pysupersensetagger

pysupersensetagger copied to clipboard

nschneid

POS-sensitive lemmatization

1

The WordNet lemmatizer has an option to use coarse POS, but not fine-grained POS, which is sometimes required to disambiguate the lemma. The [morph.py](https://github.com/nschneid/pysupersensetagger/blob/master/src/morph.py) interface to the WordNet lemmatizer recognizes...

nschneid

said2json.py script does not work with nltk 3.0.1

1

This script does not work with the 3.0.1 version, but does with the 2.0.4 version.

negrinho

As it turns out, there are some plural nouns, such as _times_ and _years_, for which WordNet has both singular and plural entries (e.g., "it was a sign of the...

nschneid

Currently, the `--cutoff` option specifies a _percept_ cutoff. But now, with the sparse representation of weights, it should be straightforward to support a _feature_ cutoff (maybe called `--fcutoff`).

nschneid

nschneid