An extraction and unification methodology for social networks data: an application to public security

Social media are invaded our daily life where millions of users are subscribed. Those online sites provide great tool for users to communicate with others from all over the world, share information, and express their opinion. Unfortunately, this ease of access and availability of information is exploited by malicious users to spread their radical ideas and prepare for terrorist attacks. In this paper, we have proposed a new methodology for data extraction, annotation, and unification in order to identify suspicious content from online social media threatening public security. To the best of our knowledge, there is no specific data collected method from social media devoted to suspicious user profile extraction. Also, implying an expert for data annotation for a security purposes remain a new research task that necessitate expertise and knowledge. In this paper, we tackle this research area by collecting suspicious content from different social media.