Gap Analysis of Business Processes using Behavioral, Structural, and Semantic Similarity Calculations

Abstract

Business process is a collection of all activities and tasks in achieving the goals of a company or an organization. This study aims to find the gap value obtained from the business process model matching technique. Process model matching can be used to identify similar processes or activities in a business process model. There are three things in finding similarities between each process model, including the similarity of label, structural similarity, and behavioral similarity. Probabilistic Latent Semantic Analysis is used to get the probability of topics. The calculation result of the probability of the document in the topic with the similarity value obtained is 0.063. Dice Coefficient is used to calculate the similarity of the structure of the business process model compared to showing the similarity results with a precision value of 0.695. Jaccard Coefficient is used to calculate the similarity of behavior of a business process model compared to showing the results of similarities with a precision value of 0.342. By looking at the precision value of the three mechanisms for measuring similarities, it can be said that business processes that are running are actually not in accordance with the business processes in the SOP.