Smart Crop Advisory System: A Machine Learning Approach to Precision Crop Recommendation Using Soil and Climatic Parameters

Abstract: Agriculture is a major contributor to the Indian economy and supports a large portion of the rural population. One of the most critical challenges faced by farmers is selecting the appropriate crop for a given season, as poor decisions can lead to reduced yields, financial losses, and soil degradation. This paper presents the Smart Crop Advisory System (SCAS), a machine learning-based approach that recommends suitable crops using key soil nutrients (N, P, K), environmental factors (temperature, humidity, rainfall), and soil pH. A Random Forest model trained on a standard crop dataset achieves an accuracy of 93.71%, outperforming other classifiers such as Decision Tree, K-Nearest Neighbours, Support Vector Machine, and Naive Bayes. The system is deployed as a RESTful API using Django REST Framework and includes a rule-based weather advisory module in English and Hindi. The proposed solution offers an accessible and efficient tool for improving crop decision-making among smallholder farmers.

Keywords: Crop Recommendation System, Random Forest, Precision Agriculture, Machine Learning, Soil Parameters, Django REST Framework, Smart Farming, Decision Support System.