The Research On Business Process Construction Method Based On Knowledge Service

Along with the science and technology innovation, the external market environment is unpredictable, and the competition between enterprises is fierce too. In order to survive, develop and grow, achieve the goal of customer-service centered and oriented, enterprise’s internal business processes have to be able to quickly build and integrate, and have the capacity of continuous optimize. Knowledge is difficult to be shared between different systems because each sector of enterprises has variety of heterogeneous systems. With the deepening of the knowledge economy, knowledge has been regarded as the primary factor in the development of the productive forces. However, as the key factor of production, the sharing and utilization of knowledge is low. So that knowledge management divorces from enterprise reality. In this paper, a new method is proposed. The knowledge service and business process management are combined, and the process-related knowledge is packaged into componentized knowledge service with web service. The novel method is advantageous to the cross-platform business collaboration and building business processes quickly and dynamically.In this paper, a novel knowledge service model by clue of business activities is proposed. The model describes enterprise knowledge resources uniformly, and packages knowledge resources associated with a specific business activity to services, and publishes to the Web for users as a medium of Web Service. The knowledge service 5-tuple is defined, and a knowledge service matching and composition algorithm based on semantic similarity is described in this paper. In the matching algorithm, QoS similarity and service function similarity are considered to sort ad services in accordance with the similarity in descending order. Then the highest similarity service is chosen as the best service. This method can improve the recall and precision. In service composition algorithm, several candidate services with high service matched-degree in the service matching phase are selected considering the perspective of global process optimization. Then specific services for each child process from these candidate services are chosen to get the best overall process. In the process, the matched-degree each child process and candidate services, the matched-degree between candidate services which correspond with adjacent sub-processes are used to turn service composition problem into a shortest (longest) path problem.