Business Process Deviance Mining with Sequential and Declarative Patterns

Abstract Business Process Deviance refers to the phenomenon where a subset of the executions of a business process deviate, in a negative or positive way, with respect to their expected or desirable outcomes. Deviant executions of a business process include ones which violate compliance rules, or executions that underachieve or exceed performance targets. Business Process Deviance Mining is concerned with uncovering the reasons for deviant executions by analyzing event logs stored by the systems which support the execution of a business process. Good characterizations of deviant executions give analysts insights concerning the causes that generate such deviance, thus providing effective process improvement solutions. In the paper, the problem of explaining deviations in business processes is investigated from a novel perspective that integrates sequential and declarative patterns with data attributes of events and traces in event logs. This integration provides process analysts with richer explanations than existing Deviance Mining approaches, thus guaranteeing a better characterization of deviant executions. The research methodology followed in the paper is the design science methodology for information systems research. Using real-life logs from multiple domains, a range of feature types and different forms of explanations are evaluated in terms of their ability to accurately discriminate between deviant and non-deviant executions of a process as well as in terms of understandability of the final outcome returned to the analysts.