An association-driven approach to composite services discovery

Nowadays, how to find composite Web services that meet user´s needs with high efficiency and accuracy has been the key issue in the service composition area. To solve this problem, we adopt service association mechanism. First, services with the same function and heterogeneous quality are gathered into one group. Each group has some associations with others according to their service function and the past execution experiences. With these associations stored in a association depository, we design an approach to not only target related service group quickly, which results in high performance, but also guarantee the success ratio. Experimental results show the effectiveness of the approach.