aiTrade

aiTrade is a program to help amateur traders make decisions by analyzing Twitter headlines from popular news companies as well as various high-influence industry leaders and politicians' tweets.

Comment

Cady Baltz, Bryant Hou, Kendra Huang

We were inspired by one of the group member's previous professors was an inspiration for this project. During his economics class, he stressed the relationship between news, politics, and finance. After research on this topic, we believe that we can take advantage of this relationship, which is why we decided to create this project.

We utilized Google Natural Language to determine tweet sentiment from 8 major sources: the New York Times, CNN Breaking News, the Economist, Guardian News, the Washington Post, Donald Trump (President of the United States), Joe Weisenthal (co-host, BloombergTV), and Vitalik Buterin (Etherium founder). We then designed an algorithm to create a score from the sentiment values, defining the score for buying, selling, and holding. The program is presented as a website using the React framework. We created the backend with Python and Google Natural Language Processing, including Sentiment Analysis.

This was our group's first time using React, so we had to familiarize ourselves with this technology to create our project. We also faced challenged in connecting the Python backend to the React frontend. We had initially used flask to connect the frontend to the backend, but this caused problems with getting information from our Python script to the React frontend. We overcame this challenge by utilizing resources and online tutorials that showed us how to call a React program with a Python script. In addition, we went through through a trial-and-error process to finally get all the code running smoothly.

Share this project:

Updates