Hybrid Collaborative Filtering Recommendation Algorithm for ALS Model Based on a Big Data Platform

Aiming at the advantages of parallel computing on a big data platform, an improved algorithm for the ALS model under a big data platform is proposed. According to the number of users' visits, the hybrid recommendation is carried out, and a similar user reconstruction matrix model is used for optimization. Experiments show that: the ALS model hybrid collaborative filtering recommendation algorithm based on big data platform can improve the recommendation accuracy compared with the traditional K-means clustering recommendation algorithm, the article based recommendation algorithm, and the tag-based recommendation algorithm, and the acceleration ratio increases significantly with the increase of nodes when the big data set are running.