Article Details
Global Watchtower
Common Sense Advisory Blogs
TAUS Releases First Version of Translation Memory Data Exchange
Posted by Donald A. DePalma on June 5, 2009  in the following blogs: Translation and Localization, Technology, Business Globalization, Best Practices
Pages | |

Calling it the "super cloud for the global translation industry," the Translation Automation User Society (TAUS) announced that it released its language data exchange for public access. This week's announcement caps an industry-wide effort that began two years ago and involves many well-known information technology suppliers. Earlier this week, we participated in a webinar and product demonstration.

The first release of the TAUS Data Association (TDA) portal offers paying members the opportunity to share their translation memory (TM) data, subject to a reciprocity agreement that limits what members can take for free to what they contribute to the repository. TAUS founder and director Jaap van der Meer characterizes TM data exchange as the heart of the TDA initiative, allowing members to improve translation memory performance, gain quicker access to new locales, and explore new business opportunities. Members can also store their own TMs at the TDA site. The project also allows anyone, paying member or not, to search the exchange database for donated translations, a potentially valuable function in validating specialized terminology.

While the TDA project plan reaches for the stars, this first version, like any initial release, is more about promise and proof of concept than a fully realized vision. We say this for several reasons:
  • A good typology seeks more content. The exchange currently has about a half billion words in 70 languages (the heaviest concentrations are in German, French, Chinese, and Japanese), contributed over the last two months by charter members. The site boasts a categorization into 16 domains. However, the lion's share of translation segments resides in the IT domain common to most charter members such as Adobe, Cisco, and Intel. Non-IT members right now are limited to Molina Healthcare and the Welsh Language Board. One of van der Meer's immediate goals is to pump up membership and data contributions from outside the information technology sector, citing efforts to bring in pharma companies and private banks to develop ontologies for their respective sectors.

  • Basic tool support will be supplemented by a future API. The initial version delivers TMX upload/download for TM sharing, account history, and search. Technology Director Yan Yu said that a future application programming interface would support integration with CAT tools and TMS products. We envision a smoother user interface with wizards for common tasks. Yu added that TAUS is talking with tool vendors about integrating their products into the TDA backbone.

  • Findings on quality and leverage will follow. Members can score the quality of the data they download using Web 2.0 collaborative ranking, but Yu showed a long to-do list of more objective criteria that would limit the possibility of a bad TM polluting the exchange. We're very interested in seeing what members learn when they begin incorporating TDA data into their own TMs, what benefits machine translation developers and users will derive from this bigger pool of data, and whether this pioneering effort has enough payback for its members.

  • Nascent partnerships need to be nurtured. Terminology consistency stands out as a potential win for the TDA, but two other communal efforts to collect terms are underway. Van der Meer said that TAUS has had several meetings with LISA to discuss their respective strategies, but hasn't spoken with the EuroTerm Bank.
Director van der Meer said that the TAUS vision for this effort is to have the TDA take a "leading position as a translation industry operating system." This first version shows that it has the bones to be a data sharing exchange, but that the initiative is still a work in progress. Growth in membership from different market sectors, more content in each of the domains, and demonstrable ROI for the data exchange are three metrics by which the market will judge its usefulness.


Post a Comment

Email address :(Your Email Address Will Not Be Displayed)

Your Comments
Enter Code given below :    

Link To This Page

Bookmark this page using the following link:http://www.commonsenseadvisory.com/Default.aspx?Contenttype=ArticleDetAD&tabID=63&Aid=586&moduleId=391

Do you have a website? You can place a link to this page by copying and pasting the code below.
Keywords: Translation, Translation management systems, Translation memory, Translation technologies

Refine Your Search
Skip Navigation Links.
Skip Navigation Links.

Terms of Use | Privacy Statement | Contact Us
Copyright © 2016 Common Sense Advisory, Inc. All Rights Reserved.