MBT(1) Memory Based Tagger

SYNOPSYS

mbt [options]

DESCRIPTION

mbt is a memory-based tagger that can tag sequences, based on training files generated by mbtg.

OPTIONS

-h or --help

show help

-s settingsfile

use a settingsfile as generated by mbtg

Or:

-l <lexiconfile>

-r <ambitagfile>

-k <known words case base>

-u <unknown words case base>

-D <loglevel>

Possible options levels are LogNormal , LogDebug , LogHeavy and LogExtreme

-e <sentence delimiter> (default '<utt>')

-E <enriched tagged testfile>

-t <testfile>

-T <tagged testfile> (default is untagged stdin)

-o <outputfile> (default stdout)

-Otimbl options


 (Note: there is NO SPACE between O and the options)
  <options>   classifier options for both known and unknown words instance bases
  K: <options>   classifier options for known words instance base
  U: <options>   classifier options for unknown words instance base
  valid  timbl options are: a d k m q v w x -

-B <beamsize for search> (default = 1)

-v di


 add distance to output

-v db


 add distribution to output

-v cf


 add confidence to output

-V or --version

show version info.

-L <file with list of frequent words>

BUGS

possibly

AUTHORS

Ko van der Sloot [email protected]

Antal van den Bosch [email protected]