Taking my first EE class at Berkeley was super difficult. I remember reading the notes multiple times and still not really understanding what was going. Often times, students need an expert to explain a concept or address misconceptions/gaps in understanding of students or else the student will never be "unstuck". While students can go to office hours, not all students have the time to go to office hours.
This app tackles 2 main issues I've seen in education:
- Students struggling to understand concepts even after reading the notes multiple times
- Instructors struggling to pinpoint what concepts students struggle with the most during the course
What it does
This app has 2 components:
- Web app that allows students to discuss various parts of a notes document.
- Data analysis platform that allows instructors to pinpoint what conceptual gaps and misconceptions students might have with course content. It also allows for instructors to understand what is unclear in the notes.
I wasn't able to fully flesh out the data analysis platform because I ran out of time but I set up the whole infrastructure via the backend: https://github.com/ethanchewy/collaborare/blob/master/mysite/main/models.py.
For the social network, I was planning to use NetworkX to construct a basic directed graph where each user is a node and if there is an edge that points between vertex u => vertex v, that means user u interacted with vertex v.
For clustering, I was planning to run k-means,a simple bags of word algorithm, or just use a word cloud to help instructors visualize what topics are visualized the most via the text mining of the concepts.
What I did
I spent most of my time figuring out how to create a selectable box and create an event ou the frontend to trigger a form to be added at the exact x,y coordinate.
Before that, I spent most of my time designing what my app would look like from a backend viewpoint: https://docs.google.com/document/d/19OlO3G4zFKcR69pnR43ikWHP6aaSGXb-SSy2tjDdN28/edit?usp=sharing
I wish I had more time to flesh this out but I tried my best!
Although I ran out of time, I structured the data in a way where instructors can easily build a clustering model to identify what topics are discussed the most, social networks between users, and concepts maps,
Moreover, if I had more time, I would be able to create a more user-friendly interface.