Optimizing the Composition of a Resource Service Chain With Interorganizational Collaboration

In collaborative tasks, the composition of distributed resource services can improve the resource utilization. From the perspective of a business process, resource services should be composed as service flow so as to serve better a business process. However, most of the existing methods neglect an important factor, the characteristic of interorganization, which should be taken into consideration. An interorganizational resource service sequence is not necessarily available to all participating organizations in a collaborative task, as each organization seeks and selects resource services independently. Therefore, resource services in sequential order are called the resource service chain (RSC). This problem is called resource service chain composition with inter-organizational collaboration (RSCCOrg). The collaborative manufacturing is taken as an instance because it is a typical collaborative task. The proposed approach here is composed of several algorithms, called the algorithms for RSCCOrg (ARSCCOrg) and can better cope with interorganizational collaboration. To begin, a model based on a weighted directed graph is presented for the resource service temporal dependences. This is convenient for describing the temporal dependences among resource services and obtaining frequent RSCs. By calculating the frequencies between every two resource services from business data, the degrees of temporal dependence between them are resolved. A set of frequent RSCs can be obtained by choosing the higher weights. Next, an algorithm is presented to extend the frequent RSCs to increase the reusability and to improve the efficiency of selection. Then, a similarity formula is presented to compare the similarities between the frequent RSCs and the extended RSCs. In particular, the characteristic of interorganization is considered in the similarity comparison. The ARSCCOrg has been tested with a practical dataset, and the results show that it is very promising.