Project for the course of Social Network Analysis

Wikipedia is the largest encyclopedia ever written in the history of mankind and it has become one of the most re- garded reference works for many people. Its online nature has allowed a system of open collaboration, where each user who can access the website can become a contributor (with a few restrictions). The structure of the website is based on categories and subcategories, and each page is usually linked to many others. The aim of this study is to analyze the net- work of pages in the category of Society through the use of network science tools, and then to use community detection techniques to see if there is a match between subcategories and the communities discovered. Moreover, the general consensus expects Wikipedia pages to be written in a neutral language: this paper aims at deter- mining through the use of text analysis techniques if there is a difference in language among categories.

• Data crawling
• Netowork characterization and comparison with synthetic networks
• Community detection
• Feature-rich network analysis (Conformity)
• Sentiment Analysis and Topic Modeling