Use this URL to cite or link to this record in EThOS: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.772090
Title: Training translators on open-source MT technology : an empirical assessment of learning using a task-based syllabus
Author: Al Sharou, Khetam Y.
ISNI:       0000 0004 7661 1270
Awarding Body: UCL (University College London)
Current Institution: University College London (University of London)
Date of Award: 2019
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Thesis embargoed until 01 Mar 2027
Access from Institution:
Abstract:
This research project tests whether an intensive course based on a newly-formulated task-based (TB) syllabus can equip Master's trainee translators with the essential skills to create their own machine translation (MT) systems. The focus is on under-resourced language pairs, and for non-European teaching settings, therefore, the syllabus was tested on two groups of Master's level English-Arabic translation students in two Arabic-speaking countries, Oman and Jordan. These two groups who joined the training course had no previous knowledge of MT. The free and open-source statistical machine translation (SMT) software, Moses, was used as a training platform. Using a form of Action Research, the learners' engagement was monitored by collecting data on their experience and their reactions to the pace of delivery and the content of the syllabus, whilst assessing their achievements in using the SMT software. For the data collection, a multi-method approach was adopted, including questionnaires, student learning logs, TB assessment, interviews, focus groups, tutor's log and classroom observations. The findings have shown that the proposed TB syllabus is a suitable introduction to enable translation students, with no previous experience of working with MT and, in a short time, to create their own MT engines to translate from English into Arabic and vice versa. The findings have also demonstrated that integrating free and open-source MT into translator training programmes is a viable option and much needed. This is, especially true in Arabic-speaking countries where students have fewer chances to have practical contact with a range of translation technologies, due to lesser support but also lesser accessibility to paid language and translation technology (such as SDL Studio's computer-aided tool platform and/or a Google Translate API in a translation memory environment). The proposed syllabus with its practical element created competent users of a freely available tool, shifting their roles from being mere evaluators/post-editors to creators of MT engines, thus expanding the learners' capabilities in terms of translation technologies. Keywords: Free and Open-source Software, Moses Toolkit, Statistical Machine Translation, English-Arabic Language Pair, Translator Training, Task-based Approach, Technical and Technological Skills, and Students' self-efficacy.
Supervisor: Federici, F. ; Williams, G. Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.772090  DOI: Not available
Share: