tca2

TCA2 is a program for sentence alignment, developed by Knut Hofland and Øystein Reigem at the UiB.

Overview

Workflow

• Workflow/Bargovuohki
• Files for alignment 2019:
  • nob2sma/admin/sd/samediggi.no
  • nob2sma/admin/sd/samediggi.no PRM-filer

Alignment

• Improving sentence alignment for pdf files
• Setting the TCA2 parameters
• How to use the graphical interface (plan)

Evaluations and notes

• fin2sme 2017
• nob2sma 2017
• 2012 FAD project
  • Goldstandard test results (empty file)
  • TMX goldstandard test with min and max values (empty file)
  • Abbr test results
  • Parallell gov. papers

Original Bergen documents

• TCA2 hompage

• [Knut Hofland and Stig Johansson: The Translation Corpus Aligner: A program for automatic alignment of parallel texts

  http://khnt.hd.uib.no/files/align3.pdf] (see also here)

• The october 2006 README file
• The original README file
• The background for TCA2
• Demo document with nice pictures

Alternatives to tca2

• We need to look at alternatives to tca2

Meetings

• 2017: 21.6.  // 4.7.

• 2012: 12.1.  // 19.1.  // 25.1.  // 1.2.  // 7.2.  // 13.2. // 17.2. // 29.2. // 12.3. // 22.3.
• 2011: 7.4.   // 25.11. // 28.11. // 8.12.  // 14.12. // 20.12.