Research On Sequence Pattern Analysis For Collaborative Task Optimization

Supported by the technology of the Internet,Internet of Things(IoT),Cloud Computing,Big Data and other emerging information technologies,the mode in which more than one organizations collaborate to complete a task,called collaborative task mode,has been applied in more and more fields.In a collaborative task,it is important to improve the entire efficiency of a business process.Accordingly,resource services should be provided to the business process as a “service flow”,so that the business process and the service flow can match each other better.However,a collaborative task generally is a cross-organizational business process,as well as a resource service flow.Resource services are independently selected and used for each organization.Therefore,the contradiction between the autonomy and global optimization restricts the selection of a resource service flow.Mining resource service flow becomes one of an important issue for a collaborative task system.In collaborative workflow,temporal relationship between resource services is analyzed,and then mining the sequence pattern of resource services is developed.Simultaneously,the characteristic of cross-organization is considered in each resource service sequence(RSS,Resource Service Sequence).An approach of analyzing and mining RSSs is proposed.Our research is carried out as follows.(1)Modeling resource service sequence.To accurately describing RSSs,based on workflow model,the temporal relationship between resource services is separated from workflow model.A method of modeling RSS is presented.(2)Mining resource service sequence.To resolve the pattern of resource service,By analyzing the temporal relationship between resource services,the degrees of temporal relationships are resolved first.Simultaneously,for the consideration of massive business data,an iterative computation approach based on Moving Average is presented to merge all distributed mining results.Then,the proposed approach is tested with a simulation experiments.(3)Resolving cross-organizational resource service sequence based on similarity.Considering the characteristic of cross-organization in a RSS,and to improve the reuse of a RSS to a workflow,RSSs are extended.First,the isomorphic RSS of an original RSS is defined.The set of original RSS is obtained by a RSS model.Then,the similarity with the information of cross-organization is defined.The set of isomorphic RSS with high degrees of similarity are regarded as the extended RSS.Also,the proposed approach is tested with a simulation experiments.Finally,a traditional collaborative task system,the cloud manufacturing service platform for small and medium-sized enterprises,is introduced.On the platform,the proposed approaches are all applied or verified to be corrected and effective.