| Category: ||E-Material |
| Subject: ||Open source language analysis package |
| From: ||Lluis Padro |
| Email: ||padro_(on)_lsi.upc.edu |
| Date received: ||17 Mar 2006 |
--- We apologize if you have received multiple copies of this
Dear list members,
We are pleased to announce the release of FreeLing version 1.3,
which improves existing functionalities of the suite, and
includes new ones, such as WN-based semantic annotation, NE
classification, and dependency parsing. Also, we are glad to
announce that this version includes two new languages (Italian
and Galician) thanks to the researchers willing to share their
data under open-source or creative-commons licences (see
"thanks" section in FreeLing web page). FreeLing is an
open-source C++ library providing language analysis services. It
is Free Software, released under Gnu LGPL. FreeLing 1.3 is being
presented and demonstrated next May at LREC-2006 in Genoa,
FreeLing is developed at TALP Research Center
<http://www.talp.upc.es>, in Universitat Politècnica de Catalunya
<http://www.upc.es>. Morphological dictionaries and grammars were
inityally developed by Centre de Llenguatge i Computació
<http://clic.fil.ub.es>, in Universitat de Barcelona
Find more information, an online demo, and download links at
http://www.lsi.upc.edu/%7Enlp (under "resources" menu)
FreeLing is designed to be used as an external library from any
application requiring language analysis services. Nevertheless,
a simple main program is also provided as a basic interface to
the library, which enables the user to analyze text files from
the command line.
The named entity classification module requires some essential
Machine Learning services such as feature extraction, and
training/classification using Adaboost models. These services are
accessible to any program linking the library, so FreeLing can be
also used as a (very) basic ML-oriented NLP development toolkit.
Features already in previous versions:
* Text tokenization.
* Sentence splitting.
* Morphological analysis.
* Named entity detection.
* Date/number/currency/ratios recognition.
* PoS tagging.
* Chart-based shallow parsing.
New features in version 1.3
* New languages: Italian and Galician
* Improved and debugged linguistic data for Spanish and
* Contraction splitting
* Improved suffix treatment, retokenization of clitic
* Physical magnitudes detection (speed, weight, temperature,
* Named entity classification.
* WordNet based sense annotation
* Dependency parsing.
Elsnet-list mailing list Elsnet-list_(on)_elsnet.org