Automated Students’ Thai Online Homework Assignment Clustering

Abstract

This paper proposes a model to cluster students' Thai online homework assignments before teachers go for further for grading, in other words Automated Students' Thai Homework Assignment Clustering. The proposed model consists of 5 parts: 1) Thai Word segmentation, 2) Stop-word removal, 3) Term Weighting, 4) Document Clustering, and 5) Performance Evaluation. The Thai Word segmentation splits sentences into individual tokens. The Stop-word removal defined as a term which is not thought to convey any meaning as a dimension in the vector space. The Term Weighting converts all tokens to vector space by TF-IDF. The Document Clustering is the process that is related to clustering algorithms computing a k-way cluster of a set of other documents. The last element of the prototype is performance evaluation of clustering validation measures in comparison between machines and human beings. The prototype was developed and tested by using Java and MySql. The experiment was conducted with 1000 undergraduate students who were assigned to complete a particular exercise in the course of Information Technology for Learning. The experimental results showed that the performance of the model could be as effective as the human performance with the 0.80 of F-measure; notwithstanding, the model apparently produced similar results significantly to human begins and faster outcome, which can facilitates teachers in terms of clustering into student groups and results, compared to other students or homework assignment grading.



Author Information
Thannicha Thongyoo, King Mongkut___s University of Technology North Bangkok (KMUTNB), Thailand
Somkid Saelee, King Mongkut___s University of Technology North Bangkok (KMUTNB), Thailand
Soradech Krootjohn, King Mongkut___s University of Technology North Bangkok (KMUTNB), Thailand

Paper Information
Conference: ACEID2016
Stream: Technology enhanced and distance learning

This paper is part of the ACEID2016 Conference Proceedings (View)
Full Paper
View / Download the full paper in a new tab/window


Posted by amp21