Tools that can be adapted to conceptual graphs

[John F. Sowa]

Many AI tools for language processing and reasoning can be adapted
to CGs by relatively straightforward methods.  Two large and mature
systems that can be adapted for processing CGs are the OpenCyc ontology
and the link grammar and parser used for grammar checking in the word
processor called AbiWord.

OpenCyc is a freely available version of the Cyc ontology and software,
which has been under active development since 1984.  The latest
version 2.0, which was recently released, has the following components:

  * The entire Cyc ontology containing hundreds of thousands of terms,
along with millions of assertions relating the terms to each other,
forming an ontology whose domain is all of human consensus reality.

  * NEW! 100,000+ "broaderTerm" assertions, in addition to the previous
generalization (subclass) and instance (member) assertions, to capture
additional relations among concepts.

  * English strings (a canonical one and alternatives) corresponding to
each concept term, to assist with search and display.

  * The Cyc Inference Engine and the Cyc Knowledge Base Browser are now
Java-based for improved performance and increased platform portability.

  * Documentation and self-paced learning materials to help users
achieve a basic- to intermediate-level understanding of the issues of
knowledge representation and application development using Cyc.

  * A specification of CycL, the language in which Cyc (and hence
OpenCyc) is written.

  * A specification of the Cyc API for application development.

  * Links between Cyc concepts and WordNet synsets.

  * NEW Links between Cyc concepts (including predicates) and the FOAF
ontology.

  * NEW! Links between Cyc concepts and Wikipedia articles.

For more information, see http://opencyc.org/

OpenCyc is available from SourceForge under the Apache license (which
permits free use, modification, and distribution for any purpose,
including commercial software).

Subsets of the OpenCyc ontology are available in OWL, but the more
expressive CycL language used for the full ontology can be translated
to and from CGIF.  That means that the OpenCyc ontology with its
inference engines can be used in conjunction with CG-based systems.

For natural language processing, parsers that generate dependency
graphs are convenient because dependency graphs can be translated
to conceptual graphs in almost a one-to-one mapping.

The dependency grammars and parsers developed at Carnegie-Mellon
university have been donated to the group that is developing the
grammar checker for the AbiWord word processor -- an open source
word processor.  That means that there is an established community
that is responsible for maintaining, developing, and extending
the grammar and parser.  They are also making it available
under the BSD license, which also supports free use for research,
education, and commercial applications.

See  http://www.abisource.com/projects/link-grammar/

The link grammar for English is the best developed and most widely
used, but link grammars compatible with the AbiWord grammar checker
have been developed for French, Russian, Persian, and Arabic.

A combination of a link grammar that generates CGs and a mapping
of the OpenCyc ontology and inference engines to and from CGIF
could make valuable additions to the CG toolkit.

John Sowa


---------------------------------------------------------------------
To unsubscribe, e-mail: cg-unsubscribe@...
For additional commands, e-mail: cg-help@...