friendly web interface where users can input their health parameters and receive instant stress predictions. The main objective of this project is to provide an early stress monitoring tool that can assist individuals in maintaining mental well-being and taking preventive health measures. This system demonstrates how artificial intelligence can be effectively used in healthcare to support stress management and lifestyle improvement.

Keywords: Artificial Intelligence, Machine Learning, Stress Detection, Health Monitoring, Stress Prediction.

Abstract: Managing personal finances effectively has become increasingly complex due to rising expenses, digital transactions, and multiple income sources. Traditional budgeting methods lack automation, predictive analysis, and intelligent insights. This paper presents the design and development of an AI Powered Personal Financial Manager, which integrates machine learning, data analytics, and intelligent recommendation systems to assist users in tracking expenses, predicting future spending patterns, and providing personalized financial advice. The system analyses transaction data, categorizes expenses, predicts monthly budgets, detects unusual spending behaviour, and generates savings recommendations. By leveraging artificial intelligence algorithms such as Linear Regression, Random Forest, and NLP-based financial advisory modules, the proposed system enhances financial decision-making, improves savings habits, and promotes financial stability.

Keywords: Personal Finance, Expense Tracking, Machine Learning, Budget Prediction, Financial Analytics, AI Recommendation System.

Abstract: The rapid digitization of public sector operations has transformed government procurement processes, but it has also introduced challenges such as information overload, fragmented systems, and limited personalization. Traditional e-procurement platforms mainly function as information repositories and require manual effort for schedule planning, compliance verification, and bid preparation, thereby reducing operational efficiency [1]. Recent advancements in artificial intelligence (AI), particularly large language models (LLMs), have demonstrated strong capabilities in automating decision-support tasks and generating intelligent recommendations in real time [3]. This paper proposes an Integrated AI-Driven E-Procurement System for Government Tenders, a web-based platform designed to streamline the tendering lifecycle through automation and conversational intelligence. Developed using the Flask framework and integrated with the Open Router API (GPT-3.5-turbo), the system enables dynamic generation of customized tender schedules, document insights, compliance guidance, and portal recommendations. Unlike conventional portals that provide static listings, the proposed system incorporates a context-aware AI assistant capable of interpreting natural language queries and generating personalized bidding strategies. The platform also includes secure user authentication with encrypted password storage using SQLite and automated email notifications upon login. Experimental validation indicates that the system reduces manual effort in procurement planning while improving transparency and decision-making quality. Future enhancements include integration with live government procurement APIs, multilingual support, and automated tender document parsing.

Keywords: E-Procurement, Government Tender, Artificial Intelligence, Large Language Model, Decision Support System, Chatbot, Flask, Open Router.

Abstract: Vehicle insurance fraud has become a significant challenge in the insurance industry, leading to financial losses and delayed claim processing. This paper presents the design and development of a Vehicle Insurance Automation System using Forensic Analysis, aimed at improving claim verification accuracy and reducing fraudulent activities. The system integrates image forensic techniques and machine learning models to automatically analyse accident-related vehicle images submitted during insurance claims. Digital forensic methods such as Error Level Analysis (ELA), metadata examination, and image consistency verification are used to detect tampering or manipulation.

Keywords: Vehicle Insurance Automation, Fraud Detection, Image Forensics, Damage Detection, Claim Verification.

Abstract: This is a fully standalone wearable fall detection and emergency alert system designed for elderly individuals. The system integrates the MPU-6050 inertial measurement unit (IMU) with an ESP32 microcontroller to detect falls using a dual-stage impact and orientation-based threshold algorithm. Unlike cloud-dependent systems, the proposed design performs all signal processing locally and transmits emergency alerts directly through a GSM module without requiring internet connectivity. A multi-stage validation mechanism combining Signal Magnitude Vector (SMV), posture angle deviation, and post-impact persistence minimizes false positives during daily activities. Upon confirmed fall detection, an SMS alert is immediately transmitted to predefined caregiver numbers using the SIM800L GSM module [9]. Experimental evaluation demonstrates 96.8% sensitivity, 95.1% specificity, and an average alert latency of 2.4 seconds. The system operates at 145 mW, enabling approximately 22 hours of continuous monitoring on a 2000 mAh Li-ion battery.

Keywords: Fall Detection; Wearable Device; ESP32; MPU-6050; GSM Alerting; Embedded Systems; Elder Safety

Abstract: In recent years, the financial services sector has undergone significant transformation due to rapid technological advancement and the growing adoption of digital investment platforms. Investment applications have become widely popular among retail investors as they provide easy market access, real-time data, and simplified trading mechanisms. However, the integration of behavioural finance tools within these platforms has also influenced how investors make financial decisions. These tools are designed based on psychological insights and aim to guide investor behaviour through features such as alerts, notifications, robo-advisory services, gamification, and behavioural nudges. This study examines the rise of behavioural finance tools in investment applications and their impact on investor decision-making patterns. The research highlights how these tools can both support and influence investor behaviour by improving financial awareness while also potentially encouraging emotional and impulsive trading. The study further evaluates how design elements within digital platforms may trigger behavioural biases such as overconfidence, herd behaviour, and excessive trading among retail investors. By analysing the role of behavioural finance mechanisms in modern investment platforms, the study provides insights into how technology is reshaping investor behaviour in the digital era. The findings emphasize the importance of responsible platform design, investor education, and regulatory awareness to ensure that behavioural finance tools enhance rational investment decision-making rather than amplifying psychological biases.

Keywords: Behavioural Finance, Investment Apps, Investor Behaviour, Digital Trading Platforms, Behavioural Biases, FinTech.

Abstract: Customer churn prediction is an essential business tool that allows organizations to predict customers who will likely stop using their services. This prediction is essential in reducing revenue loss and improving customer retention. This research work presents a machine learning-based customer churn prediction system that combines the latest resampling methods and Natural Language Processing (NLP) techniques. To handle the issue of class imbalance, Random Oversampling, Random Undersampling, SMOTE, and ADASYN are applied. In addition, BERT (Bidirectional Encoder Representations from Transformers) is applied for the extraction of contextual features from customer feedback and text data. The preprocessed features are then used as input for classification models like Random Forest, XGBoost, and Logistic Regression. The experimental results based on accuracy, precision, recall, F1-score, and ROC-AUC show improved predictive accuracy. The combination of resampling techniques and BERT-based feature extraction is highly effective in improving the accuracy of customer churn prediction.

Keywords: Customer Churn, Machine Learning, Random Forest, XGBoost, Logistic Regression, BERT, SMOTE, ADASYN, Class Imbalance, NLP.

Abstract: Artificial intelligence has rapidly evolved into a critical technological driver across numerous domains including healthcare, finance, transportation, and large-scale data analytics. Recent developments in deep learning have significantly improved the capabilities of artificial intelligence systems , particularly through the adoption of large neural network architectures that require extensive computational resources for training and deployment While these advancements have enabled improved predictive performance and broader application potential they have also resulted in substantial increases in computational power requirements and associated with the energy consumption. The growing dependence on high performance computing infrastructures has raised concerns regarding the environmental sustainability of artificial intelligence technologies. Training modern deep learning models often involves the use of multiple graphics processing units (GPUs) operating for extended durations, which leads to the considerable electricity consumption and indirectly contributes to carbon emissions depending on the energy source used by computing facilities. This study analytically investigates the relationship between computational power requirements of artificial intelligence systems and their environmental impact. The analysis focuses on key computational parameters including model complexity, hardware utilization, and training duration to estimate the energy consumption associated with AI model training processes. These energy values are subsequently translated into carbon emission estimates using carbon intensity metrics to evaluate the environmental implications of AI computation. The findings reveal that increases in model scale significantly amplify both energy consumption and carbon emissions, highlighting the need for energy-efficient artificial intelligence frameworks. The study proposes the adoption of energy-aware AI development strategies and computational optimization approaches to promote sustainable artificial intelligence systems. Keywords Artificial Intelligence, Green AI, Energy Consumption, Carbon Emissions, Sustainable Computing, Deep Learning Efficiency

Abstract: Skin diseases are among the most common health concerns worldwide, yet early diagnosis and access to dermatological care remain limited, especially in remote and underserved regions. This project presents an intelligent, mobile-friendly skin health diagnosis application designed to provide quick and accessible preliminary analysis of common skin conditions. The system leverages artificial intelligence and computer vision techniques to analyse user-provided skin images and symptoms, enabling automated detection and classification of various dermatological issues. The application is built using Python-based frameworks and integrates deep learning models trained on dermatological image datasets to ensure reliable prediction accuracy. In addition to image-based diagnosis, the system provides symptom-based suggestions, precautionary measures, and basic remedies to guide users toward informed healthcare decisions. The mobile responsive design ensures seamless accessibility across smartphones and tablets, enhancing usability and real-time interaction. Furthermore, the platform incorporates a secure user interface, fast processing, and scalable deployment using lightweight web technologies, making it suitable for both personal and clinical assistance. By combining AI-driven diagnosis with a mobile friendly interface, the proposed system aims to bridge the gap between technology and accessible skincare awareness. Future enhancements may include integration with telemedicine services, real-time dermatologist consultation, multilingual support, and continuous model improvement using larger medical datasets.

Keywords: Skin Disease Detection, Mobile Health (mHealth), Deep Learning, Convolutional Neural Network (CNN), Mobile Net, Image Processing, Telemedicine, AI-Based Diagnosis Net, Image Processing, Telemedicine, AI-Based Diagnosis.

Abstract: The Smart Solar Energy System for Lighting and Fan Control is an eco-friendly solution designed to harness solar energy for basic electrical applications, such as lighting and fan operation. With the increasing demand for electricity and frequent power outages, this system provides a reliable and sustainable alternative to conventional power supply. The system consists of a solar panel that converts sunlight into electrical energy, which is stored in a rechargeable battery for use during periods of low sunlight or at night. A controller regulates the charging and discharging of the battery, ensuring a safe and continuous supply to the connected loads. Lighting is automatically controlled using a light-dependent resistor (LDR), turning on when ambient light is insufficient and off during daylight, thereby conserving energy. The fan operates based on temperature or preset conditions, ensuring comfort while minimizing unnecessary power consumption. This system is suitable for homes, schools, hostels, and rural areas with unreliable electricity supply. It features a simple design, easy operation, low maintenance, and cost-effectiveness, making it ideal for small-scale applications. By promoting the use of renewable energy, the Smart Solar Energy System contributes to energy conservation and a cleaner, sustainable environment.

Keywords: Smart Solar Energy System, Photovoltaic Power Generation, Automatic Lighting Control, Solar-Based Fan Operation, Intelligent Energy Management, Microcontroller-Based Automation, Battery Charging and Storage, Sensor-Based Control System, Power Efficiency Optimization, Load Monitoring and Control, Sustainable Green Technology, Standalone Renewable Power System.

Abstract: The AI Medical Image Diagnosis System is a web-based application developed using Flask and Machine Learning techniques to predict diseases from uploaded medical images. The system allows users to upload medical scan images, processes them using a trained AI model, and provides instant diagnostic predictions. The application also maintains user history and provides a dashboard for monitoring previous scans. The project aims to assist in preliminary medical diagnosis using Artificial Intelligence and improve healthcare accessibility.

Keywords: AI Medical Image Diagnosis System is a web-based application, Node.js, MongoDB

Abstract: This study examines the behavioral patterns and financial performance of retail investors using the Groww application in Coimbatore City during the period 2023-2025. As digital transformation reshapes the Indian brokerage industry, discount brokers like Groww have gained significant traction among millennials and Gen Z. The research analyzes how psychological factors, app usability, and financial literacy influence investment frequency and portfolio outcomes. Based on primary data collected from 100 active users through structured questionnaires, the study applies statistical tools including Normality tests, Pearson Correlation, and Chi-Square analysis. The findings indicate a significant relationship between app interface simplicity and investment frequency, while financial literacy levels were found to be the primary driver of perceived portfolio performance. The study reveals that while the app democratizes market access, behavioral biases like over-trading remain a challenge. The research concludes with suggestions for enhancing investor education within fintech platforms to ensure long-term wealth creation.

Keywords: Investor Behaviour, Groww App, Fintech, Discount Brokerage, Financial Performance, Coimbatore, Retail Investors.

Abstract: This study examines the trends, causes, and impacts of financial frauds in the Indian banking sector during the critical period 2015-2025. The research analyses how the transition from traditional banking to digital platforms influences the frequency and nature of banking frauds, encompassing corporate loan defaults (like PNB and DHFL) and retail cyber-frauds (UPI scams, phishing, and mule accounts). The study is based on a mixed-method approach, utilizing secondary data from RBI reports alongside primary data collected from 100 respondents. It applies statistical tools such as normality test, Pearson correlation analysis, and t-test. The findings indicate a severe vulnerability among the public, with a staggering 96% of respondents having been targeted by fraudulent communications and 42% facing direct financial encounters. While awareness of basic phishing is high, knowledge regarding "Money Mules" is dangerously low (6%). The study concludes that while banks have implemented advanced technological security measures, the "Human Element"-specifically customer ignorance and poor digital hygiene-remains the weakest link, necessitating mandatory financial literacy and AI-enhanced real-time monitoring.

Keywords: Financial Fraud, Indian Banking Sector, Cyber Fraud, Corporate Governance, Money Mules, Digital Banking, UPI Scams.

Abstract: This study examines the transformation of promotional tactics within the Tamil Film Industry (Kollywood) and their subsequent impact on audience behavior during the digital decade of 2015-2025. The research analyzes how modern strategies-specifically Instagram Reels, influencer reviews, meme marketing, and controversy management-influence ticket booking decisions and the "Modern Box Office Paradox." Based on primary data from 100 respondents in the Coimbatore region, the study applies statistical tools including Percentage Analysis, Chi-Square tests, and descriptive interpretation. The findings indicate a significant shift from "Content-driven" to "Perception-driven" success, where 60% of discovery occurs via short-form video. While aggressive marketing successfully drives "First Day First Show" (FDFS) collections, the study reveals a growing "trust deficit," with 85% of audiences skeptical of manufactured hype. The study concludes that while digital tools are essential for discovery, sustainable success in the post-pandemic era requires a balance between viral "hook steps" and narrative quality.

Keywords: Tamil Film Industry, Modern Marketing, Audience Perspective, Instagram Reels, Influencer Marketing, Meme Culture, Box Office Paradox.

Abstract: Mental health disorders are becoming increasingly prevalent worldwide, and early identification of psychological risk factors remains a major challenge. Traditional clinical assessments are periodic and rely heavily on self-reporting, which may delay timely intervention. There is a growing need for intelligent systems capable of continuous monitoring while maintaining patient privacy. NeuroGuard proposes a multimodal artificial intelligence framework designed to act as a secure intermediary between patients and healthcare professionals. The system collects consent-based behavioral, textual, and emotional data through a mobile application and processes it locally to detect early signs of mental health deterioration. The framework integrates transformer-based natural language processing models, behavioral sequence analysis, and multimodal feature fusion to generate structured mental health risk scores. To preserve privacy, federated learning is employed so that raw patient data remains on the device while only encrypted model parameters are shared with the central aggregation server. An explainable AI module ensures transparency by highlighting contributing factors behind risk predictions, allowing doctors to understand clinical indicators without accessing sensitive personal data. The system provides summarized risk insights rather than full conversations or raw behavioral logs. Additionally, NeuroGuard includes an AI-powered recommendation and chatbot module that offers personalized coping strategies, educational guidance, and preventive interventions. By combining privacy preservation, explainability, and proactive risk assessment, the proposed framework presents a scalable and ethical solution for early mental health risk prediction and intervention.

Keywords: Mental Health Monitoring, Multimodal Deep Learning, Federated Learning, Explainable Artificial Intelligence (XAI), Risk Prediction, Early Intervention, Doctor-Patient Intermediary System, Privacy-Preserving AI, Transformer-Based NLP, Healthcare Chatbot.

Abstract: Carbon emissions are one of the major contributors to global climate change and environmental degradation. Monitoring and reducing carbon emissions has become an important challenge for industries, governments, and environmental organizations. The project titled "AI-Based Carbon Emission Prediction and Optimization Using Machine Learning" focuses on developing an intelligent system that can accurately predict carbon emission levels and suggest optimization strategies using advanced machine learning techniques. The main objective of this system is to analyze historical emission data and identify patterns that help in forecasting future carbon emission levels.In this system, various machine learning algorithms are used to process and analyze large datasets related to energy consumption, industrial activities, transportation usage, and environmental factors. These datasets are trained using predictive models to estimate future carbon emissions with improved accuracy. Artificial Intelligence helps in identifying trends, correlations, and hidden patterns in the data that traditional methods may fail to detect. The system also evaluates different influencing factors such as fuel consumption, electricity usage, and production levels to determine their impact on carbon emissions.The proposed system aims to provide efficient optimization techniques that help reduce carbon emissions by suggesting better energy management strategies and resource utilization methods. By implementing predictive analytics, organizations and policymakers can take preventive actions and make informed decisions to control emission levels. The system can also be integrated with smart environmental monitoring platforms to support sustainable development initiatives.Overall, this project demonstrates how Artificial Intelligence and Machine Learning can play a significant role in environmental protection by enabling accurate prediction, monitoring, and optimization of carbon emissions. The developed system contributes to creating a smarter and more sustainable future by helping industries and governments adopt eco-friendly practices and reduce their environmental impact.

Keywords: Artificial Intelligence, Machine Learning, Carbon Emission Prediction, Environmental Monitoring, Predictive Analytics, Sustainability.

Abstract: The exponential growth of internet services and digital transactions has significantly increased exposure to cyber threats, particularly phishing attacks, malware dissemination, and fraudulent web activities. Malicious URLs serve as a primary attack vector in these cyber incidents, enabling adversaries to manipulate users, extract confidential information, and compromise enterprise security infrastructures. Conventional detection mechanisms, primarily based on static blacklists and signature matching, are inadequate in identifying newly generated or zero-day malicious URLs, thereby necessitating intelligent and adaptive detection strategies. To address these limitations, this work proposes an advanced malicious URL detection framework built upon data analytics and machine learning methodologies. The system employs comprehensive feature engineering techniques to extract meaningful lexical, statistical, and structural characteristics from URLs. These features include entropy measurement, character frequency distribution, URL length, special character density, domain-related indicators, and hierarchical path depth analysis. Extracted features are standardized using structured preprocessing pipelines to ensure stability and consistency during model training and inference. The classification core of the system is implemented using the XGBoost algorithm, selected for its robustness, high predictive performance, and capability to model complex nonlinear relationships. To further enhance reliability, a heuristic-based red flag detection layer is integrated alongside the machine learning model, forming a hybrid detection architecture. This layered approach improves resilience against obfuscation techniques and stealthy phishing strategies that attempt to evade automated detection. The backend infrastructure is developed using Python, leveraging Scikit-Learn pipelines for preprocessing and model integration. An interactive dashboard interface enables real-time URL analysis, risk scoring, and visualization of feature contributions. Experimental evaluation demonstrates high classification accuracy, strong generalization capability on unseen datasets, and reduced false positive rates. Overall, the proposed system delivers a scalable, efficient, and explainable malicious URL detection solution, contributing to strengthened cybersecurity defenses in modern web environments.

Keywords: Malicious URL Detection, Machine Learning, XGBoost, Data Analytics, Cybersecurity, Phishing Detection, Feature Engineering, Entropy Analysis, Scikit-Learn, SHAP.

Abstract: Agriculture plays a crucial role in the economy and food security of many countries. Crop diseases and pest infections significantly reduce yield and quality, causing major financial losses to farmers. Early detection of plant diseases using traditional manual inspection is time-consuming, requires expertise, and is often inaccurate. This project proposes an intelligent Crop Pest and Disease Detection System that uses machine learning and image processing techniques to identify plant diseases from leaf images. A deep learning model is trained using a dataset of healthy and infected crop leaves to classify different diseases automatically. The system is integrated with a web interface where users can upload images of crop leaves and instantly receive predictions. The proposed system helps farmers and agricultural experts detect diseases quickly, reduce crop loss, and take preventive measures at the right time. This solution promotes smart agriculture by combining artificial intelligence with real-time accessibility.

Keywords: Crop Disease Detection, Pest Identification, Machine Learning, Deep Learning, Image Processing, Smart Agriculture, Leaf Image Classification, Computer Vision, Web Application, Artificial Intelligence, Precision Farming, Automated Diagnosis, Crop Health Monitoring.

Abstract: The BunBites Web App is a web-based food ordering system designed to simplify and streamline the process of ordering food online. With the increasing demand for convenient digital solutions in the food service industry, this application allows users to browse menus, select items, and place orders easily through a user-friendly interface.The system provides features such as user registration and login, menu browsing, cart management, and order tracking. It also enables restaurant administrators to manage orders efficiently by storing data in a secure database, reducing errors, and improving service speed.Developed using modern web technologies, the BunBites Web App enhances the overall customer experience and operational efficiency of restaurants. This project demonstrates the effectiveness of web-based applications in providing fast, reliable, and convenient food ordering services.

Keywords: Online Food Ordering, Web Application, User-Friendly Interface, Order Management, Database Management, Restaurant Automation.

Abstract: Loan recovery performance is a crucial indicator of banking efficiency and financial stability. Effective recovery mechanisms ensure liquidity, reduce non-performing assets (NPAs), and improve profitability. The present study examines the loan recovery performance of Canara Bank over the period 2016-2025. The study analyses recovery trends using indicators such as Gross Non-Performing Assets (GNPA), Net Non-Performing Assets (NNPA), recovery ratios, and credit growth. Secondary data were collected from annual reports of the bank, publications of the Reserve Bank of India, and other authenticated financial sources. Descriptive statistical techniques, trend analysis, and ratio analysis were employed to evaluate the performance of loan recovery over the study period. The findings reveal that the bank experienced a significant rise in NPAs between 2016 and 2019 due to stressed corporate loans and macroeconomic challenges. However, post-2019 reforms such as strengthened recovery mechanisms, asset quality review measures, and regulatory frameworks like the Insolvency and Bankruptcy Code improved recovery performance. The bank has shown a gradual decline in NPAs and better recovery ratios after 2020, indicating enhanced credit risk management. The study concludes that stronger legal frameworks, improved credit appraisal, and proactive recovery strategies have positively influenced loan recovery performance. The findings offer practical implications for policymakers and bank management in strengthening credit monitoring and recovery systems to ensure long-term financial stability.

Abstract: Managerial decision-making in complex, data-rich environments demand tools that go beyond reporting to deliver actionable intelligence. Traditional enterprise dashboards and analytics platforms are designed for data analysts, not operational managers, creating a persistent gap between insight and action. This paper introduces AI-MDE: AI-Based Micro Decision Engine for Managers, a compact, AI-powered decision support system engineered to translate raw operational data into context-aware recommendations in real time. Built with modern web technologies (React, Vite, Tailwind CSS) and powered by Google's Gemini API, AI-MDE delivers predictive analytics, risk scoring, and scenario modelling through an intuitive manager-first interface. We present the system architecture, design principles, implementation, and evaluation, demonstrating its effectiveness, usability, and scalability for mid-level and senior managers across industries. In controlled evaluations, AI-MDE achieved 89% decision recommendation accuracy, a System Usability Scale score of 84.2, and average recommendation latency of 1.76 seconds, outperforming baseline statistical models by 18 percentage points.

Keywords: Micro Decision Engine, AI-Powered DSS, Managerial Analytics, Predictive Decision Support, Gemini API, Operations Intelligence, Real-Time Risk Scoring, Workforce Planning, Scenario Modelling.

Abstract: The fitness industry has experienced rapid growth in recent years, leading to increased consumption of gym supplements such as protein powders, pre-workout formulas, and vitamins. This study aims to analyse consumer preferences towards gym supplements and understand the factors influencing their purchasing decisions. The research focuses on aspects such as brand perception, price, effectiveness, safety, and the impact of marketing strategies and social media. Primary data was collected from 100 respondents using survey methods and analysed using percentage analysis and the Chi-Square Test. The findings show that 31% of respondents use supplements mainly for muscle gain, and 43% prefer MuscleBlaze brand supplements. The analysis also reveals that monthly income has a significant association with gym workout frequency, while age group has no significant association. The study concludes that consumer preference towards gym supplements is influenced by fitness goals, affordability, marketing influence, and awareness about product quality and safety.

Keywords: gym supplements, consumer preference, muscle gain, fitness, workout

Abstract: The increasing volume of waste generated in local markets has become a major environmental and management challenge. Improper waste monitoring leads to food spoilage, financial loss, and negative environmental impact. Manual record-keeping methods are inefficient and lack proper tracking and analysis capabilities. This project proposes a web-based Waste Reduction Tracker for Market that enables systematic recording, monitoring, and analysis of waste generated in market environments. The system allows users to log waste details such as item name, quantity, and date. It provides real-time reports and graphical visualization using bar and pie charts to analyze waste patterns. By maintaining digital records and generating automated reports, the system helps market administrators identify high-waste items and take preventive measures. The proposed solution is simple, scalable, and efficient, contributing to improved waste management and environmental sustainability.

Keywords: Waste Management, Waste Tracking, Market Monitoring, Web Application, Data Visualization, Sustainability.

Abstract: This project presents a Cyber Insurance Risk Assessment Tool designed to evaluate the cybersecurity risk level of organizations and support cyber-insurance decision-making. The system collects information related to security practices, vulnerabilities, incident history, and organizational assets through a structured assessment process. Based on the collected data, the tool computes an overall cyber risk score and classifies the risk level as low, medium, or high. The application provides a user-friendly dashboard that visualizes risk scores, risk trends, and estimated financial impact of cyber incidents. It also recommends suitable insurance coverage ranges and generates detailed assessment reports for organizations and insurers. By combining risk scoring techniques and optional machine learning-based predictions, the tool helps organizations understand their cyber risk posture and assists insurers in making data-driven premium and coverage decisions. The proposed system improves transparency, accuracy, and efficiency in cyber risk evaluation and cyber insurance assessment.

Keywords: Cyber Insurance, Cyber Risk Assessment, Cybersecurity, Risk Scoring, Risk Management

Abstract: MachineryHub - A Web Platform for Industrial Equipment Sales is a web-based application designed to digitalize the buying and selling of industrial machinery through a centralized online platform. Traditional machinery stores often face challenges such as limited reach, manual inventory management, lack of real-time product availability, and inefficient customer communication. This project aims to overcome these limitations by providing a user-friendly and scalable web solution for machinery vendors and customers. The platform allows users to browse machinery categories, view detailed product specifications, compare equipment, and place purchase or enquiry requests online. Admin functionalities include machinery listing management, inventory updates, order tracking, and customer enquiry handling. The system ensures secure user authentication and efficient data handling to improve operational transparency and business efficiency. The application is developed using HTML, CSS, and JavaScript for the frontend, ensuring responsive design and smooth user interaction. The backend logic and data handling can be integrated with modern web services and databases to support real-time operations. MachineryHub enhances accessibility, reduces manual workload, and enables industrial businesses to expand their market reach through digital transformation.

Keywords: Industrial Machinery, Web-Based Application, Online Equipment Sales, Inventory Management, Digital Marketplace, Digital Transformation.

Abstract: The rapid growth of digital communication has increased the demand for efficient and accurate language translation systems. However, most existing translation tools rely heavily on internet connectivity, limiting accessibility in low-network regions, and raising significant data privacy concerns. Furthermore, traditional rule-based translation systems often produce contextually incorrect and slow translations, particularly for less-resourced languages such as Nepali and Sinhala. To address these limitations, this paper presents an AI-Based Text-to-Text Machine Translation System that translates Nepali and Sinhala text into English in an offline environment. The proposed system leverages deep learning techniques using PyTorch and HuggingFace Transformers to implement a pre-trained neural machine translation model capable of generating accurate and context-aware translations. The system is developed using Python and Django, ensuring a secure, user-friendly interface while maintaining offline functionality. The architecture consists of multiple modules including input handling, text pre-processing, neural translation engine, post-processing, and result display. By eliminating internet dependency and ensuring local processing, the system enhances privacy, accessibility, and reliability. Experimental evaluation demonstrates that the proposed solution provides efficient, grammatically coherent, and contextually meaningful English translations while maintaining complete offline usability.

Keywords: Machine Translation, Low-Resource Languages, Nepali, Sinhalese, Transformer, Transfer Learning, Sub word Tokenization

Abstract: Alzheimer's disease (AD) is a progressive neurodegenerative disorder and the leading cause of dementia worldwide, affecting approximately 50 million people with projections exceeding 150 million by 2050. Early detection of AD is critically important as interventions initiated during the prodromal phase-particularly mild cognitive impairment-have the greatest potential to slow disease progression and preserve cognitive function. Machine learning, particularly ensemble methods like Random Forest, has emerged as a powerful tool for early AD prediction by analysing complex, multimodal data including neuroimaging, genetic markers, cognitive assessments, and fluid biomarkers. This paper provides a comprehensive review of Random Forest applications in AD prediction, synthesizing findings from recent studies.The combination of Backward Elimination Feature Selection with Artificial Ant Colony Optimization has achieved 95% accuracy while reducing computation time by 81%. Key challenges including class imbalance, model interpretability, and cross-cohort generalizability are addressed through techniques such as SMOTE and SHAP analysis. This review provides researchers and clinicians with a comprehensive understanding of Random Forest's role in early AD prediction and identifies promising directions for future research.

Keywords: Alzheimer's disease, machine learning, Random Forest, early prediction, ensemble learning, feature selection, biomarker analysis, neuroimaging, cognitive assessment

Abstract: The rapid growth of digital systems has increased the need for quick, secure, and contactless access to information across various domains such as education, business, and service management. Traditional manual methods of information sharing and verification often lead to inefficiencies, delays, and data management challenges. To address these issues, this paper proposes a QR Code-Based Information and Access System, an intelligent digital platform designed to provide fast, secure, and efficient information retrieval using Quick Response (QR) code technology. The proposed system integrates QR code generation, real-time data access, web-based interfaces, and secure database management to deliver a reliable and user-friendly solution. QR codes act as a bridge between physical objects and digital information, enabling users to quickly scan codes using mobile devices to retrieve relevant data such as product details, digital documents, authentication records, or service information. The system ensures efficient data handling, reduces manual effort, and improves accessibility through instant scanning and automated data retrieval. Additionally, the platform supports dynamic QR code generation, secure data storage, and seamless interaction between frontend and backend services. The web-based interface allows administrators to generate and manage QR codes, while users can easily access the encoded information through scanning. This approach enhances operational efficiency, reduces paperwork, and minimizes the chances of data errors. System evaluation demonstrates that the QR Code-Based Information System improves speed, accessibility, accuracy, and overall user experience compared to traditional information management approaches. The proposed architecture provides a scalable and flexible solution suitable for applications in education, attendance systems, product tracking, digital verification, and service management.

Keywords: QR Code Technology, Information Access System, Secure Data Retrieval, Web-Based Platform, QR Code Generation, Digital Verification, Data Management, Real-Time Access, Information System.

Abstract: Soil health is a fundamental determinant of agricultural productivity, yet conventional testing methods remain costly and incapable of real-time feedback. This paper presents AgriSmart - a physically implemented IoT-based Soil Health Assessment and Fertilizer Recommendation System. An ESP32 Wi-Fi microcontroller (Device ID: ESP32-AGRISMART-001) is interfaced with a capacitive soil moisture sensor and a DHT11 temperature-humidity module. Soil pH is determined using the distilled-water pH paper method and entered manually via the web dashboard. Rainfall data is fetched in real time using the OpenWeatherMap API. All sensor readings are transmitted via HTTP POST in JSON format to a Django backend, stored in an SQLite database, and processed by a three-model Random Forest pipeline: Soil Type classification (71.8%), Soil Health assessment (88.2%), and Fertilizer Recommendation (93.6%). The system was validated with real soil samples and supports 12 fertilizer classes across 6 soil types and 22 crop varieties. Results confirm practical viability for precision agriculture.

Keywords: IoT, ESP32, AgriSmart, Soil Health, Fertilizer Recommendation, Random Forest, Django, SQLite, DHT11, OpenWeatherMap API, Precision Agriculture, Soil Type Classification

Abstract: This study empirically examines customer perception towards the family branding strategy of ITC Limited, one of India's most diversified conglomerates. Drawing upon primary survey data from 100 respondents across urban and semi-urban centres, and employing percentage analysis and one-way ANOVA, the research investigates consumer awareness, trust, purchase behaviour, and satisfaction associated with ITC's family brand portfolio. The findings reveal that 72% of respondents possess at least moderate awareness of ITC as a family brand, and 70% agree that ITC's family branding strengthens their trust in its products. An overall satisfaction rate of 68% affirms that ITC's family-branded products successfully deliver value and quality. ANOVA analysis demonstrates statistically significant differences in brand perception scores across age groups (F = 8.74; p = 0.001), with the 20-30 years cohort exhibiting the most positive perceptions. The study also identifies the persistent association of the ITC brand with its tobacco legacy as a potential challenge to family brand equity across non-tobacco FMCG categories.

Keywords: Family Branding, Consumer Perception, ITC Limited, Brand Equity, FMCG, Brand Trust, ANOVA.

Abstract: Early detection of dementia is essential for enabling timely clinical intervention, slowing disease progression, and improving overall patient care and quality of life. Conventional diagnostic methods such as neuropsychological assessments, brain imaging techniques, and clinical evaluations are often expensive, time-consuming, and dependent on specialized expertise, making them less accessible in rural or resource-limited settings. In response to these challenges, this paper presents an AI-based tool for early-stage dementia detection using speech analysis as a non-invasive and cost-effective alternative. Speech is a natural and information-rich medium that reflects cognitive processes, and subtle impairments in memory, attention, and executive functioning often manifest in acoustic, prosodic, and linguistic patterns during spontaneous speech production. The proposed system extracts a comprehensive set of features from recorded speech samples, including acoustic features such as Mel-Frequency Cepstral Coefficients (MFCCs), pitch, jitter, shimmer, and spectral properties; prosodic features such as speech rate, pause frequency, and pause duration; and linguistic features such as lexical diversity, syntactic complexity, word frequency distribution, and semantic coherence. The methodology follows a structured pipeline consisting of speech preprocessing (noise reduction, normalization, silence removal), feature extraction, feature selection to remove redundant and irrelevant attributes, model training, and performance evaluation. Multiple machine learning and deep learning algorithms are implemented and compared, including Support Vector Machines (SVM), Random Forest classifiers, and Long Short-Term Memory (LSTM) networks, which are particularly effective in modeling sequential and temporal dependencies in speech data. The models are trained and validated using appropriate cross-validation techniques to ensure robustness and generalization. Performance metrics such as accuracy, sensitivity, specificity, precision, recall, and F1-score are used to evaluate classification effectiveness in distinguishing early dementia cases from healthy controls. Experimental results demonstrate that the proposed AI-based tool achieves high diagnostic performance, highlighting the effectiveness of integrating acoustic and linguistic features for cognitive assessment. The findings suggest that speech-based AI systems can function as reliable, scalable, and remote screening solutions, supporting clinicians in early diagnosis and enabling proactive intervention strategies for individuals at risk of cognitive decline.

Keywords: Early Dementia Detection, Speech Analysis, Artificial Intelligence, Machine Learning, Acoustic Features

Abstract: Crime against women in India has emerged as a critical public health and social issue, demanding rigorous analysis and effective preventive strategies. This study focuses on a comprehensive examination of crimes against women, both quantitatively and qualitatively, using official data released by the National Crime Records Bureau (NCRB) and made available through the Open Government Data (OGD) Platform India. The dataset encompasses various categories of crimes, including molestation, rape, dowry-related violence, exploitation, kidnapping, and cruelty by relatives. Analytical techniques involving data manipulation and visualization were applied to uncover crime trends, patterns, and regional variations. Descriptive analysis was conducted using charts and graphs to provide meaningful insights, while predictive model was employed to forecast future crime rates based on historical data. Among the classifiers tested, Random Forest demonstrated superior performance, achieving an accuracy of 99.67% in classification and 63.32% in prediction. The findings highlight the alarming rise in crimes against women and underscore the importance of leveraging data-driven approaches to inform policy decisions, strengthen law enforcement, and promote societal awareness aimed at reducing gender-based violence.

Keywords: The scope (India, NCRB, OGD), methods (data analysis, visualization, predictive model, Random Forest), crime categories (rape, dowry, molestation, etc.), and impact (public health, policy, gender-based violence).

Abstract: Textile and garment industries need effective inventory management systems to help them smoothen the production processes and in order to avoid shortage of materials. The conventional fabric management systems like manual registers and spreadsheets usually lead to human errors, slowness in updating and inaccurate records of stock. These problems can cause the delay in production and losses. In an attempt to eliminate these shortcomings, this paper will present a proposal of the Real-Time Fabric Management System, which is software-driven solution that will be used to mechanize the processes of fabric inventory. The system captures fabric information including type, color, length and quality of the fabric and stores the same in a centralized database. It constantly updates the stock information in real time with the addition of fabric, its issuance, or its return so that all stock is monitored and with better transparency assured. It allows managing fabric flow and making decisions in a more informed way by giving the user-friendly interface to be used by managers and staff to monitor the flow of fabric.

Keywords: Inventory Automation, Real-Time Tracking, Fabric Inventory, Textile Industry, Centralized Database, Stock Management.

Abstract: The rapid evolution of user-centric design has led to the emergence of emotionally responsive digital interfaces. This project presents an Emotion-Aware Webpage Portfolio, a personalized web system that dynamically adapts its presentation based on the user's emotional state. Using real-time facial expression recognition powered by a browser-based machine learning model (TensorFlow.js), the system classifies emotions such as happiness, sadness, anger, surprise, and neutrality from live webcam input. These detected emotions are then mapped to adaptive UI components, enabling modifications in color themes, textual tone, animations, and interactive elements. The goal is to enhance user engagement by creating a portfolio interface that reacts intuitively to user affect. The project demonstrates that integrating affective computing with web technologies can significantly improve user experience, accessibility, and immersion. The results highlight strong potential for emotion-driven design in future web applications.

Keywords: Emotion detection, Affective computing, Web-based portfolio, Real-time facial expression recognition, TensorFlow.js, Adaptive user interface, Human-computer interaction, User experience personalization, Computer vision, Emotion-aware design

Abstract: The increasing number of students in educational institutions has made the process of course registration and allotment more complex and time-consuming when performed manually. Traditional manual systems often lead to several issues such as data redundancy, human errors, delays in processing, and difficulty in maintaining accurate records. To overcome these challenges, this project titled "Web-based Student Course Allotment System" proposes an automated solution to manage the course registration process efficiently. The system is developed as a web- based application that allows administrators to manage course details, seat availability, and prerequisite requirements through a centralized platform. Students can access the system through a secure login and select courses based on their eligibility and preferences. The system updates seat availability in real time, preventing over-enrollment in any course. It also verifies prerequisite conditions to ensure that students register only for courses they are qualified to take. By automating the course allocation process, the system reduces manual effort and improves the accuracy of data management. The web-based design enables easy accessibility and efficient communication between administrators and students. Furthermore, the system maintains a structured database that helps in storing and retrieving course and student information effectively. The implementation of this system significantly reduces processing time and minimizes administrative workload. It also improves transparency and reliability in the course registration process. Overall, the proposed system provides an efficient and reliable solution for managing course allotment in educational institutions.

Keywords: Course Allotment, Web-based System, Automated Registration, Database Management, Educational Software.

Abstract: Student placements play an important role in the career development of college students. However, managing student eligibility and tracking placement details manually can be difficult and time-consuming for placement officers. The Student Placement Eligibility and Tracking System is a web-based application designed to automate and simplify the placement process in educational Institutions. The system helps in identifying eligible students based on academic criteria such as CGPA, arrears, and skills. It also allows placement coordinators to track company drives, student applications, interview status, and final selections. Students can log in to check their eligibility, apply for companies, and view placement updates.This system improves efficiency, reduces manual work, and ensures transparency in the placement process. The project uses modern web technologies and a database to store and manage placement information effectively. Overall, the system helps institutions manage student placement activities in a structured and organized manner.

Keywords: Student Placement System, Eligibility Tracking, Web Application, Placement Management, Student Database

Abstract: The textile industry faces significant challenges in managing and analyzing monthly bills and invoices. Traditional methods involve manual data entry into Excel spreadsheets, making it difficult to extract insights about sales trends, pending payments, customer patterns, and business performance. Sathya Tex Web Analytics addresses these challenges by providing a comprehensive web-based platform specifically designed for textile businesses to manage bills and gain actionable insights through data analytics. The system allows textile companies to upload their monthly bills in Excel format, automatically processes and stores the data in DuckDB for high-performance analytics, and provides an AI-powered chatbot that answers business questions in natural language. Users can ask questions like 'What is my total sales this month?' or 'Which customers have pending payments?' and receive instant answers with the underlying SQL queries. The platform features a React-based dashboard displaying key performance indicators such as total sales, profit margins, pending amounts, and customer-wise breakdowns. Built using React for the frontend interface, FastAPI for backend API services, PostgreSQL for metadata storage, and DuckDB for analytical queries, the system integrates Groq large language model through LangChain to convert natural language questions into SQL queries. The AI component automatically generates safe read-only queries, executes them against DuckDB, and presents results in conversational format. This eliminates the need for textile business owners to learn SQL or hire data analysts. By combining modern web technologies with AI-powered analytics, Sathya Tex Web Analytics transforms raw billing data into strategic business insights, enabling textile companies to make data-driven decisions and improve operational efficiency.

Keywords: Textile Analytics, Bill Management, AI Chatbot, Natural Language Queries, DuckDB, FastAPI, Business Intelligence

Abstract: Antibiotic resistance is one of the most serious global health challenges, where bacteria evolve mechanisms to resist the effects of antibiotics. This leads to treatment failure, prolonged illness, increased healthcare costs, and higher mortality rates. Traditional laboratory-based methods for detecting antibiotic resistance require significant time and manual interpretation. This project proposes an Artificial Intelligence (AI)-based system that analyzes patient laboratory data such as bacterial type, antibiotic tested, and Minimum Inhibitory Concentration (MIC) values to predict whether the bacteria are resistant or sensitive to specific antibiotics. Machine Learning algorithms are used to automate the detection process, enabling faster and more accurate clinical decision-making. The system helps healthcare professionals select appropriate antibiotics and reduces misuse, thereby contributing to better patient outcomes and combating antimicrobial resistance.

Keywords: Antibiotic Resistance, Machine Learning, Healthcare AI, MIC Analysis, Clinical Decision Support.

Abstract: Language learning is an essential skill in the modern globalized world. Many learners struggle to find real-time speaking partners to practice their target language. This project, Language Exchange Explorer, is a full-stack web application designed to connect users who want to mutually exchange languages. The system allows users to register, create profiles, specify their native and learning languages, and automatically match with suitable partners.The application is developed using frontend technologies such as HTML, CSS, and JavaScript, and backend technologies including Node.js, Express.js, and MongoDB. The system implements authentication, profile management, matching logic, and chat functionality.This platform promotes peer-to-peer learning, real-time interaction, and cost-effective language improvement. The project demonstrates practical implementation of full-stack development concepts including REST APIs, database integration, user authentication, and system design using Data Flow Diagrams (DFD).

Keywords: Language Exchange, Full Stack Development, Web Application, Node.js, Express.js, MongoDB, HTML, CSS, JavaScript

Abstract: Krit Text Suite is an advanced word analysis mobile application built using Flutter, offering users real-time, in-depth insights into word meanings and linguistic properties. By integrating multiple API endpoints via RapidAPI, the app fetches and categorizes word-related data, including definitions, synonyms, antonyms, pronunciation, syllables, examples, and rhymes, presenting the results in a structured and user-friendly interface. The application ensures seamless real-time data retrieval with intelligent error handling, allowing it to skip non-responsive APIs while displaying successfully retrieved information. Secure authentication is implemented via Firebase Authentication for OTP-based phone number verification and AWS Lambda for email-based login and signup, with MongoDB Atlas managing user credentials and search history. Krit Text Suite is designed for linguists, students, writers, and educators, offering an intuitive and efficient platform for vocabulary expansion and language learning. The app is hosted on Firebase for web accessibility, ensuring scalability and security.

Keywords: Flutter, Dart, Firebase Authentication, AWS Lambda, MongoDB Atlas, RapidAPI, Word Analysis, Mobile Application, OTP Authentication, Real-Time API.

Abstract: Speech Emotion Recognition (SER) is an important area of research in human-computer interaction that aims to identify human emotions from speech signals. Accurate detection of emotions such as happiness, sadness, anger, fear, and neutrality can significantly enhance applications in virtual assistants, mental health monitoring, and customer service systems. Traditional emotion recognition systems relied on handcrafted acoustic features and conventional machine learning techniques, which often struggled to capture complex patterns in speech data. In this project, a deep learning-based approach is proposed to automatically recognize emotions from speech signals. The system processes audio inputs by extracting relevant acoustic features such as Mel-Frequency Cepstral Coefficients (MFCCs), pitch, and energy. These features are then used to train a deep learning model capable of learning emotional patterns from speech data. The proposed model improves classification accuracy by effectively capturing temporal and spectral characteristics of speech signals. The developed system is evaluated using labeled speech emotion datasets and demonstrates promising performance in recognizing multiple emotional states. The results show that deep learning models can significantly enhance emotion recognition accuracy compared to traditional approaches. This work highlights the potential of speech emotion recognition systems in building more natural and emotionally aware human-machine interactions.

Keywords: Artificial Intelligence, Machine Learning, emotion Detection,

Abstract: Water quality is critical for public health, agriculture, and industrial processes. Traditional monitoring relies on manual sampling and laboratory analysis, which is time-consuming, costly, and unable to provide real-time data. This project presents a Smart Water Quality Monitoring System using IoT technology, integrating a pH sensor and flow sensor with an ESP32 microcontroller. Sensor data is displayed on an LCD and transmitted via WiFi to a cloud platform for real-time remote monitoring. The system is cost-effective, automated, and scalable, enabling continuous monitoring, early contamination detection, and proactive water management for domestic, industrial, and municipal use.

Keywords: IoT, ESP32, pH Sensor, Flow Sensor, Water Quality, Real-Time Monitoring, Cloud Platform, Blynk.

Abstract: In traditional user authentication systems, identity verification is commonly based on textual inputs like usernames, passwords, or personal IDs, which are often vulnerable to misuse, forgetfulness, or theft. Existing systems that attempt face-based recognition either store static images or depend on manual matching, lacking automation and real-time detection capabilities. These approaches do not support live face capturing and fail to fetch dynamic user-related information from a database, resulting in inefficient or insecure identification mechanisms. The proposed system addresses these limitations by introducing a live face capturing and identification framework integrated with OpenCV. When a user registers, their face is captured in real-time using the system's web interface connected to the webcam through OpenCV. The captured image is then stored securely in the server's file system or database. During login or verification, the system again uses OpenCV to capture a fresh face image and compares it with stored images using face similarity techniques. Upon successful matching, all relevant user information is dynamically fetched from the database and displayed securely on the interface. This system ensures that only the rightful user gains access to sensitive data, enhancing both security and user experience. The modular architecture seamlessly integrates live camera input via OpenCV, secure data storage, and efficient face comparison logic, offering an intelligent and real-time authentication mechanism suitable for modern web applications.

Abstract: Given the sensitivity of personal health information and the rising prevalence of data breaches, healthcare analytics faces a significant challenge in ensuring the privacy of sensitive data while simultaneously providing valuable insights. By incorporating privacy-preserving parameters, zero-knowledge proofs (zk-SNARKs), blockchain technology, and a multi-tenant cloud environment, the secure framework presented in this paper addresses these issues. The framework ensures that healthcare records remain protected during analytics computations without exposing raw data by employing cutting-edge cryptographic methods, particularly zk-SNARKs. In order to validate computations, the privacy-preserving analytics engine makes use of anonymized healthcare records and generates zk-SNARKs. When these proofs are incorporated into a blockchain network, they produce a transparent, tamper-proof ledger that guarantees safe healthcare transactions. This strategy is absolutely necessary in circumstances like telemedicine, where secure data sharing and computation are of the utmost importance. By demonstrating its application in a telemedicine app, the framework provides a scalable and secure solution to a pressing issue, demonstrating its practical significance in healthcare analytics.

Keywords: Healthcare Data, Security, Blockchain Technology, Cloud Computing, Multi-Tenant, Cloud Environment, Secure Data Sharing, Healthcare Data Analytics, Secure Data Processing.

Abstract: This article presents the results of a study of the chemical composition of essential oils of Artemisia absinthium L. and Artemisia vulgaris L., grown in the Samarkand region of the Republic of Uzbekistan. The relevance of this study is due to the widespread use of Artemisia species in the pharmaceutical, food, and cosmetic industries, as well as the need to study the regional characteristics of their chemical profile. The essential oils were obtained by hydrodistillation from the above-ground parts of plants during the flowering phase. Qualitative and quantitative analysis of their component composition was performed using gas chromatography-mass spectrometry (GC-MS). Differences in the ratios of the dominant components of the essential oils of the studied species were identified, which are related to the species and the ecological and climatic conditions of the growing region.

Keywords: Artemisia absinthium L., Artemisia vulgaris L., essential oil, medicinal plants, component composition, chromatograph mass spectrometry, hydrosteam distillation, wormwood, common wormwood.

Abstract: In modern digital communication systems, ensuring the secure transmission of sensitive data is a major challenge due to increasing cyber threats and unauthorized access. This paper proposes a secure data protection framework that integrates One-Time Password (OTP) generation with an RSA-based key exchange scheme. In the proposed system, whenever a data owner initiates a data-sharing process, a unique OTP is dynamically generated to authenticate the user. The generated OTP is then encrypted using the RSA encryption algorithm before transmission. The RSA key exchange mechanism ensures that only the intended recipient possessing the correct private key can decrypt the OTP and access the shared data. This approach provides a dual layer of security by combining dynamic authentication with strong asymmetric encryption. As a result, the system significantly enhances data confidentiality, integrity, and protection against interception or unauthorized access during transmission. The proposed scheme can be effectively applied in secure communication systems, cloud data sharing platforms, and authentication-based applications.

Keywords: One-Time Password (OTP), RSA Encryption, Key Exchange, Data Security, Secure Communication.

Abstract: Electronic voting systems must guarantee security, transparency, and voter privacy simultaneously. This paper presents a secure and transparent e-voting system that integrates blockchain technology, Ethereum smart contracts, differential privacy mechanisms, and email-based One-Time Password (OTP) voter authentication. The proposed system leverages the immutability and decentralization of blockchain to ensure tamper-proof vote recording, while Ethereum smart contracts automate the electoral process without requiring a trusted central authority. Differential privacy is applied to protect individual voter choices from statistical inference attacks, ensuring that aggregate vote counts cannot be traced back to individual voters. Email-based OTP authentication provides a lightweight yet effective two-factor verification mechanism that prevents unauthorized access and double voting. The system architecture includes a voter registration module, a secure login interface, a real-time vote casting and tallying smart contract, and a transparent audit trail accessible to all stakeholders. Security analysis demonstrates resistance to common threats including Sybil attacks, replay attacks, and ballot stuffing. Experimental evaluation shows the system achieves high throughput, low latency, and strong privacy guarantees. This approach addresses key limitations of existing e-voting platforms and contributes a practical framework for conducting elections that are simultaneously verifiable, anonymous, and resilient to adversarial manipulation.

Keywords: Blockchain, Differential Privacy, E-Voting, Ethereum, Smart Contracts, Voter Authentication

Abstract: As the Internet of Things (IoT) technology is becoming more popular, the information that is being generated is in vast quantity and rich with sensitive information and that information has to be shared and processed safely within the cloud infrastructure. However, privacy preserving methods like secret sharing and secure multi-party computation need several encrypted shares to be exchanged with various nodes making the communication overhead of IoT devices large. This paper presents a safe data sharing model of the IoT systems, that is effective in reducing the overhead of communication. The combination of the performance advantages of the Elliptic Curve Cryptography (ECC) on secure key exchange, NTRU cryptosystem on efficient post-quantum cryptography, and a threshold based secret sharing scheme on distributed key management has been achieved in this respect. proxy re-encryption in this system also ensures that the key shares are distributed safely to the authorized stakeholders without causing harm to the original secret information. Thereby only one time the data are encrypted and therefore several key fragments are delivered to the receiver rather than several ciphertexts, thereby, reducing the overhead of communications but at the same time having high data confidentiality.

Keywords: Internet of Things, Data Sharing, Elliptic Curve Cryptography, NTRU Cryptosystem, Secret Sharing, Proxy Re Encryption.

Abstract: Phishing represents a critical cybersecurity threat where attackers deploy fraudulent websites to deceive users into disclosing sensitive personal information, such as banking credentials and passwords. Traditional blacklist-based detection methods are largely ineffective against new and rapidly changing phishing URLs. This research proposes an automated, intelligent detection system that utilizes machine learning to identify both known and unknown phishing websites by analysing 30 structural URL characteristics. A dataset of labelled legitimate and phishing URLs was sourced from the UCI Machine Learning Repository and PhishTank to train the system. Multiple classification algorithms were evaluated, including Logistic Regression, Decision Tree, and Random Forest, with the XGBoost (Extreme Gradient Boosting) classifier emerging as the optimal model, achieving a peak accuracy of 96.88%. The final trained model was integrated into a real-time web application developed using the Streamlet framework, providing a scalable and efficient solution for proactive cybersecurity.

Keywords: Phishing Detection, Machine Learning, XGBoost, Cybersecurity, URL Analysis, Feature Extraction.

Abstract: In response to the need for strengthened mathematics instruction under recent curriculum reforms, this study examined how the National Mathematics Program (NMP) supports the effective implementation of the Matatag Curriculum in public secondary schools. Employing a mixed-method explanatory sequential design, the research involved 388 Grade 7 students from 103 secondary schools in Capiz during SY 2024-2025, supplemented by Focus Group Discussions. Results showed that the overall and top five areas of NMP implementation, barriers, impacts, and achievement were perceived at a moderate level. Significant differences in achievement were found across varying levels of program implementation, barriers, and impacts. Moreover, implementation, barriers, and impacts were significantly interrelated and demonstrated meaningful relationships with student achievement. Regression analysis identified implementation quality, prevailing barriers, and program impact as significant predictors of achievement. Based on these findings, an Enhanced Teacher Training Program was developed to strengthen pedagogical competence, contextualized instruction, assessment strategies, and instructional resource utilization, thereby reducing barriers and improving mathematics achievement under the Matatag Curriculum.

Abstract: This paper presents an ElGamal-based asymmetric updatable encryption scheme designed to tackle the challenges associated with secure key rotation in cryptographic systems. The proposed method allows ciphertexts encrypted under a previous key to be securely and efficiently transitioned to a new key without the need for decryption, thereby preserving data confidentiality and integrity. By exploiting the mathematical characteristics inherent in ElGamal encryption, the scheme supports unlimited key update iterations, asymmetric encryption functionality, and is independent of specific ciphertext formats. Lightweight pseudorandom generators (PRGs) are employed to ensure secure and efficient handling of the random values necessary during encryption and re-encryption operations. The approach guarantees strong forward and backward security, protecting against data leakage even if keys are compromised. Extensive performance assessments demonstrate its efficiency, showing minimal computational and communication overhead, making it well-suited for both large-scale infrastructures and environments with limited resources. Additionally, comparative studies confirm its advantages over existing methods in terms of encryption speed, ciphertext update duration, and scalability. Overall, this work offers a practical and secure solution for frequent key management across various applications, including cloud storage, Internet of Things (IoT) devices, and secure communication networks.

Keywords: Blockchain, Differential Privacy, E-Voting, Ethereum, Smart Contracts, Voter Authentication

Abstract: Pesticides play a crucial role in contemporary agricultural practices, serving to manage pests, weeds, and diseases that can diminish crop yields. The application of pesticides is crucial for augmenting agricultural productivity and ensuring food security; however, their overuse or misuse has engendered significant apprehension concerning environmental contamination and public health. Pesticide residues can spread through soil, water, air, and food. This can potentially harm the environment and reduce biodiversity. People exposed to pesticides can experience immediate poisoning, neurological problems, reproductive issues, and an increased risk of cancer. This review analyses the environmental and public health effects of pesticides, based on a thorough examination of existing research. The text dives into the mechanisms of pesticide toxicity, the pathways of exposure, and sustainable management strategies, such as integrated pest management and biological control. The investigation underscores a pressing need for safer ways to apply pesticides, along with tougher rules and a shift toward more environmentally friendly options. These measures are essential to mitigate environmental pollution and protect public health.

Keywords: Pesticides, Environmental pollution, Human health, Toxicity, Integrated pest management

Abstract: The growing demand for compact, eco-friendly, and portable refrigeration systems has encouraged the development of thermoelectric cooling technology. This project presents a design and development of a thermoelectric refrigerator using a TEC1-12706 Peltier module for small-scale cooling applications. The system operates based on the Peltier effect, where heat is transferred from one side of the module to the other when electric current passes through it.The proposed setup consists of a TEC1-12706 Peltier cooler kit, 12V cooling fans, and a 12V 200W SMPS power supply to drive the thermoelectric module efficiently. A conductive type 12V 100W PTC ceramic air heater with fan is integrated on the hot side to regulate and dissipate heat effectively. A mechanically fabricated insulated box is designed to maintain the cooling chamber and improve thermal efficiency. To monitor system performance, two electronic temperature detectors (thermometers) are installed-one to measure the temperature at the hot side and another at the cold side of the Peltier module.When power is supplied, the Peltier module creates a temperature difference, causing one side to become cold while the opposite side becomes hot. The cooling fan helps circulate cold air inside the chamber, while the heater-assisted heat dissipation unit ensures proper heat removal from the hot side. The system provides a compact, vibration-free, and environmentally friendly refrigeration solution without the use of refrigerants.This thermoelectric refrigeration system is suitable for portable cooling applications, food preservation, medical storage, and small electronic cooling systems, offering advantages such as low maintenance, simple construction, and eco-friendly operation.

Keywords: Thermoelectric Refrigeration, Peltier Effect, TEC1-12706 Peltier Module, Thermoelectric Cooling System, Compact Refrigerator, Solid State Cooling, PTC Ceramic Heater, Temperature Monitoring System, Electronic Temperature Detector, SMPS Power Supply, Heat Dissipation System, Portable Cooling Device.

Abstract: Cardiovascular diseases remain the leading cause of mortality worldwide, necessitating accurate early prediction systems. This paper proposes HybridBoost, a novel XGBoost-SMOTE ensemble framework designed to address class imbalance in the UCI Heart Disease dataset through strategic oversampling and feature optimization. Unlike existing approaches that achieve approximately 92-94% accuracy, HybridBoost attains 96.8% accuracy, 0.96 F1-score, and 0.98 AUC-ROC through 5-fold cross-validation. The proposed methodology integrates Recursive Feature Elimination (RFE) with SMOTE oversampling (1:1 ratio) prior to classification using XGBoost (n_estimators = 100, max_depth = 6). Comparative analysis against Random Forest, AutoML, and SMOTE-ENN-XGBoost demonstrates a 3-5% improvement in performance. Feature importance analysis identifies chest pain type (cp), maximum heart rate (thalach), and ST depression (oldpeak) as the primary predictors of heart disease. These findings are consistent with established clinical indicators reported in previous studies. HybridBoost advances precise binary heart disease classification, moving beyond multiclass approaches, heart failure-specific models, and generic ensemble methods. The results highlight its potential for clinical decision support and future deployment in healthcare environments.

Keywords: Heart disease prediction, XGBoost, SMOTE, ensemble learning, class imbalance, UCI dataset, feature selection, cardiovascular risk assessment

Abstract: Choosing an appropriate career path after completing secondary education is one of the most challenging decisions for students. Many students face confusion due to the lack of proper guidance and centralized information regarding courses, colleges, entrance examinations, and career opportunities. Traditional career counseling methods are often limited in accessibility and may not provide structured guidance to all students. To address this issue, this paper presents a web-based platform called Higher Education Mentor, which provides a structured educational roadmap for students after the completion of 10th class or Intermediate education. The proposed system integrates information about courses, career paths, entrance examinations, and colleges into a single platform to help students explore their academic opportunities effectively. The platform is developed using modern web technologies including Next.js and React for the frontend, Spring Boot for the backend, and MongoDB for database management. The system allows users to select their educational background and interests and then generates a step-by-step roadmap that guides them toward suitable higher education options. The developed platform offers a user-friendly interface, scalable architecture, and efficient information retrieval system. By providing centralized and structured educational guidance, the system helps students make informed decisions regarding their future academic and career paths.

Keywords: Career Guidance System, Higher Education Mentor, Educational Roadmap, Web Application, Next.js, Spring Boot, MongoDB.

Abstract: Parkinson's disease is the second most common progressive neurodegenerative disorder, after Alzheimer's disease caused by the selective loss of dopaminergic neurons in the substantia nigra region of midrain. Clinical symptoms include resting tremor, bradykinesia, postural instability and rigidity. Although ageing, genetic and environmental factors have been suggested as the putative risk factors for the progression of the disease, the exact molecular mechanisms involved in the pathogenesis of the disease remains elusive. Mitochondria are the power house of the cell where generation of the ATP takes place. Recent studies have shown the importance of the mitochondrial dysfunction in the pathogenesis of the disease. A number of the mitochondrial neurotoxin like MPTP, paraquat, rotenone etc has been shown to induce the PD like features. Dysfunctional mitochondria release the cytochrome c in the cytosol which results in the death of the neurons through the apoptosis. Thus clearance of the dysfunctioned mitochondria can be regarded as a key event in the prevention of the PD. A number of mechanisms including fission, fusion, mitochondrial derived-vesicles formation have been reported in the maintainance of the mitochondrial quality of which mitophagy is most important in the clearance of the defective mitochondria. Mitophagy refers to the removal of the defective mitochondria through autophagy. Parkin and PINK 1 are two proteins function cooperatively in the clearance of the defective mitochondria during autophagy. A strong link of PINK 1/Parkin has been suggested with mitochondrial dysfunction, mitochondrial vesicular trafficking, mitochondrial dynamics, quality control and mitophagy. Pink1 is a ser- thr kinase which fuctions as sensor and detects the dysfunctioned mitochondria, after which it recruits cytosolic Parkin having E3 Ubiquitin ligase activity to depolarized mitochondria. Parkin then ubiquitinates the dysfunctiond mithochondria and directs them towards the autophagy for the clearance. Defects in the lysosomal-autophagy pathway or mutations in the PINK/ Parkin may result in poor clearing of the dysfunctioned mitochondria that ultimately results in the generation of the ROS generation and PD pathogenesis.

Keywords: Parkinson disease, Mitochondria, Autophagy, PINK1, Parkin, Mitochondria, Mitophagy, VDAC1, MDVs, UPS.

Abstract: The rapid growth of digital commerce has significantly transformed consumer purchasing behaviour, particularly in the fashion retail sector. Online platforms provide convenience, variety, and attractive promotional strategies that often stimulate impulse purchases among consumers. This study examines the factors influencing impulse buying behaviour among young consumers in online fashion retail. The research focuses on elements such as promotional offers, social media influence, website design, and ease of payment. Data were collected from young consumers through a structured questionnaire and analysed using descriptive statistical methods. The findings reveal that digital marketing strategies, time-limited offers, and social media exposure strongly influence impulse buying tendencies. The study provides insights for fashion retailers to design effective marketing strategies to attract young online shoppers.

Keywords: Impulse buying, online fashion retail, consumer behaviour, digital marketing, young consumers.

Abstract: Zero-Knowledge Proofs (ZKP) present a transformative approach to data privacy, enabling secure data sharing without revealing the underlying information. This paper explores the application of ZKP in secure data sharing, emphasizing its role in enhancing privacy, reducing data breaches, and ensuring compliance with modern data protection regulations. By analyzing real-world scenarios and conducting simulations, this study highlights the potential of ZKP in establishing trust in data ecosystems.

Abstract: The stock market plays an important role in economic development by facilitating capital formation and providing investment opportunities. Stock market indices serve as key indicators of overall market performance. In India, benchmark indices such as the Nifty 50 and BSE Sensex are widely used to evaluate trends and movements in the equity market. These indices represent leading large-cap companies listed on the National Stock Exchange and the Bombay Stock Exchange and are considered major barometers of the Indian stock market. This study aims to evaluate and compare the performance of Nifty 50 and Sensex by examining their risk-return characteristics. The study is based on secondary data collected from the official databases of NSE and BSE. Daily closing price data for the period 1 January 2021 to 31 December 2025 were used for the analysis. Statistical tools such as mean return, standard deviation, coefficient of variation, and correlation analysis were applied to assess the performance and volatility of the indices. The results indicate that Nifty 50 recorded slightly higher mean returns compared to Sensex during the study period. The standard deviation results show that both indices exhibit similar levels of volatility, although Sensex experienced marginally higher fluctuations. The coefficient of variation suggests that Nifty 50 provides better risk-return efficiency. In addition, correlation analysis reveals a very strong positive relationship between the two indices, indicating that they move closely together. Overall, the findings suggest that both indices effectively represent the performance of the Indian equity market.

Keywords: Nifty 50, BSE Sensex, Risk-Return Analysis, Volatility, Indian Stock Market.

Abstract: Artificial Intelligence (AI) has rapidly evolved from a niche technological innovation into a mainstream global phenomenon. This paper examines the growth of AI adoption worldwide using secondary data from international reports, institutional surveys, and industry analyses. Findings indicate exponential growth in both individual and organizational adoption, with approximately one in six people globally using AI tools by 2025. The study highlights key drivers such as digital transformation, productivity gains, and generative AI breakthroughs, while also addressing regional disparities and challenges. The paper concludes with implications for policymakers, businesses, and future research.

Keywords: Artificial Intelligence, Growth Trends, Key drivers, AI Tools, AI Adoption Trends

Abstract: The practical necessity of child-centred education system has led to a major change in the education system all over the world. The curriculum is being presented to the children in line with real life, so that they can play an active role and participate in their own progress. Keeping this kind of modern education system in mind, the National Education Policy 2020 has emphasized hands-on practice from the foundation level of the education system. By prioritizing children's thinking, curiosity, creative exploration, interest, etc., elements like experiential learning and critical thinking have been given place in the education policy. In this study, an attempt has been made to explain how the child's spontaneous and active participation helps in cognitive development, reflective observation being one of the important elements of child-centred education. Through this element, the student can become aware of his problems and how he can improve his problem-solving skills. Finally, it explains how to activate the classroom and knowledge system by combining both experiential learning and critical thinking to make teamwork fruitful. This study reveals that, active engagement has a positive effect on increasing the confidence, curiosity, and creativity of students. Hands on activities make children possess the power of reflective observation. That is, active engagement directly and indirectly helps in acquiring problem solving skills. Critical thinking is necessary to find multifaceted solutions to an incident or problem by going deep into that problem for the purpose of finding a suitable solution.

Keywords: experiential learning, critical thinking, rote learning, NEP-2020, reflective observation.

Abstract: Managing academic class schedules and teaching assignments is a combinatorially complex problem that continues to challenge educational institutions worldwide. For school administrators, optimally assigning faculty to courses---while simultaneously balancing expertise, availability, and workload constraints---can require weeks of iterative trial and error. Although digital tools exist, many institutions still depend on fragmented spreadsheet-based processes. This paper presents a unified, web-based Integrated Timetable Scheduling and Faculty Workload Management System designed to address these challenges through formal Constraint Satisfaction Problem (CSP) modelling. The system leverages intelligent constraint enforcement, Minimum Remaining Value (MRV) heuristics, and forward-checking search strategies to generate conflict-free, balanced timetables efficiently. A real-time dynamic substitution algorithm handles faculty absences automatically by identifying optimal replacements based on subject expertise and current workload. The system was implemented as a fully functional web application --- the Academic ERP --- and validated through both algorithmic simulation and live system testing. Experimental evaluations confirm that the system eliminates scheduling conflicts entirely, reduces workload variance by approximately 85%, and generates timetables in under ten seconds on standard institutional hardware.

Keywords: Timetable Scheduling, Faculty Workload Management, Constraint Satisfaction Problem, CSP, Backtracking Search, MRV Heuristic, Substitute Allocation, Academic ERP, Educational Technology.

Abstract: Sustainable agriculture plays a vital role in ensuring global food security, environmental protection, and economic stability. Accurate crop yield prediction enables farmers and policymakers to make informed decisions regarding crop planning and resource management. This paper proposes a machine learning approach for crop yield prediction using the Extreme Learning Machine (ELM), a fast-neural network model with strong generalization capability. The dataset includes soil nutrients such as Nitrogen, Phosphorous, and Potassium along with environmental parameters like temperature, humidity, and soil type. Categorical variables are encoded and missing values are handled using median replacement. The ELM model uses a Single Layer Feedforward Network with randomly generated input weights and output weights calculated using the Moore-Penrose pseudoinverse. Performance is evaluated using RMSE and R² Score. A Streamlit-based interface is developed to allow users to input parameters and receive instant crop yield predictions.

Keywords: Crop Yield Prediction, Extreme Learning Machine (ELM), Machine Learning, Sustainable Agriculture, Soil Parameters, Climatic Factors, Precision Agriculture, Agricultural Technology, Streamlit, Real-time Prediction

Abstract: Anthropometric characteristics play a vital role in determining physical fitness and sports performance. Among these characteristics, chest muscle circumference reflects upper body muscular development and respiratory efficiency. The present study aimed to compare chest muscle circumference between tribal and non-tribal sportsmen of Goa with reference to two age groups (21-25 years and 26-30 years). A total of 100 male sportsmen were selected for the study, comprising 50 tribal and 50 non-tribal players from different sports disciplines in Goa. The subjects were further categorized into two age groups. Chest muscle circumference was measured using a standard steel measuring tape following accepted anthropometric procedures. The collected data were analyzed using mean, standard deviation, and independent t-test to determine the significance of differences between groups. The findings of the study revealed that non-tribal sportsmen demonstrated slightly higher mean values of chest muscle circumference compared to tribal sportsmen in both age groups; however, the differences were not statistically significant at the 0.05 level of significance. The results suggest that participation in sports activities contributes to similar levels of muscular development among both tribal and non-tribal athletes regardless of age category. The study highlights that sports training and physical activity may minimize anthropometric differences between different ethnic groups.

Keywords: Anthropometry, Tribal athletes, Non-tribal athletes, Chest muscle circumference, Age groups, Sports science

Abstract: Dyes are extensively used in industries such as textiles, food, leather, and plastics and their discharge into the environment causes serious ecological hazards. Dye-contaminated water alters color, depletes dissolved oxygen, and disrupts aquatic ecosystems. This study investigates the adsorption potential of Moringa oleifera seed and flower powders, in both raw and activated forms, for the removal of textile dyes from industrial effluents. These plant-derived biosorbents were prepared and thermally activated to enhance surface characteristics and functional group availability, thereby improving dye interaction. Adsorption experiments were systematically performed under varying pH, adsorbent dosage, and contact time to establish optimal operational conditions. Results indicated that the activated biosorbents exhibited notably higher dye removal efficiency than their raw counterparts, attributed to enhanced surface area, porosity, and active functional sites. Total Dissolved Solids (TDS) analysis further validated the results, showing a marked reduction in dissolved contaminants and improvement in water quality after treatment. This study thus demonstrates that thermal activation significantly improved the adsorption performance and surface reactivity of Moringa oleifera-based biosorbents. Due to their low cost, renewability, and environmental safety, these biosorbents present a sustainable and effective alternative for treating dye-contaminated textile effluents, contributing to eco-friendly wastewater management and sustainable industrial practices.

Keywords: Moringa oleifera, Biosorbents, Activated Carbon, Dye removal, Textile effluent, Adsorption

Abstract: Road safety is still a major global concern, with many accidents being caused by inexperienced drivers,risky driving practices and lack of knowledge of traffic laws.Textbooks and lectures in the classroom are the mainstays of traditional driver education approaches, which frequently fall short of offering interesting and useful learning opportunities. In order to enhance traffic rule comprehension through interactive simulation and game-based learning, this study introduces the Road Rules Simulator, a gamified learner licensing training system. To provide an interesting training environment, the system combines scenario-driven driving simulations created with the Unity engine and C# programming with quiz-based learning. To improve student motivation and knowledge retention, the suggested system includes gamification components like levels, scoring, badges, awards, and real-time feedback. Additionally, it has analytics and progress tracking components that analyze student performance, identify infractions, and offer remedial advice. Without the dangers of the real world, the simulator offers a secure virtual setting where students can hone their driving abilities and follow traffic laws. Combining simulation and gamification enhances decision-making skills, engagement, and rule comprehension, according to experimental application. For contemporary driver education, the system offers an affordable, scalable, and accessible option.

Keywords: Road Rules Simulator, Driving Simulation, Gamified Learning, Traffic Rule Education, Driver Training System.

Abstract: Heart disease is among the top causes of death around the world, and catching it at an early stage can make a real difference in how patients are treated. The problem, though, is that many machine learning models built for this purpose do not work well because they are trained on limited data, and often the dataset itself is skewed - meaning there are far more healthy patients than sick ones. In this work, we tried to fix this by bringing together two ideas: synthetic data generation using CTGAN, and a stacking ensemble classifier. We first used CTGAN to produce new, realistic patient records that mirror the original data's patterns, then trained a stacking model - XGBoost, Random Forest, and Gradient Boosting as base learners, with Logistic Regression on top - on the combined real and synthetic dataset. When we tested it, the model hit 92% accuracy and beat the basic XGBoost model on every metric we checked, including precision, recall, and AUC. The takeaway is simple: adding synthetic data and stacking classifiers together noticeably strengthens cardiovascular risk prediction.

Keywords: Cardiovascular Disease, CTGAN, Synthetic Data Augmentation, Stacking Ensemble, Machine Learning

Abstract: The rapid growth of digital fashion retail platforms has created a growing demand for recommendation engines capable of addressing the unique cultural, regional, and aesthetic diversity of Indian traditional attire. Existing systems predominantly rely on Western clothing datasets and fail to capture the nuanced interplay between Indian skin tones, body morphologies, regional textile traditions, and occasion-specific norms. This paper presents a comprehensive AI-based multimodal fashion recommendation system specifically designed for Indian traditional and regional clothing. The proposed architecture integrates a multimodal computer vision API to extract key biometric attributes - specifically skin tone classification from face photographs and body morphology estimation from full-body images - into a five-class skin tone schema and a seven-category body shape taxonomy. A custom rule-based content-based filtering engine, conditioned on an eight-dimensional user profile vector, then maps extracted biometric features to regionally appropriate Indian outfits across four geographic zones and eight occasion categories. An AI-assisted natural language generation module further enriches outputs with culturally contextualised descriptions, styling guidance, and personalised notes. The system is deployed as a full-stack web application using a React (Vite) frontend and a Python Flask backend communicating over a REST API. Experimental evaluation confirms sub-150 millisecond offline recommendation latency, strict cultural accuracy enforced by regional boundary filters, and positive user acceptance in preliminary testing. Comparative observations confirm that the decoupled biometric-extraction and recommendation architecture substantially reduces cultural misclassification over general-purpose generative approaches.

Keywords: Indian Fashion Recommendation; Computer Vision API; Content-Based Filtering; Skin Tone Detection; Body Morphology Classification; Regional Outfit Database; Flask; React; Multimodal AI; Cultural Recommendation System.

Abstract: Traditional attendance systems are time-consuming and prone to manual errors. This paper proposes an automated one-click attendance system using deep learning-based face recognition. The system utilizes dlib's pre-trained ResNet face embedding model to convert facial images into 128-dimensional numerical vectors. Euclidean distance is applied to measure similarity between stored student embeddings and real-time classroom images. A tolerance threshold is used to classify students as present or absent. The system integrates a Streamlit-based user interface for image upload and attendance visualization. Experimental results demonstrate reliable recognition performance under controlled lighting conditions. The proposed framework provides a fast, accurate, and scalable solution for automated classroom attendance management.

Keywords: Face Recognition, Attendance System, Deep Metric Learning, Euclidean Distance, dlib, Streamlit.

Abstract: DocCrypt is a decentralized document management system that integrates Blockchain technology, InterPlanetary File System (IPFS), and Natural Language Processing (NLP). The system allows users to securely upload documents, generate cryptographic hashes, and store files in a decentralized IPFS network while recording metadata on the Aptos blockchain to ensure transparency and tamper-proof verification. Additionally, NLP models enable users to query documents using natural language and receive contextual answers from document content. This approach improves document security, accessibility, and verification compared to traditional centralized storage systems.

Keywords: Blockchain, IPFS, Document Security, Natural Language Processing, Decentralized Storage.

Abstract: The transportation and storage of hazardous chemicals in ISO tankers pose significant safety risks, particularly in coastal port environments where accidental releases can affect both industrial facilities and nearby populations. This paper includes, consequence and dispersion analysis of the effects of accidents such as leaks of Styrene Monomer and n-Hexane in the area through the ALOHA (Areal Locations of Hazardous Atmospheres) computer modeling software. On a representative meteorological coastal environment, the Terminals located at the Kamarajar Port at India was simulated. The Acute Exposure Guideline Levels (AEGLs) was used to assess toxic threat areas, whereas lower explosive limit (LEL) levels and vapor cloud explosion overpressure levels were used to determine flammability and explosion dangers. It is observed that Styrene Monomer is unsafe mainly due to localized yet serious toxic risks because it is low in volatility and heavy, but n-Hexane exhibits widespread flammable and explosion risks because it is of high volatility and evaporates quickly. The comparative analysis indicates the effects of chemical properties and the environment on the dispersion behavior. This paper reveals the efficiency of ALOHA in supporting emergency preparations, risk prevention, and safety control within the scope of the chemical transportation and port operations.

Keywords: ALOHA; Chemical Dispersion Modeling; ISO Tanker; Styrene Monomer; n-Hexane; Risk Assessment; AEGL; Flammable Vapor Cloud; Port Safety.

Abstract: A commonly used distillate fuel in domestic, industrial and aviation associated applications is kerosene; nevertheless, the fact that it is only naturally conductive to electricity means that the potential of accumulation of static charges is very high during manipulations and pumping and transfer processes. Low-conductivity fuels may have a risk of fire and explosion due to the nature of the static discharge. This paper encompasses a comparative experimental research on five additive categories Stadis 450, AOT (dioctyl sodium sulfosuccinate), PEG-400, an ionic improver salt and carbon black nanoparticles in enhancing the electrical conductivity of kerosene. ASTM D2624 was utilized in conductivity measurement; ASTM D445 was used to measure viscosity and ASTM D56 was used to measure flash point. There was also a 30 days stability test to determine the performance after a long period. Findings have revealed the carbon black nanoparticles(5 ppm) to have the most significant effect on conductivity enhancement(220 pS/m) followed by Stadis 450 (180 pS/m) and ionic improver salt(150 pS/m) to have the second and third highest effect, respectively, in moving the kerosene to the safe conductivity range. Notably, all of the additives did not lead to major variations in viscosity or flash point. The results obtained prove that proper conductivity enhancers can significantly increase kerosene safety, without sacrificing the basic fuel characteristics, which can be useful in the management of fuel storage, transportation, and industrial safety.

Keywords: Electrical Conductivity; Kerosene; Static Electricity; Conductivity Improvers; Carbon Black Nanoparticles; ASTM Standards; Fuel Safety

Abstract: Efficient material utilization is a critical challenge in sheet metal manufacturing industries, particularly in CNC-based cutting environments. Traditional manual nesting approaches often result in suboptimal placement of irregular parts, leading to excessive material wastage, higher production costs, and increased operational time. This project presents an intelligent web-based sheet metal optimization system that automates geometric extraction and implements AI-driven heuristic nesting algorithms to improve sheet utilization efficiency. The proposed system accepts DXF (Drawing Exchange Format) files as input, extracts geometric entities, computes bounding boxes, and performs collision-free placement using optimized space partitioning logic. The AI-based nesting engine arranges irregular shapes within predefined sheet dimensions while dynamically allocating additional sheets when necessary. The system provides real-time visualization of part placement, material efficiency computation, and detailed wastage reports. Experimental validation demonstrates significant improvement in sheet utilization and reduction in material waste compared to conventional manual nesting practices. The framework is particularly beneficial for small and medium scale industries.

Keywords: Web-Based Manufacturing Optimisation, Sheet Metal Nesting, DXF Geometry Extraction, Heuristic Optimisation, Irregular Shape Packing, Material Utilisation, and Collision Detection.

Abstract: This paper presents an Industry 4.0-based Smart Yarn Monitoring and Alert System designed to enhance efficiency, accuracy, and automation in textile production environments. The proposed system utilizes an ESP32 microcontroller with integrated Wi-Fi capability to enable real-time monitoring and remote data access. An E18-D80 infrared (IR) sensor is employed for continuous yarn thread detection, allowing immediate identification of thread breakage. Additionally, a proximity sensor is used to count roller rotations, enabling indirect estimation of yarn weight through calibrated per-rotation measurements.The system provides both local and remote monitoring functionalities. A 16×2 LCD with I2C interface displays real-time parameters such as yarn status, rotation count, and estimated weight. Simultaneously, data is transmitted to the Blynk IoT platform, facilitating live monitoring and instant alert notifications via a mobile application. The hardware is powered by a 12V LiFePO₄ battery, with an LM2596 buck converter ensuring stable voltage regulation for system components.The proposed solution offers a cost-effective, reliable, and scalable approach for automated yarn production monitoring. It significantly reduces manual intervention, minimizes material wastage, and improves operational productivity. The system demonstrates strong potential for deployment in smart textile industries, aligning with Industry 4.0 principles of digitalization, real-time analytics, and remote accessibility.

Keywords: ESP32, Internet of Things (IoT), Yarn Monitoring System, Industry 4.0, Infrared Sensor, Proximity Sensor, Real-Time Monitoring, Smart Textile System, Weight Estimation, Blynk Platform, Automation, Wireless Monitoring, Fault Detection.

Abstract: Carbon storage facilities are essential in carbon sequestration and mitigation of climate change but the storage of carbonaceous substances brings with them high risk of fire occurrences in the form of smoldering combustion, spontaneous ignition and explosions of combustible dusts. The traditional fire detectors usually are not able to give early alerts because the smoke takes time to develop and the ventilation in storage facilities is complicated. The paper describes the improved fire detection system of carbon storage facilities based on Fire Dynamics Simulator (FDS). The simulations are carried out using computational fluid dynamics where the growth of the fire, propagation of smoke, temperature and the evolution of the gas species are analyzed under different fire conditions. The detector response behavior is tested by combining multi-sensor models, which consist of smoke, heat, and gas sensors to determine the best sensor placement and detection threshold. The outcome of the simulations indicate that using gas based and multi sensor detection schemes would save a great deal of time, relative to the conventional smoke only systems in terms of detection time. The suggested solution offers a performance-based model of efficient and reliable fire detection systems design in carbon storage processes.

Keywords: Fire Dynamics Simulator (FDS), Carbon Storage, Fire Detection Systems, Computational Fluid Dynamics (CFD), Multi-Sensor Detection, Combustible Dust Fire

Abstract: Power Quality Analyzers (PQA) are essential for monitoring and maintaining electrical networks. These assist in identifying changes in electrical measurements due to load fluctuations and other power quality factors. Traditional monitoring systems are mostly real-time, however they do not have automated alerts which would help in fault detection and corrective actions. The system is built on top of a Raspberry Pi Pico W microcontroller to which the voltage and current sensing modules, OLED display, and communication interfaces are connected. Wi-Fi gateway is used for passing the data obtained by different sensors. The transmitted and stored Data can be processed in further with an opensource cloud platform such as ThingSpeak to help monitor different aspects like IR status, temperature etc. SIM900A GSM module is used to alert the users in real-time as soon as any abnormal power event or voltage fluctuation occurred over the system in order to ensure higher reliability of the system. This ensures timely action and aids in avoiding the destruction of electrical devices. Experimental Results: The proposed system performance is compared with conventional measuring apparatuses like FPGA-based Power Quality Analyzers and state-of-the-the art Fluke meters under different load conditions. The results collected indicate that the developed system enables correct measurements and monitoring. The low-cost IoT design-based alternative proposed represents a simple method for managing the analysis and enhancement of power quality in home, business, and industrial electrical applications.

Keywords: Internet of Things (IoT), Raspberry Pi Pico W, Voltage and Current Sensors, GSM Alert System, Remote Data Monitoring, Cloud Platform (ThingSpeak)

Abstract: Lithium-ion batteries have been used in batteries of electric vehicles, consumer electronics, and industrial test devices in a large degree because of their large energy density and their charging ability for many times. Cyclic and repetitive testing Environmental conditions the Charge Discharge Charge (CDC) testing environments expose the batteries to repeated cycling, large current loads, intense and many state-of-charge transitions, and very high thermal and electrochemical stress levels. Such conditions enhance a threat of thermal runaway that is a severe safety issue that involves the uncontrolled heat generation resulting in a fire or an explosion. In this paper, the phenomenon of thermal runaway in the work of lithium-ion batteries in CDC condition has been developed in detail. The experimental assessment was then repeated at various C-rates and ambient temperatures using real-time interactive thermal imaging, embedded thermocouples, and philosophical voltagecurrent in determining the initial signs of C - indication of thermal inconsistency. The paper also suggests a whole set of passive temperature control and active change on the battery management system (BMS), such as the dynamical feedback control on the temperature, active current control and the outlier forecasting etc. The findings indicate that the traditional BMS protection plans cannot be used in long-term situations of CDC testing. The suggested multilayer prevention approach is an efficient tool that will eliminate the threat of thermal runaway and enhance operational safety and test reliability. The results set a practical idea on how to test batteries more safely, and their practical advice can be used in longer-term solutions to a larger-scale approach to improving thermal performance in real-world lithium-ion based battery systems.

Keywords: Lithium-ion batteries; Thermal runaway; Charge-discharge cycling; Battery management system; Thermal monitoring; Energy storage safety.

Abstract: In recent years, Environmental, Social, and Governance (ESG) investing has gained significant traction as investors increasingly prioritize ethical considerations alongside financial returns. This study investigates the impact of ESG criteria on mutual fund performance in the Indian financial market, with a focus on return, risk, and risk-adjusted metrics. Using secondary data from selected Indian ESG mutual funds, the study applies correlation and regression analysis to evaluate the relationship between ESG ratings and fund performance. The findings indicate a weak correlation between ESG ratings and short-term returns, but a strong positive correlation with long-term (3-year) returns. Regression analysis further suggests that higher ESG ratings are associated with improved risk-adjusted returns, as measured by the Sharpe Ratio. While no consistent linear relationship was found between ESG scores and risk indicators such as beta or standard deviation, ESG integration appears to contribute to more stable fund performance over time. This research provides empirical evidence supporting ESG investment as a sustainable and potentially rewarding strategy in the Indian context.

Keywords: ESG Investing, Mutual Funds, Fund Performance, Risk-Adjusted Returns, Sharpe Ratio, ESG Ratings, Sustainable Finance, Volatility, India, Ethical Investment

Abstract: This paper introduces an AI-based personalized learning platform that aims to enhance student performance and support better teaching decisions using data-driven techniques. The system is developed as an Android application using Java/XML and is connected with Firebase Real time Database for efficient data handling. It uses Generative AI to customize study content, tests, and revision plans based on each student's learning progress. The platform continuously monitors quiz results and user activity to provide accurate recommendations and identify areas where students need improvement. Faculty members are supported with dashboards that offer insights into student performance, learning gaps, and potential academic risks. In addition, a 24×7 AI chat bot is available to solve student queries instantly, improving the overall learning experience. Interactive dashboards also help visualize progress in a simple way. Overall, the system provides a smart, scalable, and efficient solution for modern digital education by combining personalization, analytics, and AI technologies.

Keywords: Artificial Intelligence, Personalized Learning, Adaptive Education, Generative AI Chatbot, Firebase Real time Database, AI Recommendation System, Student Analytics, Android E Learning Platform

Abstract: This study investigates the design, fabrication, and performance of a small-scale horizontal-axis windmill integrated with dual-purpose functionality: electricity generation and water pumping. By leveraging renewable wind energy, the system captures kinetic energy through a 3D-printcd aerodynamic rotor and converts it into mechanical energy, which is subsequently transformed into regulated electrical power. The system incorporates an energy storage and conditioning unit to ensure continuous operation under Fluctuating wind conditions, The stored energy is utilized to drive a diaphragm pump for water lifting, while simultaneously providing il regulated electrical supply for external louds, "The proposed system demonstrates a cost-effective, decentralized energy solution tailored for rural and off-grid regions where access to conventional power and water infrastructure is limited. Experimental evaluations confirm the feasibility of the integrated architecture, achieving stable power conversion and reliable water pumping performance. 'The study underscores the significance of modern fabrication techniques, such as 3D printing, in enhancing the modularity, accessibility, and scalability of small-scale renewable energy systems, The results provide a practical demonstration of how small-scale renewable infrastructure can contribute to sustainable development and resource management in resource -constrained areas.

Keywords: Wind Energy, 3D Printing, Renewable Energy, Electricity Generation, Water Pumping, Horizontal Axis Wind-mill, Decentralized Energy.

Abstract: Air pollution is a major global environmental and health concern due to rising levels of particulate matter and harmful gases. These pollutants cause serious respiratory and cardiovascular diseases. Conventional air purifiers depend on grid electricity, increasing cost and indirect pollution. This project proposes a Solar Air Purifier with an AQI Monitoring System. A solar panel generates energy, stored in a battery via a charge controller. The system powers a DC fan and a multi-stage filtration unit. It includes a pre-filter, HEPA filter, and activated carbon filter. The HEPA filter removes fine particles with high efficiency, while the carbon filter removes gases and odors. An air quality sensor monitors pollution levels and displays real-time AQI on an LCD. The system is eco-friendly, cost-effective, and suitable for various environments. Experimental results show improved air quality, making it a sustainable solution for healthier living.

Keywords: Solar Air Purifier, Air Quality Index (AQI), Renewable Energy, HEPA Filtration, Activated Carbon Filter, Air Pollution Control and Environmental Monitoring