A Concept Map-Based Learning Paths Automatic Generation Algorithm for Adaptive Learning Systems

Abstract

A concept map is an important knowledge visualization tool in adaptive learning systems. The weak concepts of students can be identified by analyzing the concept maps to provide learning guidance and teaching suggestions to students and teachers. However, in the current researches, it is difficult to analyze effective information from the concept maps when the number of concepts is large and the associations between concepts are complicated. It rarely reflects the learning performance of different student groups, which do not reflect the characteristics of the adaptive learning systems. A learning path automatic generation algorithm, Learning paths generation (LPG) algorithm, for adaptive learning systems is proposed. The LPG algorithm fully considers the learning performance of different student groups. The concept maps with different learning features are generated by the clustering algorithm and association rules mining, and several simplified learning paths are generated by using the topological sorting algorithm. Experiments show that the LPG algorithm has a good discrimination for the student groups and can generate different learning paths according to the students' learning performance.