Name
Abstract | ||
|---|---|---|
The ability to proactively monitor business processes is one of the main differentiators for firms to remain competitive. Process execution logs generated by Process Aware Information Systems (PAIS) help to make various business process specific predictions. This enables a proactive situational awareness related to the execution of business processes. The goal of the approach proposed in the current paper is to predict the next business process event, considering the past activities in the running process instance, based on the execution log data from previously completed process instances. By predicting the business process events, companies can initiate timely interventions to address undesired deviations from the desired workflow. In our study, we propose a multi-stage deep learning approach which formulates the next business process event prediction problem as a classification problem and applies deep feedforward multilayer neural networks after extracting features with feature hashing and deep stacked autoencoders. The experiments conducted on a variety of business process log datasets reveal that the proposed multi-stage deep learning approach provides promising results. The results are compared against existing deep recurrent neural networks and other approaches as well. |
Year | Venue | Field |
|---|---|---|
2017 | CBI | Artifact-centric business process model,Business process management,Data mining,Business process,Computer science,Process modeling,Artificial intelligence,Business process modeling,Business process discovery,Machine learning,Business rule,Process mining |
DocType | Citations | PageRank |
Conference | 2 | 0.35 |
References | Authors | |
22 | 3 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Nijat Mehdiyev | 1 | 57 | 7.75 |
| Joerg Evermann | 2 | 447 | 35.82 |
| Peter Fettke | 3 | 812 | 78.37 |