Intelligent Chatbot Assistant in Agriculture Domain

Abstract

Agriculture is known as the economic game changer of India. It is the primary driver of GDP growth because of India’s robust agricultural industry, and proper knowledge about agriculture techniques help increase crop yield. So, answering the different types of crop-related queries is essential. We proposed the intelligent chatbot application in the agriculture domain so that farmers can get the correct information about farming practices. Our system is farmer-friendly and capable enough to instantly answer farm-related queries from the knowledge base, such as plant protection, fertilizer uses, government schemes, and many others. We used the agriculture-related data in question-answer format and implemented the pre-trained model of the Sentence-Transformer approach to answer providing. We also deployed the TF-IDF and Bag-of-Words method but achieved a reasonable accuracy rate for the test data in the sentence transformer pre-trained model. With the help of API services, our system also shows the crop’s latest mandi (market) rate and current weather information. So, the proposed chatbot system will keep the contribution for farmer’s cost savings. Overall, our chatbot system is straightforward and more efficient for the farmer to make better decisions.