Nowadays, a lot of data regarding business process executions are maintained in event logs. The next activity prediction task exploits such event logs to predict how process executions will unfold up until their completion. The present paper proposes a new approach to address this task: instead of using traces to perform predictions, we propose to use the instance graphs derived from traces. To make the most out of such representation we train a message passing neural network, specifically a Deep Graph Convolutional Neural Network to predict the next activity that will be performed in the process execution. The experiments performed show promising performance hinting that exploiting information about parallelism among activities in a process can induce a performance improvement in highly parallel process.
Exploiting Instance Graphs and Graph Neural Networks for next activity prediction / Chiorrini, Andrea; Diamantini, Claudia; Mircoli, Alex; Potena, Domenico. - (2022). (Intervento presentato al convegno Second international workshop on leveraging machine learning in process mining).
Exploiting Instance Graphs and Graph Neural Networks for next activity prediction
Andrea Chiorrini;Claudia Diamantini;Alex Mircoli;Domenico Potena
2022-01-01
Abstract
Nowadays, a lot of data regarding business process executions are maintained in event logs. The next activity prediction task exploits such event logs to predict how process executions will unfold up until their completion. The present paper proposes a new approach to address this task: instead of using traces to perform predictions, we propose to use the instance graphs derived from traces. To make the most out of such representation we train a message passing neural network, specifically a Deep Graph Convolutional Neural Network to predict the next activity that will be performed in the process execution. The experiments performed show promising performance hinting that exploiting information about parallelism among activities in a process can induce a performance improvement in highly parallel process.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.