Crop Yield Management System Using Machine Learning Techniques

Abstract

Farming is the backbone of agriculture country like India. Farmers lose their yield due to lack of knowledge about new technologies and plantation parameters which help them to increase their yield. The proposed system, Aruvi, performs machine learning analysis and applies Ontology-based mapping to assist the farmers in order to increase their yield. Aruvi is basically a chatbot that can mimic a virtual conversation with user (farmer) using regional language (Tamil). Aruvi is trained and made to learn on its own using ontology based mapping. Based on the user query it gives relevant answers, which is more useful for farmers in remote places. Using the proposed system, the user can know about the crops, their atmospheric conditions and suitable soil by querying the system in their own regional language to the chatbot. Another advantage of Aruvi, users can converse with it apart through menus or buttons via text or speech on websites or through mobile apps. Based on the trained dataset and real time scenario the accuracy of the system is 83.25%. This can be improved by collecting the real time conditions in that particular region and training Aruvi using them.