RSF: A recommendation system for farmers

Abstract

In this paper, we present a recommendation system named as RSF for farmers, which can recommend farmers most suitable crops to produce in different areas. The system first detects a user's location and works with different agro-ecological and agro-climatic data in upazila level to calculate similarity between upazilas using Pearson co-relation similarity algorithm. Then it selects top-n similar upazilas. Finally, utilizing the seasonal information and crop production rates of each crop of the similar upazilas, it recommends top-fc crops to a user of a upazila. The system has been evaluated with real data and we have found a reasonable and satisfactory accuracy. The system can help farmers in producing appropriate crops. As a result, they can increase their lifestyle and can contribute more to the society. We provided the system layout both in Bangla and English so that the farmers as well as the people who works with farmers can get help easily from this.