[Corpora-List] DKPro WSD 0.9.1 released
Tristan Miller
miller at ukp.informatik.tu-darmstadt.de
Mon Oct 14 14:19:05 UTC 2013
We are pleased to announce the first major release of DKPro WSD, version
0.9.1.
DKPro WSD is a modular, extensible Java framework for word sense
disambiguation. It provides UIMA components which encapsulate corpus
readers, linguistic annotators, lexical semantic resources,
disambiguation algorithms, and evaluation and reporting tools.
To obtain the software, please visit its websites on Google Code:
https://code.google.com/p/dkpro-wsd/
https://code.google.com/p/dkpro-wsd-gpl/
More about DKPro WSD
--------------------
DKPro WSD supports rapid design and evaluation of word sense
disambiguation algorithms and pipelines. With the framework you can:
* evaluate any number of algorithms simultaneously
* compute common evaluation metrics: coverage, precision, recall, F1
* produce confusion matrices to compare pairs of algorithms
* filter or break down results by part of speech, etc.
* show results with and without user-specific backoff algorithms
* rescore results with a user-specified sense clustering
* visualize disambiguation algorithms in action
* perform parameter sweeping to find optimal system configurations
* run related tasks such as entity linking and word sense induction
There is out-of-the-box support for many different data sets (Senseval,
SemEval, SemCor, AIDA CoNLL-YAGO, TAC KBP, MASC, etc.) and lexical
semantic resources (WordNet, UBY, Wiktionary, etc.).
DKPro WSD is free software, released under the Apache License and GNU
General Public License.
Changelog
---------
Changes since the preview release (0.9.0):
* upgraded to DKPro Core 1.5.0, uimaFIT 2.0.0, UBY 0.4.0, and TWSI 1.0.1
* added module for word sense induction
* moved Wikipedia-specific graph algorithms to a separate module
--
Tristan Miller, Doctoral Researcher
Ubiquitous Knowledge Processing Lab (UKP-TUDA)
Department of Computer Science, Technische Universität Darmstadt
Tel: +49 6151 16 6166 | Web: http://www.ukp.tu-darmstadt.de/
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 263 bytes
Desc: OpenPGP digital signature
URL: <http://listserv.linguistlist.org/pipermail/corpora/attachments/20131014/730af63c/attachment-0001.sig>
-------------- next part --------------
_______________________________________________
UNSUBSCRIBE from this page: http://mailman.uib.no/options/corpora
Corpora mailing list
Corpora at uib.no
http://mailman.uib.no/listinfo/corpora
More information about the Corpora
mailing list