As you may know, the stock market crashed this year in March due to COVID-19. So we decided to get into investing this spring to take advantage of the lower prices. Now the first question most people ask is what do you buy? Looking at technical charts full of complicated numbers and indicators quickly become very overwhelming and intimidating, so we created Due Diligence for Dummies!
What it does
Due Diligence for Dummies lets a brand new investor look up any stock and see for themselves, in plain English, how regular investors like you and I, and large investing firms are feeling about a particular stock. We’re not giving out any advice, but it rules out any complicated numbers and calculations in a regular chart and gives the new investor an opportunity to get a high level glance of the public’s opinion of a stock and potentially come to a decision on their own.
How we built it
Due Diligence for Dummies works by providing our backend with a timeframe and stock symbol, which scrapes the internet in real-time for articles and posts about a stock and uses our natural language processing machine learning models to generate a sentiment analysis and summary of the articles scraped. Alongside that, we used React to display all the data for our frontend, and Redux and Express to communicate to our server which performs the machine learning processing.
Challenges we ran into
Setting up model parameters and web scraping sources
Accomplishments that we're proud of
Utilizing web scraping and API's to produce our own datasets to perform sentiment analysis on and then rendering all in real-time!
What we learned
How to leverage NLP models to integrate them into applications to solve interesting problems! We also loved working with data that we produced by ourselves.
What's next for Due Diligence for Dummies
In the future we want to be able to scrape for events that occurred within that stocks lifetime that had a large impact on it’s price and map it onto a chart. This way the user can easily find and understand correlations and be on the lookout for future events such as new product announcements and earnings reports.