Mobile-Based Hybrid Experience Sampling for Real-Time Emotion Detection and Mental Health Insights

Abstract

The present study uses mobile technology to improve data dependability and application in mental health research by investigating the deployment of the Experience Sampling Method (ESM) for real-time analysis of emotional reactions. Designed using a cross-platform mobile app that combines audio, visual, and self-report survey data gathered via both random and time-based random sampling techniques. Over a 14-day period, participants received eight daily messages that prompted replies to structured questions gauging instantaneous emotional states. This hybrid data collection and processing system allows continuous monitoring of emotional dynamics, therefore offering high temporal resolution insights into users' psychological states. A convolutional neural network (CNN) architecture was employed to process data for emotion classification, achieving a 75% accuracy rate. With implications for individualized mHealth applications aiming at psychological health, the results show the potential of ESM combined with deep learning models in enhancing real-time emotional monitoring.

Cite this article as: G. Doğan, E. Yıldırım, B. K., Ö. Şengel, and F. P. Akbulut, “Mobile-based hybrid experience sampling for real-time emotion detection and mental health insights,” Electrica, 26, 0291, 2026. doi: 10.5152/electrica.2026.25291.