Personalized Sliding Window Recommendation Algorithm Based on Sequence Alignment

With the explosive growth of the amount of information in social networks, the recommendation system, as an application of social networks, has attracted widespread attention in recent years on how to obtain user-interested content in massive data. At present, in the process of algorithm design of the recommending system, most methods ignore structural relationships between users. Therefore, in this paper, we designed a personalized sliding window for different users by combining timing information and network topology information, then extracted the information sequence of each user in the sliding window and obtained the similarity between users through sequence alignment. The algorithm only needs to extract part of the data in the original dataset, and the time series comparison shows that our method is superior to the traditional algorithm in recommendation Accuracy, Popularity, and Diversity.

Automated summary

In this paper, a personalized sliding window is designed for different users by combining timing information and network topology information. The information sequence of each user in the sliding window is extracted and user similarity is obtained through sequence alignment. The results show that our method outperforms traditional algorithms in recommendation accuracy, popularity, and diversity.