Research on the Application of User Tags Based on Hybrid Recommendation of Agricultural Products Features
With the development of the Internet,the competition for agricultural products in the e-commerce market has become increasingly fierce,and users are unable to find suitable preferences from the numerous agricultural products information.In response to this issue,this article aimed to study the semantic retrieval of ontological knowledge of agricultural products features,constructed a specific feature knowledge system of agricultural products,and deeply analyzed the dimensions and scenarios of user label construction and user profile characterization.Based on this,the combination and association of user labels and user profile for agricultural products feature factors were summarized.Improving TF-IDF weight calculation was put forward based on collaborative filtering by collecting user behavior and using live tags;emphasizing the design of a hybrid recommendation system framework for agricultural products features based on user interest preference recommendation algorithms,filtering out the implicit association content between users and items,recalling the agricultural products that users prefer,and then using the logistic regression method to rank the recommended agricultural products features,thereby recommending the Top N agricultural products to users.This algorithm integrates the knowledge ontology of agricultural products features and combines a hybrid recommendation method of content and matrix decomposition to achieve more accurate agricultural products recommendations for consumers in the shortest possible time,and it provided significant application references for further deepening the promotion of the competitiveness and efficiency of the rural primary industry,increasing farmers'income,and expanding the development of agricultural products informatization.
user tagscharacteristics of agricultural productsagricultural products e-commercerecommended application of agricultural productsfarmers'income
万方数据
维普
