Accurate and Transparent Path Prediction Using Process Mining

Abstract

Anticipating the next events of an ongoing series of activities has many compelling applications in various industries. It can be used to improve customer satisfaction, to enhance operational efficiency, and to streamline health-care services, to name a few. In this work, we propose an algorithm that predicts the next events by leveraging business process models obtained using process mining techniques. Because we are using business process models to build the predictions, it allows business analysts to interpret and alter the predictions. We tested our approach with more than 30 synthetic datasets as well as 6 real datasets. The results have superior accuracy compared to using neural networks while being orders of magnitude faster.

Business Process Event Prediction Through Scalable Online Learning

“…Recently, the use of process trees has been proposed in order to predict future activities in [29]. However, this approach implies a more complex treatment, meaning a barrier for scalability.…”

Discovering Business Area Effects to Process Mining Analysis Using Clustering and Influence Analysis

“…Also, process mining models obtained using process mining techniques themselves can be used as a model for prediction. In [3] the authors first generate a process tree using the inductive miner algorithm, after which this process tree is used to predict the future path of ongoing cases. This approach does not take any additional event-or case attributes into account.…”

“…To be able to perform them as efficiently as possible, we needed to implement support for the XES file format 2 . This support was made official in November 2018, when QPR ProcessAnalyzer 2018.6 received the official XES Certification for the XES WG 3 . This has allowed customers to use public and, at least in the process mining community, widely known datasets for training and experimenting with our product.…”

“…Lately, also process mining industry has recognized the possibilities of Introduction machine learning techniques as several commercial process mining tools, such as QPR ProcessAnalyzer 1 , Celonis 2 , and myInvenio 3 , have started to incorporate more and more predictive analytics functionalities into them. Machine learning-based models do not require domain experts and their ability to provide a full picture of the as-is process as a prescriptive process model, which could be highly biased and cannot as easily be automated.…”

Exploiting Event Log Event Attributes in RNN Based Prediction

“…Also, process mining models obtained using process mining techniques themselves can be used as a model for prediction. In [1] the authors first generate a process tree using the inductive miner algorithm, after which this process tree is used to predict the future path of ongoing cases. This approach does not take any additional event-or case attributes into account.…”