ARC Discovery Project: “Improving Business Decision-Making via Liquid Process Model Collections”
This project will develop an innovative approach to create and update as necessary the large collection of business process models that represent a complex organisation, so that this collection captures the actual way in which the organisation performs its business processes. Deploying theoretical, conceptual and empirical research, this project will capitalize on the value hidden in large process data, as recorded in event logs. The approach will be implemented in an open-source technology to facilitate advanced investigations and predictions that can ultimately lead to better strategic decision-making. This technology also has the potential to become a research-enabling tool for the large research community in business process management.
Work Package I: Foundations of liquid process model collections
This package will establish the foundational notions behind liquid process model collections, such as that of overall alignment score between an event log and a process model collection, and operationalise these notions by means of software tools.
Work Package II: Transformation of static process model collections
This package will make existing, static process model collections liquid. This will be achieved by devising a novel algorithm for identifying and applying appropriate perturbations to an existing process model collection (e.g. adding or removing an activity or a decision point), in order to improve the overall alignment score, as defined in WP1.
Work Package III: Domain-driven discovery of liquid process model collections
This package, complementary to WP2, will develop a set of parametric algorithms for the semi-automatic discovery of liquid process model collections from scratch.
Work Package IV: Continuous realignment of liquid process model collections
This package will design and implement a set of algorithms for retaining model alignment with real-world processes, as the latter evolve over time.
Work Package V: Process analytics
This package will design an extensible analytics framework that exploits the richness of information available in liquid process model collections and linked logs, to enable a range of analyses based on historical processes (descriptive analytics) and forecasts of future processes (predictive analytics). These analyses will provide insights into different organisational aspects from performance issues to potential compliance violations or fraudulent activities. The results will take the form of charts, tables as well new and/or annotated process models, which will be generated based on specific stakeholders’ demands.
Australian Research Council: AUD$847,000
- 2015: $260,000
- 2016: $277,000
- 2017: $310,000
Queensland University of Technology (QUT)
- Prof. Marcello La Rosa
- Prof. Arthur H.M. ter Hofstede
- Prof. Michael Rosemann
- Dr Moe Wynn
- Dr Chun Ouyang
- Dr Michael Adams