Twitter Data Clustering and Visualization

Twitter Data Clustering and Visualization ■

Andrei Sechelea, Andrei Sechelea, Tien Do Huu, Tien Do Huu, Evangelos Zimos, Evangelos Zimos, Nikos Deligiannis, Nikos Deligiannis

Abstract ■

We present a system for the acquisition, analysis and visualisation of Twitter data. Twitter messages are harvested and stored in a distributed cluster, and the data is processed using algorithms implemented in a MapReduce framework. We present a clustering algorithm capable of identifying the main topics of interest in a tweet data set. Also, we designed a visualization method which allows to follow the intensity of twitter activity at a given geographical location.