Personalized presentation of multimedia objects for home healthcare environments: a peer-based intelligent tutoring approach.
John Champaign, Robin A Cohen · International Conference on User Modeling, Adaptation, and Personalization · 2012
In this paper we present an approach for reasoning about which media content from an existing repository should be presented to users. We elaborate on our technique by considering students within an e-health intelligent tutoring environment. Our approach models the benefits in socially connected learning gained by peers in order to then recommend those objects predicted to offer the best gains in knowledge for the student. This is achieved in a framework where the past learning gains of peers are modeled and recorded with the objects in the repository. We previously confirmed the value of the approach by simulating student learning. From here, we then conduct a user study comparing the learning achieved by students presented with objects selected by our algorithms, compared to a less principled approach for curriculum sequencing; this is performed for the application of home healthcare (assisting caregivers of autistic children). We provide compelling evidence for the value of our proposed vision for achieving effective peer-based tutoring: through past experiences of peers in an extensive repository.