Skip site navigation (1)Skip section navigation (2)

FreeBSD Manual Pages


home | help
WordNet::Similarity::GUserFContributed PerlWordNet::Similarity::GlossFinder(3)

       WordNet::Similarity::GlossFinder	- module to implement gloss finding
       methods for WordNet::Similarity measures	of semantic relatedness
       (specifically, lesk and vector)

	 use WordNet::QueryData;
	 my $wn	= WordNet::QueryData->new;
	 defined $wn or	die "Construction of WordNet::QueryData	failed";

	 use WordNet::Similarity::GlossFinder;
	 my $obj = WordNet::Similarity::GlossFinder->new ($wn);
	 my ($err, $errString) = $obj->getError	();
	 $err and die $errString;

	 my $wps1 = 'england#n#1';
	 my $wps2 = 'winston_churchill#n#1';

	 # get the glosses of these two	synsets, since we are not using	a
	 # configuation	file to	specify	relations, we will only	get the
	 # immediate glosses of	the two	wps entries. The default weight	and
	 # relation appear in $weight and $relation - these can	be modified
	 # via a configuration file.

	 my ($wps1gloss, $wps2gloss, $weight, $relation	) = $obj -> getSuperGlosses ($wps1, $wps2);
	 print "$wps1gloss->[0]\n";
	 print "$wps2gloss->[0]\n";
	 print "$weight->[0]\n";
	 print "$relation->[0]\n";

       This class is derived from (i.e., is a sub-class	of)
       WordNet::Similarity. Two	of the measures	of similarity, provided	in
       this package, viz. WordNet::Similarity::lesk and
       WordNet::Similarity::vector deal	with WordNet glosses. This module
       provides	methods	for easy access	to the required	glosses.

       This module inherits all	the methods of WordNet::Similarity.
       Additionally, the following methods are also defined.

       Public methods

	   Specifies the parts of speech that measures derived from this
	   module support (namely, nouns, verbs, adjectives and	adverbs).

	   parameters: none

	   returns: true

	   Overrides method of same name in WordNet::Similarity.  Prints
	   module-specific configuration options to the	trace string (if
	   tracing is on).  GlossFinder	supports module	specific options:
	   relation, stop and stem.

	   Parameters: none

	   Returns: nothing

	   Overrides the configure method in WordNet::Similarity. This method
	   loads various data files, such as the stop words and	relations.

	   Parameters: $file --	path of	the configuration file.

	   Returns: nothing

       $self->getSuperGlosses($wps1, $wps2)
	   This	method returns a list of large blocks of concatenated glosses
	   (super-gloss) for each specified synset. A super-gloss is the block
	   of text formed by concatenating the glosses of a synset with
	   glosses of synsets related to it in WordNet.	"Related" synsets are
	   identified by specific relations specified in the "relations" file.
	   If no relations file	was specified in the configuration, only the
	   gloss of that synset	is returned.

	   Parameters: wps1 and	wps2 --	two synsets.

	   Returns: List of superglosses for both synsets (2-D array).

       Private Methods

	   This	method loads relations from a relation file.

	   Parameters: none

	   Returns: nothing

       Many of the methods in this module can work with	either offsets or wps
       strings internally.  There are several interesting consequences of each

       1.  An offset is	not a unique identifier	for a synset, but neither is a
	   wps string.	An offset only indicates a byte	offset in one of the
	   WordNet data	files (data.noun, data.verb, etc. on Unix-like
	   systems).  An offset	along with a part of speech, however, does
	   uniquely identify a synset.

	   A word#pos#sense string, on the other hand, is the opposite
	   extreme.  A word#pos#sense string is	an identifier for a unique
	   word	sense.	A synset can have several word senses in it (i.e., a
	   synset is a set of word senses that are synonymous).	 The synset
	   {beer_mug#n#1, stein#n#1} has two word senses.  The wps strings
	   'beer_mug#n#1' and 'stein#n#1' can both be used to refer to the
	   synset.  For	simplicity, we usually just use	the first wps string
	   when	referring to the synset.  N.B.,	the wps	representation was
	   developed by	WordNet::QueryData.

       2.  Early versions of WordNet::Similarity::* used offsets internally
	   for finding paths, hypernym trees, subsumers, etc.  The module
	   WordNet::QueryData that is used by Similarity, however, accepts
	   only	wps strings as input to	its querySense method, which is	used
	   to find hypernyms.  We have found that it is	more efficient
	   (faster) to use wps strings internally.

	Ted Pedersen, University of Minnesota Duluth
	tpederse at

	Siddharth Patwardhan, University of Utah, Salt Lake City
	sidd at


       WordNet::Similarity(3) WordNet::Similarity::vector(3)

       Copyright (c) 2005, Ted Pedersen	and Siddharth Patwardhan

       This program is free software; you can redistribute it and/or modify it
       under the terms of the GNU General Public License as published by the
       Free Software Foundation; either	version	2 of the License, or (at your
       option) any later version.

       This program is distributed in the hope that it will be useful, but
       WITHOUT ANY WARRANTY; without even the implied warranty of
       General Public License for more details.

       You should have received	a copy of the GNU General Public License along
       with this program; if not, write	to

	   The Free Software Foundation, Inc.,
	   59 Temple Place - Suite 330,
	   Boston, MA  02111-1307, USA.

       Note: a copy of the GNU General Public License is available on the web
       at <>	and is included	in this
       distribution as GPL.txt.

perl v5.24.1			  2008-03-2WordNet::Similarity::GlossFinder(3)


Want to link to this manual page? Use this URL:

home | help