Extended PrefixSpan for Efficient Sequential Pattern Mining in a Game-based Learning Environment

Abstract

This paper proposed an extended version of PrefixSpan as a better sequential pattern mining for a game-based learning environment (GBLE). The extended version of PrefixSpan evolved on integrating time interval constraints, clustering valued actions and extracting the closed sequences. These three concepts were derived after a previous work showed limitations of PrefixSpan in generating sequence patterns that can be used in tutoring services of a GBLE. The extended PrefixSpan underwent two phases of evaluation, performance evaluation and analyzing the quality of generated sequence patterns. The evaluation results showed that the extended versions provided a significant improvement in terms of execution time and the number of generated sequence patterns. Lastly, it shows significant improvement in the quality of sequence patterns generated as shown in better tutoring service it provided after integrating it to the GBLE.