Datawarehouse-based approach for the analysis of terrorism-related activities in social networks

Abstract

Social network analysis techniques for activities related to terrorism are mainly based on data mining techniques. These techniques do not take into account the various axes of analysis allowing a study according to several facets. In this article we propose a comparison study of these techniques. We present our approach of analyzing these activities, based on Data warehouse and OLAP analysis. We aim to improve the analysis of these cyber threats. OLAP analysis allows us to explore social networks to detect dangerous content in the direction of targeted cyber threats. Our approach is based on five-tier architecture: (1) data sources; (2) ETL; (3) Data warehouse; (4) Analysis; (5) Presentation. In our experimentations, we used Twitter to detect and analyze the incitement to terrorism and determine the users supported the terrorism. We proposed a datamart with a metric named score, calculated using a data mining technique. Also, we used OLAP analysis techniques, based on the history of positive scores, to determine the users inciting terrorism, their locations and their retweets.