Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

ZzzGPT: An Interactive GPT Approach to Enhance Sleep Quality

Yonchanok Khaokaew, Kaixin Ji, Thuc Hanh Nguyen, Hiruni Kegalle, Marwah Alaofi, Hao Xue, Flora D. Salim

TL;DR

The paper tackles turning rich wearable sleep data into actionable guidance by framing sleep quality as a regression problem for sleep efficiency and proposing a two-stage framework that leverages LLMs for data augmentation and an interactive interface. Stage One trains multiple regressors on GLOBEM data, augmented with synthetic samples generated by LLMs to address class imbalance, and identifies a compact 20-feature subset that yields strong predictive performance with CatBoost and ensemble methods. Stage Two provides an interactive, LLM-powered chatbot that offers sleep predictions, data acquisition, and personalized recommendations with visual, manipulable interface elements. The work demonstrates that LLM-based data augmentation can improve predictive accuracy and that integrating explanatory, user-centric tools can enhance adoption of wearable sleep-monitoring technologies, moving toward a personalized sleep-health dashboard.

Abstract

This paper explores the intersection of technology and sleep pattern comprehension, presenting a cutting-edge two-stage framework that harnesses the power of Large Language Models (LLMs). The primary objective is to deliver precise sleep predictions paired with actionable feedback, addressing the limitations of existing solutions. This innovative approach involves leveraging the GLOBEM dataset alongside synthetic data generated by LLMs. The results highlight significant improvements, underlining the efficacy of merging advanced machine-learning techniques with a user-centric design ethos. Through this exploration, we bridge the gap between technological sophistication and user-friendly design, ensuring that our framework yields accurate predictions and translates them into actionable insights.

ZzzGPT: An Interactive GPT Approach to Enhance Sleep Quality

TL;DR

The paper tackles turning rich wearable sleep data into actionable guidance by framing sleep quality as a regression problem for sleep efficiency and proposing a two-stage framework that leverages LLMs for data augmentation and an interactive interface. Stage One trains multiple regressors on GLOBEM data, augmented with synthetic samples generated by LLMs to address class imbalance, and identifies a compact 20-feature subset that yields strong predictive performance with CatBoost and ensemble methods. Stage Two provides an interactive, LLM-powered chatbot that offers sleep predictions, data acquisition, and personalized recommendations with visual, manipulable interface elements. The work demonstrates that LLM-based data augmentation can improve predictive accuracy and that integrating explanatory, user-centric tools can enhance adoption of wearable sleep-monitoring technologies, moving toward a personalized sleep-health dashboard.

Abstract

This paper explores the intersection of technology and sleep pattern comprehension, presenting a cutting-edge two-stage framework that harnesses the power of Large Language Models (LLMs). The primary objective is to deliver precise sleep predictions paired with actionable feedback, addressing the limitations of existing solutions. This innovative approach involves leveraging the GLOBEM dataset alongside synthetic data generated by LLMs. The results highlight significant improvements, underlining the efficacy of merging advanced machine-learning techniques with a user-centric design ethos. Through this exploration, we bridge the gap between technological sophistication and user-friendly design, ensuring that our framework yields accurate predictions and translates them into actionable insights.
Paper Structure (16 sections, 1 equation, 6 figures, 1 table, 1 algorithm)

This paper contains 16 sections, 1 equation, 6 figures, 1 table, 1 algorithm.

Table of Contents

  1. Introduction
  2. Dataset & Problem Definition
  3. Stage One -- Predictive Model
  4. Regression Models
  5. Feature Selection
  6. Data Pre-processing
  7. Data Generation with LLMs
  8. Experiments
  9. Experiment setting
  10. Baseline Models.
  11. Results
  12. Feature Performance Comparison.
  13. Model Performance Comparison.
  14. Stage Two -- Interactive Interface
  15. Conclusion
  16. ...and 1 more sections

Figures (6)

  • Figure 1: Wordcloud of Quoras quora-question-pairs and Reddit redditqa datasets
  • Figure 2: Overview of the proposed two-stage framework
  • Figure 3: Sleep Efficiency distribution throughout 4 years of GLOBEM dataset (2018 to 2021)
  • Figure 4: Prompt for generating additional data
  • Figure 5: The performance of the proposed model with other baseline models, proposed by 10.1145/3351274, 7764553, and info:doi/10.2196/mhealth.5960.
  • ...and 1 more figures