MABED (Mention-Anomaly-Based Event Detection) is a statistical method for automatically detecting significant events that most interest Twitter users from the stream of tweets they publish. In contrast with existing methods, it doesn't only focus on the textual content of tweets but also leverages the frequency of social interactions that occur between users (i.e. mentions). MABED also differs from the literature in that it dynamically estimates the period of time during which each event is discussed rather than assuming a predefined fixed duration for all events.
The experiments we conducted on both English and French Twitter corpora show that MABED has a linear runtime in the corpus size. They also demonstrate that the mention-anomaly-based approach leads to more accurate event detection and improved robustness in presence of noisy Twitter content. Furthermore, we note that MABED helps with the interpretation of detected events by providing clear temporal and textual descriptions.
Adrien Guille, Cécile Favre (2015) Event detection, tracking, and visualization in Twitter: a mention-anomaly-based approach Springer Social Network Analysis and Mining, vol. 5, iss. 1, art. 18
Adrien Guille, Cécile Favre (2014) Mention-anomaly-based Event Detection and Tracking in Twitter Proceedings of the 2014 ACM/IEEE International Conference on Advances in Social Network Mining and Analysis (ASONAM), pp. 375-382
Bellow are two visualizations that were automatically generated with MABED from a corpus of 1.5 million tweets published in November 2009 by 52,494 US-based users.