TAUS - Enabling better automated translation

Friday
Jul 30th
Text size
  • Increase font size
  • Default font size
  • Decrease font size

Systran and Statistics

E-mail Print

Andrew JoscelyneSYSTRAN is the world’s longest serving commercial supplier of machine translation systems, with roots going back to the earliest days of MT research. The company recently announced a major new release (SYSTRAN Enterprise Server 7) that integrates statistical techniques into its rule-based core. This technology briefing looks at these latest innovations and how SYSTRAN intends to leverage them competitively.

TAUS members, download the report.

SYSTRAN recognizes that the growing availability of free online translation services and the deployment of services based on statistical machine translation has clearly increased competition and boosted market dynamics for MT. For the desktop segment, the challenge is to provide more value than simply “raw translation”. And in the corporate market, “free online translation services” do not really impact the need for customized translation, yet demand is growing, partly as a result of the higher visibility of MT. As a result, the enterprise segment is growing at a sustained pace.

It has therefore refocused much of its technology offering on enterprise requirements, adding value through a ‘hybrid’ technology that mixes the output fluency of statistical methods with the adequacy and predictability of rule-based processing.

This new offering has been in the pipeline for some time, as it has been difficult for the company to both meet regular deadlines for new product releases, while at the same time carrying out fundamental development on the hybrid model. So SYSTRAN 7 reached the market in June 2009, with versions 7.1 scheduled for February and 7.2 for March 2010. The new release 8 is currently in R&D.

In Release 7 it has adopted a corpus-based approach to solving the dictionary problem by introducing term harvesting to populate its dictionaries, and also the disambiguation problem by finding the appropriate incidence of a potentially ambiguous expression or structure from the data by harnessing the statistics of usage.

Readers of this TAUS Technology Briefing will gain unique insight into how one of the most durable and technologically savvy company’s in the sector is building on top of its existing rule based technology with statistical techniques.

TABLE OF CONTENTS

  1. THE COMPANY TODAY
  2. THE HYBRID PHILOSOPHY
  3. SYSTRAN 7 – HYBRID MT IN ACTION
  4. THE ROLE OF CORPORA
  5. TRAINING AND CUSTOMIZATION PROCESSES
  6. MEASURING QUALITY
  7. POSTEDITING
  8. LANGUAGE PAIRS AND DICTIONARIES
  9. PRICING AND ROI
  10. TECHNOLOGY ROADMAP

Full report exclusively available to TAUS members >> TAUS TECHNOLOGY BRIEFING: SYSTRAN AND STATISTICS

Become a TAUS Member

 

JOIN OUR MAILING LIST

Reports

 

Postediting in Practice

 

Implement Open Source MT

 

Increase Your Leveraging

Members

 

Pangeanic, Manuel Herranz

A mechanical engineer at a quality assurance depar...

 

McAfee, Paul Walsh

Paul Walsh is the Vice President for Localization ...

 

SDL, Jeremy Harpham

Jeremy Harpham is Senior Product Marketing Manager...