Development of the representation of cause-effect dependencies for the knowledge base of the business process management system

Abstract

The problem of constructing knowledge representation in the process control system for knowledge-capacious business processes in the aspect of reflecting the cause-effect relationships between the context of performance of actions and the actions of the business process is studied. It is shown that the general approach to solving this problem is related to the isolation of causal dependencies on the basis of an analysis of the event log containing records of the behavior of the business process. When solving the problem, problems of clarifying the structure of the representation of the cause-effect aspect of knowledge are formulated in accordance with the features of the log of the information system of process control; the construction of logical facts and rules in accordance with the structure of log events; formalization of knowledge representation taking into account facts, rules and a priori restrictions. It is established that the connection between the artifacts of the context and the events of the business process log is carried out through common attributes. Between artifacts and attributes and between events and attributes, there is one-to-many relationship. The logical component of the knowledge base is structured in the form of logical facts and rules. Logical facts determine the state of the business process at discrete points in time based on the values of the properties of artifacts. The output rules determine the change in the state of the business process. A knowledge representation is proposed that takes into account the state of the context for the execution of business process actions in the form of weighted logical facts, as well as output rules that support the selection of actions taking into account the current state of the context. The difference between the proposed representation is that when defining facts, the attributes of events are taken into account, and the rules are the structure and sequence of events of the business process log. A priori knowledge of the subject area in the form of constraints is also tak en into account. The practical importance of the knowledge representation is the ability to automatically identify the cause-effect relationships between the actions of the business process in accordance with the information in its log.