WMBT

02 Jan 2012

11:31 WMBT is able to tag unknown words

The previous version of WMBT could not recover tags for tokens where morph analyser failed. The current version (as i... Adam Radziszewski

29 Dec 2011

11:20 Zadanie #3561 (Zamknięty): Clean-up configs & wiki description

Leave only configs with affix features.
Configs for no guessing, for guessing (kipi+nkjp)
rm old configs: *-1, *-... Adam Radziszewski

21 Dec 2011

16:33 Zadanie #3353: Treatment of unknown words

-A works more-or-less ok (outputs loads of possible tags, but is correct), training in progress Adam Radziszewski

16:17 Zadanie #3353: Treatment of unknown words

Guessing seems done. New default config (unktagfreq=3) yields... Adam Radziszewski

14 Dec 2011

15:23 Zadanie #3353: Treatment of unknown words

First is more universal, makes sense to use affixes & regexes as features to get sort of morpho analysis.
Adam Radziszewski

15:21 Zadanie #3353 (Zamknięty): Treatment of unknown words

Two options here:
1. Find hapaxes, train a model for them, add completely guess new tags instead of igns.
2. Gather... Adam Radziszewski

15:19 Zadanie #3352 (Zamknięty): Extend with CRF++

Add secondary implementation for CRF++. Note that it needs feature description files and different feature vectors (C... Adam Radziszewski