As college students, education is very important to us. Some of us have parents who teach, so we know how frustrating the disconnect between students and professors can be for everybody involved. Students want to learn and teachers want their students to be successful. However, feeling embarrassed can keep students from asking questions or reporting how much they actually understand. Education professionals have struggled to solve this problem for years so we decided to craft our solution, Clasroom.ai, during Knight Hacks 2019.
What it does
Confidence Levels Classroom.ai allows students to self-report their confidence levels during class. It anonymously shares this data with teachers so they can adjust mid-lecture or look back later to see what confused students most. This data is also used to create a personalized model that recognizes individual student's faces and determines whether they are feeling confused, neutral, or confident. Our hope is that the model would eventually be strong enough that students would no longer have to self-report their confidence levels.
Other Features Along with reporting confidence levels, students can also send anonymous questions or requests to the teacher that they might be too embarrassed to ask otherwise. If other students have the same question, they can upvote the message, letting the teacher know exactly how many students want this explained.
How we built it
Challenges we ran into
Making an accurate model takes time. It was hard to collect the necessary amount of data and train the model within the time constraints of Knight Hacks. In order to make the most accurate model possible, we embraced the hackathon spirit and got very little sleep. When one member succumbed to exhaustion, it was up to the other members to pick up where they left off. This required us to learn languages we hadn't used before. Anyways, we made it through :).
Accomplishments that we're proud of
We're most proud of being able to develop a project that combines each of our strengths and interests.