Research And Application Of Collaborative Filtering Recommendation Algorithm For Agricultural Products Based On Spectral Clustering And Seasonal Function

It is worthwhile and beneficial to have a recommendation of agricultural product by studying effective and precise techniques.An effective agricultural product recommendation platform can better help agricultural growers sell agricultural products,better serve consumers to select appropriate and satisfactory agricultural products,and better serve the National Rural Revitalization Strategy and agricultural digitization construction.This study optimizes the user-based collaborative filtering recommendation algorithm in order to increase the effect of agricultural product recommendation by combining the characteristics of agricultural products.The following are the key research findings:(1)Spectral clustering technology is used to optimize the effect of agricultural product recommendation.This thesis adopts spectral clustering technology to optimize the collaborative filtering recommendation algorithm to improve the impact of matrix sparsity in the application of Collaborative Filtering Recommendation Algorithm in the agricultural products market.The spectrum clustering technology is used to cluster users and find the same cluster as the target users,so as to construct user-score matrix.In this way,the scale of user-score matrix can be effectively reduced,so as to reduce the sparsity of the user-score matrix and improve the recommendation effect.(2)Based on the seasonal characteristics of agricultural products,a seasonal function is constructed by combining the month factor and sales duration,and a collaborative filtering(USS-CF)recommendation algorithm combining spectral clustering and seasonal function is proposed to improve the recommendation effect of agricultural products.In this thesis,because of the prominent seasonal characteristics of agricultural products,in view of the prominent seasonal characteristics of agricultural products,this thesis constructs a seasonal function of agricultural products based on the relationship between peak sales hours and circulation month of a certain type of agricultural products and uses it in the recommendation algorithm,so as to improve the recommendation effect of agricultural product recommendation algorithm.Experiments are conducted on real data sets.By contrast experiments between the proposed USS-CF recommendation algorithm and the traditional user-based collaborative filtering recommendation algorithm,the effectiveness of the recommendation algorithm is verified.(3)The development of an edible agricultural products We Chat platform based on the USS-CF recommendation algorithm is carried out.Based on the USS-CF recommendation algorithm and according to the software engineering method,the development of the edible agricultural products We Chat platform is carried out.The requirements analysis,system design,and system implementation are carried out.The platform provides fast and convenient recommendation services for transactions between consumers and merchants through the We Chat mini program,and better meets the need of relevant consumers to know information about agricultural products.