2008 | OriginalPaper | Chapter
PACTOLE: A Methodology and a System for Semi-automatically Enriching an Ontology from a Collection of Texts
Authors : Rokia Bendaoud, Yannick Toussaint, Amedeo Napoli
Published in: Conceptual Structures: Knowledge Visualization and Reasoning
Publisher: Springer Berlin Heidelberg
Activate our intelligent search to find suitable subject content or patents.
Select sections of text to find matching patents with Artificial Intelligence. powered by
Select sections of text to find additional relevant content using AI-assisted search. powered by
PACTOLE stands for “Property And Class characterization from Text for OntoLogy Enrichment” and is a semi-automatic methodology for enriching an initial ontology from a collection of texts in a given domain. PACTOLE is also the name of the associated system relying on Formal Concept Analysis (FCA). In this way, PACTOLE is able to derive a concept lattice from a formal context, consisting of a binary table describing a set of individuals with their properties. Given a domain ontology and a set of objects with their properties (extracted from a collection of texts), the PACTOLE system builds two concept lattices: the first corresponding to the restriction of the ontology schema to the considered objects and the second to the extracted pairs (object, property). As they are based on the same set of individuals, the two ontologies are merged using context apposition. The resulting final concept lattice is analyzed and a number of knowledge units can be extracted and furthermore used for enriching the initial ontology. Finally, the final concept lattice is mapped within the
${\cal FLE}$
KR formalism. The paper introduces and explains in details the PACTOLE methodology with the help of an example in the domain of astronomy.