Workflow-based assessment of student online activities with topic and dialogue role classification

Jun Ma, Jeon Hyung Kang, Erin Shaw, Jihie Kim

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Scopus citations

Abstract

The Pedagogical Assessment Workflow System (PAWS) is a new workflow-based pedagogical assessment framework that enables the efficient and robust integration of diverse datasets for the purposes of student assessment. The paper highlights two particular e-learning workflows supported by PAWS. The first workflow correlates student performance, as measured by project grades, with different dialogue roles, information seeker and information provider, that students take on in project-based discussion forums. The second workflow identifies the distribution of question topics within student discussions. Both workflows employ state of the art natural language processing techniques and machine learning algorithms for dialogue classification tasks. Workflow results were reviewed with a course instructor and feedback regarding the analysis and its fidelity are reported.

Original languageEnglish
Title of host publicationArtificial Intelligence in Education - 15th International Conference, AIED 2011
Pages187-195
Number of pages9
DOIs
StatePublished - 2011
Event15th International Conference on Artificial Intelligence in Education, AIED 2011 - Auckland, New Zealand
Duration: 28 Jun 20111 Jul 2011

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume6738 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference15th International Conference on Artificial Intelligence in Education, AIED 2011
Country/TerritoryNew Zealand
CityAuckland
Period28/06/111/07/11

Keywords

  • Discourse analysis
  • discussion assessment
  • workflow technology

Fingerprint

Dive into the research topics of 'Workflow-based assessment of student online activities with topic and dialogue role classification'. Together they form a unique fingerprint.

Cite this