We won the Bloomberg API prize at HackPrinceton Fall 2014 - http://challengepost.com/software/context-yu2ia

• Taking a news article that the user provides

  • Understanding the information behind that article
    • Using methods to find the item that is of the highest probability
  • Giving contextual background on that particular topic using certain sources
    • Contextual background could be: A timeline, news articles that relate, and definitions of words that are important
  • Good UI that takes an article, and displays it in a nice way with all this contextual information.

• Second build ontop:

  • Taking a term, and giving you context on that term.

• We can build initial term base by scraping articles, and seeing things that are relevent.

Majority of our work was in clustering algorithms, natural language processing, sentimental analysis, and using these tools to do machine learning and output them in a great UI.

Our finishing API stack

• bloomberg
• twitter
• firebase
• new york times
• imgur
• wikipedia
• google charts
• reddit

Our finishing tech stck

• python
• mongo
• flask
• scikit-learn
• nltk