Trump's tweet about Lockheed-Martin cuts $4 billion in value. Musk's $420 price per share announcement cost 11% shaved off of Tesla's price. Studies have demonstrated Twitter chatter strongly correlates with short-term stock market activity. When combined with sentiment-based trading methods, Twitter analysis may be a powerful tool for successful day-traders.

What it does

Twitter Trader is a powerful real-time natural language processing application that we’ve designed to scrape Twitter for the sentiment from politicians, executives, and the media that could affect short-term equity prices.

How we built it

Twitter API, Stock API, NLP Library, pandas.

Challenges we ran into

While we had a group of four, two of our group members were more business oriented so a lot of the coding was done by two people, while the other two directed the path of our coding as well as working on the presentation/logo. It was our first time trying to implement RESTful APIs, as well as the email automation through a python script that made us hit major roadblocks.

Accomplishments that we're proud of

Many times we were uncertain if we could get through the roadblocks we encountered. However, we were able to get a working product within the short amount of time with more implementations with stock info and email automation included. We really believe our idea is innovative and it was a great learning experience.

What we learned

  • Implementation of RESTful APIs using the Python packaged requests and Tweepy.
  • Implementation of email automation using smtplib package.
  • Implementation of a BackgroundScheduler to automate our program to run a check for tweets frequently.
  • Entrepreneurship/teamwork skills to create a program and presentation with people from different backgrounds (Edward & Sejin: Computer Science, Harrison: Psychology/Business, Oliver: Finance)

What's next for Twitter Trader

  • A web GUI that allows users to change their profile. (i.e. add more usernames to follow etc)
  • Mobile App integration for push notifications.
  • Refined and specialized NLP capabilities.
  • Machine learning powered regressions for price predictions over time.
  • Automated mutual fund offerings.

Built With

Share this project: