;Process mining technology has been widely used in business process management. At present many studies have indicated the use of process mining deduced from the event logs show that if the business process and business process models expected are inconsistent, it may be a risk of internal mismanagement. In the process of deduced business processes, the event log has a different levels result in different of grains. As the result, in this study we try to resolve the differences business process deduced from the different grain in the same database.
On the other hand, business process derives from the event logs usually use the frequency of the activities interaction to determine the direction of the business processes. But this method ignores the few times and high-value business processes, the heuristic algorithms cannot identify the real business operations business process
This study presents a new method to replace to derive business processes use the frequency, against quantity and processes value to derive business processes.
In this study, we use the directly dependent coefficient of the adjusted heuristic algorithm to solve the problem of the difference consistency, and propose a new way to adjust directly dependent coefficients, which can effectively improve the accuracy and solving speed of the heuristic algorithm.
In this study, we use the stone industry for case validation, comparing two business processes deduced from different grain, give one of the business processes a directly dependent coefficient threshold value, then use of heuristic algorithms for solving the iv
corresponding direct dependent coefficient threshold value. This method can solve the inconsistent business processes from the different grain in the same database, and identify the corresponding consistent business processes.