Development of a method for the probabilistic inference of sequences of a business process activities to support the business process management

Models of temporal rules of execution of the business process actions were proposed for the use in absence in the process model of complete information on the reasons for execution of these actions caused by interference of the work executors. The rules are formed on the basis of analysis of the sequence of events in the business process log which makes it possible to determine temporal conditions and constraints on execution of the corresponding actions. The rule models can be applied as an element of knowledge representation for the process management system since they reflect experience of the business process execution recorded in the log. The use of rules allows one to limit the number of possible versions of execution of the business process taking into account its current state. As a result, the time of making decisions on the process management is reduced for the case of contradiction between the current version of the business process and the model. A new method of probabilistic inference was proposed that uses the presented rules to form new, admissible sequences of actions in an atypical situation that arose as a result of adjustment of the business process by its executors. The method applies knowledge representations based on the Markov logic network which makes it possible to arrange new sequences of actions according to the probability of their execution using weighed temporal rules. Use of a combination of rules for pairs of sequential and spaced in time actions ensures higher accuracy of calculating the probability of execution of new business process versions. The proposed method takes into account information from the event log when rules are supplemented. This enables continuous supplementing of rules in execution of the business process. The above enables practical real­time application of the method in automated formation and expansion of knowledge bases for the process management systems.

[1]  I. Shostak,et al.  A retrospective analysis technology of the Green Software Ecosystems development on the parametric identification of the Brown's model , 2018, 2018 IEEE 9th International Conference on Dependable Systems, Services and Technologies (DESSERT).

[2]  Pedro M. Domingos,et al.  Efficient Weight Learning for Markov Logic Networks , 2007, PKDD.

[3]  Matthew Richardson,et al.  Markov logic networks , 2006, Machine Learning.

[4]  Igor Shostak,et al.  Ontology based approach for green software ecosystem formalization , 2018, 2018 IEEE 9th International Conference on Dependable Systems, Services and Technologies (DESSERT).

[5]  Viktor Levykin,et al.  METHOD OF AUTOMATED CONSTRUCTION AND EXPANSION OF THE KNOWLEDGE BASE OF THE BUSINESS PROCESS MANAGEMENT SYSTEM , 2018 .

[6]  Mathias Weske,et al.  Business Process Model Abstraction , 2015, Handbook on Business Process Management.

[7]  Savas Konur,et al.  A survey on temporal logics for specifying and verifying real-time systems , 2013, Frontiers of Computer Science.

[8]  Igor Shostak,et al.  Information support for business processes at virtual enterprises with multi-agent technologies , 2018, 2018 IEEE 9th International Conference on Dependable Systems, Services and Technologies (DESSERT).

[9]  Yevgeniy V. Bodyanskiy,et al.  Implementation of search mechanism for implicit dependences in process mining , 2013, 2013 IEEE 7th International Conference on Intelligent Data Acquisition and Advanced Computing Systems (IDAACS).

[10]  Manfred Reichert,et al.  Data-Driven Modeling and Coordination of Large Process Structures , 2007, OTM Conferences.

[11]  Chalyi Sergii,et al.  Causality-based model checking in business process management tasks , 2018, 2018 IEEE 9th International Conference on Dependable Systems, Services and Technologies (DESSERT).

[12]  Norbert Gronau,et al.  A Proposal to Model Knowledge in Knowledge-Intensive Business Processes , 2016, BMSD 2016.