Our team was trying to think of a way to get young people informed and excited about investing. We believe that not only can investing in real companies be stressful, but also that it requires a lot of learning about companies and keeping up with current events to really learn from the experience. While these are good skills for young people to build, the majority of them will likely be scared off by this as a barrier to entry. What we proposed and built is a market based around a topic young people are very familiar with and would have to learn very little about to get started investing: memes.
What it does
Our project analyzes meme popularity over time by scraping popular meme sites such as imgur and iFunny and determines a price for the stock based on the number of occurrences of that type of meme. We use Selenium to scrape the web pages and the Clarifai API for image processing and recognition to determine if the scraped memes fall into the 5 categories of memes that we currently support: Bongo Cat, Distracted Boyfriend, Expanding Brain, Is This A Pigeon, and Skeleton Doot (in celebration of Halloween). We also used Google trends to acquire search data on these memes over the last 2 years.
How we built it
The web scraper and image recognizer for our projects were written in Python. We used a MySQL database hosted on the Google Compute Engine as well as a PHP web server also on GCE to store, process, and present our project's data.
Challenges we ran into
We ran into issues with setting up the database as well as learning the Clarifai API for the first time but were able to resolve both issues relatively quickly.
Accomplishments that we're proud of
We are proud of our ability to fully implement the trading feature in as little time as we had and that we were able to work on a project we all really enjoyed.
What we learned
We learned that there was a lot more that went into this project than we originally imagined and that we had to sacrifice certain goals in order to be able to complete the project on time.
What's next for Meme Market
In the future, we would definitely work on improving the UX/UI of our site to make it more appealing to our target audience, improve the image recognition model so it is even more accurate, and scrape more sites for memes to get a more accurate representation of their popularity.