Continuous Monitoring and Development of an Intelligent Tutoring System through Data Mining

Abstract

Continuous improvement of an intelligent tutoring system (ITS) involves fine-tuning its workings from time to time. We performed a series of data-driven analyses on student log data from Mindspark to understand usage patterns and identify ways to make the ITS more effective. We found that eighty percent of students did not complete all topics on Mindspark as they spent less than the recommended time on Mindspark and repeated concepts. The repetition of concepts was not because they used Mindspark insincerely. We are making improvements to the system based on the findings.