Autoencoders for improving quality of process event logs

Process Model Discovery from Unlabeled Event Logs by Using Non-Overlapping Sequential Distinct Event Patterns

“…In [42] the authors presented a method based on autoencoders to improve the quality of process event logs by considering anomalous and missing values of attributes in event logs.…”

Process Discovery in Business Process Management Optimization

“…Some authors (e.g., Reference [14]) raise the significant problem of the quality of event logs. They notice that processes discovered using logs of low quality will also be of low quality, resulting in incorrect business decisions taken upon them.…”

“…They notice that processes discovered using logs of low quality will also be of low quality, resulting in incorrect business decisions taken upon them. Authors of Reference [14] propose the use of methods based on autoencoders, which are a class of neural networks, to reconstruct the leaking values in logs. They report a significant improvement of the model quality after applying that method.…”

Online Anomaly Detection Using Statistical Leverage for Streaming Business Process Events

“…To this end, the research field of event log anomaly detection (or event log cleaning) has emerged recently, providing methods to detect anomalies at trace level [1,2,9,10,13], i.e., concerning the order and occurrence of activities in a process, and at event level [15,16,18], i.e., concerning the value of attributes of events, using a variety of different approaches. Note that event log anomaly detection is normally (process) model-agnostic, that is, it does not assume the existence of a process model or clean traces from which a model can be extracted.…”

“…Online settings also prevent the application of machine learning reconstructive techniques for anomaly detection, e.g. [16,17]. These, in fact, normally rely on deep learning models, which require a high number of data points (complete process traces in this scenario) to be trained effectively.…”