Fairness-Aware Behavioral Clustering for Self-Regulated Learning in Virtual Learning Environments (Poster 30)

• In Event: Sunday Poster Session 3:05 pm
  In Poster Session: Studying and Self-Regulated Learning SIG Poster Session

Abstract

Virtual learning environments (VLEs) are widely used in K-12 education; however, students struggle to engage in self-regulatory learning processes, drawing on the need to provide adaptive and personalized support. Although students’ behavioral patterns in VLEs were studied to design effective SRL support, very limited attention has been paid to the algorithm bias and the fairness of clustering results. We examined six behavioral patterns of SRL identified by fair-capacitated clustering (FCC), an algorithm that incorporates constraints to ensure fairness in the assignment of data points, from a sample of 14,251 secondary school learners with 20,297,075 rows of log data in a virtual math learning environment. The preliminary design ideas for adaptive support customized for each cluster are discussed.

Authors

• Yukyeong Song, University of Tennessee

• Chenglu Li, University of Utah

• Wanli Xing, University of Florida

• Hakeoung Hannah Lee, University of Virginia

• Shan Li, Lehigh University

• Yingbo Ma, University of Florida