Abstract:

Tags are always utilized to represent the interest and property of microblog users. We propose an improved microblog user interest modeling method based on multitag semantic correlation via analyzing the tag characteristics of microblog users and the limitations of existing microblog recommendation methods. Firstly, the co-occurrence frequency of tag pairs in the micro-blog user set is calculated to obtain the inner correlation between tag pairs. Secondly, the path is constructed based on the link tags for each tag pair and the outer correlation of tag pairs is obtained via the shared entropy. Finally, we combine the above two correlations to acquire the semantic correlation relation matrix, based on which the user tag matrix can be updated, thus the microblog user interest model based on multitag semantic correlation can be constructed. We evaluate our method through a series of experiments based on a dataset crawled from the open API of Sina Weibo and the results are analyzed. The results show that our method outperforms traditional user interest discovering methods.

Key words: