Social Big Data Mining Framework for Extremist Content Detection in Social Networks

摘要

Social networks provide the platform for flows of ideas and affordable and global online communications. Many people use social networks to communicate and express their opinions in supporting or opposing different causes, with most of this user-generated content being textual information. As there are a lot of raw data of people posting real time messages about their opinions on a variety of topics in daily life, it is a worthwhile research endeavor to collect and analyze these data, which may be useful for government to make informed decisions or to monitor public opinion. Data available in social media is obviously only one type of information that can be of interest when trying to detect a possible terrorist or radical group, there are several cases for example in which the social media has been used by radical thinkers to act as influencers and encourage fanatics with the same radical views to take violent action. Therefore, in this paper, we propose a framework for opinion mining and extremist content detection in on line social media data. Social media data targeted in this work to analyze, is the public text post on Facebook, the most popular social networking site. With this framework, machines can learn how to automatically extract the set of messages from Facebook public pages, using API graph calls, filter out non-opinion messages. determine their sentiment regarding the issue of interest directions (i.e. positive, negative) and detect violent or extremist content. The purpose of this model is to build a Big Data application that gets stream of public data from Facebook social network, which can help law enforcement and cybercrime analysts with analyzing and monitoring social media, in the search of digital trace of violence or radicalism, that can be exploited in further digital forensic investigation.