VOLUME 12, ISSUE 8, AUGUST 2025

Abstract: The aims of this research are: 1) to determine the effect of the dose of liquid organic fertilizer from coconut water with bamboo shoots on the growth of chili plants, 2) to find out the best dose of liquid organic fertilizer from coconut water with bamboo shoots on the growth of chili plants, and 3) to determine whether or not the results of this research are suitable for use. as a learning resource in the form of Student Worksheets (LKPD) in growth material. This type of research is experimental research using a Completely Randomized Design (CRD). This research contained 4 treatments and 8 replications. Control with no dose of liquid organic fertilizer with coconut water with bamboo shoots, 3 treatments with different doses of liquid organic fertilizer with coconut water with bamboo shoots, namely, P1 (30 ml of liquid organic fertilizer with coconut water with bamboo shoots), P2 (40 ml of liquid organic fertilizer with water coconut with bamboo shoots), P3 (60 ml liquid organic fertilizer coconut water with bamboo shoots). The parameters observed in this research were the height and branch growth of chili plants. Data were analyzed using One-Way ANOVA (Normality, Homogeneity, Hypothesis and BNJ Test). Based on the research results, there is an effect of giving a dose of liquid organic fertilizer from coconut water with bamboo shoots on the growth of chili plants. Hypothesis test results show Fhit > F daf. The BNJ test further showed that P3 treatment had the best effect on the growth of chili plants. Based on the validation analysis of learning resources, this research is suitable for biology learning in the form of Student Worksheets (LKPD).

Keywords: Growth, Liquid, Learning Resources, LKPD.

Abstract: In response to the growing need for digitally competent educators in higher education, this study presents the development of a modular digital pedagogy curriculum designed to enhance teaching and learning effectiveness. The curriculum was created within the framework of a transnational Erasmus+ project, incorporating the contributions of academic experts from multiple institutions. It addresses key areas such as digital literacy, e-learning design, online as-sessment, and emerging technologies in education. Organized into flexible, competency-based modules, the curriculum supports both initial teacher training and continuing professional development. The design process was grounded in evi-dence-based practices, needs analysis, and iterative peer feedback. This article outlines the structure, rationale, and peda-gogical underpinnings of the curriculum, highlighting its potential to equip educators with the skills necessary for navi-gating digitally enriched teaching environments in higher education. The modular format allows for adaptive use across disciplines, institutions, and learning contexts, fostering scalable and sustainable integration of digital pedagogy.

Keywords: Digital pedagogy, modular curriculum, higher education, teaching competence, educational innovation.

Abstract: The rapid transformation of science, technology, engineering, and mathematics (STEM) industries has intensified the demand for agile, future ready professionals. Organizations now face the dual challenge of identifying competency gaps within their workforce and aligning training opportunities with emerging skills. Traditional professional development (PD) models are often static, generic, and unable to capture the dynamic nature of evolving STEM roles (Lent et al., 2017; Ainslie & Huffman, 2019; Bryson & Zimmermann, 2020). This paper proposes an AI driven competency gap analysis model that leverages accessible GPT-class language models to extract skills from employee records and industry role requirements, analyze competency gaps, and recommend targeted micro credentials for continuous professional development. Grounded in Social Cognitive Career Theory (SCCT) (Bandura, 1986; Lent et al., 2017) and building on established research demonstrating the impact of organizational support for development on workforce commitment (Tansky & Cohen, 2001; Chick & Vance, 2025), the framework utilizes GPT's natural language processing capabilities for comprehensive competency analysis and generates human-readable development plans aligned with identified skill gaps (Burke, 2002; Boud & Jorre de St Jorre, 2021). By connecting individualized skill gap insights to scalable learning solutions through an accessible, single-platform approach, this study offers a replicable model that contributes to workforce development theory and practice. The findings highlight how GPT-supported continuous professional development can drive skill relevance, employee engagement, and organizational retention in fast evolving technical fields while maintaining practical implementation feasibility.

Keywords: AI-driven competency analysis; GPT-based workforce development; accessible professional development systems; STEM skills gap analysis

Abstract: Music has long been recognized as a powerful tool for influencing emotional states and enhancing psychological well-being. With the advent of artificial intelligence and computer vision, it is now possible to tailor music experiences dynamically based on a user's current mood. This paper presents a novel music recommendation system that leverages facial emotion recognition to make accurate emotion-specific music suggestions. The system utilizes the CK+48 dataset, which comprises grayscale facial images classified into seven emotional states: anger, contempt, disgust, fear, happiness, sadness, and surprise. Two deep learning approaches were integrated: a Convolutional Neural Network (CNN) optimized for real-time webcam input and a ResNet-based transfer learning model for image uploads. The CNN model achieved an accuracy of 99.49%, whereas the ResNet model achieved 97.46%. Built with a Flask backend and responsive web frontend, the system enables seamless emotion detection and music playback. The proposed solution offers a more empathetic and context-aware alternative to conventional music players by aligning the musical output with the user's emotions in real-time.

Keywords: Facial Emotion Recognition, Music Recommendation System, Deep Learning, Convolutional Neural Network (CNN), Transfer Learning, Affective Computing.

Abstract: The value of user experience is greatly increased by personalized recommendations of content in the digital era. The conventional recommender systems are based on tastes or a long history of past users that can never be updated to reflect the current emotional desires in real-time. This study introduces an innovative system, which takes into consideration facial emotion recognition to provide dynamic and emotion-based recommendation related to the three categories of content, music, movies and books. This system allows providing contents that are harmonious with the emotional contexts of users by recognizing happiness, sadness, anger, and neutrality with the help of computer vision and deep learning.

Keywords: Facial Emotion Recognition (FER), Multi-Content Recommendation System,Deep Learning, Convolutional Neural Network (CNN), FER-2013 Dataset, Real-Time Emotion Detection, Content Personalization, Computer Vision, Adaptive User Experience.

Abstract: Air pollution is a global concern that has severe effects on the environment and public health. This research seeks to discuss various indoor and outdoor pollutants which are all predicted using the Auto Regressive Integrated Moving Average (ARIMA) model. An ARIMA model can accurately predict pollutant levels thus helping in making interferences aimed at improving air quality by employing historical information from specific sites within Agra City. Various error metrics determine the effectiveness of the ARIMA model in increasing awareness levels and the need for immediate action toward the reduction of harmful substances. The results indicate promising outcomes with Root Mean Square Error (RMSE) values around 1.358 for NO2, 2.2615 for SO2, and 1.2501 for PM10, respectively, hence suggesting that the predictions are highly accurate regarding the amount of pollutants present in the air. These findings have a significant effect on employing data-driven approaches to prevent air pollution as well as promoting environmental sustainability.

Keywords: Air pollution, ARIMA, World Health Organization (WHO), Pollutants, health risks

Abstract: Accurate diagnosis of lung and breast cancer is crucial for effective patient treatment and management. This study presents a novel framework that integrates hybrid clustering and Convolutional Neural Network (CNN) based classification for improved diagnosis of lung and breast cancer. The integration of hybrid clustering allows for the identification of intricate patterns within the lung and breast cancer datasets, while CNN ensures effective feature extraction and classification. The results verified the effectiveness of the proposed approach in accurately clustering and classifying lung and breast cancer instances. Classification results reveal a high level of accuracy for both lung and breast cancer datasets, with lung cancer achieving an accuracy score of 0.9847 and breast cancer reaching an accuracy score of 0.9986. Precision, recall, and F1 scores further validate the robustness of the approach. The proposed approach demonstrates promising potential for accurate cancer diagnosis and prognosis.

Keywords: Breast Cancer, Lung Cancer, Clustering, Classification, Data Mining

Abstract: This study investigates the natural convection flow and heat transfer characteristics of Powell-Eyring non-Newtonian fluids in a porous medium, incorporating the effects of internal heat generation or absorption. The governing equations-continuity, momentum, and energy-are formulated for an incompressible, laminar flow regime and transformed into a dimensionless form using similarity variables derived through group symmetry analysis. The Powell-Eyring constitutive relation introduces non-linear rheological behavior, transitioning smoothly to Newtonian fluid behavior under specific parameter limits. The resulting coupled non-linear ordinary differential equations are solved numerically using established methods such as the Runge-Kutta shooting technique. Parametric analyses are performed to assess the influence of key dimensionless parameters, including the Prandtl number, porous drag coefficient, heat source/sink strength, and Powell-Eyring fluid constants, on velocity and temperature profiles. The findings provide insights into thermal boundary layer behavior in complex rheological fluids and are applicable to engineering systems involving porous media heat transfer, such as geothermal reservoirs, polymer processing, and energy storage devices.

Keywords: Natural convection; Powell-Eyring fluid; Porous media; Heat source/sink; Similarity transformation; Non-Newtonian fluids; Boundary layer theory; Numerical solution.

Abstract: The paradigm of patient care is undergoing a significant transformation, moving from episodic, in-clinic assessments to continuous, remote monitoring facilitated by the Internet of Things (IoT). This shift is driven by a pressing need to manage patient care more efficiently, particularly for vulnerable populations and those with chronic illnesses, as conventional healthcare systems face increasing strain from limited resources and suboptimal nurse-to-patient ratios. This paper presents the design, implementation, and validation of a comprehensive, IoT-enabled solution for continuous health monitoring. The system integrates a suite of non-invasive biosensors, including Electrocardiogram (ECG), pulse oximetry, and temperature sensors, with a NodeMCU microcontroller for data acquisition and processing. Utilizing the Arduino IoT Cloud platform, the system transmits vital signs data wirelessly for real-time visualization and storage. Key functionalities include a robust alerting mechanism to notify caregivers of critical health abnormalities and a user-friendly, cloud-based interface accessible via mobile and web applications. The successful implementation demonstrates a practical and accessible solution that enables remote patient oversight, enhances the reliability of care, and supports proactive health management.

Keywords: Internet of Things (IoT), Remote Health Monitoring, Biosensors, NodeMCU, Arduino IoT Cloud, Vital Signs Monitoring.

Abstract: The deployment of machine learning models in critical decision making requires reliable explanations that remain stable under varying data conditions. While SHapley Additive exPlanations (SHAP) provides theoretically grounded feature importance rankings, the stability of these explanations when models encounter corrupted or degraded data remains poorly understood. This study investigates the robustness of SHAP feature importance rankings under controlled data corruption scenarios across three classification algorithms and datasets of varying complexity. The methodology employs optimally regularized Logistic Regression, Random Forest, and XGBoost models trained on medical, financial, and text classification datasets. Controlled corruption mechanisms combining 5% random sample removal and Gaussian noise injection with standard deviation equal to 0.1 times feature standard deviation simulate realistic data quality degradation. Stability metrics including Spearman correlation, Kendall tau, and top k feature overlap quantify ranking preservation. Results demonstrate that properly regularized models maintain substantial SHAP stability, with Spearman correlations exceeding 0.89 across all configurations. Random Forest exhibits superior stability with near perfect correlation (0.999) on structured data, while maintaining correlations above 0.95 across all scenarios. The findings establish that appropriate regularization and model selection enable reliable SHAP explanations even under moderate data corruption, providing practical guidelines for deploying interpretable machine learning in production environments where data quality cannot be guaranteed.

Keywords: SHAP, explainable AI, feature importance, model interpretability, data corruption, robustness analysis.

Abstract: Air pollution remains one of the most pressing environmental challenges of the 21st century, it comes with severe consequences for both public health and ecological balance. Prolonged exposure to pollutants such as particulate matter, carbon dioxide, and volatile organic compounds has been linked to respiratory illnesses, cardiovascular diseases, and even premature mortality. Despite these risks, conventional air quality monitoring systems are often limited by high costs, fixed infrastructures, and restricted accessibility, leaving large populations without adequate real-time information. To address this gap, this study presents Vento Aureo, an IoT and Artificial Intelligence (AI)-based framework designed for real-time air quality monitoring and forecasting. The system leverages portable IoT sensors to collect pollutant data, which is later transmitted to the cloud for analysis. Machine learning algorithms are employed to identify patterns and predict short-term air quality trends, enabling proactive responses to hazardous conditions. Data visualization and user interaction are facilitated through a mobile application that delivers live readings and predictions directly to end-users, supporting informed decision-making in daily life. Furthermore, the framework holds potential to aid policymakers and urban planners by providing accessible, large-scale insights into pollution dynamics. By integrating portability, affordability, and predictive intelligence, Vento Aureo offers a practical step toward mitigating the harmful effects of poor air quality and promoting healthier urban environments.

Keywords: Air Quality Monitoring, Internet of Things (IoT), Machine Learning, Noise Pollution, Real-Time Data, Cloud Integration, Smart Environment.

Abstract: This paper presents a decade-long analysis of Canara Bank's financial performance using the CAMEL framework-Capital Adequacy, Asset Quality, Management Efficiency, Earnings Quality, and Liquidity-from FY 2015-16 to FY 2024-25. The study assesses how the 2020 merger with Syndicate Bank influenced its financial soundness. Secondary data were drawn from Canara Bank's annual reports, RBI publications, and financial databases. The results demonstrate significant improvement in capital buffers, asset quality, operational efficiency, and profitability post-merger. Liquidity remained consistently above the regulatory norms throughout the decade. The CAMEL analysis provides comprehensive insights into the impact of structural reforms in public sector banks. The study offers practical implications for policymakers, bank management, and investors aiming for sustainable financial performance.

Keywords: CAMEL Model, Canara Bank, Financial Analysis, Bank Merger, Asset Quality, Public Sector Banks

Abstract: Since digital communication has advanced so quickly, data transfer is now essential to many applications, such as multimedia streaming, telemedicine, and monitoring. The security of video communication is still a major concern, though, because of growing data breaches, illegal access, and cyberthreats. Traditional encryption algorithms such as AES and RSA are commonly applied in securing video communication, but they often face limitations like high computational load, complex key handling, and potential weaknesses against quantum-based attacks. To overcome these challenges, this research introduces a Hybrid Multi-Key Cryptography approach that integrates Elliptic Curve Cryptography (ECC) to enhance both security and efficiency during video transmission. The proposed system begins with video segmentation, where the input video is divided into frames and blocks. Each block is then encrypted using a two-layer mechanism: a lightweight symmetric cipher paired with an ECC-driven key exchange. Here, ECC is employed for key distribution and management, ensuring a reliable and secure transfer of encryption keys between the sender and receiver. Then, for quick and effective video frame encryption, the symmetric encryption algorithm is used. The suggested method improves confidentiality and guards against cryptanalysis attacks by dynamically altering the encryption keys at various intervals of time.

Keywords: cryptanalysis attacks, confidentiality, enhanced security.

Abstract: One of the few questions that India, after 78 years of independence still grappling with is the role of religion in politics and what kind of relation politics and religion should have? To find out this answer, this paper will delve into the latter decades of Nineteen century early decades of the 20th century, for these two time periods were very tumultuous in terms of shaping the future of Indian nationalism. The Hindu Sangthan movement, that was begins to reform the Hindu religion, gradually takes a rigid form. This paper also sheds some light on this journey.

Keywords: Reforms. Hinduism. Islam. Communalism. Nationalism.

Abstract: The postpartum period places significant nutritional demands on mothers, particularly during lactation. Iron and vitamin D are two essential micronutrients whose deficiencies are commonly observed in breastfeeding women, often with serious health implications. This review examines the dual challenge of iron and vitamin D deficiencies among lactating mothers, focusing on their prevalence, underlying causes, health impacts, and current management practices. Iron deficiency, frequently stemming from childbirth-related blood loss and inadequate dietary intake, can result in anemia, fatigue, reduced cognitive performance, and weakened immunity. At the same time, vitamin D deficiency, often due to low sun exposure and insufficient intake, affects calcium metabolism, bone health, and immune function. The simultaneous occurrence of both deficiencies may amplify negative health effects, especially in low-resource settings where access to proper nutrition and medical care is limited. Addressing these deficiencies is crucial for safeguarding maternal health and supporting infant growth and development. The paper concludes by emphasizing the need for continued research and stronger health during lactation and micronutrient deficiency.

Keywords: iron, vitamin D, maternal behavior.

Abstract: This synopsis presents a practical and secure online banking prototype that blends a traditional relational database with a private blockchain ledger. The goal is to improve integrity, transparency, and auditability of banking transactions without compromising usability or performance. The work starts by tracing the evolution of digital banking-from branch-led paper systems to ATMs, online and mobile banking-highlighting how centralization remains a single point of failure and a lucrative target for attackers. It then introduces blockchain as a distributed, append-only, tamper-evident ledger, and explains how its built-in cryptography and consensus mechanisms create a powerful audit trail for financial records. The proposed system anchors every transaction to a private blockchain (for immutability) while keeping sensitive customer data in an ACID-compliant PostgreSQL database (for privacy and performance). A two-password security flow-static MPIN for login and OTP for authorizing each transaction-helps reduce account takeover risk. The prototype is built with Python (Flask), uses Proof of Work (PoW) for demonstration, and includes an admin dashboard with a blockchain explorer. Testing shows responsive user interactions via asynchronous mining, and the roadmap outlines a move to permissioned consensus (PBFT/IBFT) for scale. Overall, this hybrid approach offers banks a credible, evolutionary path to stronger security, more efficient auditing, and faster reconciliation with familiar tools and user experiences.

Abstract: The expansion of digital communication has made data transmission a cornerstone of modern applications such as telemedicine, multimedia streaming, and surveillance. Despite its importance, securing video communication remains a major concern in the face of cyberattacks, unauthorized access, and frequent data breaches. Conventional encryption schemes like AES and RSA have long been employed for safeguarding video content, but they often introduce high computational demands, complex key management, and limited resilience against emerging quantum-based attacks. To overcome these limitations, this study introduces a Hybrid Multi-Key Cryptography framework that incorporates Elliptic Curve Cryptography (ECC) to strengthen security while improving efficiency.for enhanced security and efficiency in video transmission. The encryption process begins with video segmentation, where the original video is divided into frames and blocks. Each block undergoes a dual-layer encryption mechanism, incorporating ECC-based key exchange and a lightweight symmetric encryption scheme. The ECC-based approach is used for key management and distribution, ensuring secure key exchange between sender and receiver. The symmetric encryption algorithm is then employed for fast and efficient encryption of video frames. By dynamically changing the encryption keys at different time intervals, the proposed technique prevents cryptanalysis attacks and enhances confidentiality.

Keywords: cryptanalysis attacks, confidentiality, enhanced security.

Abstract: With growing cybersecurity threats like phishing, spam and ransomware are increasing rapidly, causing major security risk and data loss, it is difficult to accurately detect them. This paper presents an integrated detection system which uses rule-based heuristics and machine learning models deployed within web-based platform. The system consists of modules like Phishing, URLs, SMS spam, Email spam and Ransomware Detection. The design include user authentication and login history, helps in real-time deployment. We tested the system on standard datasets and found it to be highly accurate and reliable, mainly high accuracy is seen in spam and ransomware classification. These results show that the proposed approach is both scalable and effective for real-time threat detection.

Keywords: Phishing detection, spam filtering, ransomware detection, machine learning, cybersecurity, XGBoost, Random Forest, Naïve Bayes, hybrid models.

Abstract: Assessment of mental health is challenging in the changing and digitally networked world today. Mental map navigating hidden emotions is a web-based platform that can integrate machine learning models across the three key modalities such as text, speech and behavioral inputs to detect the underlying emotional states and the possible risks. It uses natural language processing methods to process user with journal text converting raw text into labels using Sentence-BERT embeddings and logistic regression classification. The speech module uses audio signal processing drawing the MFCC features from user uploaded and recorded audio sample applying random forest for the emotions detected with associated labels. Behavior inputs having lifestyle activity are tracked through surveys with custom mapping logics and classifier for in depth risk. It supports muti-step process with the user information showing the result as diagnosis, confidence and suggestions.

Keywords: Random forest classifier, Sentence-BERT, MFCC, logistic regression.

Abstract: The Drug Inventory and Supply Chain Tracking System is a comprehensive solution designed to streamline the management of pharmaceutical products across the supply chain. From manufacturers to end-users, this system ensures real-time visibility, traceability, and control over drug inventory, minimizing the risks of stockouts, overstocking, counterfeiting, and inefficiencies. Utilizing modern technologies such as barcode/RFID scanning, cloud-based databases, and data analytics, the system allows stakeholders to monitor inventory levels, track shipments, and forecast demand with high accuracy. Additionally, the implementation of secure authentication and transaction logging enhances the transparency and reliability of the supply chain, ensuring that only authorized and genuine products reach consumers. This system not only improves operational efficiency but also plays a critical role in public health by ensuring the timely availability of safe and effective medications.

Keywords: Drug inventory, Supply chain, Stock management, Real-time tracking, Barcode/RFID, Expiry tracking, Batch monitoring, Reorder alerts, Compliance, Distribution.

Abstract: Modern security teams in busy public spaces often struggle to keep up with ever-growing video surveillance demands, missing crucial incidents due to human fatigue and excessive false alarms. This project puts forward a smart, automated surveillance tool that uses only the movement patterns captured by cameras-never faces or identifying features-to spot genuinely suspicious activities like fights or panic situations in crowds. Built using open-source tools on regular computers, the system breaks up video footage, analyzes how people move, and recognizes the difference between everyday strolls and dangerous behavior, all while preserving privacy. Its machine learning algorithms learn from thousands of real-life examples, so it's accurate and alert only when it really matters, drawing attention to anomalies with clear boxes and labels on the screen. In essence, this research marks a step toward smarter, fairer public safety, making advanced surveillance accessible and ethical for crowded places everywhere.

Keywords: Abnormal Behaviour Detection, Computer Vision, Optical Flow Analysis, Machine Learning, Crowd Surveillance, SVM (Support Vector Machine), KNN (K-Nearest Neighbours), Logistic Regression.

Abstract: The rapid growth of social media platforms has given rise to new opportunities for communication, networking, and information sharing across the globe, but it has also given rise to fraudulent and spoof accounts that compromise user trust, spread misinformation, and facilitate malicious activities. Detecting such accounts accurately is challenging due to their dynamic behavior and the complexity of social interactions online. This project presents an automated A framework that makes use of machine learning techniques to analyze data and generate intelligent prediction.to identify fraudulent and spoof accounts in social media. The system integrates multiple models, including XG Boost, Random Forest, and Naïve Bayes, to analyze account features and behavioral patterns effectively. A web-based platform has been developed to provide real-time detection, user authentication, and historical logging for enhanced usability and security. The framework was tested on standard datasets, and the results demonstrate high accuracy in identifying and separating authentic users from fraudulent or impersonated accounts. The study highlights the suggested approach ensures that the system is more efficient and reliable, solution is both scalable and adaptable, making it a reliable approach for strengthening security in social.

Keywords: Cybersecurity, Spam Filtering, Machine Learning, Random Forest, XG Boost, Naïve Bayes, Hybrid Modeling.

Abstract: Effective inventory management and accurate demand forecasting remain central to the success of modern retail and supply chain systems. Industries must continually put up a delicate balance between holding enough stock to meet customer needs while avoiding excess that leads to high costs and waste. This paper provides a comprehensive review of the methods and models developed to address this challenge. It explores traditional approaches such as statistical and econometric models, as well as newer techniques based on machine learning and also hybrid frameworks that combine the strengths of different models to achieve greater accuracy and adaptability. Recent research has explored a wide range of approaches ranging from classical deterministic models such as Economic Order Quantity (EOQ) [2], to statistical methods like ARIMA [5], and modern machine learning and deep learning approaches including LSTM and CNN-LSTM [15]. This survey provides insights from thirty scholarly works, examining methodological advancements and applications across retail, manufacturing, and food industries. The review highlights key contributions, compares model performances, and discusses practical challenges in data preprocessing, model selection, and performance evaluation. Findings indicate that while traditional models remain useful for structured environments, data-driven and hybrid models offer superior adaptability in uncertain, dynamic markets. Future work emphasizes integrating explainable AI with real-time optimization to bridge the gap between theoretical models and industrial practice.

Keywords: Inventory Management; Demand Forecasting; Supply Chain Systems; Machine Learning; Deep Learning; Hybrid Models; ARIMA; LSTM; Explainable AI; Real-time Optimization.

Abstract: Detecting Distributed Denial of Service (DDoS) attacks in Software Defined Networks (SDNs) is crucial for safeguarding network infrastructure from malicious disruptions. This study utilizes the CICIDS dataset to evaluate and compare various machine learning (ML) and deep learning (DL) methods for anomaly detection. The models assessed include Recurrent Neural Networks (RNN), Long Short-Term Memory (LSTM), Gated Recurrent Units (GRU), Bidirectional LSTM (BiLSTM), Convolutional Neural Networks (CNN), a CNN-BiLSTM hybrid, Support Vector Machines (SVM), Random Forest, AdaBoost, XGBoost, Decision Trees, Logistic Regression, K-Nearest Neighbors (KNN), and an ensemble Voting Classifier. Among these, the Voting Classifier produced the best outcome, reaching 93% accuracy with strong precision, recall, and F1-score. These findings highlight the enhanced accuracy offered by ensemble learning in DDoS detection in SDNs and position the Voting Classifier as a strong candidate for future developments in anomaly detection.

Keywords: Software Defined Networks (SDN), Distributed Denial of Service (DDoS), Anomaly detection, Machine learning, Ensemble learning (voting Classifier).

Abstract: Manual segmentation of customers consumes a lot of time, in some cases months, even years to break down information and track down patterns in it. Customer Segmentation done through machine learning models result in quick identification of the ideal customers. This research paper focuses on the tourism industry to target the right customers for their business. By using the tourism dataset of customers, the research paper aims to produce a better decision making visualization patterns through histogram, pie charts, and heatmaps. Moreover, the use of Bayesian Inference Model, Descriptive Basic Analysis and Linear Regression Analysis only on the important attributes makes the decision making for the tourism business quite easy. Finally, the use of clustering unsupervised machine learning models on the dataset generates the primary, secondary, and tertiary group of customers that the company can target for the sale of their tourism packages. Clustering models will generate clusters as the output where each cluster showcases a group of customers. The clustering models employed under this research are K-means, DBSCAN, Affinity Propagation, Mini Batch K-means and Optics Algorithm. The result showed that the Mini Batch K-means algorithm had a better accuracy score for the segmentation than other algorithms used.

Keywords: segmentation, Bayesian, regression, unsupervised, propagation, accuracy.

Abstract: Obesity is the major global health concern in the 21st century, distinguished by high accumulation of body fat and linked to chronic illness and mortality. Breakfast which provides the human body with majority of energy requirements (20-35%), plays a important role in weight management and appetite regulation. However modern lifestyle factors, such as time constraints and shift work, have led to breakfast skipping which subsequently contributes to obesity risk. Diet rich in protein, especially in breakfast, were found to increases satiety due to release of anorexigenic and orexigenic signals. Research evidence provide protein sources and amount affect appetite modulation, with high-protein breakfasts satisfying hunger, increasing feelings of fullness, and altering specifically in improved hormonal responses as compared to carbohydrate-based breakfast or skipping breakfast altogether. Tools to measure subjective appetite and eating behavior are variously validated, such as the Visual Analogue Scales (VAS) or the Three-Factor Eating Questionnaire (TFEQ). According to the literature, the implementation of high-protein breakfast as part of the diet strategy may be the effective method to address obesity due to its ability to modulate both hormonal responses to appetite and behavioral regulation of it. This review article highlights appetite regulation and satiety, including the roles of leptin, ghrelin, PYY, GLP-1, and cholecystokinin (CCK), the impact of protein on satiety, the effects of high-protein breakfasts on appetite control, methods for hunger assessment, and existing research gaps.

Keywords: Appetite hormones, Satiety, Obesity, Protein breakfast, Hunger assessment

Abstract: The rise of digital technology has increased both the frequency and complexity of cybercrimes, demanding efficient methods for uncovering digital evidence. Digital Artifact Extractors address this need by automating the identification, extraction, and preservation of artifacts left by user activity, system operations, or applications. Unlike manual processes that are time-consuming and error-prone, automation improves accuracy, reduces investigation time, and allows investigators to focus on analysis and case resolution. This paper highlights the functionality, significance, and applications of digital artifact extractors in modern cyber forensics and cybersecurity.

Keywords: Digital Forensics, Cybercrime, Artifact Extraction, Automation, Cybersecurity

Abstract: This study investigates phenological changes in plants of the Indian Himalayan region in response to global warming. Phenology, the study of recurring biological events in relation to climate, has emerged as one of the most sensitive bioindicators of climate change. Using four decades of data (1980-2020) derived from field monitoring, herbarium records, citizen science datasets, and remote sensing, we analyzed shifts in flowering, fruiting, and leaf senescence across six altitudinal zones. Results show that flowering and fruiting have advanced by 2-4 days per decade, while autumn senescence has been delayed, effectively extending the growing season. Regression analyses revealed strong correlations between temperature anomalies and earlier flowering, with each 1 °C increase advancing phenophases by 3-4 days. Early-blooming species were more responsive than late-flowering perennials, reflecting interspecific variability in climatic sensitivity. These findings align with global trends and highlight risks of ecological mismatches with pollinators, altered plant competition, and agricultural vulnerabilities such as frost damage and pest emergence. The study underscores the urgency of long-term monitoring and adaptive strategies to safeguard biodiversity, agricultural productivity, and ecosystem resilience in a warming climate.

Keywords: Phenology; Climate Change; Global Warming; Indian Himalayas; Flowering Shifts; Fruiting; Leaf Senescence; Bioindicators; Agriculture; Ecosystem Resilience

Abstract: This paper introduces a deep learning-powered system designed to detect potholes from road images, assess their severity, and map their location. At the core of the system is a Convolutional Neural Network (CNN) trained on a curated dataset of road images containing both potholes and non-potholes. The model is able to accurately identify potholes and further categorize their severity into three levels-Minor, Moderate, or Major-based on its confidence score. To demonstrate the system's functionality, detected potholes are assigned random locations within Bengaluru, and a detailed PDF report is generated. The report includes the detection results, supporting images, a severity distribution chart, and location information. The entire solution is deployed as a Flask-based web application, offering a simple interface where users can upload road images, receive real-time predictions, and download comprehensive reports.

Keywords: Pothole Detection, Deep Learning, Convolutional Neural Networks (CNN), Image Classification, Road Safety, Web Application (Flask).

Abstract: The rising threat of rice diseases such as blast, brown spot, sheath blight, and bacterial leaf blight has emphasized the need for early, accurate, and scalable detection methods. This research brings together insights from 30 peer-reviewed studies to provide a consolidated view of how computational intelligence is being applied in rice disease diagnosis. Recent advancements highlight the use of deep learning models-including CNNs, ResNet, DenseNet, EfficientNet, Vision Transformers, and object detection frameworks like YOLO-alongside traditional machine learning techniques such as Support Vector Machines (SVM) and Random Forests (RF). In certain cases, hybrid systems (e.g., CNN-SVM) have demonstrated almost flawless classification performance, while advanced architectures like DenseNet201 and Vision Transformers have reported accuracy levels exceeding 98%. Other studies focus on practical aspects such as IoT-enabled monitoring, mobile-based deployment, and image segmentation for precise localization of disease-affected regions. Although these developments are highly encouraging, challenges remain in transitioning from controlled research environments to real-world applications. Key issues include generalizing across diverse field conditions, maintaining computational efficiency on resource-limited devices, and ensuring models perform reliably across different geographic and climatic variations. This paper provides a critical evaluation of available methodologies, summarizes key achievements, and discusses limitations that researchers and practitioners must address. Finally, it outlines potential future directions for developing more adaptive, resource-efficient, and farmer-friendly rice disease detection systems that can support sustainable agricultural practices.

Keywords: Rice disease detection, CNN, precision agriculture, machine learning, IoT in farming.

Abstract: This research aims to improve road safety by developing a deep learning-based system capable of detecting and recognizing tail lights, brake lights, and driver hand gestures in Indian traffic conditions. Using the YOLOv8 object detection architecture, the system has been designed and implemented in two key phases: training and inference. During the training phase, a comprehensive custom dataset of traffic images - carefully labelled with seven distinct classes - was used to train the YOLOv8 model. This dataset includes real-world conditions such as varying lighting, complex backgrounds, and occlusions. In the inference phase, the trained model processes new images or videos, automatically detecting the presence of vehicle signals and hand gestures. The output consists of bounding boxes and class labels that are visually rendered and saved, ensuring traceability and ease of analysis.

Keywords: Brake Light Detection, Tail Light Detection, Hand Gesture Recognition, Object Detection.

Abstract: Through the combination of artificial intelligence (AI) and the Internet of Things (IoT), smart agriculture has emerged as a revolutionary step to increase crop yield and sustainability in recent years. This integration has made it possible to continuously monitor farms and automatically assess the health of crops. However, a number of issues plagued the current Convolutional Neural Network (CNN)- based smart agriculture system, including limited applicability in remote agricultural regions, inconsistent data collecting in a variety of field conditions, and poor generalization on single crop diseases. A smart agriculture system based on Vision Transformers (ViT) is suggested as a solution to these problems. The first of four layers in this system architecture is the data acquisition layer, which is equipped with sensor nodes and cameras to gather environmental data and photos of plant leaves. The communication layer follows, which is in charge of carrying out data transfer to the following layer, The processing layer comes next, when threshold evaluation and Simple Moving Average (SMA) filtering are used to preprocess environmental data. Additionally, bilinear interpolation is used to scale and normalize the image data. The pretrained ViT model is then fed this pre processed data in order to classify plants with multiple crops and diseases. Finally, the user receives an interactive web-based dashboard with the output, which comprises the illness kinds that were identified and their confidence levels. The suggested ViT-based system beat the current method and achieved greater accuracy, according to experimental results.

Keywords: Vision Transformer, CNN, IoT, KNN, Simple Moving Average

Abstract: Mental health issues like anxiety and depression and even stress have become major issues in society and contribute to a lowering standard of living, productivity levels have decreased, and individuals are becoming more aware of their struggles. Unfortunately, many cannot access professional therapy due to stigma, cost, and a lack of trained professionals. This research seeks to outline the design and development of an AI based Mental Health Chatbot using Python, Flask and Natural Language Processing (NLP) to aid individuals in mental health support 24/7 in a confidential, professional and unbiased manner. The system encapsulates a classifier that was trained on a larger dataset which includes many intents related to both mental and physical health, which enables the chatbot to discover stress, anxiety, and depression but also larger physical health challenges like a fever or cold. It comes with user authentication, chat history, carries on conversation, real time prediction accuracy. The experimental results showcase ~92% accuracy with intent classification, ensuring the responses were empathetic and trustworthy.

Keywords: Mental health chatbot, natural language processing, Flask, python, supervised learning, health informatics, AI-based therapy, intent classification.

Abstract: This project presents a cutting-edge e-library system driven by AI that aims to update how consumers engage with digital information. The system seeks to provide a more customized and interesting reading experience by using cutting-edge technology including text-to-speech capabilities, machine learning, artificial intelligence, and natural language processing. Features like AI-generated summaries for speedy content review, tailored book suggestions based on reading history, and a text-to-audio tool to make books accessible to visually impaired users are all available to users. In addition to offering a smooth borrowing and return procedure with automatic due date monitoring, the portal lets users search for books by title, author, ISBN, or category. To enhance library operations, the administrative console provides library managers with an effective means of managing material, monitoring user activity, and producing reports. A rising user base and a developing content library may be accommodated by the system without affecting performance because it was developed with scalability in mind. A safe and secure user experience is ensured with a strong emphasis on security, with user data encrypted and safeguarded in accordance with privacy regulations such as GDPR. When it comes to improving the traditional library experience, our AI-powered e-library system provides a complete and flexible answer. Setting a new benchmark for digital libraries and advancing knowledge management in the digital era, its dynamic features, strong security features, and accessibility-focused approach make it an interesting and useful tool for administrators and users alike.

Keywords: AI, e-library, recommendation system, summarization, sentiment analysis, personalization.

Abstract: This paper presents a multi-modal deepfake detection system capable of analyzing images, videos, and audio for signs of manipulation. The project addresses the limitations of single-modality detection systems by combining various machine learning and deep learning techniques to identify complex forgeries. The proposed system utilizes CNN and ResNet for detecting spatial inconsistencies in images, an LSTM on frame sequences for identifying temporal anomalies in videos, and a combination of Librosa and a Random Forest classifier for detecting synthetic audio patterns. The system aims to be resilient against adversarial attacks and provide accurate, real-time results through a user-friendly, Flask-based web interface. The anticipated outcomes include superior detection accuracy, balanced precision and recall, and enhanced generalization across diverse datasets.

Keywords: Deepfake Detection, Multi-modal Detection, Machine Learning, Deep Learning, Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM).

Abstract: Knee osteoarthritis (OA) is one of the leading causes of disability, affecting millions of people and limiting their ability to carry out daily activities. Diagnosis is usually performed by analyzing X-ray or MRI scans, but the manual process depends heavily on expert interpretation, which can vary between clinicians and take significant time. To overcome these challenges, this project introduces an automated system that applies deep learning methods, specifically Convolutional Neural Networks (CNNs) enhanced with AlexNet, to identify and grade knee OA. The system is designed to process medical images through steps of preprocessing, augmentation, and feature extraction before classification. It further predicts OA severity using the Kellgren-Lawrence (KL) grading scale. Performance is assessed with metrics such as accuracy, precision, recall, F1-score, and ROC-AUC. By aiming for an accuracy of at least 90%, this model seeks to provide a reliable decision-support tool that reduces subjectivity, speeds up diagnosis, and supports early treatment planning in clinical practice.

Keywords: Knee Osteoarthritis, Convolutional Neural Network, AlexNet, Deep Learning, Medical Imaging, Feature Extraction, Kellgren-Lawrence Scale, Automated Diagnosis, Transfer Learning, Clinical Support System.

Abstract: This study focuses on the thermal analysis of chemical reactions characterized by exponential heat generation, modeled using a reaction-diffusion equation. The non-linear nature of the heat generation, driven by temperature, reflects the rapid acceleration of reaction rates at higher temperatures. A double interpolation process is used to improve the precision of temperature distribution solutions over space and time. Starting from the classical heat equation with a reaction term, the analysis employs Dirichlet boundary conditions and a Gaussian initial temperature profile. The double interpolation method enhances solution accuracy, particularly in capturing steep temperature gradients caused by exponential heat generation. This approach is especially valuable for high-temperature chemical reactions, such as combustion, where precise thermal control is essential. The study's results provide insights into the thermal behavior of reactive systems, making it useful for applications like industrial reactors and heat-sensitive processes.

Keywords: Thermal analysis, chemical reactions, exponential heat generation, double interpolation process, reaction kinetics

Abstract: Civil engineers (CEs) are instrumental in shaping the built environment, often managing demanding schedules and physically intensive fieldwork. In India, the dietary habits of practicing CEs offer insights into how professional stress, lifestyle demands, and cultural factors influence daily nutrition. This study provides an overview of prevailing food consumption patterns among CEs in India and their impact, examining aspects such as meal frequency, dietary choices, meal timing, and typical dining locations. The analysis is further supported by observed trends and preferences identified through a structured sample survey.

Keywords: Civil Engineers, Dietary Habits, Food Consumption Patterns, Lifestyle and Nutrition, Meal Frequency and Timing, India, Survey Analysis

Abstract: This paper presents a machine learning -based system for detecting vitamin deficiencies by combining facial image analysis with symptom -based evaluation. Using a Convolutional Neural Network (CNN) trained on a curated dataset of facial images labeled with various vitamin deficiencies, the model predicts the likelihood of deficiency with high accuracy. The system further refines its prediction by incorporating results from a symptom questionnaire, providing a final deficiency classification along with a confidence score. Additionally, th e application generates a comprehensive PDF report containing the detection results, annotated facial images, a deficiency probability graph, and dietary recommendations. The proposed solution is implemented in Google Colab, integrating the trained model with an interactive interface for image upload, symptom entry, real -time prediction, and report generation.

Keywords: Vitamin Deficiency Detection, CNN -Based Image Classification, Medical Image Analysis, Symptom -Based Diagnosis, Deep Learning for Healthcare, Convolutional Neural Network, Image and Symptom Integration.

Abstract: This study focuses on the laminar boundary layer flow of Williamson fluids over a wedge, utilizing similarity transformation techniques to reduce the governing nonlinear partial differential equations into ordinary differential equations. Scaling group transformation technique is applied to derive similarity solutions, and numerical computations are carried out using MATLAB's ODE solver to explore the effects of Williamson fluid parameters on flow characteristics. Graphical representations of velocity profiles and their slopes are generated for varying dimensionless parameters, demonstrating significant influence on fluid behavior.

Keywords: Scaling group transformation, Williamson fluid, MATLAB ODE solver, Boundary layer flow.

Abstract: In order to assess water quality in Mysore district lakes, a study would likely look at both physical parameters (temperature, pH, turbidity, etc.). The study conduct through lakes are Karanji lake, Hebbal lake and Hadinaru lake. Karanji Lake is a lake located in the city of Mysore in the state of Karnataka, India. The lake is surrounded by a nature park consisting of a butterfly park and a walk-through aviary. This aviary is the biggest 'walk-through aviary' in India. Hebbal Lake, located near Mysore, India, is a scenic water body known for its natural beauty and recreational activities. Hadinaru Lake, also known as Doddalake, is a freshwater lake located in the Nanjangud taluk of Mysuru district, approximately 33 km from Mysore city. This study highlights the quality of water analysed with reference to various physico-chemical parameters in selected lakes of Mysore District, Karnataka.

Keywords: Lakes, Water Quality, Physical and Chemical parameters, Development.

Abstract: This paper presents a hybrid framework that integrates intrusion detection and congestion prediction for 5G networks using supervised and unsupervised machine learning techniques. Unlike traditional approaches that focus only on congestion modeling, this work combines anomaly detection with congestion-aware forecasting to enhance both network reliability and security. The system is trained on the NSL-KDD dataset for identifying malicious traffic and on synthetic 5G congestion parameters for predicting potential bottlenecks. Supervised classifiers such as Random Forest and SVM are used to recognize labeled patterns, while clustering and anomaly detection methods capture emerging traffic behaviors without prior labels. The model is deployed through a Flask-based web platform that provides interactive dashboards, correlation heat-maps, attack category distributions, probability-based predictions, and real-time visualization of congestion risk. Additional features such as anomaly timelines, automated PDF reporting, and educational support pages make the system suitable for both cyber security learning environments and lightweight deployment in real networks. By combining predictive congestion control with intelligent intrusion detection, this framework offers a proactive, interpretable, and scalable solution for modern 5G communication infrastructures.

Keywords: Network Anomaly Detection, Intrusion Detection System, Flask, Machine Learning, NSL-KDD

Abstract: Tribals and forest dwellers in Odisha are bestowed with various kinds of non-timber forest products (NTFP). Most of these NTFPs include medicinal and dietary need-based forest products. Wild edible mushrooms are a group of such products which have culinary, therapeutic, and commercial values but are underrated. Forests of Western Odisha are mostly dry deciduous and moist deciduous in nature best suitable environment for the availability of fungi, general and macro fungi in particular. Identification and biochemical properties of wild mushrooms of western Odisha are not properly estimated, leading to casualties in some or different parts of this area. Sometimes, the cooking ingredients also make a difference in the toxicity level. Therefore, this attempt has been made to estimate pH, which is an indicator of food quality under different cooking conditions in 13 wild mushrooms found in western Odisha, which are either consumed or sold in markets. Results show that raw mushrooms cooked for a longer time reduce pH and make the dish acidic, but the addition of spices and condiments increases the pH and makes the food healthier by making it neutral or less acidic.

Keywords: Macro fungi, mushroom, western Odisha, tribal food, Toxicity.

Abstract: Green audit of the institution is important in consonance to assess the environmental performance of educational institutions and to consider potential options for turning the educational campus into an eco-campus. The MGS varsity campus, Bikaner Rajasthan, has undergone a green audit to evaluate its environmental impact. The main focus of this green audit is on the consumption of energy in terms of electricity, soil and water quality, vegetation, waste management procedures, and the campus carbon footprin Arora,P.(2017).To learn more about the resources on campus and their consumption, a questionnaire survey was first carried out.. The collection of various information from different unit of varsity was sorted, tallied, and examined to give a report on the environment with recommendations (IGBC 2021).

Keywords: Green Audit, Carbon Footprint, environmental conservation and sustainability

Abstract: The growing significance of sustainability has compelled the FMCG and consumer goods firms to consider using social media as one of the main means of reaching consumers and advocating environmentally friendly behavior. This paper researches how the Social Media Engagement, User-Generated Content and Influencer Persuasiveness affect consumer Purchase Intention when the mediating variable is Brand Trust. Structured questionnaire was used to gather data on 202 subjects in Chennai city. Regression-based mediation was used according to Baron and Kenny method, and the value of indirect effects was tested with the help of the Sobel test. The findings demonstrate that Social Media Engagement and User-Generated Content have a significant role in Purchase Intention both directly and indirectly, mediated by Brand Trust, which implies partial mediation. Influencer Persuasiveness, in its turn, has a direct impact on Purchase Intention, and does not affect Brand Trust. The predictive power of Brand Trust itself was established as high. The paper emphasizes the role of integrating trust-building, including consumer engagement and peer content, and persuasive influencer-based approaches to enhance sustainable branding activities. The results offer practical implications to FMCG companies in developing a viable social media strategy that create credibility, trust, and sustainable consumer behavior.

Keywords: Sustainable branding, FMCG, Social media, Brand trust, Influencer persuasiveness, User-generatedcontent, Purchase intention, Chennai

Abstract: Cloud computing has revolutionized the provision of IT services via elastic models like Software as a Service (SaaS), Platform as a Service (PaaS), and Infrastructure as a Service (IaaS). These models, though, come with an array of security threats that keep changing with technology. This review analyses and classifies widespread security attacks in the cloud service models-focusing on prevalent attacks like SQL injection in SaaS, unauthorized access in PaaS, and data breach in IaaS. It integrates current defence mechanisms, with a particular emphasis on machine learning methods and cryptographic mechanisms, and stresses the increasing importance of joint security efforts by cloud users and providers. Moreover, the paper summarizes actual cloud-based attack vectors in real life, categorizing them based on severity to facilitate risk prioritization. The assessment also discusses the specific challenges brought about by cloud integration into industrial SCADA systems, detailing their primary vulnerabilities and categorizing related threats into types such as hardware-level, protocol-based, and insider attacks. Lastly, it talks about changing trends and best practices and highlights the move from ad hoc security reactions to formal, risk-defined cloud security strategies.

Keywords: Cloud Security, Cloud Computing, Threats & Vulnerabilities (or "Cloud Threats"), Cloud Service Providers (CSPs), Cloud Deployment Models (SaaS, PaaS, IaaS), Zero Trust Architecture (ZTA)

Abstract: This study employs Failure Mode and Effects Analysis (FMEA) to assess risks across five construction levels (Basement, Lintel, Roof, Casing, and Finishing). Surveying 750 stakeholders (architects, designers, engineers, supervisors, workers), we identify 12 critical failure modes, their root causes & severity. Key findings reveal supervisors and workers as most accountable (26-39% responsibility), with "Unsafe working conditions" & "No training" as dominant causes. We propose targeted interventions to mitigate risks and enhance quality.

Keywords: Failure Mode and Effects Analysis, Unsafe working condition, No training, Intervention