Premium
A model‐driven approach for semantic web service modeling using web service modeling languages
Author(s) -
Mohseni Mohsen,
Sohrabi Mohammad Karim,
Dorrigiv Morteza
Publication year - 2021
Publication title -
journal of software: evolution and process
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.371
H-Index - 29
eISSN - 2047-7481
pISSN - 2047-7473
DOI - 10.1002/smr.2364
Subject(s) - computer science , unified modeling language , class diagram , applications of uml , metamodeling , object constraint language , modeling language , web service , uml tool , activity diagram , composability , software engineering , programming language , data mining , software
Abstract The service‐oriented software engineering approach has been increasingly utilized to design and develop complex distributed systems. Exploiting semantic web technologies to increase utilization of the web services at the semantic level leads to create semantic web services (SWSs). This paper proposes a model‐driven architecture (MDA) to model the SWS and to transform it from a high‐level modeling language, such as the Unified Modeling Language (UML), to a low‐level semantic description, such as the Web Service Modeling Language (WSML). To annotate all aspects of the SWS, a UML profile is provided by extending the UML metamodel. Some stereotypes and tagged values are also defined to support WSML. Therefore, the structure aspects of the SWS are modeled with a class diagram based on stereotypes that are defined in the UML profile. The logical and behavioral aspects of the SWS are modeled and expressed using activity diagrams, sequence diagrams, and Object Constraint Language (OCL). Experimental results show that the proposed approach not only increases the level of independence, accuracy, expressiveness, understandability, and machine processability but also significantly reduces the complexity and heterogeneity. Moreover, the proposed method is also evaluated in comparison with the previous methods of the literature, and the results will show that the proposed method outperforms the other methods in terms of expressiveness, understandability, scalability, level of independence, and accuracy.