Date of Publication

2013

Document Type

Dissertation

Degree Name

Doctor of Philosophy in Computer Science

Subject Categories

Computer Sciences

College

College of Computer Studies

Department/Unit

Computer Science

Thesis Adviser

Raymund Sison

Defense Panel Chair

Rachel Edita Roxas

Defense Panel Member

Remedios Bulos
Merlin Teodosia Suarez
Raymund Sison
Nelson Marcos

Abstract/Summary

Although Asynchronous Online Discussions (AOD) offer many features that benefit the learners (e.g. asynchronism, anonymity, time and space independence), some problems still remain due to the lack of effective support towards the instructors task of monitoring and assessing the students interaction. This work outlines the theoretical underpinnings, experiments, results, and implications for the design of a novel and alternative computational method that can be used to analyze educational AOD messages. The method implements a hypothetical approach where representations of the manifest and latent features of text are combined to form a lexical network structure called LexNet. We argue that the structure of LexNet provides a natural mechanism to represent dependencies between concepts or words that describe the topic of text documents. The proposed method is intended to be used for developing discussion support tools that can provide evaluative feedback to instructors with regards the topical progress of online discussions, along with estimated measures of the students understanding of specific concepts relative to the domain and/or topic of discourse. To demonstrate the feasibility of the proposed method, a web-based prototype discussion support tool called ReLATED (Resource-Lean Analyzer of Transcripts of Educational Discourse) is presented. The evaluation of ReLATED using actual educational discourse transcripts has shown that the design of LexNet facilitates comparisons between the topical contexts of words and is potentially useful in monitoring the topic focus and assessing the appropriacy of concepts embedded in forum messages. Initial analysis and projections also showed that the time and memory requirements of the prototype system can be supported in an AOD environment without causing considerable delay in the students interaction. Finally, the designs incorporated into the proposed method exemplify how quality software attributes (e.g., usability, topic portability, and scrutability and adaptability) in future discussion support tools can be achieved with minimal tension, these attributes offer an extra value that can potentially encourage widespread and successful use of the system among teachers.

Abstract Format

html

Language

English

Format

Print

Accession Number

CDTG005460

Shelf Location

Archives, The Learning Commons, 12F Henry Sy Sr. Hall

Keywords

Computer-aided transcription systems

Upload Full Text

wf_yes

Share

COinS