Fire now, fire later: alarm-based systems for prescriptive process monitoring

Abstract

Predictive process monitoring is a family of techniques to analyze events produced during the execution of a business process in order to predict the future state or the final outcome of running process instances. Existing techniques in this field are able to predict, at each step of a process instance, the likelihood that it will lead to an undesired outcome. These techniques, however, focus on generating predictions and do not prescribe when and how process workers should intervene to decrease the cost of undesired outcomes. This paper proposes a framework for prescriptive process monitoring, which extends predictive monitoring with the ability to generate alarms that trigger interventions to prevent an undesired outcome or mitigate its effect. The framework incorporates a parameterized cost model to assess the cost–benefit trade-off of generating alarms. We show how to optimize the generation of alarms given an event log of past process executions and a set of cost model parameters. The proposed approaches are empirically evaluated using a range of real-life event logs. The experimental results show that the net cost of undesired outcomes can be minimized by changing the threshold for generating alarms, as the process instance progresses. Moreover, introducing delays for triggering alarms, instead of triggering them as soon as the probability of an undesired outcome exceeds a threshold, leads to lower net costs.

When to Intervene? Prescriptive Process Monitoring Under Uncertainty and Resource Constraints

“…To verify the effectiveness and relevance of the proposed method, we experimentally investigate whether we can learn when to trigger an intervention to maximize the total gain at the run time, considering the tradeoff between intervening now or later. We compare our results to baselines that consider either predictive models without quantifying the prediction uncertainty [6,14] or only the current case's execution state scores [2,18] as state-of-the-art baselines by addressing the following research questions:…”

“…Recently, various PrPM methods have been proposed [2,6,14,18]. These methods, however, implement intervention policies based on predictions of negative outcomes without considering the uncertainty of these predictions.…”

“…Metzger et al [14] suggest using ongoing cases prediction scores and their reliability estimate with a reinforcement learning technique to discover when to trigger runtime interventions. Another work by Fahrenkrog et al [6] proposes triggering one or more alarms when cases are more likely to end negatively, followed by an intervention.…”

“…Prescriptive Process monitoring (PrPM) is a family of process mining methods that trigger runtime actions to optimize a process's performance [6,14]. PrPM methods use event logs describing past business process executions to train machine learning (ML) algorithms for two goals.…”

Prescriptive process monitoring: Quo vadis?

“…This intervention could e.g. , come in the form of an action performed by a process worker, such as calling a customer, that helps to mitigate or prevent the negative outcome from materializing ( Fahrenkrog-Petersen et al, 2022 ). Thus, it is essential to define a policy for when a prescription is generated.…”