Abstract
The opportunity to reduce human translation costs has always been the most prominent value proposition for the continued investment in commercial machine translation. But what about the value to large, multilingual and multicultural enterprises? What value would reducing such costs have for the company with worldwide technical support in Germany that must support a new team in Brazil with real-time text chat? Similarly, what value would it have for the hotelier with hotels in China who needs to integrate language into their world-renowned loyalty program? This chapter reviews some of the history of machine translation in the enterprise, where it has succeeded and where it has failed, and what the future looks like for enterprise use of machine translation.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
Kelly, Nataly, DePalma, Donald A. &. Stewart, Robert G. (2012) “The Language Services Market: 2012” Common Sense Advisory.
- 2.
DePalma, Donald A. (2007) “Machine Translation Attracts Eyeballs, Not Software Revenue.” December 20, 2007 http://www.commonsenseadvisory.com
- 3.
A typical example of a “gist” translation from the 1990s would have looked like “The parties of opposition took the Treaty, which includes 95 dates with the first action of Peña Nieto by Mexico.” This example is, but it gives you enough, if you know the domain of the topic, to gather that one of the first actions of the new Peña Nieto Government in Mexico was to commit to 95 milestones of some sort. Certainly enough knowledge, if you follow Mexican politics, to know that you want to understand more.
- 4.
Moore’s law is the observation that the number of transistors on integrated circuits doubles approximately every 2 years.
- 5.
This was the case while the second author was involved in Globalink. That period extended through the sale of Globalink to Lernout and Hauspie in 1998.
- 6.
A morpheme is an atomic morphological unit of a language that cannot be further divided.
- 7.
Despite all the resources that have been invested in natural language technologies surprisingly little effort has been put into streamlining the way the linguists work. While their peers in other areas of computing have complex IDEs with predictive input and visual diagramming tools, linguists working on machine translation still often have to hand-code formalisms developed decades ago in plain-text editors. This is comparable to attempting to build a skyscraper with a shovel, a hammer, and a chisel. It can be done, but in the age of earth-moving machines this is not a sound approach.
- 8.
All the customer scenarios date from 2010 and later.
- 9.
Grimes, Seth. (2011) “Text-Analytics Demand Approaches $1 Billion” Information Week Software. May 12, 2011. http://www.informationweek.com/software/business-intelligence/text-analytics-demand-approaches-1-billi/229500096
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2013 Springer Science+Business Media New York
About this chapter
Cite this chapter
Garr, B., Berman, V. (2013). Machine Translation: The Enterprise Point of View. In: Neustein, A., Markowitz, J. (eds) Where Humans Meet Machines. Springer, New York, NY. https://doi.org/10.1007/978-1-4614-6934-6_3
Download citation
DOI: https://doi.org/10.1007/978-1-4614-6934-6_3
Published:
Publisher Name: Springer, New York, NY
Print ISBN: 978-1-4614-6933-9
Online ISBN: 978-1-4614-6934-6
eBook Packages: EngineeringEngineering (R0)