Inspiration
Most stock prediction applications use previous stock trends to predict whether a company's stock will rise or fall. However, there's so much information being uploaded to the internet every second that it would be a waste not to use it.
What it does
This platform uses Amazon Kinesis to process the realtime stream of Twitter data from the Tweepy API. It runs the tweets through a basic algorithm I wrote to determine if a tweet written about a given company is positive or negative. It then congregates the data to determine if there will be a significant rise or fall in the stock for the company.
How I built it
I built it using CodeStar on Amazon Web Services (AWS). This created a template for the website built on Python Django. Python Django was useful because I could implement my Python scripts into the website seamlessly. Python is essential for its large catalog of data analysis libraries.
Challenges I ran into
Amazon Web Services is a bit overwhelming to use at first. Most of their services are built to scaling up for a large user base, so it felt silly using it for a project I just started. There's still a lot I have to learn to utilize all the services it has to offer. I spent a lot of time learning the basics of AWS so I didn't have time to connect all of the components of the platform.
Accomplishments that I'm proud of
Even though it's not finished, I'm content that I at least have a foundation for an actual platform. I don't have a strong background on working on backend components for applications so this was a good introduction to that.
What I learned
I leaned how to use AWS to start a web app built on the Python Django framework. This will be useful for future projects and I'm looking forward to learning more.
What's next for big data.
There's a lot that could be done for big data to turn it into a viable platform, like making the front end actually connect to the backend. Besides the obvious stuff of finishing the basic functionality, it would be interesting to train the algorithm on past data. For example, measure the overall connotation of past articles and compare how the stock prices changed around the time it was written. This would increase the accuracy of the algorithm and make it more reliable to base stock decisions on.
Log in or sign up for Devpost to join the conversation.