Inspiration

We wanted to provide users with more information about their investment decisions than just analyst predictions and stock trends.

What it does

We pull twitter data and use sentiment analysis to determine if a company is being spoken about in a positive or negative way. Using the number of followers, retweets, likes and the sentiment of the tweet, we make a recommendation to the user about buying, holding, or selling a stock.

How we built it

First, we used tweepy in order to pull data from twitter about a specific company/stock by making API calls. This includes both real-time tweets and tweets in past based on the users requested time frame. Then, using the json library, we then handed the tweets over to sentiment analysis. Sentiment analysis was performed using the NLTK library, then combined with other factors about the tweet was put into a mathematical formula. Then using HTML, CSS, javascript and flask, we created a website to relay the information about the stock pick back to the user.

Challenges we ran into

Use of tweepy and NLTK in conjunction with each other was challenging, and then pushing it to a website for easy user access.

Accomplishments that we're proud of

Using Tweepy in conjunction with NLTK, and providing users with an easy-to-use and unique application to influence investments decisions.

What we learned

We learned how to use tweepy, NLTK, flask to create a web application.

What's next for Stock Recommendations based on Social Media Analysis

There are multiple improvements we can make. First: Using other social media sites aside from twitter. Second: analyzing live data from multiple sites. Third: Upload the program to a domain for accessibility. Fourth: Using a more sophisticated mathematical equation that uses more data. Five: Conducting analysis on a far greater number of posts.

Share this project:

Updates