Evaluation of a method for partial ontology alignment in multi-agent system
by Ademir Roberto Freddo, Cesar Augusto Tacla
International Journal of Reasoning-based Intelligent Systems (IJRIS), Vol. 1, No. 3/4, 2009

Abstract: In this paper, we present a method to partially align ontologies in dialogues of agents. The method is a hybrid approach based on terminological, structural and extensional techniques. Differently from other methods, the proposed approach tries to find similarities between two concepts examining only their properties (or roles in description logic), producing property signature vectors for each concept, examining role/property restrictions for each pair of concepts and computing the distances between different vectors in a high dimensional space. This approach allows for identifying similarity between concepts where other methods cannot find similarities, for instance, between disjoint concepts. The method has two steps, property similarity computation and concepts similarity computation, and can also be used as a pre-processing phase to semantic methods that need initial equivalence relations between concepts in order to work. We describe the method and show how it works with a case study. We perform an evaluation of our method using some real ontologies of the bibliography domain from the Ontology Alignment Evaluation Initiative (OAEI).

Online publication date: Thu, 27-Aug-2009

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Reasoning-based Intelligent Systems (IJRIS):
Login with your Inderscience username and password:

    Username:        Password:         

Forgotten your password?


Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email subs@inderscience.com