PRIMs Cognitive Architecture
Skills can be a powerful concept in applications in education. According to the skill approach, in order to master a particular course or topic, a set of skill needs to be learned.
Not all individual students will master the skills in the same order, and sometimes skills have other skills as prerequisit.
The right set of skills can be determined by a thorough task analysis, and validated through experimentation. Several successful cognitive tutors have been built using that methodology. However, this is a time-consuming process, and in many domains it is not so easy to isolate the particular skills.
An alternative approach is to use data from students on assignments that are relevant for the course as a source to determine the underlying skills. We have developed a method to deduce skill graphs from data, in which assignments are placed into nodes in the graph.
Arrows between nodes indicate prerequisite knowledge. The skill graph can be used to provide feedback to the student and teacher on the student's progress, or it can be used as input for a tutoring system to decide what material is most suitable for the student at this moment.