Contract-based Diagnosis for Business Process Instances using Business Compliance Rules

Abstract

In order to increase the quality of business pro cesses when they are automated, the correctness of the activities can be checked by means of an analysis of the corresponding business compli ance rules. By analyzing the trace of an instance of a business process, it is possible to detect the correctness of the process and to determine which activity is faulty. Each activity or set of activities is related to a set of business compliance rules, which work as contracts that the activities must satisfy throughout the dataflow. In order to diagnose a business process instance, not all the activities participate in every single execution, since there are control flows that per mit the execution of several branches for a varied number of times. We propose to automate the di agnosis of these executions of a business process taking into account the involved activities and their business compliance rules. Our main contri butions are related to the construction of the cor responding framework using several techniques related to the constraint programming paradigm to obtain the incorrect activities. The two differ ent proposals consider the tradeoff between the obtaining of the minimal diagnosis and the per formance