Analysis of blogs for cyber security threat detection

Abstract

Cyber Security threats and attacks are becoming more common and sophisticated over the years. In order to prevent them, the investigator first needs to retrieve, store and analyse a large pool of information. Blogs, since their first appearance in the 1990s, have become very popular, thus become a very rich source of information for the investigators. However, we need an efficient data mining and information retrieval technology to extract the useful information from the semi-structured text of the blogs. Our project aims to create a Web application that could analyse blogs for cyber security threat detection and at the same time evaluate the suitability of using mining techniques known as Latent Semantic Analysis (LSA) and Latent Dirichlet Allocation (LDA).