Data-driven Behavioural and Affective Nudging of Online Learner -- 2

Abstract. In this work-in-progress paper, we describe the architecture of a system that can automatically sense an online learner’s situation and context (affective-cognitive state, fatigue, cognitive load, and physical environment), analyse the needs for intervention, and react through an intelligent agent to shape the learner’s self-regulated learning strategies. The paper describes the system concept and its software architecture and design: what sensory data are captured and how they are processed, analysed, and integrated; what intervention decision will follow and what behavioural and affective nudges will be given.