Recognizing Farmers' Requirements with Visual Analytics of Inquiries

Abstract

Aims: To identify the trending queries raised in Kisan Call Centers by employing visual analytics. Study Design:  Analyzing secondary data on farmers’ queries by employing various visual analytics techniques. Methodology: The study was conducted using various visual analytics techniques. The choropleth map was used to visualize the district-wise spatial distribution of queries within Telangana state. The line graph was used to understand the temporal trend in queries raised during the study period. The word clouds, bigrams and network maps were used to identify the words that are often appeared in the query text. Before visual analytics, the raw data was processed using various text cleansing steps such as removing all non-alphanumeric characters, punctuation, and extraneous whitespace, lowercasing and tokenizing. Several popular packages available in the R programming language were employed to carry out the analysis. Results: From choropleth map, it was found out that most queries during the study period were raised in the eastern districts of Telangana, highest being raised at Mahabubabad (>17000). The average number of queries raised in a month was found to be around 8000 queries per month. The line graph indicated that the queries raised during the second wave of COVID-19 were less. The word cloud indicated that the word ‘management’ has appeared most number of times in the query text followed by ‘weather’ and ‘paddy’. Conclusion: From this study, we found that the maximum queries on farming in Telangana are raised in Mahabubabad district. It was also found that the queries on ‘nutrient management in paddy’ and ‘pm kisan samman’ were highest. This highlights the need for educating farm tele advisors on the trending queries for efficient delivery of information needed by the farmers.