Semantics Based Construction And Intelligent Management Technology For Business Process Model

Business Process Management(BPM)plays an important role in modern enterprise management activities,which has gradually become a necessary means to enhance the core competence of enterprises.Many enterprises have accumulated lots of process models during the realization of business process automation,which leads to the formation of process repositories.How to mining the value of process repository to further enhance the level of business process management and promote the intellectualization of process model management,became a new subject for process management in the era of knowledge economy,which also caused a new wave of research in both industry and academic.However,there exists some problems that lead to enormous challenge.Due to the variant of business process modeling technology,the business models built by different technology are heterogeneous,and most of models are deficient in knowledge expression completeness.In the current research field of business process model management,most studies focus on certain point rather than provide an overall solution for the realization of intelligent management for business process model.In light of above situation,aiming at the implementation of intelligent management for process model,this paper studies on the overall framework logic with the global consideration of key related technologies,which includes the research details of business process modeling,process similarity comparison and process retrieval.The main contents of this paper are summarized as follows:(1)By comprehensive analysis of the state of art on the process management,the framework for intelligent management of process model is presented based on the elaboration of some key issues,which provides an overview of key technologies for the realization of intelligent management for process model.At the same time,the direction and specific research content of key technologies are clarified.(2)A multi-perspective semantic process modeling technology is presented to make up the deficiency of current process modeling technology for building complete process model with information of two perspectives,which involves both business semantics and control flow structure knowledge.By introducing the cognitive hierarchy to process modeling,fine grain information extraction is fulfilled.In order to prove the correctness of process model,twofactor verification mechanism for both business semantics and process structure is established.(3)A process similarity computation metric is proposed based on the multi-perspective semantic process model.After finishing the formal definition of the multi-perspective process model,the process similarity is calculated by means of feature extraction on different perspectives,where analytic hierarchy process is employed for weight allocation.The principle of calculation is illustrated by an example analysis,and experiments are carried out to compare the performance with current used methods,such as graph edit distance and behavior profile process similarity metric,which proves the validness,adaptiveness and time efficiency of the new proposed similarity metric.(4)In light of the accuracy and response speed in application of knowledge query from process repository,a new retrieval schema with capabilities of user perspective perception,process repository cluster and case-based reasoning is put forward.Hierarchical agglomerative clustering is selected as target clustering method by comparing the adaptiveness with k-means on process clustering.The combination of two level indexes on clustering knowledge and CBR ensures the low time cost and high retrieval quality of the retrieval schema.Also,the retrieval process and the final similarity result are optimized by adjusting the weights on different perspectives by use of user perspective perception.The investigation on time efficiency,accuracy and recall proves the superiority of the retrieval schema.(5)Based on the research results of related key technologies,a prototype system for intelligent management of business process model is designed and implemented.An overall system technical architecture is introduced by the analysis of core functional requirement,and the key technology dedicated to each function is full described with the function roadmap.The main interaction scenarios are studied and the corresponding implemention results are displayed.As a practical proof of above research results,the implemention of the prototype system provides a reference from theory to practice for further study in intelligent management of business process model.