|
|
|
|
| Bachelorprojekt - Softwareteknologi | Project No. 0244: Online Process Mining and Complex Event Processing |
| Aktuelle | Tidligere |
Process mining is a data analysis method that aims to discover, monitor, and improve real processes (not assumed ones) by extracting knowledge available from event logs recorded from systems within an organization. Classical process mining assumes that data is stored in ßtatic" event logs. In online process mining, we assume that data comes as a continuous and potentially infinite stream of events. Each event contains the same set of information compared to static logs but the data is consumed immediately after it is produced. After that, the event is discarded and traces are kept only as an äggregated" information (for example, it is stored within a trained machine learning model). Examples of real-time data processing systems are banks' ATMs or air traffic control systems where events need to be processed (and decisions need to be made) within extremely short deadlines. Online process mining refers to the processing of event data (coming from the execution of processes) immediately after the events are observed and to deliver the decision immediately afterward (for example, in the case of conformance checking, the system has to immediately inform whether the execution is "compliant" with the system or not).
The project consists of working with such online process mining algorithms, improving, implementing and evaluating them. The reference platform where the algorithms is developed and maintained by the SPE section.
| Prerequisites: |
Some level of familiarity with programming languages (such as Java) is required. Basic knowledge of data science can also be beneficial.
|
Supervisor(s) Andrea Burratin
| Sidst opdateret: Nov 17, 2020 af Hans Henrik Løvengreen |