Developing a 'Neo-Metric' for the evaluation of translation memory

Watkins, Gareth 2016. Developing a 'Neo-Metric' for the evaluation of translation memory. New Voices in Translation Studies 15 , pp. 31-62.

Preview

PDF - Accepted Post-Print Version Download (579kB) | Preview

Abstract

A novel system by which Translation Memory (TM) tools can be evaluated would be desirable for translation industry stakeholders. Such evaluation systems are commonly known as metrics. While metrics devised to evaluate Machine Translation (MT) are well developed, there has been limited work on the creation of a TM tool metric. This paper discusses the creation of a ‘Neo-Metric’ designed to help translators, translation professionals and translation industry stakeholders as they evaluate TM software. Having identified a need for such a metric, a brief literature review is conducted. Variables to be included in the Neo-Metric are discussed, and an initial metric is introduced. This first version of the metric is briefly tested, weaknesses are identified, and based on these initial tests, a refined metric is developed for more in-depth testing along with testing in the field, both to be discussed in future papers.

Item Type:	Article
Date Type:	Publication
Status:	Published
Schools:	English, Communication and Philosophy
Publisher:	IATIS
Date of First Compliant Deposit:	4 August 2017
Date of Acceptance:	19 September 2016
Last Modified:	15 Apr 2024 03:35
URI:	https://orca.cardiff.ac.uk/id/eprint/103310

Actions (repository staff only)

Edit Item