Tranz is an experimental Machine Translation and Translation Memory tool.  It consists of the following modules:
  • Extraction: To extract text for translation from Microsoft Office, Open, HTML, and text files.  (Apple iWork support coming soon).
  • Segmentation: To split extracted blocks of text into sentence-level segments for translation.
  • Tokenization: To split segments into individual word-level tokens for further analysis.
  • Re-Mix Engine: A Unique feature of the Tranz engine to mix and match the input tokens into different fragments which are taken as candidates for translation.
  • Pattern Engine: Compliments the remix engine by providing template (example) based translation analysis.
  • Coverage Analyzer: Analysis of candidate fragments for translation coverage, gaps, and coverage quality.
Currently Supported Languages include English and Japanese.  Chinese support is coming soon.

Tranz is still experimental and thus not available to the general public.  If you would like more information, or are interested in evaluation the technology, please contact us.