QoS-Based Filters in Web Service Compositions: Utilizing Multi-Criteria Decision Analysis Methods

During the past decade, Web services (WS) have gained attention, as they provided the means for interoperability between loosely coupled distributed systems over the Web. Regardless of the platforms and technologies deployed in their development, by using WS, the aforementioned systems can seamlessly exchange messages and, thus, engage into business transactions over the global network. Nevertheless, these opportunities resulted in a wide plethora of available WS, making it hard for end‐users and composition engines to select the optimal service between two or more services with seemingly identical operational features. As a result, the WS composition engines often make decisions on the basis of the QoS criteria. In modern literature, many researchers have tried to confront the issue of QoS‐based composition, by applying multi‐criteria decision analysis (MCDA) methods often showing mixed results. This is due to the different characteristics of each MCDA method. In this work, we tried to identify the key criteria that should be taken into account when deciding the optimal MCDA method to use in a given scenario, on the basis of its unique business context, and in addition, we provide an empirical study in order to compare the advantages and disadvantages of the mostly used MCDA methods in the field of WS selection and composition.