Key Action Extraction for Learning Analytics

Abstract

Abstract. Analogous to keywords describing the important and relevant content of a document we extract key actions from learners' usage data assuming that they represent important and relevant parts of their learning behaviour. These key actions enable the teachers to better understand the dynamics in their classes and the problems that occur while learning. Based on these insights, teachers can intervene directly as well as improve the quality of their learning material and learning design. We test our approach on usage data collected in a large introductory C programming course at a university and discuss the results based on the feedback of the teachers.

Using Keystroke Analytics to Improve Pass-Fail Classifiers

“…Scheffel et al (2012) describe a method of data distillation, namely the extraction of key actions and key action sequences in order to leave behind meaningful data. The authors outline how the contextualized attention metadata (CAM) from a substantial university course in C programming is collected and then distilled using TF-IDF.…”

Monitoring, awareness and reflection in blended technology enhanced learning: a systematic review

“…As shown in Figure 2, the majority of papers considered blended learning as the combination of either face-to-face and distance learning (21), or face-to-face and computer-mediated interaction (20), or both of them. Only a few cases (Ram et al, 2011;Scheffel et al, 2012) presented blended learning as the mixture of formal and non-formal learning. Despite people constantly learning anywhere and anytime, 95% of the reviewed papers were devoted to formal learning and only one paper focused explicitly on non-formal settings (Ram et al, 2011).…”

Theory-led Design of Instruments and Representations in Learning Analytics: Developing a Novel Tool for Orchestration of Online Collaborative Learning

“…ADA has the potential to aid researchers to study learning, learners, and learning environments (Thompson, Ashe, Carvalho, et al, 2013); aid students by developing student-facing LA for meta-cognitive feedback and formative assessment (De Liddo et al, 2011;Lockyer & Dawson, 2011;Buckingham Shum & Ferguson, 2012;Wise, Zhao, & Hausknecht, 2014); and aid educators with educator-facing analytics, some of which can help facilitate orchestration (De Liddo et al, 2011;Trausan-Matu et al, 2014). Researchers utilizing ADA techniques for LA are concerned not just with mining data for analysis, but with using this analysis to enhance learning or learning environments in the subsequent redesign of learning tasks or for learner-facing analytics (e.g., Scheffel et al, 2012).…”