Scholarlink

This study presents a temporal modelling pipeline that combines unsupervised clustering, a first-order Markov model and sequential pattern mining to identify psychosocial states, including stress, focus, and relaxation, using multimodal digital biomarker data. The results reveal distinct psychosocial state trajectories, along with user-specific switching patterns and transition dynamics regarding inferred psychosocial factors to inform just-in-time adaptive interventions for intelligent timing and dynamic personalisation.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Temporal Inference of Psychosocial States from Digital Biomarkers for Just-in-Time Adaptive Interventions

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Aishah Shah,
  • Vladimir Tomberg

摘要

This study presents a temporal modelling pipeline that combines unsupervised clustering, a first-order Markov model and sequential pattern mining to identify psychosocial states, including stress, focus, and relaxation, using multimodal digital biomarker data. The results reveal distinct psychosocial state trajectories, along with user-specific switching patterns and transition dynamics regarding inferred psychosocial factors to inform just-in-time adaptive interventions for intelligent timing and dynamic personalisation.