Theory-based Learning Analytics: Using Formal Concept Analysis for Intelligent Student Modelling

摘要

Learning Analytics is one of the most important fields for driving (educational) artificial intelligence. In this paper we briefly introduce one of Europe's key research initiatives to make a step from primarily statistics and data mining driven ways to do Learning Analytics, towards more theory-grounded, psycho-pedagogically inspired analyses. One approach we suggest is Formal Concept Analysis, which is a strong theoretical backbone in technology-enhanced learning settings For educators, FCA may provide Al-based analyses of 'big data' sets and intuitive representations for a number of pedagogical questions concerning the performance of students on the individual- as well as the class-level.