Complex network and data analytics to classify and group users from a social networking site to help security agencies in Bangladesh

Abstract

Complex Networks can be used to model different types of relationships in different domains ranging from bioinformatics and computational biology through social networking sites to real life occurrences. On the other hand, very briefly, the term Data Analytics refers to analysis of data from any domain with a goal of discovering information that is useful for the task at hand. Data Analytics have become ubiquitous and can relate to almost every branch of Science and Engineering. This thesis aims to exploit both Complex Network and Data Analytics to assist the Security Agencies in Bangladesh in gathering intelligence and discovering information from the Social Networking Sites(SNS). In recent times, SNS like Facebook have become a major platform of social interactions. Facebook allows users to share ideas, thoughts, opinions etc. with friends and acquaintances. It allows groups to be formed- both public and private, pages to be created and so on. Similar minded users are often found to belong to similar groups and/or ‘like’ (an action of Facebook) similar pages and posts. Also, Facebook friends are often influenced by each other in how they will behave with respect to a group, page, post or comment. This raises the question, whether a user can be classified based on her affiliation to a group or connection to a page or registered emotion on a post or comment. This thesis therefore aims to capture meaningful data from Facebook, identify relationships from the captured data to construct a Complex Network of relevant users and then apply Network and Data Analytics to identify different types of users from the perspective of the security agencies of Bangladesh.