Inspiration

Anyone who has ever used BuzzFeed before knows that there is too much content released on a daily basis to be able to look at everything, every day. We were inspired to create a simpler interface to find interesting articles from BuzzFeed sorted by the general sentiment from the comments. This would allow us to easily find controversial or fun articles and not have to sift through the thousand or so articles released on a daily basis.

How it works

BuzzFeed queries BuzzFeed's private API and retrieves all of the comments from the articles that were released each day. From there, it uses HP's IDOL OnDemand API to determine the general sentiment of the comments and will display that information over a period of time.

Challenges we ran into

BuzzFinder uses extensive caching to prevent exceeding the relatively low limits of the BuzzFeed API. While our original plan was to provide a real-time analysis of individual articles, this ended up being our major roadblock that forced us to switch to a historical perspective.

Built With

  • buzzfeed
  • flask
  • hp-idol-ondemand-api
  • python
  • sentiment-analysis
Share this project:

Updates