Content Sentiment Analysis on the Dark Web

The Dark Web is a term used to describe the unsavory side of the Internet. The content on the Dark Web is diverse and widely dispersed. As a result, accessing and analyzing content have become a growing concern. In this thesis, we perform content sentiment analysis on various Dark Web sites and extract useful insights. In particular, a crawler was developed to traverse onion links and then perform content sentiment analysis on each link. Crawling reveals essential information and provides risk scores for the crawled URL. The crawler searches for additional links and then performs content sentiment analysis on those links too. Extracting the relationship between these links is the central focus and developing a meaningful connection between the links demonstrates that anonymity can be broken. The crawler then creates a graphical illustration that connects these links based on their risk scores. The aim of this project is to create a process that can assess the content of any website on the dark web. This purpose is fulfilled by a combination of two classifiers SVM, and Naïve Bayes as well as the sentiment analysis.