Personalized recommendation system based on social relationships and historical behaviors

Yan-Li Lee,Tao Zhou,Kexin Yang,Yajun Du,Liming Pan

DOI: https://doi.org/10.48550/arXiv.2206.13072

2022-07-14

Abstract:Previous studies show that recommendation algorithms based on historical behaviors of users can provide satisfactory recommendation performance. Many of these algorithms pay attention to the interest of users, while ignore the influence of social relationships on user behaviors. Social relationships not only carry intrinsic information of similar consumption tastes or behaviors, but also imply the influence of individual to its neighbors. In this paper, we assume that social relationships and historical behaviors of users are related to the same factors. Based on this assumption, we propose an algorithm to focus on social relationships useful for recommendation systems through mutual constraints from both types of information. We test the performance of our algorithm on four types of users, including all users, active users, inactive users and cold-start users. Results show that the proposed algorithm outperforms benchmarks in four types of scenarios subject to recommendation accuracy and diversity metrics. We further design a randomization model to explore the contribution of social relationships to recommendation performance, and the result shows that the contribution of social relationships in the proposed algorithm depends on the coupling strength of social relationships and historical behaviors.

Social and Information Networks

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to combine users' social relationships and historical behaviors more effectively in the recommendation system to improve the accuracy and diversity of recommendations. Specifically, most of the existing recommendation algorithms focus on users' historical behaviors while ignoring the influence of social relationships on user behaviors. Social relationships not only contain the inherent information of users' similar consumption tastes or behaviors, but also imply the influence of individuals on their social network neighbors. Therefore, the paper proposes a new algorithm that jointly utilizes social relationships and historical behaviors through linear optimization techniques, aiming to solve the following problems:

1. **Improve recommendation accuracy**: By considering social relationships and historical behaviors simultaneously, improve the prediction accuracy of the recommendation system, especially the performance when dealing with active users, inactive users, and cold - start users.
2. **Enhance recommendation diversity**: While improving the recommendation accuracy, maintain the diversity of the recommendation list and avoid recommending overly homogeneous items.
3. **Deal with data sparsity**: In the case of insufficient user historical behavior data, improve the recommendation effect by introducing social relationship information, especially for the recommendation of new users or new items.

To verify the effectiveness of the proposed algorithm, the author conducted experiments on four different types of user groups, including all users, active users, inactive users, and cold - start users, and used multiple evaluation metrics, such as Precision, Recall, F - score, Area Under the Average Precision - Recall Curve (AUPR), Intra - similarity, Hamming Distance, and Popularity. The experimental results show that the proposed algorithm outperforms the benchmark algorithms in multiple metrics, especially the performance on inactive users and cold - start users is particularly prominent.