Data-Driven Process Discovery and Analysis

Complex technical systems, industrial systems or infrastructure systems are rich of customizable features and raise high demands on quality and safety-critical aspects. To create complete, valid and reliable planning and customization process data for a product deployment, an overarching engineering process is crucial for the successful completion of a project. In this paper, we introduce a framework for process management in complex engineering projects which are subject to a large amount of constraints and make use of heterogeneous data sources. In addition, we propose solutions for the framework components and describe a proof-of-concept implementation of the framework as an extension of a well-known BPMS.

Business Process Intelligence (BPI) is an emerging topic that has gained popularity in the last decade. It is driven by the need for analysis techniques that allow businesses to understand and improve their processes. One of the most common applications of BPI is reporting, which consists on the structured generation of information (i.e., reports) from raw data. In this article, state-of-the-art process mining techniques are used to periodically produce automated reports that relate the actual performance of students of a Dutch University to their studying behavior. To avoid the tedious manual repetition of the same process mining procedure for each course, we have designed a workflow calling various process mining techniques using RapidProM. To ensure that the actual students’ behavior is related to their actual performance (i.e., grades for courses), our analytic workflows approach leverages on process cubes, which enable the dataset to be sliced and diced based on courses and grades. The article discusses how the approach has been operationalized and what is the structure and concrete results of the reports that have been automatically generated. Two evaluations were performed with lecturers using the real reports. During the second evaluation round, the reports were restructured based on the feedback from the first evaluation round. Also, we analyzed an example report to show the range of insights that they provide.

Real-life business processes are complex and often exhibit a high degree of variability. Additionally, due to changing conditions and circumstances, these processes continuously evolve over time. For example, in the healthcare domain, advances in medicine trigger changes in diagnoses and treatment processes. Case data (e.g. treating physician, patient age) also influence how processes are executed. Existing process mining techniques assume processes to be static and therefore are less suited for the analysis of contemporary, flexible business processes. This paper presents a novel comparative case clustering approach that is able to expose changes in behavior. Valuable insights can be gained and process improvements can be made by finding those points in time where behavior changed and the reasons why. Evaluation using both synthetic and real-life event data shows our technique can provide these insights.

Process discovery algorithms typically aim at discovering process models from event logs. Most algorithms achieve this by solely using an event log, without allowing the domain expert to influence the discovery in any way. However, the user may have certain domain expertise which should be exploited to create better process models. In this paper, we address this issue of incorporating domain knowledge to improve the discovered process model. First, we present a verification algorithm to verify the presence of certain constraints in a process model. Then, we present three modification algorithms to modify the process model. The outcome of our approach is a Pareto front of process models based on the constraints specified by the domain expert and common quality dimensions of process mining.

Business process models are intensively used in organizations with various persons being involved in their creation. One of the challenges is the usage of a consistent terminology to label the activities of these process models. To support this task, prior research has proposed quality metrics to support the usage of consistent terms, mainly based on linguistic relations such as synonymy or homonymy. In this paper, we propose a new approach that utilizes hypernym hierarchies. We use these hierarchies to define a measure of abstractness which helps users to align the level of detail within one process model. Moreover, we define two techniques to detect specific terminology defects, namely process hierarchy defects and object hierarchy defects, and give recommendations to align them with hypernym hierarchies. We evaluate our approach on three process model collections from practice.

Operational support as an area of process mining aims to predict the performance of individual cases and the overall business process. Although seasonal effects, delays and performance trends are well-known to exist for business processes, there is up until now no prediction model available that explicitly captures seasonality. In this paper, we introduce time series Petri net models. These models integrate the control flow perspective of Petri nets with time series prediction. Our evaluation on the basis of our prototypical implementation demonstrates the merits of this model in terms of better accuracy in the presence of time series effects.

Event data or traces of activities often exhibit unexpected behavior and complex relations. Thus, before and during the application of automated analysis methods, such as process mining algorithms, the analyst needs to investigate and understand the data at hand in order to decide which analysis methods might be appropriate. Visual analytics integrates the outstanding capabilities of humans in terms of visual information exploration with the enormous processing power of computers to form a powerful knowledge discovery environment. The combination of visual data exploration with process mining algorithms makes complex information structures more comprehensible and facilitates new insights. In this position paper I portray various concepts of interactive visual support for process mining, focusing on the challenges, but also the great opportunities for analyzing process data with visual analytics methods.

Multidimensional process mining adopts the concept of data cubes to split event data into a set of homogenous sublogs according to case and event attributes. For each sublog, a separated process model is discovered and compared to other models to identify group-specific differences for the process. For an effective explorative process analysis, performance is vital due to the explorative characteristics of the analysis. We propose to adopt well-established approaches from the data warehouse domain based on relational databases to provide acceptable performance. In this paper, we present the underlying relational concepts of PMCube, a data-warehouse-based approach for multidimensional process mining. Based on a relational database schema, we introduce generic query patterns which map OLAP queries onto SQL to push the operations (i.e. aggregation and filtering) to the database management system. We evaluate the run-time behavior of our approach by a number of experiments. The results show that our approach provides a significantly better performance than the state-of-the-art for multidimensional process mining and scales up linearly with the number of events.