[Python-announce] NLTK 3.6.4 released

[Ilia Kurenkov]

We are making our debut on this mailing list with a release that fixes
a security
vulnerability <https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2021-3828>.


This release is intended to be a drop-in replacement. It contains the
following changes:

* deprecate nltk.usage(obj) in favor of help(obj)
* resolve ReDoS vulnerability in Corpus Reader
* solidify performance tests
* improve phone number recognition in tweet tokenizer
* refactor CISTEM stemmer for German
* identify NLTK Team as the author
* replace travis badge with github actions badge
* add SECURITY.md

Thanks to the following contributors to 3.6.4:
Tom Aarsen, Mohaned Mashaly, Dimitri Papadopoulos Orfanos, purificant,
Danny Sepler

What is NLTK?
The acronym stands for "Natural Language ToolKit". NLTK is educational
software for those interested in natural language processing, though some
have also incorporated it into commercial applications (and its License
supports this). A key use case is processing standard datasets using the
library's corpus readers, to support ML experiments in other packages.

Find out more on http://www.nltk.org/

--
Ilia