Dynamics of Wikipedia Subnetworks in the Field of Computer Science

My undergrad dissertation, this was a lot of fun :)

It explores the structural dynamics of Wikipedia articles within the field of Computer Science. It looks at how Wikipedia’s hyperlink and categorisation systems reveal patterns in knowledge representation and collective memory. The project constructs a graph-based model using Doc2Vec embeddings of article content and reduces dimensionality using t-SNE for visualisation. A categorisation method based on centrality is applied to clean Wikipedia’s noisy category hierarchy.

The final interactive web application enables users to explore this semantic graph, view traffic patterns using Wikipedia’s Pageview API, and analyse relationships between topics like AI, sorting algorithms, and cybersecurity. Through case studies such as AlphaGo and Cambridge Analytica, the project also examines how public attention shifts across related articles.

Read the report here.