What it does

It accepts a few search terms, hash tags, to:/from: account names and performs a twitter search call. The returned tweets are then run through IBM's bluemix sentiment analysis, categorized on the basis of positive/neutral/negative and compiled into a bunch of important statistics, insights and plots using R.

How I built it

I used Twitter's search API for fetching data, IBM's bluemix API to analyze it, R for plotting and Python to bring it all together.

Challenges I ran into

Getting rid of twitter #spam from results.

What I learned

Graph plotting in R, Twitter's powerful APIs, Guts of sentiment analysis and it's potential impact

What's next for HotorNot

Real-time statistics using Twitter's stream API.

Share this project: