Preference-based AI Planning for Web Service Composition

A web service is a technology that allows data to be exchanged between applications using open protocols and standards. Web services generally offer basic functions that solve specific problems. The composition of services occurs when more complex problems need to be solved using successive invocations to other services. The automatic composition of web services is a current problem. Due to the continuous creation and updating of services, the manual composition of new services becomes impractical. Planning algorithms are an alternative to automate the process of web services composition. In particular, the planning algorithms with preferences allow users to guide the algorithm to find a solution in line with their requirements (e.g. prioritizing the use of services according to properties such as availability, cost, response time, etc.). This paper introduces a Planning approach with preferences to compose and run complex services from a set of web services given as input. To evaluate the approach, different preferences about the input services were used to obtain composition of services. As a result, we measured the feasibility of the proposed approach and the impact of the use of preferences on the composition result.