Mining Evolutionary Topic Patterns in Community Question Answering Systems

Abstract

Community Question Answering (CQA) is becoming a popular Web 2.0 application. By analyzing evolutionary topic patterns from CQA applications, one can gain insights into user interests and user responses to external events. This paper proposes a novel evolutionary topic pattern mining approach. This approach consists of three components: 1) extraction of the topics being discussed through a temporal analysis; 2) discovery of topic evolutions and construction of evolutionary graphs of extracted topics; and 3) life cycle modeling of the extracted topics. We show empirically the effectiveness of our approach using two real-world data sets.