Multi-Attribute Decision Web Service Selection based on Optimal Ant Colony Algorithm

This paper focuses on issues related to the selection of Web services in multi-attribute states. With the rapid development of the Internet, how to choose a quality service for users under the constraints of multi-attribute factors is a question worth thinking about. To select the optimum combination of services for the user among a large number of multi-attribute services is our goal. User requirements can be viewed as a workflow, and a complete requirements workflow contains multiple tasks. Each task consists of different services, and different tasks are used to satisfy different user requirements. How to select the optimal services from each task and combine them to form a complete service portfolio is the work we need to accomplish. In this work, we optimize the candidate service attributes and calculate the attribute weights using a weighted arithmetic mean operator, and then select the optimal service combination using an optimal ant colony algorithm. Compared with the original method in the service selection process, MDM based on OACO Algorithm can solve the multi-attribute service selection problem, while the service data after the weight calculation avoids the constraints of multiple attributes on the service selection results. The complexity of the algorithm traversal is greatly decreases by using the Cartesian product idea to group the services based on the same attribute values and then using the ant colony algorithm. For the experimental results, the multi-attribute optimization based on the optimized ant colony algorithm has significant benefits over the original algorithm in terms of time performance.