E-commerce Recommendation Algorithm of Agricultural Products based on Knowledge Graph

To effectively address information overload and knowledge redundancy, search and recommendation have emerged as an effective way for users to improve retrieval efficiency in massive amounts of information. Traditional ecommerce recommendation systems cannot recommend agricultural products based on their unique attributes. The knowledge graph stores rich semantic information on nodes and edges, enabling deep mining and analysis of entities and relationships. This article adopts a top-down approach to construct a knowledge graph of agricultural e-commerce, integrates attention mechanisms to construct a recommendation model, and updates the embedding of its own nodes by recursively propagating the embedding of neighboring nodes on the knowledge graph. Additionally, attention mechanisms are used to distinguish the importance of neighboring nodes and more accurately capture user interests. The research results provide a solution for constructing an agricultural product ecommerce system. By typing keywords in search engines, users can quickly find agricultural product information related to their own needs, providing more accurate and personalized agricultural product recommendation services.