Krushi – The Farmer Chatbot | IEEE Conference Publication | IEEE Xplore

Krushi – The Farmer Chatbot


Abstract:

As per the reports of November 2020, around 58% of India's population earns a primary source of living from agriculture. But, close to 10,000 farmers every year succumb t...Show More

Abstract:

As per the reports of November 2020, around 58% of India's population earns a primary source of living from agriculture. But, close to 10,000 farmers every year succumb to the harsh conditions in the agricultural sector. These conditions arise when the product of crops is not as desired, spoilt crops, which leads to large loans on the farmers.In this paper, we present an Artificial Intelligence (Al) chatbot that assists the farmers by providing solutions to agricultural queries. Some of the questions are concerning crop sowing, crop diseases, seasons related issues. Thus, benefit the farmers in making the right decisions regarding their crops, thereby increasing their yield.Krushi - The Farmer Chatbot is an end-to-end trainable learning model to create a conversational system with minimum error and answer questions about current conditions. The chatbot is build using Artificial Intelligence (AI) and Machine Learning (ML) techniques. The dataset for the chatbot is used from Kisan Call Centre (KCC). The proposed system, answers queries related to weather, plant protection, animal husbandry, market price, fertilizer uses, government schemes, soil testing with an overall accuracy of 96.1% using RASA X.
Date of Conference: 25-27 June 2021
Date Added to IEEE Xplore: 12 August 2021
ISBN Information:
Conference Location: Mumbai, India

I. Introduction

Agriculture is one of the major sources of employment for a large number of people in the world. However, there are millions of small scale and marginal farmers who have a low level of awareness as they live in remote areas. Conventionally the field officers visit the fields and interact with farmers to provide supply training and advisory on practices that the farmers can use in farming and aspects of agriculture. The farmers that are probably main users of the traditional agricultural data about rainfall and crop production which is collected by the government in its raw form are unable to utilize it. To make it useful for the farmers this raw data has to be analyzed and fed to a system that would provide relational trends. The agricultural field is growing at a rapid pace not only in the technological aspects but also in the ways of production of crops. Several software are being developed to educate and instruct the farmers with this innovative technology. Basic information about farming is provided, they require a sizable amount of research to generate accurate information. The proposed framework defeats the downsides by giving a User Interface, where the farmers or different clients can communicate adequately to get desired responses with a lesser number of steps.

Contact IEEE to Subscribe

References

References is not available for this document.