Abstract
As part of an ongoing project where SWoRD, a Web-based reciprocal peer review system, is used to support disciplinary writing, this study reports machine learning classifications of student comments on peer writing collected in the SWoRD system. The student comments on technical lab reports were first manually decomposed and coded as praise, criticism, problem detection, solution suggestion, summary, or off-task. Then TagHelper 2.0 was used to classify the codes, using three frequently used algorithms: Naïve Bayes, Support Vector Machine, and a Decision Tree. It was found that Support Vector machine performed best in terms of Cohen's Kappa.
| Original language | English |
|---|---|
| Title of host publication | Educational Data Mining 2008 - 1st International Conference on Educational Data Mining, Proceedings |
| Pages | 192-196 |
| Number of pages | 5 |
| Publication status | Published - 2008 Dec 1 |
| Event | 1st International Conference on Educational Data Mining, EDM 2008 - Montreal, QC, Canada Duration: 2008 Jun 20 → 2008 Jun 21 |
Other
| Other | 1st International Conference on Educational Data Mining, EDM 2008 |
|---|---|
| Country | Canada |
| City | Montreal, QC |
| Period | 08/6/20 → 08/6/21 |
Fingerprint
All Science Journal Classification (ASJC) codes
- Computer Science Applications
- Information Systems
Cite this
}
Machine classification of peer comments in physics. / Cho, Kwangsu.
Educational Data Mining 2008 - 1st International Conference on Educational Data Mining, Proceedings. 2008. p. 192-196.Research output: Chapter in Book/Report/Conference proceeding › Conference contribution
TY - GEN
T1 - Machine classification of peer comments in physics
AU - Cho, Kwangsu
PY - 2008/12/1
Y1 - 2008/12/1
N2 - As part of an ongoing project where SWoRD, a Web-based reciprocal peer review system, is used to support disciplinary writing, this study reports machine learning classifications of student comments on peer writing collected in the SWoRD system. The student comments on technical lab reports were first manually decomposed and coded as praise, criticism, problem detection, solution suggestion, summary, or off-task. Then TagHelper 2.0 was used to classify the codes, using three frequently used algorithms: Naïve Bayes, Support Vector Machine, and a Decision Tree. It was found that Support Vector machine performed best in terms of Cohen's Kappa.
AB - As part of an ongoing project where SWoRD, a Web-based reciprocal peer review system, is used to support disciplinary writing, this study reports machine learning classifications of student comments on peer writing collected in the SWoRD system. The student comments on technical lab reports were first manually decomposed and coded as praise, criticism, problem detection, solution suggestion, summary, or off-task. Then TagHelper 2.0 was used to classify the codes, using three frequently used algorithms: Naïve Bayes, Support Vector Machine, and a Decision Tree. It was found that Support Vector machine performed best in terms of Cohen's Kappa.
UR - http://www.scopus.com/inward/record.url?scp=80052273105&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=80052273105&partnerID=8YFLogxK
M3 - Conference contribution
SN - 9780615306292
SP - 192
EP - 196
BT - Educational Data Mining 2008 - 1st International Conference on Educational Data Mining, Proceedings
ER -