Heuristic Theme Clustering for Online Social Media
Xiaole Qi, Xiaodong Wang, Shaohe Lv · 2019
In this paper, we propose heuristic theme clustering based on Word2Vec. Through the classification standard of ODP (Open Directory Project), the category center words are set, then the priori information of the ODP is used to guide the theme clustering. In this way, it can avoid the problem that the clustering label of the traditional unsupervised learning is not clear enough, and difficult to understand, and it can avoid the requirement of a large amount of annotation data for the classification method of supervised learning. So that heuristic theme clustering can be applied to the current online social media. The results show that our method has a good amount of theme information, and has a high precision.