EDM
No Single Metric Tells the Whole Story: Interpreting Multiple Pattern Mining Metrics
Wed Jul 1, 2:00 PM–2:25 PM · North 202
Part of AI Methods & Metrics
★ Notable speakers
Ryan S. Baker
★★
— Educational data mining; BROMP observational protocol; student disengagement detection
Jaclyn Ocumpaugh
★
— Automated detection of student affect and engagement; co-developer of the BROMP field-observation protocol
Luc Paquette
★
— Educational data mining, student behavior modeling, hint usage analysis, self-regulated learning in ITS
Argues that no single metric fully captures sequential pattern quality, and demonstrates how interpreting multiple pattern mining metrics together yields richer insights.
Authors
Andres Felipe Zambrano, Jaclyn Ocumpaugh, Ryan S. Baker, Zhanlan Wei, Xiner Liu, Hyeongjo Kim, Qianhui Liu, Jeffrey Ginger, Luc Paquette, Amanda Barany