A Novel Approach to Detect Anomalies in Business Process Event Logs Using Deep Learning Algorithm

Abstract

Enterprises whose businesses are driven by web-based or cloud-based applications contain thousands of business processes involved. Due to the dynamic runtime environments and distributed nature of business processes and dependencies, there is possibility of noise and anomalies. Moreover, naturally, businesses are interested in finding anomalies in business processes and rectify them for improving quality of service (QoS). Especially, as part of process mining, anomaly detection has become an important research area in the contemporary era. Many anomaly detection methods came into existence based on machine learning techniques. There are attempts made using autoencoders for business process anomaly detection. However, from the literature, it is understood that there is need for a deep learning-based autoencoder with unsupervised learning approach for efficient detection of anomalies by analysing business process event logs. Towards this end, in this paper, we proposed a methodology and defined an algorithm known as deep learning encoder-based anomaly detection (DLE-AD) for enhancing the ability of anomaly detection. From the experiments, it is revealed that deep learning-based anomaly detection showed better performance over the traditional approaches. The proposed algorithm is evaluated against state of the art and found that it outperforms the existing methods.