Applying Automated Language Translation at a Global Enterprise Level
Rychtyckyj, Nestor, Plesco, Craig, AI Magazine
We have been applying AI and machine-translation (MT) technology at Ford Motor Company since the late 1990s. Our initial goal was to utilize MT to translate vehicle build instructions from English to the native languages in the countries and regions where our assembly plants are located. The source text utilized a controlled language that we developed, called Standard Language, and we initially thought that applying MT technology would be a straightforward process. Controlled languages, such as Standard language, restrict the complexity and ambiguity of human languages by restricting syntax and terminology (Huijsen 1998). As such, they have been utilized in a number of different industrial applications (Godden 2000). However, there were many issues dealing with technical terminology, ungrammatical aspects of Standard Language, Ford-specific terminology, and the need to process uncontrolled text that needed to be addressed. We partnered with Systran Software Incorporated and with AppTek (now SAIC) to use their machine-translation technology and also incorporated natural language processing (NLP) algorithms within our artificial intelligence environment to analyze terminology and modify the source text to improve translation accuracy (Rychtyckyj 2007). The need to support manufacturing expansion in non-English speaking countries in Eastern Europe and Asia (such as in Russian and Chinese) led us to add additional language capability and to develop translation glossaries for all of the supported languages. The automated language translation for manufacturing work continues and will expand as Ford's global manufacturing footprint increases. However, the international growth within the company was not limited to manufacturing only and we found that there are many different groups within the company that need some type of machine-translation solution. Therefore, in 2010 we deployed an internal web-based machine-translation solution that sought to leverage our work in manufacturing and make automated translation a reality for the entire company. In the following sections we will describe the process of delivering a new technology across an entire company and lessons that we learned.
Machine translation has become ubiquitous in the last few years. Since the advent of the first MT systems in the 1960s the technology has been supported by a few specialized vendors and the cost to develop machine-translation systems was significant. This situation has changed dramatically in the last few years as the introduction of statistical machine translation has decreased the development time and subsequently large companies such as Google and Microsoft have become heavily involved in machine translation. The main result of this is that most users have had some experience with the technology and will likely have some type of preconceived bias (either positive or negative) when they are introduced to it as part of their daily work.
Unfortunately, many users still treat machine translation as a "black box" technology and expect to receive high-quality translations suitable for their specific purposes (conversational, business unit jargon, and so on) given any sort of input without having to do any other work. Other users have had bad experiences and do not believe that machine translation can work well in any instance. A large part of our work is to educate and manage these user expectations so that they can use the technology effectively. For example, a very common request that we have is to translate screen headings and labels into another language as part of a conversion process. These headings are usually one or two word phrases that often contain abbreviations and acronyms. This type of translation is difficult for machine (and human) translation because there is very little context available and these phrases may be ambiguous in many cases without a detailed knowledge of the application. In these cases, it is critical …
Questia, a part of Gale, Cengage Learning. www.questia.com
Publication information: Article title: Applying Automated Language Translation at a Global Enterprise Level. Contributors: Rychtyckyj, Nestor - Author, Plesco, Craig - Author. Magazine title: AI Magazine. Volume: 34. Issue: 1 Publication date: Spring 2013. Page number: 43+. © 2009 American Association for Artificial Intelligence. COPYRIGHT 2013 Gale Group.