Authors: Miss . Chintapalli Lakshmi, Mr. kunjam Nageshwar Rao, P.Mohan rao 3
Abstract: India, as an agro-based economy, continues to have a substantial share of its population dependent on agriculture as the primary source of livelihood. However, productivity is often constrained by challenges such as limited access to timely information, difficulty in diagnosing plant diseases, and inadequate awareness of government schemes and market dynamics. Traditional reliance on manual methods or intermediaries frequently results in delays and misinformation, further hindering agricultural efficiency. To address these limitations, this paper presents an AI-powered Chatbot for Farmers, designed to deliver real-time, accurate, and accessible assistance. The system integrates Natural Language Processing (NLP) for query understanding, Convolutional Neural Networks (CNNs) with fine-tuned VGG-16 for plant disease detection, and machine learning models for crop recommendation and decision support. Furthermore, the chatbot incorporates multilingual support via translation APIs, enabling seamless interaction in regional languages and ensuring inclusivity across diverse farming communities. The proposed chatbot provides a wide range of services, including query resolution, crop suggestion, disease diagnosis from leaf images, and dissemination of critical updates on weather, market prices, and government policies. Experimental results demonstrate an accuracy of nearly 96% in disease classification and high precision in intent recognition, establishing the reliability and robustness of the system. By functioning as a virtual agricultural assistant, the solution empowers farmers with expert-level, user-friendly guidance, thereby enhancing decision-making, reducing losses, and ultimately improving agricultural productivity.
DOI: https://doi.org/10.5281/zenodo.17062303
