02.04.2013 Views

CONTENTS

CONTENTS

CONTENTS

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

KNOWLEDGE ENGINEERING: PRINCIPLES AND TECHNIQUES<br />

Proceedings of the International Conference on Knowledge Engineering,<br />

Principles and Techniques, KEPT2009<br />

Cluj-Napoca (Romania), July 2–4, 2009, pp. 219–222<br />

A FORMAL CONCEPT ANALYSIS APPROACH TO ONTOLOGY<br />

SEARCH<br />

IOAN ALFRED LETIA (1) AND MIHAI COSTIN (2)<br />

Abstract. Finding the most suited ontology for a given domain or problem is<br />

not an easy task but through our work here we are trying to ease this endeavor<br />

by providing a semiautomated search mechanism. Semiautomated, because we<br />

need our search target to be defined by an expert. Once that target defined, by<br />

combining the processing power of MAS with FCA, we analyse the search space<br />

in order to find the most compatible ontology with the initial target.<br />

1. Motivation and Goal<br />

The main step for solving a certain problem has always been finding or creating<br />

the best tool for the job at hand, no matter what the domain was.<br />

In information science, our domain of interest, the tool in question is an ontology<br />

capable of mapping the problem domain with success. Finding the right ontology for<br />

a certain task has been even from the begining one of the theoretical strong-points<br />

of the semantic web but things are not as easy as it seems [8]. The right ontology is<br />

never easy to find and creating one is even harder.<br />

This, without any doubt, involves some sort of intelligent ontology search, a<br />

guided search that has an array of ontologies as search space. For this paper, we are<br />

using a predefined set of existing ontologies as search space. The process and the<br />

results themself can later on be applied to a larger scale, all being reduced afterwards<br />

to a scalability problem.<br />

We mentioned guided search because our process starts from a well defined domain.<br />

This domain must be described by some expert in the field using a set of<br />

concepts, concepts that will represent the search target. Our system uses this set<br />

of concepts in order to find the best suited ontology by expanding and matching it<br />

against the concepts found in the search space.<br />

2. Proposed Solution<br />

Having the initial domain defined by an expert in the field, as to direct our search,<br />

the search agents can then start trying to find the concepts that match the initial set.<br />

2000 Mathematics Subject Classification. 68T30, 94A15.<br />

Key words and phrases. Ontology, Formal Concept Analisys, Multiagent Systems, Search<br />

methods.<br />

219<br />

c○2009 Babe¸s-Bolyai University, Cluj-Napoca

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!