| dc.description.abstract | Courses are widely offered on the World Wide Web. Teachers must manage learning activities in the web learning system to create valuable learning effect. Recently, educators have increasing interest in portfolio assessment to actively assess and make effective strategies to promote students’ learning performance. Although the web leaning system can log web learning behaviors in web logs, the web log is not designed for pedagogical purposes. Teachers can not easily obtain the relationship between various learning behaviors and strategies. Consequently, teachers encounter difficulties to make decision to regulate effective learning activities or strategies by inferring students’ activity performance from historical information. This study describes the methodology to assist teachers analyze web portfolios, which contain in detail students’ learning behaviors, classroom processes, peer interaction history, for perceiving student activity performance in web classrooms. Database and data mining technologies are herein employed to faithfully record learning activity, analyse portfolios for pedagogical strategy regulation, and strategy enactment in web classrooms. With the proposed methodology, teachers can go online and efficiently use the web portfolios to monitor students’ learning status and perceive relationship among learning behaviors, pedagogical strategies, and peer interactions, rather than knowledge state of individual students. Therefore, teachers can make decisions to manage learning activities to promote learning effect in web learning system. | en_US |