Optimization of web services composition using artificial bee colony algorithm

In the open network environment, Web service has a strong dynamic nature and the optimal service combination scheme that produced in design stage may become invalid. Therefore, the single optimal service combination scheme is difficult to meet the individual needs of users, which will reduce the utilization of resource and the satisfaction of users. To solve this problem, this paper improves nectar selection strategy of the artificial bee colony algorithm. In addition, the paper designs a new neighborhood search formula and scout bee operation strategy, which effectively prevents the artificial bee colony algorithm from converging prematurely. After that, combined with Pareto strategy, it improves a Web services combination optimization method that is based on Pareto multi-objective artificial bee colony algorithm. This method will recommend a group of Pareto optimal solutions to users instead of recommending a single optimal solution to users. In this way, it can deal with the instability of combinational services in the dynamic environment and the different needs of users. Finally, the paper uses the relevant experiments to verify the feasibility and effectiveness of service combination optimization method improved in this paper.