PACIS 2022 Proceedings

Paper Number

1256

Abstract

Predictive business process monitoring (PBPM) techniques aim at predicting future process behavior. A trend in PBPM is to use deep learning (DL), more concretely deep neural networks, to capture the entire process information within one predictive model. In most DL-based PBPM techniques, one single model is trained to predict the future behavior of a running process instance. However, especially resource information influences the efficiency and effectiveness of a process as paths through the process model can strongly depend on the resource executing the activities. Thus, a "one-model-fits-all" approach might not result in high predictive performance across all resources, such as humans or machines. Therefore, we design a novel DL-based method for resource-specific next activity predictions. In our preliminary evaluation, we present promising results based on two real-life event logs. Ultimately, we discuss our future research plans.

Comments

Paper Number 1256

Download

Share

COinS
 

When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.