Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
487153 | Procedia Computer Science | 2015 | 8 Pages |
One problem of process discovery is the duplicated task, a task that appears multiple times in one process model. This paper presents “Slice and Connect,” an automated tri-dimensional process discovery technique considering duplicated task. It adds the notion of pool as a group of tasks that belongs to a single entity (e.g. department, organization, country). The algorithm entails two steps: first, the event log is sliced according to the pool attribute, and then the pool dependency network (PDN) is discovered. Second, the inter-pool dependency is calculated to connect the PDNs, the result of which is a process model represented as an integrated dependency network (IDN). Duplicated task is only allowed within a different pool to clearly separate the activity context in IDN. The complexity of the process model thereby is increased; therefore, the comprehensibility of the process model is also increased.