Synthetic event log dataset on a Loan Application scenario
Synthetic event logs generated starting from a Declare process model related to a Loan Application scenario composed of positive and negative traces
The synthetic dataset related to a Loan Application process is composed of:
- a log of positive traces (declarative_traces_positives.xml)
- a first log of negative traces (declarative_traces_negatives1.xml)
- a second log of negative traces (declarative_traces_negatives2.xml)
- a third log of negative traces (declarative_traces_negatives3.xml)
The positive traces are directly generated from a Declare process model, while each of the negative event logs has been obtained by negating one of the constraints of the same Declare model.
Additional information:
The four event logs are generated starting from a Declare model using a tool [1] based on Abductive Logic Programming. The Declare model has been inspired by the Loan Application process reported in [2]. The positive and negative event logs can be used for discovering process models starting from positive and negative examples.
[1] D. Loreti, F. Chesani, A. Ciampolini, and P. Mello, “Generating synthetic positive and negative business process traces through
abduction,” Knowl. Inf. Syst., vol. 62, no. 2, pp. 813–839, 2020
[2] M. Dumas, M. L. Rosa, J. Mendling, and H. A. Reijers, Fundamentals of Business Process Management, Second Edition.
Springer, 2018
The asset has been developed for the microproject "Discovering Temporal Logic Patterns as Binary Supervised Learning" that involves University of Bologna and Fondazione Bruno Kessler, as well as the Free University of Bolzano as external partner.