K-MEANS METHOD OF OUTER TUTORING LOOP GENERATION APPLICATION FOR COMPUTER TUTORING PROGRAMS ON SQL

A.G. Chukhray, Ie.S. Vagin, Z.V. Tomchenko, P. Anzenberger

A method of tasks clustering in computer SQL tutoring systems for outer tutoring loop organizing purposes is proposed. Solution of the problem relates to the cluster analysis, in particular – to the k-means algorithm application. One of the approaches based on competence components sets difference for the clustering of tutoring tasks is considered. A method of clusters and tasks sorting based on their complexity comparing is proposed. Some problems of authoring tools for intelligent tutoring systems architecture development are analized by SQLTOR development process example. Usage of solutions implemented in SQLTOR reduces manual and time efforts of a domain specialist (a teacher) during SQL tutoring course compiling. Main directions of future SQLTOR development are given.

Key words: Authoring tool for ITS, SQL learning, clustering, k-means, Hamming distance, competence components in tutoring.