This is a prototype used for a team project in the Human-Computer Interaction course (ITE3062) at Hanyang University, Fall 2024.
This is also an extension of my side project.
In the app, the feature to record and revisit unsolved questions had been less-used feature in the app.
To addreess this, we designed a new feature that helps user recording, to encourage users to record unsolved questions using the app.
The proposed feature in this project is the keyword extraction, which identifies important keywords from the uploaded question image.
Our ultimate goal was to develop this feature and evaluate its effectiveness through a user study.
We recruited users to participate in a prototype test and survey.
Participants were given pre- and post- surveys before and after testing the prototype.
Each survey consisted of 4 questions about app usage and study habits, with responses measured on 5-point likert scale.
After collecting responses, we conducted paired t-test to evaluate the improvement for each question and the overall mean score, determining whether the improvements were statistically
significant.
- Pix2Text
- Sentence Transformers
- FastAPI
- Next.js
(Group 3)