Modeling the spread of ideas is a very abstract concept, and we wanted to be able to quantify it, both in analyzing real world data, and in using said data to develop semi-realistic simulations.
What it does and How we built it
It takes a chosen number of subreddit communities of reddit.com, and uses known coupled differential equations for the spread of infectious diseases to simulate the spread of ideas within these communities. In turn, ideas can spread between communities using the conditional probabilistic principles of Markov Chains.(See the page for a more in depth look at the mechanisms).
Challenges we ran into
It was hard to gather real, usable data to match models to than expected. It was also difficult to obtain useful documentation for a number of python libraries we needed.
Accomplishments that we're proud of
Successful, semi-realistic simulations for a fairly large number of interconnected subreddits, with theoretical models that we developed, and dynamic, interesting graphics to boot, all on one page.
What we learned
We learned how to use gitHub, how to create dynamic graphics for HTML using python, how to gather public information from APIs, and much more.
What's next for Simulating the Spread of Ideas with Epidemiology
We plan to finish and polish the webpage, and eventually refine the algorithmic and modeling process.
We also wanted to upload our final .gif dynamic graphical results, but they were too big, so we'll just show them in person! https://github.com/funkshun/fysix