Generating synthetic positive and negative business process traces through abduction

Abstract

As recent years have seen the rise of a new discipline commonly addressed as process mining, focused on the management of business processes, two tasks have gained increasing attention in research: process discovery and compliance monitoring. In both these fields, the demand for event log benchmarks with predefined characteristics has determined the design of various methodologies and tools for synthetic log generation. However, artificially created as well as real-life logs often contain positive examples only (i.e. process instances deemed as compliant w.r.t. the model), while the presence of negative process instances (i.e. non-compliant traces) can be crucial to correctly evaluate the performance and robustness of a novel process discovery or conformance checking technique. In this work, we investigate positive and negative trace generation in case of both declarative and procedural model specifications and we present our abduction-based approach to log synthesis. The theoretical study is concretely applied in a software prototype for log generation, which takes as input a declarative or structured workflow model and emits logs containing positive and negative traces. The approach provides both a highly expressive notation for the description of the business model and the ability to generate logs with various customizable features. The final comparative study of other existing log generators reveals several advantages of the proposed approach and draws the direction of future improvements.