Conformance Checking to Evaluate Business Process Models using Modified Time-based Heuristics Miner Algorithm

Business processes are the collections of activities which produce some specific service for organizations. A lot of services are produced in the organization, making them difficult to document. Information systems were built to perform the business processes and produce event logs as the documented business processes. The event log will be analyzed using process discovery technique to mine the process model by implementing some well-known algorithms such as Alpha Miner in deterministic algorithms and Heuristics Miner in heuristic algorithms. All of the algorithms have their own excellence and limitations in analyzing and discovering the event log into process models. Modified Time-based Heuristics Miner is the latest improvement of Heuristics Miner algorithm which can be mined the process model and define sequential and parallel (AND, OR, and XOR) relation. Comparing to the original, Heuristics Miner algorithm, is only able to discover sequential and parallel (AND and XOR) relation. To know the differences between original Heuristics Miner and Modified Time-based Heuristics Miner including the process model and its relations, the evaluation using conformance checking consist of fitness, precision, simplicity, and generalization was done in this research. This paper will be focused on the results of evaluation in conformance checking from both of the algorithms. The results presented that the conformance checking of Modified Time-based Heuristics Miner was higher than that of the original Heuristics Miner algorithm in those four criteria.