Paper: Online topic model for Twitter considering dynamics of user interests and topic trends

ACL ID D14-1212
Title Online topic model for Twitter considering dynamics of user interests and topic trends
Venue Conference on Empirical Methods in Natural Language Processing
Session Main Conference
Year 2014
Authors

Latent Dirichlet allocation (LDA) is a topic model that has been applied to var- ious fields, including user profiling and event summarization on Twitter. When LDA is applied to tweet collections, it gen- erally treats all aggregated tweets of a user as a single document. Twitter-LDA, which assumes a single tweet consists of a single topic, has been proposed and has shown that it is superior in topic semantic coher- ence. However, Twitter-LDA is not capa- ble of online inference. In this study, we extend Twitter-LDA in the following two ways. First, we model the generation pro- cess of tweets more accurately by estimat- ing the ratio between topic words and gen- eral words for each user. Second, we en- able it to estimate the dynamics of user in- terests and topic trends online based on th...