FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W185119131> ?p ?o ?g. }

• W185119131 abstract "An increasing number of studies have been conducted to explore the application of automated writing evaluation (AWE) systems in English as a Foreign Language (EFL) writing instruction. Results from these studies show that current AWE systems pay insufficient or no attention to local coherence, an important element in textual coherence. This inability of local coherence analysis reduces the applicability of AWE systems in classroom-based essay tutoring and validity of these systems. It is therefore important to examine whether the computer can analyse local coherence in EFL essays automatically and, if so, how helpful such a computer system can be in EFL writing tutoring in contrast to human raters. Inspired by these two questions, this thesis discussed the development of a local coherence analysing engine (CTutor), and validated it in a setting of EFL writing pedagogical practice. CTutor is based on Centering Theory, which proposes that a discourse is a collection of utterances and that transition types between utterances are relevant to coherence of a discourse. This thesis utilized state-of-the-art natural language processing (NLP) technology to apply Centering Theory in developing CTutor. The evaluation of CTutor was both classroom-based and corpora-based. The classroom-based evaluation recruited 60 second-year university students and randomly placed them into a CTutor group receiving feedback from this software tool and a rater group having feedback from human raters. An initial questionnaire was conducted to investigate participants’ computer literacy and their understanding of coherence. Both groups needed to complete three writing tasks: one summary and two essays; and each writing task involved submissions of the first drafts and revised versions. Participants used feedback on their first drafts to do revision. The feedback that they obtained for revising was only about local coherence quality and in the form of underlining sentences containing possible coherence breakdowns. After completing all the writing tasks, participants of both groups were asked to do a follow-up questionnaire which attempted to explore participants’ attitudes towards the feedback that they received and to investigate their understanding of coherence after completing all the writing activities. In addition, participants with the most negative, neutral and the positive idea about CTutor were interviewed to further explore their perceptions of computer-generated feedback. The results from the classroom-based evaluation included questionnaire, interview, and written data. The questionnaire data showed that both groups in this study had equivalent computer literacy and that both groups had improvement in understanding of the concept of coherence and strategy usage in maintaining coherence after they finished all the writing tasks. In the questionnaire data it was noticed that participants of both groups held positive attitudes towards the feedback on local coherence quality. More importantly, the interview data revealed that participants of the CTutor group expect the software tool to provide further details on local coherence and information on global coherence. The written data, including participants’ first and second submissions, presented that both groups have similar revision patterns. The feedback from CTutor helped the CTutor group as effectively as that from human raters did the rater group. Findings from the classroom-based evaluation indicated that CTutor can produce acceptable and useful feedback on local coherence quality for EFL essay tutoring. The corpora-based evaluation was intended to check whether local coherence analysis conducted by CTutor can predict overall quality of essays. This evaluation validated CTutor with three groups of human-rated EFL essays by Chinese English learners at different proficiency levels, from high school(NMET), to college English-major second-year (TEM 4), and to college English-major forth-year (TEM8). Coherence scores generated by CTutor and overall scores given by human raters were put in regression calculation (with SPSS V.16). Results suggested that coherence scores from CTutor can significantly predict overall quality of essays. The findings in this evaluation further provided the answers with regard to the two initial research questions of this thesis. The application of Centering Theory showed the potential of automated local coherence analysis in the setting of EFL writing teaching The findings from the two studies of this thesis suggested that Centering Theory can be applied into CTutor for analysing local coherence in EFL essays. Additionally, CTutor offers effective feedback to help participants’ revising. Nonetheless, the issues, such as a shortage of detailed information on local coherence and the lack of global coherence analysis, are expected to be addressed in the further research. Grosz, B. J., Weinstein, S., & Joshi, A. K. (1995). Centering: A framework for modeling the local coherence of discourse. Computational Linguistics, 21(2), 203-225." @default.
• W185119131 created "2016-06-24" @default.
• W185119131 creator A5036240291 @default.
• W185119131 date "2014-11-21" @default.
• W185119131 modified "2023-09-27" @default.
• W185119131 title "Computer-Assisted EFL Essay Local Coherence Tutoring: An application of Centering Theory" @default.
• W185119131 hasPublicationYear "2014" @default.
• W185119131 type Work @default.
• W185119131 sameAs 185119131 @default.
• W185119131 citedByCount "0" @default.
• W185119131 crossrefType "book" @default.
• W185119131 hasAuthorship W185119131A5036240291 @default.
• W185119131 hasConcept C105795698 @default.
• W185119131 hasConcept C138885662 @default.
• W185119131 hasConcept C145420912 @default.
• W185119131 hasConcept C154945302 @default.
• W185119131 hasConcept C15744967 @default.
• W185119131 hasConcept C204321447 @default.
• W185119131 hasConcept C2781181686 @default.
• W185119131 hasConcept C2986378755 @default.
• W185119131 hasConcept C33923547 @default.
• W185119131 hasConcept C41008148 @default.
• W185119131 hasConcept C41895202 @default.
• W185119131 hasConceptScore W185119131C105795698 @default.
• W185119131 hasConceptScore W185119131C138885662 @default.
• W185119131 hasConceptScore W185119131C145420912 @default.
• W185119131 hasConceptScore W185119131C154945302 @default.
• W185119131 hasConceptScore W185119131C15744967 @default.
• W185119131 hasConceptScore W185119131C204321447 @default.
• W185119131 hasConceptScore W185119131C2781181686 @default.
• W185119131 hasConceptScore W185119131C2986378755 @default.
• W185119131 hasConceptScore W185119131C33923547 @default.
• W185119131 hasConceptScore W185119131C41008148 @default.
• W185119131 hasConceptScore W185119131C41895202 @default.
• W185119131 hasLocation W1851191311 @default.
• W185119131 hasOpenAccess W185119131 @default.
• W185119131 hasPrimaryLocation W1851191311 @default.
• W185119131 hasRelatedWork W1919525645 @default.
• W185119131 hasRelatedWork W2137641526 @default.
• W185119131 hasRelatedWork W2462058254 @default.
• W185119131 hasRelatedWork W2736590803 @default.
• W185119131 hasRelatedWork W2739976240 @default.
• W185119131 hasRelatedWork W2767033731 @default.
• W185119131 hasRelatedWork W2889185756 @default.
• W185119131 hasRelatedWork W2950885993 @default.
• W185119131 hasRelatedWork W2992721065 @default.
• W185119131 hasRelatedWork W3028269575 @default.
• W185119131 hasRelatedWork W3045985348 @default.
• W185119131 hasRelatedWork W3096343153 @default.
• W185119131 hasRelatedWork W3097109642 @default.
• W185119131 hasRelatedWork W3114162736 @default.
• W185119131 hasRelatedWork W3135884202 @default.
• W185119131 hasRelatedWork W3168405756 @default.
• W185119131 hasRelatedWork W3194524842 @default.
• W185119131 hasRelatedWork W3200464929 @default.
• W185119131 hasRelatedWork W3202558929 @default.
• W185119131 hasRelatedWork W3112595131 @default.
• W185119131 isParatext "false" @default.
• W185119131 isRetracted "false" @default.
• W185119131 magId "185119131" @default.
• W185119131 workType "book" @default.