nschneid / pysupersensetagger

AMALGrAM, an English supersense tagger written in Python

nschneid/pysupersensetagger Issues

mweval.ph: gold_pred_Ii statistic doesn't count strength-free tags I and i
Updated 9 years ago
POS-sensitive lemmatization
Updated 9 years ago1
Script to list MWEs in data
Closed 9 years ago9
tags2sst breaking
Closed 9 years ago4
nltk version dependencies need to be corrected in README.md
Closed 9 years ago
said2json.py script does not work with nltk 3.0.1
Updated 9 years ago1
Make the module callable from Python user code on a per-sentence basis
Closed 10 years ago
For each round of learning or decoding, display time and number of tokens processed
Closed 10 years ago
Lemmatization of 'times' and 'years'
Updated 10 years ago
Add proper support for *feature* cutoff
Updated 10 years ago
Store weights in a sparse data structure to avoid iterating over 0's in decoding
Closed 10 years ago2
In decoding, consider iterating over percepts (outer loop), scoring all labels at once
Closed 10 years ago
Don't create a feature for /dev/null passed as a lexicon (--lex)
Updated 10 years ago
Warn about edit conflicts
Closed 10 years ago2
sentence IDs
Closed 10 years ago
--predict
Closed 10 years ago4