Assessment of queries of farmers at Kisan Call Center using natural language processing

To provide agricultural information and the latest updates to the farmers, the Ministry of Agriculture and Farmers Welfare, Government of India launched a helpline service “Kisan Call Center (KCC)” in 2004. A brief summary of each query is recorded in the KCC database for later reference and use. This repository enables us to analyze, interpret and infer knowledge for preparing policies to address the farmer's concern and problems. Researchers, planners, policymakers and government officials need the required datasets for devising informed decision rules and policies. In this study, queries of five years duration i.e., 2017 to 2021of Uttar Pradesh state from the KCC repository have been extracted and analysed to explore the insights exist in the repository. The data collection process has been automated by developing a web-scraping tool with user-friendly interface. The standard procedure of Natural Language Processing with N-gram techniques for extracting the features from the KCC repository was applied. Exploratory data analysis has been performed to visualize the hidden insights, trends and patterns. There were 3.6 million queries and 0.51 million queries were related to plant protection. It is observed from the study that paddy and wheat crops were the most prevalent crops having concern among the farmers for their protection to avoid severe damage due to pest attacks. This study will be useful for extracting the data in a user-friendly manner and helpful for policymakers in deciding towards the distribution of input resources required to minimize the yield loss due to various pests attack in a particular crop or cropping system in that region.