Parameterized Algorithms for Matching and Ranking Web Services

摘要

The paper presents two parameterized and customizable algorithms for matching and ranking Web services. Given a user query and a set of available Web services, the matching algorithm performs a logic-based semantic matchmaking to select services that functionally match the query and maintains those which fully verify the constraints specified by the user. The ranking algorithm takes as input the matching Web services, assigns to each one a score in the range 0-1 and finally rank them based on the score values. The algorithms have been implemented, evaluated and compared to iSEM and SPARQLent. Results show that the algorithms behave globally well in comparison to these frameworks.