Time based Discovery of parallel business processes

Process mining for discovering concurrent activities is important since there are many of them contained in business processes. The concurrency is formed by AND parallel or OR conditional. However, most of the existing process mining algorithm discover only concurrency formed by AND parallel. Substituting OR conditional with AND parallel does not always discover the real business processes. Also, the existing process mining algorithms use linear dependence principle; therefore, they require complete event logs which are difficult to be provided since there are many possible traces. In this regard, this paper proposes Time based Discovery algorithm which utilizes non-linear dependence principle. The proposed algorithm can effectively distinguish AND parallel and OR conditional. The experimental results show that the proposed algorithm can discover the concurrent business processes formed by AND parallel or OR conditional.