Skip to content

Latest commit

 

History

History
54 lines (41 loc) · 2.25 KB

README.md

File metadata and controls

54 lines (41 loc) · 2.25 KB

Internet Thoughts

alt text

internetthoughts.net

P.S we built our sentiment analyis engine into a neat little REST API It's super easy to use :)

Ever wonder what the internet thinks about a certain topic, issue, or person? Well, what better way to gauge the internet's popular opinion than by mass analyzing things said about that topic? Just enter a search query and a sentiment analysis of tons of data in relation to that topic will be returned to you. Find out where you stand amongst the internet! Internet-thoughts was built for YHack 2016 at Yale University by Tash-had Saqif, Robert Xu and Jonathan Sun.

Demo for Query "Internet"

Gif Demo

The magic behind it all

A custom sentiment analysis engine

  • We built our own engine to analyze sentiment in the data we fetched
  • Data is analyzed using a corpus
  • Engine accounts for slang, emoji's, compound words

Twitter, Reddit & Tumblr API's

  • Fetched the most popular tweets, reddit comments and Tumblr posts matching the query, from the last 7 days.
  • Parsed the data from the respective platforms for use with sentiment analysis engine
  • Platforms chosen specifically to ensure diverse demographic (age groups)

Technologies

  • NodeJs
  • JavaScript
    • jQuery
    • Embedded JavaScript (EJS)
    • PaperScript
    • Apiheap.js
  • HTMl, CSS
  • Heroku

TODO

  • Save data to database for analysis of change over time
  • Frequent words Wordmap
  • Context
    • Count exclamation marks
    • Count consecutively repeated letters
    • Compare with long term data from database
    • Use data from movie reviews, other sources
    • Use data from surely positive/surely negative sources (positive news, etc)
    • Use letter cases before converting to lowercase
  • Markov chains to predict search queries
  • Stability
    • Implement Travic CI
    • Implement SSL

License

Internet-thoughts is licensed under MIT