| Popis: |
Sina Weibo is one of the most influential social platforms in China. Recommendation system helps user to find celebrities that they may interest in and thus helps to attract more users. User's Weibo contents reflect their personal preferences. In this paper we proposed an LDA topic modeling based recommendation method which can discover topics of user's Weibo contents and recommend celebrities that users are interest in. The comparison result shows that our method outperforms tf-idf-based recommendation method. |
