Skip to content
andrely edited this page Dec 28, 2010 · 6 revisions

The disambiguator requires Ruby 1.8.7.

Run the disambiguator by piping output from the Oslo-Bergen tagger to the the disambiguator.

If you have installed the Gem you should be able to do:

<Oslo-Bergen tagger> | run_obt_stat.rb > <output file>

If you have the zip file, cd into the new directory and do:

<Oslo-Bergen tagger> | bin/run_obt_stat.rb > <output file>

The default output is a copy of the input but with only one reading remaining for each word. You may specify a reduced vertical table format (VRT) by using the option --output vrt.

The default encoding is utf-8. Note that you need to ensure that the encoding is correct or the disambiguator will fail ungracefully. Latin1/ISO-8859-1 encoding may be specified with the option --format latin1.

##Internal notes

Memory profiling

Clone this wiki locally