5G marketing system based on multi-sensor collaborative filtering algorithm

Abstract

Massive user interaction information provides more data choices for recommender systems. With the increase of user choices, the sparsity of user interaction data has become an important challenge. For traditional recommendation algorithms, how to extract effective features to solve the problem of data sparsity has always been a hot research direction in the field of recommendation systems. In view of the above-mentioned problems, this paper proposes a recommender based on enhanced collaborative filtering algorithm for 5G marketing system. Specifically, by broadening the input vector field and fusing the user's social information and personal interaction information, the problem of multisource information collaborative modeling is effectively solved. Further, in order to achieve the enhancement of vector structure, we introduce a social regularization term and an internal regularization term to alleviate the overfitting problem of most recommendation algorithms. Experiments on a large number of real-world datasets show that our proposed method effectively improves the performance of the model on sparse data.