Comparison of behavioral similarity use TARs and Naïve algorithm for calculating similarity in business process model

Every organization needs one or more business processes to support the analysis, redesign, and implementation of an activity. Problems that occur when one or more business processes have in common that should identify the effectiveness and efficiency of the models from different processes. A common process model is determined by calculating the similarity of some business process model using a specific algorithm. This observation aims to compare two algorithms to calculate the value of behavioral similarity using Transition Adjacency Relations (TARs) and Naïve. From that comparison, we will find a suitable algorithm to calculate the value of that behavioral similarity. In this experiment, the authors will analyze the similarity of process models based on their behavior by comparing them using TARs and Naïve. The result of the behavioral similarity calculation with TARs algorithm is 0.36. Whereas the similarity calculated using naïve algorithm is 0.3016. Looking at the calculation of the similarity between these two algorithms, the value of the naïve algorithm is lower. From that, we concluded that between TARs and Naïve, TARs is superior. There are many possible causes for a low similarity value in Naïve, one of them is because the two models of business process is different when viewed by behavioral standpoint. Furthermore, it is likely influenced by the structure of a more complex business process model.