This project is a unique approach to tailoring an educational experience with ML. While others focus on ‘student assistants’ which operate on a combination of student responses, guided exercises, and NLP we have a completely new and disruptive idea. Give students what they really want. Break down the vast, often overwhelming media resources available for learning into small segments (call them learning bytes) that directly relate to the problem at hand (say a homework problem). Think of learning bytes as handy, perfectly relevant bits of media; a minute or two of video, a few lines from a textbook or website, a relevant figure, etc. We’ve done it manually and amassed a substantial learning byte database. We deliver them to students on-demand in a completely unique user interface with positive reviews from students. ML could rapidly expand the system. With the help of the Discovery Team, we hope to develop ML approaches which identify and curate learning bytes from existing, openly available media. The goal is to provide improved support for students and a new learning system for a start-up in the educational marketplace.
