Student Portfolio Analysis by Data Cube Technology for Decision Support of Web-Based Classroom Teacher

Abstract

As learners increasingly use Web-based distance learning systems over years, large amounts of learning logs are generated. An instructor needs analysis tools to manage the logs and discover patterns within them to help improve instruction. A variety of analysis tools, including descriptive statistics, statistical inference, prediction, etc., can be utilized to analyze the effects of a teaching strategy from the logs. However, logs of a Web server, as learners' portfolios, cannot satisfy the requirements of these analysis tools. To resolve this problem, a data cube model is proposed as the infrastructure to store learning logs for analysis. We also describe the method of using query language to retrieve information from a database to construct the data cube. Furthermore, user-friendly operations for manipulating a data cube can retrieve statistical information from the data cube. Although statistical tools for managing Web logs exist, none specifically address the needs of the distance learning instructor. This article uses data cubes and database technology as fundamental analysis tools to satisfy a distance learning instructor's requirements for managing and analyzing learning logs.