Commerce Product Recommendation Algorithm Based on Collaborative Filtering

Abstract

To acquire more ideal e-commerce recommendation results, this paper proposes an e-commerce recommendation strategy combined with big data technology and based on collaborative filtering. The algorithm puts forward an overall framework of personalized recommendation algorithm based on Hadoop platform, and then improves the traditional collaborative filtering algorithm. A number of users who are less different from the target users and have more co-occurrence times as the nearest neighbors of the target users and generate recommendations to fill the sparse user item scoring candidate set more reasonably is selected. The experimental results show that the collaborative filtering algorithm based on big data effectively solves the problem of new project recommendation, verifies the effectiveness of the evaluation index system of the distributed recommendation system, and shows better recommendation quality and scalability than the traditional recommendation algorithm.