Our final project for NETS 150 at the University of Pennsylvania is broken down into two parts:
We used concepts from information retrieval and document search to scrape Kanye West's twitter feed as well as a lyrics database for all of Kanye’s song lyrics. To scrape Kanye’s Twitter profile, we used the Twitter4j library. For the database, we made calls to a RESTful API. We then used the Vector Space Model to compare tweets and album lyrics, and album lyrics with other album lyrics.
We also used web scraping methods to collect all of Kanye’s collaborators for his Grammy nominations. We then created a graph of collaborations using our custom graph implementation and then we ran a neighborhood overlap analysis on the data to look for cliques and other features.
The [pdf](/My Beautiful Dark Twisted Triadic Closure.pdf) outlines the analyses and discoveries made in analyzing our model.
