A Collaborative Recommendation Model of Agricultural Planting Technology Based on User Characteristics

Abstract

In view of the lack of informatization in the field of agricultural planting production, the traditional collaborative filtering recommendation algorithm has problems such as cold start, sparse scoring matrix, and poor scalability, resulting in poor recommendation quality. A personalized recommendation model of collaborative filtering agricultural planting technology that integrates user characteristics is proposed. First, the user’s initial characteristics are constructed according to the user’s geographic location, planting occupation, and main crops, and user behavior information is used to update user characteristics. Then combine the user characteristic similarity model and the rating matrix similarity model, and reconcile the weighting factors to form the user’s comprehensive similarity. Finally, use Top-N for personalized recommendation. By integrating user characteristics, the recommendation model is more suitable for agricultural planting scenarios, and the recommendation results are more flexible and reasonable. Experimental results show that compared with user-based collaborative filtering recommendation and user-characteristic-based recommendation, the proposed algorithm improves precision by 2% and 7%, recall rate by 2% and 9%, and F1 value is 71%. The effectiveness of the proposed method is verified.