Paper: A Simple Bayesian Modelling Approach to Event Extraction from Twitter

ACL ID P14-2114
Title A Simple Bayesian Modelling Approach to Event Extraction from Twitter
Venue Annual Meeting of the Association of Computational Linguistics
Session Main Conference
Year 2014
Authors

With the proliferation of social media sites, social streams have proven to con- tain the most up-to-date information on current events. Therefore, it is crucial to extract events from the social streams such as tweets. However, it is not straight- forward to adapt the existing event ex- traction systems since texts in social me- dia are fragmented and noisy. In this pa- per we propose a simple and yet effec- tive Bayesian model, called Latent Event Model (LEM), to extract structured rep- resentation of events from social media. LEM is fully unsupervised and does not require annotated data for training. We evaluate LEM on a Twitter corpus. Ex- perimental results show that the proposed model achieves 83% in F-measure, and outperforms the state-of-the-art baseline by over 7%.