Mining Data-Logs From Intelligent Tutors to Create Learning Profiles of Students

Abstract

Intelligent tutoring systems are being used to teach mathematics due to the fact that they can provide individual tutoring, instant feedback on learning and flexibility to learn at own pace. In such computer supported learning environments, it is necessary that students develop the ability to learn independently. This paper reports the result obtained after mining the data logs generated by the intelligent tutoring system, ALEKS. Three clusters were generated after using two-step cluster analysis based on a variable which represents students' ability to learn independently. It was further found that students' marks in the coursework and in the final exam differed significantly among the three clusters.

Student's Patterns of Interaction with a Mathematics Intelligent Tutor: Learning Analytics Application

“…This ratio is represented by the variable mtop (which is an abbreviation of mastered_to_practiced). It was found that there is a moderately strong, positive and significant correlation between the value of mtop and student's final exam marks (r=0.466, sig=0.0) [10]. A high value of this variable indicates that a student is able to master most of the topics she is attempting to master, whereas a low value indicates contrary.…”

“…The objective of learning analytics is to derive information which can reveal how students use the intelligent tutoring or other e-learning systems and identify potential determinants of academic achievement. [17,20,10].Application of methods of learning analytics can be a powerful means to inform and support learners, teachers and their institutions in better understanding and predicting personal learning needs and performance [20,44]. This analysis can reveal detailed information about different patterns of learning activities, such as the rate at which learner is progressing in the learning environment, under which circumstances their progress is accelerated or decelerated.…”

“…Cluster analysis techniques group cases and hence can be used as a tool for exploratory data analysis. But this exploratory analysis is done without any subjectivity, hence it provides unambiguous profiles of learning behavior [5,10].…”

“…It was found that though single variable mtop is a significant predictor of final exam marks, it only explains 16% of changes in the final exam. [10]. Therefore other predictors were examined through correlation analysis and prior knowledge was found to be another significant predictor.…”

“…Learning analytics is a research discipline which provides framework for analysis of these system-generated large data logs in order to understand learning activities [17,39,41]. Use of such systems and reports generated by methods of learning analytics empower faculty to engage students in authentic learning opportunities and can increase student participation and motivation [10].…”