| Grant number: | 17/26491-1 |
| Support Opportunities: | Scholarships abroad - Research |
| Start date: | August 01, 2018 |
| End date: | July 31, 2019 |
| Field of knowledge: | Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques |
| Principal Investigator: | Marcelo Fantinato |
| Grantee: | Marcelo Fantinato |
| Host Investigator: | Hajo Alexander Reijers |
| Host Institution: | Escola de Artes, Ciências e Humanidades (EACH). Universidade de São Paulo (USP). São Paulo , SP, Brazil |
| Institution abroad: | University Amsterdam (VU), Netherlands |
Abstract The combination of Business Process Management (BPM) and data mining has established a new research field -- known as process mining. The goal of process mining is to extract knowledge about data obtained from the work carried out at different stages of the BPM life-cycle. Process mining seeks to improve business processes by discovering links between variables and behavioral (or misbehavioral) patterns. The data to be mined are usually formed of event logs produced by the information systems used by organizations. Although there has already been a significant evolution regarding the specific techniques required for process mining, they are still unsuitable for unstructured processes, which are in fact those most often found in real organizations. The execution flow of unstructured processes has a weak causal dependence on its activities, i.e., these flows largely depend on occasional decisions made by their participants, which makes the execution of the instances essentially different from each other. This high degree of irregular behavior leads to considerable complexity and represents a challenge for current process mining techniques. Some studies have sought the support of advanced data mining techniques to assist in handling this type of scenario, including the use of genetic algorithms. However, even with the aid of genetic algorithms, the problem of how to discover unstructured business process models has not yet been satisfactorily resolved. This project adopts two advanced strategies: cooperative coevolution and the island model. Cooperative coevolution makes a subjective fitness assessment of individuals by determining whether or not they work well together; the island model gives rise to the evolution of subpopulations. It is expected that better solutions will be found for the discovery of unstructured process models. | |
| News published in Agência FAPESP Newsletter about the scholarship: | |
| More itemsLess items | |
| TITULO | |
| Articles published in other media outlets ( ): | |
| More itemsLess items | |
| VEICULO: TITULO (DATA) | |
| VEICULO: TITULO (DATA) | |