Predicting Collaboration based on Students' Pauses in Online CSCL Conversations

Sibel Denisleam (Molomer), Mihai Dascalu, Stefan Trausan-Matu

Research output: Contribution to journalArticleAcademicpeer-review

10 Downloads (Pure)

Abstract

As Computer Supported Collaborative Learning (CSCL) gains a broader usage as a viable alternative to traditional educational scenarios, the need for automated tools capable of evaluating active participation and collaboration among peers in online discussions increases. In this study, we validate a quantitative model of predicting involvement in CSCL chats based on student’s pauses throughout the timeline of the conversation. Starting from a corpus of 10 chat conversations, our proposed model explains 55% of the variance in terms of student participation and 42% in terms of collaboration, although relying on simple quantitative indices.
Original languageEnglish
Pages (from-to)83-92
JournalPolytechnical University of Bucharest. Scientific Bulletin. Series C: Electrical Engineering and Computer Science
Volume79
Issue number2
Publication statusPublished - Oct 2017
Externally publishedYes

    Fingerprint

Keywords

  • Computer Supported Collaborative Learning
  • pause analysis
  • fluency
  • speed
  • automatic evaluation of participation and collaboration

Cite this