Hansoo Lee

Hansoo Lee

Auto-generated profile, based on publicly available information

Hansoo Lee

Dyson School of Design Engineering, Imperial College London, London, United Kingdom and Intelligence and Interaction Research Center, Korea Institute of Science and Technology, Seoul, Republic of Korea

Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications

2 December 2025 in Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies
Verified
Journal Article
Open access
HLHansoo Lee
Hansoo Lee
TPTaehyeon Park
Taehyeon Park
YKYoungji Koh
Youngji Koh
JLJae-Gil Lee
Jae-Gil Lee
Uichin LeeUichin Lee
Uichin Lee
With advancements in mobile sensing technologies, there is a growing need for scalable and interpretable stress monitoring solutions that remain robust over time. Existing smartphone passive sensing approaches rely on statistical app usage features or multi-modal sensor data, making them susceptible to distribution shift and feature evolution (e.g., schema drift), and adding model complexity. To address these challenges, we propose the Smartphone Human Interaction-based Routine Behavior Task Mining, Modeling, and Feature Extraction (SHIRBT-MMF) framework, which models stress-related behaviors by mining finegrained interaction routine tasks rather than aggregated app usage patterns. SHIRBT-MMF leverages multi-level sequential pattern mining and large language model-based automated task modeling to extract interpretable and stable features from within-app UI state transitions. Unlike traditional methods that require hundreds of apps, SHIRBT focuses on a small, consistent set of routine-based tasks, mitigating covariate shift and feature evolution while improving model robustness. We validated SHIRBT-MMF through one- and four-month in-the-wild datasets with 26 participants, demonstrating that the SHIRBT-based personalized model achieves an average accuracy of 75%, outperforming baseline models by 5% while using only 3-6% of app types. Additionally, SHIRBT features remain stable over time, reducing covariate shift and ensuring reliable performance. With its expandability to other mental health, interpretability, and privacy-conscious design, the SHIRBT-MMF framework lays the foundation for personalized digital mental health monitoring.

5 October 2024
Verified
Conference Paper
Open access
HLHansoo Lee
Hansoo Lee
TPTaehyeon Park
Taehyeon Park
Uichin LeeUichin Lee
Uichin Lee
Smartphone data-driven mental health detection approach is widely used. However, existing methods focus primarily on enhancing model performance through numerous superficial smartphone usage features that do not effectively capture human behaviors, rather than interpreting the model's explanatory power. This study proposes a methodology of smartphone-based human behavior task modeling to advance "Explainable AI (X-AI)" for detecting mental health states. Our approach utilizes in-app UI-level trajectory data to extract more fine-grained human behavior task mining and modeling. We introduced novel concepts to investigate the nature and scope of smartphone-based human behavior tasks. This study discusses how these concepts can enhance the trustworthiness and transparency of models for both users and system providers. Furthermore, we expect that our approach could reduce data collection, thereby reducing memory usage and mitigating privacy concerns by enabling on-device learning.

13 April 2026
Verified
Conference Paper
Open access
HLHansoo Lee
Hansoo Lee
YCYoonjae Cho
Yoonjae Cho
SKSonya S Kwak
Sonya S Kwak
RCRafael A Calvo
Rafael A Calvo
Sleep is vital for health, yet access to data alone does not guarantee improvement. While wearables and health apps enable tracking, users face a “Data-Action Gap,” struggling to interpret metrics and translate them into action. Current interventions fail to bridge this: static dashboards lack context, rule-based agents rely on rigid scripts, and LLM-agents lack grounding in personal data, causing trust issues. We propose SAGE (Sensor-Augmented Grounding Engine) for an LLM-powered sleep care agent. SAGE normalizes continuous sleep, physiological, and activity data from the sensors into a queryable time-series layer. It supports (1) selective system-initiated monitoring that triggers notifications only upon detecting meaningful deviations against personal baselines to reduce alert fatigue, and (2) user-initiated Q&A where natural language is translated into executable database queries. By ensuring responses are grounded in precise period, comparison, and metric data, SAGE aims to enhance personalization, traceability, and trust, articulating a novel design space for evidence-based messaging in sleep care.

27 September 2025
Verified
Conference Paper
Open access
HLHansoo Lee
Hansoo Lee
EKEunseo Kim
Eunseo Kim
JLJeongmin Lee
Jeongmin Lee
HJHyunsoo Joo
Hyunsoo Joo
SKSonya S Kwak
Sonya S Kwak

20 February 2026 in JMIR Preprints
Verified
Preprint
Open access
SPSomin Park
Somin Park
Uichin LeeUichin Lee
Uichin Lee
HLHansoo Lee
Hansoo Lee
BACKGROUND Sleep health plays a critical role in emotional regulation and social motivation, which shapes how individuals engage with others in daily life. Insufficient or disrupted sleep has been associated with loneliness and social withdrawal, suggesting that sleep may influence patterns of social engagement. As much of young adults’ social interaction now occurs through smartphones, digital social engagement has become an integral component of everyday social behavior. However, little is known about how day-to-day fluctuations in objectively measured sleep relate to same-day digital social engagement in everyday life. OBJECTIVE This study aims to investigate daily, within-person associations between objective, multidimensional sleep health indicators and smartphone-based digital social engagement. Using longitudinal passive sensing data from young adults, we examine how day-to-day variations in sleep health relate to three domains of digital social engagement via smartphones: (1) app usage, (2) typing behavior, and (3) notification responsiveness. By analyzing dense, repeated measures, we seek to characterize fine-grained temporal relationships between sleep and social behavior in naturalistic settings. METHODS We conducted a longitudinal observational study of 14 young adults (mean age 25.21 years, SD 2.55), followed for a mean of 88.47 days (SD = 45.74). Sleep was monitored using a contactless under-mattress sensor (Withings Sleep [59]), capturing 6 sleep health indicators: sleep duration, sleep efficiency, wake after sleep onset (WASO), rapid eye movement (REM) sleep ratio, number of sleep sessions, and sleep regularity index (SRI). Smartphone interaction data were passively collected via a custom Android app. Digital social engagement was assessed using indicators of app usage, typing behavior, and notification responsiveness. Within-person associations were examined using generalized linear mixed-effects models with person-mean centering, with Benjamini-Hochberg correction applied for multiple comparisons. RESULTS A total of 1,327 analytical day observations were analyzed. Sleep duration emerged as the most consistent predictor of digital social engagement, showing significant within-person associations with 6 of 9 outcome variables. Longer sleep duration was associated with higher adjusted app usage duration (β=0.11, 95% CI 0.07-0.14; adjusted P<.001), increased app usage frequency (β=0.05, 95% CI 0.02-0.08; adjusted P<.001), more keystrokes (β=0.11, 95% CI 0.06-0.15; adjusted P <.001), longer typing duration (β=0.11, 95% CI 0.07-0.16; adjusted P <.001), and higher notification response frequency (β=0.06, 95% CI 0.02-0.1; adjusted P=.02). However, longer sleep was associated with slower typing speed (β=-0.1, 95% CI -0.15 to -0.05; adjusted P<.001). Higher sleep efficiency was associated with decreased app usage duration (β=-0.07, 95% CI -0.10 to -0.03; adjusted P<.001) and faster notification response time (β=-0.07, 95% CI -0.12 to -0.02; adjusted P=0.04). Higher REM sleep ratio predicted increased notification response frequency (β=0.10, 95% CI 0.06-0.14; adjusted P<.001). CONCLUSIONS This study demonstrates that daily fluctuations in sleep health, particularly sleep duration, are associated with same-day digital social engagement patterns in young adults. These findings suggest that adequate sleep may enhance capacity or motivation for digital social interaction, while the inverse relationship between sleep duration and typing speed indicates potential nonlinear effects. Passive sensing of sleep and smartphone behavior offers a promising approach for understanding the temporal dynamics between sleep health and digital social engagement in naturalistic settings. Future research should incorporate additional contextual data to better interpret the broader contextual meaning of digital social engagement patterns.