Location-based emerging event detection in social networks

Abstract

This paper proposes a system for the early detection of emerging events by grouping micro-blog messages into events and using the message-mentioned locations to identify the locations of events. In our research we correlate user locations with event locations in order to identify the strong correlations between locations and events that are emerging. We have evaluated our approach on a real-world Twitter dataset with different granularity of location levels. Our experiments show that the proposed approach can effectively detect the top-k ranked emerging events with respect to the locations of the users in the different granularity of location scales. © 2013 Springer-Verlag.