We were inspired by Target Alpha and wanted to create something that would assist us to make a more educated decision to stock purchases based on public opinions and emotions related to a specific company. A great deal can be learned from how the general public feels about a specific company and we believed that this could be extended to even make financial decisions.

What it does

SentiTweets uses sentiment analysis from Twitter Tweets to predict stock performance using Google's Natural Vision Machine Learning API. Upon inputting a NASDAQ ticker, the general public opinion is gauged and used to make a decision about whether to buy or sell stocks.

How we built it

Backend was built using: Twitter API, Google Natural Vision Machine Learning API Frontend was built using: HTML, CSS, Node.js

Challenges we ran into

The time period for this Hackathon was only 5 hours, learning and implementing the APIs within a short time period was a challenge.

Accomplishments that we're proud of

It works! During the time of this hackathon, GM had announced that they would be closing a major manufacturing location in Oshawa, this received negative feedback from critics and the general public. Our program was able to successfully conclude that at this time, stock prices would fall and stocks should not be bought.

What we learned

How to use sentiment analysis and gain comprehensive results.

What's next for SentiTweets

Improved UI. We can also apply a similar process to other social media platforms. Last, we can integrate a "user profile" feature and integrate machine learning algorithms to buy and sell a specific users stocks based on public opinion about companies.

Share this project: