Prescriptive process monitoring methods seek to enhance business process performance by triggering real-time interventions, such as offering discounts to increase the likelihood of a positive outcome (e.g., a purchase). At the core of a prescriptive process monitoring method lies an intervention policy, which determines under which conditions and when to trigger an intervention. While state-of-the-art prescriptive process monitoring approaches rely on black-box intervention policies derived through reinforcement learning , algorithmic decision-making requirements sometimes dictate that the business stakeholders must be able to understand, justify, and adjust these intervention policies manually. To address this requirement, this article proposes WB-PrPM (White-Box Prescriptive Process Monitoring), a framework that enables stakeholders to define intervention policies in business processes. WB-PrPM is a rule-based system that helps decision-makers balance the demand for effective interventions with the imperatives of limited resource capacity. The framework incorporates an automated method for tuning the parameters of the intervention policies to optimize a total gain function. An evaluation is presented using real-life datasets to examine the tradeoffs among various parameters. The evaluation reveals that different variants of the proposed framework outperform existing baselines in terms of total gain, even when default parameter values are used. Additionally, the automated parameter optimization approach further enhances the total gain.

Abstract Increasing the success rate of a process, i.e. the percentage of cases that end in a positive outcome, is a recurrent process improvement goal. At runtime, there are often certain actions (a.k.a. treatments) that workers may execute to lift the probability that a case ends in a positive outcome. For example, in a loan origination process, a possible treatment is to issue multiple loan offers to increase the probability that the customer takes a loan. Each treatment has a cost. Thus, when defining policies for prescribing treatments to cases, managers need to consider the net gain of the treatments. Also, the effect of a treatment varies over time: treating a case earlier may be more effective than later in a case. This paper presents a prescriptive monitoring method that automates this decision-making task. The method combines causal inference and reinforcement learning to learn treatment policies that maximize the net gain. The method leverages a conformal prediction technique to speed up the convergence of the reinforcement learning mechanism by separating cases that are likely to end up in a positive or negative outcome, from uncertain cases. An evaluation on two real-life datasets shows that the proposed method outperforms a state-of-the-art baseline.