Adaptable and Reusable Query Patterns for Trace-Based Learner Modelling

摘要

This paper defines a framework to describe Learner Modelling (LM) process based on interactions traces. This framework includes an RDF-Based representation of knowledge models that can be used by a LM designer. The first model enables the LM designer to describe observations about learner's interactions with a TEL-system. The second model enables the LM designer to describe the structure of learner's profile. This framework supports also the description of reusable and adaptable SPARQL-based query patterns. These patterns enable the LM designer to calculate and infer learner profile elements for different TEL systems. We define the notion of query pattern and illustrate its application in the context of two TEL systems.