Explainable Recommendation for Self-Regulated Learning

Recent years have seen rapid advances in intelligent technology to support online learning, but these have primarily targeted formal educational contexts such as classrooms and e-courses. In contrast, the predominant form of adult learning in the workplace is informal and self-directed. Learners selfassess competency, set goals, find relevant learning resources, and initiate learning activities covering many topics at different depths at different points in time. Our approach to supporting selfregulated learning is embodied in PERLS, a mobile personal assistant application that serves as a virtual mentor for informal learning. A key component of PERLS is its recommendation system, designed to adaptively co-construct a path towards desired learning outcomes with the learner. Recommendations are made largely on the basis of value propositions, each a persuasive explanation for taking a particular learning action to advance along a particular learning path. In this paper, we present a process model of self-regulated learning used in PERLS and our approach to generating explained recommendations and using them to co-construct learning paths.