Arvid Lepsien

Process mining analyzes processes based on event logs, which are generally assumed to be complete and trustworthy. These assumptions are often violated in real-world contexts, raising uncertainty about result reliability. Uncertainty-aware process mining promises to address this by explicitly leveraging uncertainty in calculations. We perform a structured survey of uncertainty-aware process mining (34 papers), which illustrates the currently dominant conceptualization of uncertainty in process mining and reveals the need for a conceptual foundation and integration with related research that handles uncertainty implicitly (e.g., process data quality). To address this, we construct an integrative conceptual framework for uncertainty in process mining by extending the core survey findings through an iterative hermeneutic review of seminal works from related research. The framework consists of two complementary lenses, detailing the (i) possible sources of and (ii) viable treatment strategies for uncertainty. Application of the framework shows both its integrative power for (future) research and its utility for guiding the handling of uncertainty in practical application.

Process mining and other forms of event data analytics have shown to be valuable tools for supporting the management and execution of business processes. For this, process recordings in the form of event logs are required. Digitalization efforts have led to an increased availability of event data for many previously paper-based processes. This has also inspired the extension of process analytics beyond classical business processes. One example of this is judicial processes, which are not necessarily bound by typical business constraints, but nonetheless face issues related to, e.g., increasing efficiency or improving resource allocation. Thus, this paper proposes a structured approach that combines domain expert knowledge with data-driven analysis to explore the usefulness of process analytics for judicial processes. Using an exemplary use case from a German social court, we show how event logs can be extracted from digitalized court files and, using our approach, identify bottlenecks. Using this approach, we are able to derive statistically grounded and expert validated bottlenecks as well as four actionable insights for reducing case durations. Thereby, we show that process analytics is a promising tool for facilitating the optimization of judicial processes.

Process mining allows the discovery of business processes from an event log. However, event logs are rapidly increasing in size and process mining algorithms struggle with the computational load when efficient processing is required. This calls for methods that decrease the event log size while still preserving the representativeness of the event log. This paper presents two new algorithms for sampling event logs. The first algorithm called chooses traces from an event log above a threshold and subsequently selects traces with underrepresented Directly Follows Relations. The second sampling algorithm called selects samples that have a high intersection of Directly Follows Relations with the original event log. Usually, is complemented with for a more accurate sample representation. They perform well for conformance checking and excel in certain scenarios for process discovery. Thus, both algorithms outperform existing sampling algorithms.

Various kinds of uncertainty can occur in event logs, e.g., due to flawed recording, data quality issues, or the use of probabilistic models for activity recognition. Stochastically known event logs make these uncertainties transparent by encoding multiple possible realizations for events. However, the number of realizations encoded by a stochastically known log grows exponentially with its size, making exhaustive exploration infeasible even for moderately sized event logs. Thus, considering only the top-K most probable realizations has been proposed in the literature. In this paper, we implement an efficient algorithm to calculate a top-K realization ranking of an event log under event independence within O(Kn), where n is the number of uncertain events in the log. This algorithm is used to investigate the benefit of top-K rankings over top-1 interpretations of stochastically known event logs. Specifically, we analyze the usefulness of top-K rankings against different properties of the input data. We show that the benefit of a top-K ranking depends on the length of the input event log and the distribution of the event probabilities. The results highlight the potential of top-K rankings to enhance uncertainty-aware process mining techniques.

This study aimed to demonstrate the application of process mining on video data of pigs, facilitating the analysis of behavioral patterns. Video data were collected over a period of 5 days from a pig pen in a mechanically ventilated barn and used for analysis. The approach in this study relies on a series of individual steps to allow process mining on this data set. These steps include object detection and tracking, spatiotemporal activity recognition in video data, and process model analysis. Each step gives insights into pig behavior at different time points and locations within the pen, offering increasing levels of detail to describe typical pig behavior up to process models reflecting different behavior sequences for clustered datasets. Our data-driven approach proves suitable for the comprehensive analysis of behavioral sequences in conventional pig farming.