Supplier Recommendation Method Based on Semantic Matching and Bipartite Graph Network Collaborative Filtering Algorithm

In recent years, e-commerce has achieved rapid growth through the use of big data technology. With the increasing number of participants in emerging markets, e-commerce intermediaries play a key role in finding the best match between demand and supply. This article relies on data mining techniques to analyze 14,330 inquiry orders and 40,131 quotation records, revealing hidden relationships, patterns, and trends in the order data, and constructed a personalized and target supplier recommendation system. Firstly, a bipartite graph collaborative filtering recommendation algorithm is adopted for supplier recommendation. Then, TF-IDF and Word2Vec pre-trained models are used to calculate cosine similarity of commodity text data to identify buyer's product demands. Finally, a bipartite graph collaborative filtering algorithm and semantic matching algorithm are used to provide a list of recommended suppliers for the target users. This article explores the opportunities brought by big data to the supplier recommendation system in the commodity market.