Student Performance Assessment Using Bayesian Network and Web Portfolios

Abstract

Web-based curriculum development allows students to develop their learning portfolio and interact with peers on a Web learning system. This system contains Web portfolios that record in detail students' learning activities, peer interaction, and knowledge progress. However, teachers cannot easily diagnose students' learning processes and regulate effective strategies according to student activity performance without the model of activity performance on Web learning systems. This study proposes a novel methodology that employs Bayesian network software to assist teachers in efficiently deriving and utilizing the student model of activity performance from Web portfolios in an online manner. Teachers can assess and diagnose performances with the model of learning activity on Web learning systems. The model of activity performance also allows teachers to manage various activity performances in Web learning systems so that desired strategies can be achieved to promote learning effectiveness.