Fundamental Data Mining Techniques for Declarative Process Mining

Beschreibung

Process mining is a business process management technique that uses business process execution data for analysis. By providing the data in so-called event logs, process mining tools generate process models that describe the executions as precisely as possible. This can result in either graph-based notations such as Petri nets or BPMN or declarative notations such as Declare. One hypothesis of the thesis is that declarative constraints can increase the comprehensibility of models when process mining produces large, confusing "spaghetti diagrams". This paper presents an approach, including a prototype implementation, to apply association and sequence pattern analysis to event logs with the aim of producing declarative process models. Preprocessing steps and the translation of rules and patterns to declare constraints are explicitly addressed. This provides analysts with transparent insights into the basis of the entire declarative model.

Autor*in

Nico Grohmann

Themen in »Fundamental Data Mining Techniques for Declarative Process Mining«

Apriori BPM BPMN Data FP-Growth Flow GSP Horus Information Integration Petri Transformation Translation abstraction activation

Stimmen zu »Fundamental Data Mining Techniques for Declarative Process Mining«

Details

Link teilen