Graph-Based Modeling for Adaptive Control in Assistance Systems
The topic of this contribution is characterization and analysis of assistance systems in order to enable adaptivity, i.e., as personalized adaptive systems. The research question of this article is how to facilitate the modeling efforts in adaptive e-learning assistance systems. Adaptivity here means to personalize the usage experience to the individual needs of the users and their current working context. For that, adaptive systems need usage models and user models. The problem statement is that expert knowledge and recurrent efforts are needed to create and update these types of models. Data driven and graph analytics approaches can help here, in particular when looking at standardized interaction data and models which encode sequences such as interaction paths or learning paths. This article studies how to make use of interaction usage data to create sequence-typed domain and user models for their use in adaptive assistance systems. The main contribution of this work is an innovative concept and implementation framework to dynamically create Ideal Paths Models (IPM) as reference models for adaptive control in adaptive assistance systems.