Parameterized semantic matchmaking and ranking framework for web service composition

Abstract

The matchmaking is a crucial operation in Web service composition. The objective of matchmaking is to discover and select the most appropriate (i.e., that responds better to the user request) Web service among the different available candidates. Several matchmaking frameworks are now available in the literature. However, most of these frameworks present at least one of the following shortcomings: (i) use of strict syntactic matching, which generally leads to low recall and low precision of the retrieved services; (ii) use of capability-based matchmaking, which is proven to be inadequate in practice; (iii) lack of customization support; and (iv) lack of accurate ranking of matching Web services, especially within semantic-based matching. In this research project, we propose several conceptual and algorithmic solutions to jointly deal with these shortcomings. More precisely, the solutions proposed in this research project permit to fully overcome the first, third and fourth shortcomings of semantic matchmaking frameworks. The second shortcoming is partially addressed in this research project. A prototype called PMRF (Parametrized Matching and Ranking Framework) supporting the proposed algorithms has been implemented. The performance analysis shows that the algorithms behave globally well in comparison to similar existing ones.

Date of Award	2015
Original language	English
Supervisor	Nadia Yacoubi-Ayadi (Supervisor), Afef Bahri (Supervisor) & Salem Chakhar (Supervisor)