Learner-driven Subgoal Labeling on Math ProblemsAs an undergrad researcher of KIXLAB, I worked on building a system that helps math learners to better understand solution structures. I led this research project and presented a Late-breaking-work at CHI 2019 .
- Desgined UX and developed jQuery-base front-end
- Implemented an API server that uses Word2vec and Flask for data procesing
- Devised an algorithm for clustering sequential data
- Designed an optimal structure for Firebase Realtime DB
- Led the whole research and made a research result
Tech StackjQuery, Express, Node.js, Flask, Firebase Realtime DB
Human-machine Collaboration for QnA AutomationMy team for the CS project course developed a QnA system that answers programming beginners' questions on function usage. With 14 participants we conducted two experiments and found that the system is more effective to a function search than online searches. My team summarized the results and submitted a paper to L@S 2019 Work-in-progress.
- Designed and developed the QnA window in a form of Chrome extension
- Implemented a server notification feature on Chrome through background.js
- Developed a prototypical coding platform by using React.js and Semantic UI library
- Evaluated the system with a A/B testing
Tech StackChrome extension, React.js, node.js, AWS, Firebase Realtime DB
Community-driven Music Composition
For the course project of the social computing class, I developed a crowd composing system which helps and organizes people to compose a melody together. During a 1-week deployment in school, the system gathered 80 users, and my team received the "Best Crowdsourcer" award.
- Designed and developed the whole system UI and UX with jQuery and Bootstrap library
- Introduced a ranking system for social gamification
- Designed the database structure
- Designed the mechanism to aggregate crowdsourced data
Tech StackjQuery, Bootstrap, Firebase Realtime DB
Realtime QnA Platform for Lecturers
This is the course project for the intro to Human-Computer Interaction class. This project aims to build a QnA system that encourages questioning during lectures and help a lecturer check students' level of understanding. The system was introduced to KAIST professors, and they gave positive comments and ideas for improvement.
- Designed and developed the student-side interface
- Developed responsive interface by using the Bootstrap grid system
- Synchronized the student-side and lecturer-side systems through the DB polling method
- Devised a system feature to encourage questioning