Process Mining - Classification

Classification

There are three classes of process mining techniques. This classification is based on whether there is a prior model and, if so, how it is used.

• Discovery: There is no a priori model, i.e., based on an event log some model is constructed a process model can be discovered based on low-level events. For example, using the alpha algorithm, which is a didactically driven approach, where the authors state the lack of analytic capability for large event data volumes with such simple method. There exist many techniques to automatically construct process models (e.g., in terms of a Petri net) based some event log. Recently, process mining research also started to target the other perspectives (e.g., data, resources, time, etc.). For example, the technique described in (Aalst, Reijers, & Song, 2005) can be used to construct a social network.

• Conformance analysis: There is an a priori model. This model is compared with the event log and discrepancies between the log and the model are analyzed. For example, there may be a process model indicating that purchase orders of more than 1 million euro require two checks. Another example is the checking of the so-called “four-eyes” principle. Conformance checking may be used to detect deviations to enrich the model. An example is the extension of a process model with performance data, i.e., some a priori process model is used to project the bottlenecks on. Another example is the decision miner described in (Rozinat & Aalst, 2006b) which takes an a priori process model and analyzes every choice in the process model. For each choice the event log is consulted to see which information is typically available the moment the choice is made. Then classical data mining techniques are used to see which data elements influence the choice. As a result, a decision tree is generated for each choice in the process.

• Extension: There is a prior model also. This model is extended with a new aspect or perspective, i.e., the goal is not to check conformance but to enrich the model. An example is the extension of a process model with performance data, i.e., some prior process model dynamically annotated with performance data (e.g., bottlenecks are shown by coloring parts of the process model).

See the book Process Mining: Discovery, Conformance and Enhancement of Business Processes by Wil van der Aalst for details.