The article presents a new approach to understanding human behavior through mobile sensing. The researchers propose an automated strategy that uses smartphone sensor data to study behaviors like mood or sleep quality. This strategy includes selecting appropriate sensors, constructing features, and selecting models. The process is designed to be less time-consuming and burdensome than traditional methods, and it does not require expertise in multiple domains. The resulting sensing framework is also more generalizable, making it applicable to different scenarios.
Publication date: 13 Nov 2023
Project Page: https://doi.org/10.1145/1122445.1122456
Paper: https://arxiv.org/pdf/2311.05457
