Kermanshah Earthquake Event Tracking Through Persian Tweets

摘要

By expanding online social networks during the last decade, online event tracking has become one of the vital issues among researchers. In this study, Persian tweets which exchange among Iranian users have been considered in order to track hidden patterns of evolutionary Kermanshah earthquake event. A well-known dynamic community detection algorithm has been applied over key-graphs which are generated from extracted keywords of tweets. The key-graphs are generated from KNTUPT dataset which consists of entire Persian tweets between November 24^th, 2017 and December 8^th, 2017. The experimental study reveals the dynamic community detection algorithm had a superior result in terms of modularity as community quality and execution time as performance.