Taehyeon Park
Github: TaehyeonPark
Research Engineer I'm an undergraduate student majoring in Computer Science and Electrical Engineering at KAIST. I've mainly worked on BE development, infra management, algorithm development, and framework implementation in both research and engineering projects, focusing on building reliable and efficient systems.
Experience
ICLab@KAIST
Yuseong-gu, Daejeon, South Korea
Dec 2023 - Sep 2024
Research Intern
Data Mining, ML, HCI
SPARCS
Yuseong-gu, Daejeon, South Korea
Mar 2023 - Now
Developer
BE, Infra, Data Analysis
Education
KAIST (Bachelor)
Yuseong-gu, Daejeon, South Korea
Feb 2023 - Now
Double Major CS & EE
Mainly interested in Data Science, AI and System Programming.
Projects
Developed the backend for Taxi for KAIST, a web and mobile service enabling KAIST members to easily find taxi ride-shares based on time and location. Designed and implemented RESTful APIs, real-time chat using WebSocket, and integrated MongoDB and Redis for efficient data handling and caching.
TAXI-Analytics | Private Repository
Designed and led Taxi-Analytics, an ongoing data analysis project (2024–Present) that evaluates the impact of events, promotional campaigns, and overall service usage of Taxi for KAIST. Utilized statistical methods such as histograms, KDE, and Regression Analysis to assess service effectiveness and provide actionable insights for future strategic directions.
Dev Center | Wiki, Github
Led the design and development of Taxi Dev Center, a Docker-based platform for Taxi for KAIST that provides isolated, scalable dev environments via VPN and private DNS server. Architected the system to streamline developer onboarding and deployment.
Misinfo Public Impact | Paper, Wiki, Github
Conducted a case study analyzing the role of digital media and misinformation during the 2024 U.S. presidential election, focusing on political engagement driven by celebrity and CEO endorsements. Utilized LLM for fact-checking Trump’s tweets, revealing that misinformation posts received significantly higher public engagement than factual ones.
Research Highlights
Accepted conference paper (UbiComp '24) | DOI
Smartphone-based Human Behavior Task Modeling for Explainable Mental Health Detection Model
Under reviewing conference paper 2025 |
Developed a multi-level sequential pattern mining algorithm and an LLM-based SHIRBT (Smartphone Human Interaction Routine Behavior Task) modeling and feature extraction framework for daily stress monitoring. This approach enabled more fine-grained and robust behavioral features. Post-deployment analysis demonstrated reduced covariate shift and schema drift, indicating improved model stability and generalization.
SKILLS
Infra Management, Web Development, Data Mining & Engineering, Data Analysis, AI Development
Interests
Programming, coding and coding.