Inspiration

Studies have shown that Twitter tweets can accurately predict coronary heart disease based on their sentiments. Going off of this, we looked into IBM's Alchemy API and started hacking from there with the intention of measuring sentiments of different hashtags in different locations.

What it does

SentiMeter takes in a Twitter hashtag and displays recent tweets, along with their sentiments from a scale of -1.0 to 1.0. Negative values represent negative sentiments and positive values represent positive sentiments.

How I built it

We used the Meteor framework to build a WebApp that integrates the Alchemy API, Twitter API with a simple front-end design. We started with a simple UI to input words and interact with the Alchemy API. Then we integrated the twitter API and polished the UI.

Challenges I ran into

We were new to working with Meteor framework, so we had some trouble figuring out how to make API calls to Alchemy and Twitter. We struggled with synchronous and asynchronous calls and how to have Twitter JSON objects recognized in the Meteor framework.

Accomplishments that I'm proud of

Successfully finishing an app using multiple APIs.

What I learned

We learned how to use the Meteor framework-- we created templates, collections, and functions. We also learned how to make REST calls to the Alchemy API while parsing the JSON information given back. We learned how to plan out a hack and execute it step by step in manageable chunks.

What's next for SentiMeter

Averaging sentiment for similar-context tweets, more tweets, better visualization and integrating tweet locations.

Built With

Share this project:

Updates