Title: An improved user interest modelling by considering user social relationship

Authors: Yanru Wang

Addresses: School of Computer Science and Technology, China University of Mining and Technology, Xuzhou 221116, China

Abstract: User interest modelling has always been an important direction in the research of microblog social network. In previous studies, people paid more attention to the expansion of tags from microblog short texts, but ignored the information contained in the relationship between users. We improved the user interest model based on multi-tag semantic correlation by considering user social relationship (MTSC-SR). This model is built on the basis of the previous improved microblog user-tag matrix and integrates the social similarity matrix of users. The construction of user social similarity matrix is used to represent the similarity of two users by considering the static and dynamic interest information of users. We verify the validity of our method through experiments, and the data set is captured by the open API. The results show that the proposed algorithm has good performance in extracting users' interest features.

Keywords: user interest model; user-tag matrix; microblog social network; social similarity.

DOI: 10.1504/IJCI.2021.122705

International Journal of Collaborative Intelligence, 2021 Vol.2 No.3, pp.191 - 204

Received: 15 Aug 2020
Accepted: 17 Sep 2020
Published online: 06 May 2022 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article