Toward an efficient web service composition based on an improved BTLBO algorithm

The composition of Web services represents a critical challenge in the development of business process. It consists of creating new functionalities by selecting individual services from a set of abstract activities and merging them into massive services. With the increasing number of available similar functionalities, the quality-of-service parameters, provided by Web services, have become criteria for selecting a composite web service plan that provides best possible service quality values among other possible compositions. As, the problem of Web service selection is a time-consuming optimization problem, we present, in this work, an enhanced version of the Balanced Teaching-Learning-Based optimization algorithm, called Discrete Balanced Teaching-Learning-Based optimization algorithm 'DBTLBO', which supports the discrete nature of the web service selection problem, by selecting a specific composition of web services from a set of candidate services, while meeting user needs and QoS preferences. several experiments were performed in a simulated environment. The obtained results were compared with those of TLBO, GA and PSO algorithms under the same conditions and clearly demonstrated the effectiveness of the proposed algorithm compared to the other algorithms mentioned above.