A Novel Business Process Prediction Model Using a Deep Learning Method

Abstract

The ability to proactively monitor business processes is a main competitive differentiator for firms. Process execution logs generated by process aware information systems help to make process specific predictions for enabling a proactive situational awareness. The goal of the proposed approach is to predict the next process event from the completed activities of the running process instance, based on the execution log data from previously completed process instances. By predicting process events, companies can initiate timely interventions to address undesired deviations from the desired workflow. The paper proposes a multi-stage deep learning approach that formulates the next event prediction problem as a classification problem. Following a feature pre-processing stage with n-grams and feature hashing, a deep learning model consisting of an unsupervised pre-training component with stacked autoencoders and a supervised fine-tuning component is applied. Experiments on a variety of business process log datasets show that the multi-stage deep learning approach provides promising results. The study also compared the results to existing deep recurrent neural networks and conventional classification approaches. Furthermore, the paper addresses the identification of suitable hyperparameters for the proposed approach, and the handling of the imbalanced nature of business process event datasets.

A systematic literature review on state-of-the-art deep learning methods for process prediction

“…Metzger and Neubauer (2018) used Matthews correlation coefficient instead of accuracy because the accuracy is susceptible to imbalanced data, but they do not fit their training to this insight. Mehdiyev et al (2020) compared their aggregated prediction results on balanced and imbalanced training data and report a steep increase in performance, when the training data is balanced. The most commonly used datasets Helpdesk and BPIC 12 are heavily imbalanced.…”

“…andPasquadibisceglie et al (2019) solved this restriction by padding all shorter prefixes to the maximum possible length Ezpeleta et al (2018). trained a separate model for multiple prefix lengths andMehdiyev et al (2020) used n-grams as inputs to model the temporal dependencies while preserving a fixed input size. Theis and Darabi (2019) mined a petri net for the event logs and use this information to model the control-flow within an FFNN input.…”

Logistical Models of Competitive Development in Conditions of the Digital Transformation in Production and Marketing Business Processes

“…al. determine modeling monitoring of business processes as an important competitive advantage of a company in the system of its strategic planning [8].…”

Decay Replay Mining to Predict Next Process Events

“…All of the metrics are used to compare against the earlier-introduced next event prediction techniques. The subsequent definitions of metrics are based on [18], [63], [64].…”