PM4ILP: An Approach for the Identification and Improvement of Unstructured and Loosely Specified Processes

Process improvement is a key concept in business process management (BPM). It ensures the quality of the model and promotes better process performance in terms of execution time, resource cost and process flexibility. In fact, this concept of flexibility stands for the ability of a process to adapt to its environment depending on the context in which it is executed. However, it is still difficult to predict how a process is performed: some parts of the process may not be detailed until the moment of its execution, or incomplete due to the lack of real-world data, events, data from process instances and resource availability. This paper focuses on proposing an approach for identifying these types of processes that could be classified as loosely specified or unstructured. It will also bring forward ideas for improvement using process mining techniques and tools which makes it possible to analyze the execution traces of a process and modify/complete the model according to the results deduced from the analysis. To validate the efficiency of the proposed approach, a case study was conducted within the health care domain from which we were able to apply the different phases of the approach on the COVID-19 process.