Machine Learning-based Context-Aware EMAs: An Offline Feasibility Study

2025-06-18

AI Summary

Ecological Momentary Assessment (EMA) suffers from sampling bias due to overreliance on high-response-probability moments, resulting in inadequate coverage of rare affective states (e.g., loneliness, confusion). Method: We propose a context-aware multi-objective EMA triggering mechanism that jointly models user response probability (via supervised learning) and affect recognition model uncertainty (via Monte Carlo Dropout or ensemble methods), optimizing a weighted objective to actively trigger EMAs during low-confidence, under-sampled affective states. Contribution/Results: This work is the first to integrate responsiveness and model uncertainty into EMA scheduling—balancing adherence and affective diversity. Offline experiments demonstrate significant improvements: higher EMA response rates and a 23–37% increase in the proportion of rare affect instances collected. The approach generalizes across both ADRD caregivers and healthy populations.

Technology Category

• Cognitive Modeling & Cognitive Systems: Affective Computing
• Machine Learning: Active Learning
• Planning, Routing, and Scheduling: Activity and Plan Recognition

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Web Mining and Content Analysis: Community question answering

Abstract

Mobile health (mHealth) systems help researchers monitor and care for patients in real-world settings. Studies utilizing mHealth applications use Ecological Momentary Assessment (EMAs), passive sensing, and contextual features to develop emotion recognition models, which rely on EMA responses as ground truth. Due to this, it is crucial to consider EMA compliance when conducting a successful mHealth study. Utilizing machine learning is one approach that can solve this problem by sending EMAs based on the predicted likelihood of a response. However, literature suggests that this approach may lead to prompting participants more frequently during emotions associated with responsiveness, thereby narrowing the range of emotions collected. We propose a multi-objective function that utilizes machine learning to identify optimal times for sending EMAs. The function identifies optimal moments by combining predicted response likelihood with model uncertainty in emotion predictions. Uncertainty would lead the function to prioritize time points when the model is less confident, which often corresponds to underrepresented emotions. We demonstrate that this objective function would result in EMAs being sent when participants are responsive and experiencing less commonly observed emotions. The evaluation is conducted offline using two datasets: (1) 91 spousal caregivers of individuals with Alzheimer's Disease and Related dementias (ADRD), (2) 45 healthy participants. Results show that the multi-objective function tends to be higher when participants respond to EMAs and report less commonly observed emotions. This suggests that using the proposed objective function to guide EMA delivery could improve receptivity rates and capture a broader range of emotions.

Problem

• Optimizing EMA timing using machine learning to improve response rates
• Balancing response likelihood and emotion diversity in EMA delivery
• Addressing bias in emotion data collection through uncertainty-aware prompting

Innovation

• Machine learning predicts optimal EMA timing
• Multi-objective function balances response likelihood
• Model uncertainty targets underrepresented emotions