Creation Method and Evolution Evaluation of Concept Knowledge Maps

Abstract

A concept knowledge map provides dependency among related knowledge and is used to help learners perform study navigation. The diversity of knowledge in the field of education and the discrepancy between subjects and learners extremely complicates the dependency among related knowledge. The rapid emergence of a number of educational information resources has significantly increased the size of knowledge maps. Moreover, the internal logical relations in various types of knowledge become less clear. This paper proposes a new knowledge unit centrality algorithm and evaluation method for the evolution model of concept knowledge maps. Results show that the dependency among related types of knowledge has an essential feature of complex network link directivity. Link prediction and maximum likelihood means are introduced to prove the existence of an ideal concept knowledge map. A method for quantifying the dependency among related knowledge is also developed to analyse the influence mechanism on learning efficiency. This method provides a new way for the comprehensive integration of information technology and educational science.