Book

Process Mining: Data Science in Action

Authors:

Abstract

This is the second edition of Wil van der Aalst’s seminal book on process mining, which now discusses the field also in the broader context of data science and big data approaches. It includes several additions and updates, e.g. on inductive mining techniques, the notion of alignments, a considerably expanded section on software tools and a completely new chapter of process mining in the large. It is self-contained, while at the same time covering the entire process-mining spectrum from process discovery to predictive analytics. After a general introduction to data science and process mining in Part I, Part II provides the basics of business process modeling and data mining necessary to understand the remainder of the book. Next, Part III focuses on process discovery as the most important process mining task, while Part IV moves beyond discovering the control flow of processes, highlighting conformance checking, and organizational and time perspectives. Part V offers a guide to successfully applying process mining in practice, including an introduction to the widely used open-source tool ProM and several commercial products. Lastly, Part VI takes a step back, reflecting on the material presented and the key open challenges. Overall, this book provides a comprehensive overview of the state of the art in process mining. It is intended for business process analysts, business consultants, process managers, graduate students, and BPM researchers.

Chapters (6)

In recent years, data science emerged as a new and important discipline. It can be viewed as an amalgamation of classical disciplines like statistics, data mining, databases, and distributed systems. Existing approaches need to be combined to turn abundantly available data into value for individuals, organizations, and society. Moreover, new challenges have emerged, not just in terms of size (“Big Data”) but also in terms of the questions to be answered. This book focuses on the analysis of behavior based on event data. Process mining techniques use event data to discover processes, check compliance, analyze bottlenecks, compare process variants, and suggest improvements. In later chapters, we will show that process mining provides powerful tools for today’s data scientist. However, before introducing the main topic of the book, we provide an overview of the data science discipline.
Information systems are becoming more and more intertwined with the operational processes they support. As discussed in the previous chapter, multitudes of events are recorded by today’s information systems. Nevertheless, organizations have problems extracting value from these data. The goal of process mining is to use event data to extract process-related information, e.g., to automatically discover a process model by observing events recorded by some enterprise system. A small example is used to explain the basic concepts. These concepts will be elaborated in later chapters.
Process mining is impossible without proper event logs. This chapter describes the information that should be present in such event logs. Depending on the process mining technique used, these requirements may vary. The challenge is to extract such data from a variety of data sources, e.g., databases, flat files, message logs, transaction logs, ERP systems, and document management systems. When merging and extracting data, both syntax and semantics play an important role. Moreover, depending on the questions one seeks to answer, different views on the available data are needed. Process mining, like any other data-driven analysis approach, needs to deal with data quality problems. We discuss typical data quality challenges encountered in reality. The insights provided in this chapter help to get the event data assumed to be present in later chapters.
After covering control-flow discovery in depth in Part III, this chapter looks at the situation in which both a process model and an event log are given. The model may have been constructed by hand or may have been discovered. Moreover, the model may be normative or descriptive. Conformance checking relates events in the event log to activities in the process model and compares both. The goal is to find commonalities and discrepancies between the modeled behavior and the observed behavior. Conformance checking is relevant for business alignment and auditing. For example, the event log can be replayed on top of the process model to find undesirable deviations suggesting fraud or inefficiencies. Moreover, conformance checking techniques can also be used for measuring the performance of process discovery algorithms and to repair models that are not aligned well with reality.
The successful application of process mining relies on good tool support. Traditional Business Intelligence (BI) tools are data-centric and focus on rather simplistic forms of analysis. Mainstream data mining and machine learning tools provide more sophisticated forms of analysis, but are also not tailored towards the analysis and improvement of processes. Fortunately, there are dedicated process mining tools able to transform event data into actionable process-related insights. For example, ProM is an open-source process mining tool supporting all of the techniques mentioned in this book. Process discovery, conformance checking, social network analysis, organizational mining, clustering, decision mining, prediction, and recommendation are all supported by ProM plug-ins. However, the usability of the hundreds of available plug-ins varies and the complexity of the tool may be overwhelming for end-users. In recent years, several vendors released dedicated process mining tools (e.g., Celonis, Disco, EDS, Fujitsu, Minit, myInvenio, Perceptive, PPM, QPR, Rialto, and SNP). These tools typically provide less functionality than ProM, but are easier to use while focusing on data extraction, performance analysis and scalability. This chapter provides an overview of available tools and trends.
Process mining provides the technology to leverage the ever-increasing amounts of event data in modern organizations and societies. Despite the growing capabilities of modern computing infrastructures, event logs may be too large or too complex to be handled using conventional approaches. This chapter focuses on handling “Big Event Data” and relates process mining to Big Data technologies. Moreover, it is shown that process mining problems can be decomposed in two ways, case-based decomposition and activity-based decomposition. Many of the analysis techniques described can be made scalable using such decompositions. Also other performance-related topics such as streaming process mining and process cubes are discussed. The chapter shows that the lion’s share of process mining techniques can be “applied in the large” by using the right infrastructure and approach.
... Both Disco and Glyph do not scale well, as their visualizations become explosively cluttered with less readability. Even though Disco includes a process mining interface to filter out patterns that are not frequent [16], it does not offer ways of interacting with the data to compare different patterns or inspect anomalies. To address these problems, we developed INSPECT and verified its capabilities by subsequent evaluations. ...
... In addition to visualization, business approaches explored the development of process visualizations and journey maps based on a data mining approach called process mining [16]. The process resulting from this process mining algorithm is often visualized in various forms, one of these approaches is a journey map [16]. ...
... In addition to visualization, business approaches explored the development of process visualizations and journey maps based on a data mining approach called process mining [16]. The process resulting from this process mining algorithm is often visualized in various forms, one of these approaches is a journey map [16]. A good example of this work is Disco, a commercially available system that uses event logs to generate visual representation of process models [6], which was not explicitly applied to game research before. ...
... To detect deviations in trading systems, we consider conformance checking [1,2]. Conformance checking is a process mining technique to search for differences between models describing expected behavior of processes and event logs that record real behavior of such processes [1]. ...
... To detect deviations in trading systems, we consider conformance checking [1,2]. Conformance checking is a process mining technique to search for differences between models describing expected behavior of processes and event logs that record real behavior of such processes [1]. Event logs consist of traces related to runs of processes; a trace is a sequence of events, where each event indicates an activity executed. ...
... For each object r = (r (1) , ..., r (n) ) in an event e = (a, R(e)), its components r (1) , ..., r (n) represent the state of r after the execution of a. We assume that the first component of r, r (1) , is the object identifier which cannot be modified; id(r) = r (1) denotes the identifier of r. We consider that objects in a trace can be distinguished. ...
Preprint
Full-text available
Trading systems are software platforms that support the exchange of securities (e.g., company shares) between participants. In this paper, we present a method to search for deviations in trading systems by checking conformance between colored Petri nets and event logs. Colored Petri nets (CPNs) are an extension of Petri nets, a formalism for modeling of distributed systems. CPNs allow us to describe an expected causal ordering between system activities and how data attributes of domain-related objects (e.g., orders to trade) must be transformed. Event logs consist of traces corresponding to runs of a real system. By comparing CPNs and event logs, different types of deviations can be detected. Using this method, we report the validation of a real-life trading system.
... Prescriptive Process Monitoring [1,2] is a branch of Process Mining [3] that, leveraging historical process data recorded in an event log, aims at providing users with recommendations that, when followed during the execution of a business process, improve the probability of avoiding negative outcomes, or optimizing performance indicators. For example, a Prescriptive Process Monitoring system might recommend the interventions to carry on, or the activities to execute in order to minimize the likelihood of a patient going to intensive care, or the time required for dismissing a patient from a hospital. ...
... The main basic concept in Process Mining [3] is the event record (or simply event) that represents the occurrence of an activity in a business process. An event is associated with three mandatory attributes: the event class (or activity name) that states the name of the activity the event refers to, the timestamp that specifies when the event occurred and the case id, which is an identifier of the case of the business process in which the event occurred. ...
... The training of the DT has been performed with a grid search to tune the hyperparameters with 5-fold cross-validation on L train . The range of values used for the hyperparameters are: i) the Gini index or the entropy criterion for the computation of the impurity; ii) [4, 6, 8, 10, ∞] for the maximum depth of the DT; iii) the use of class weights or not during the training to avoid poor performance due to the imbalance of the datasets (see, for example, hospital billing 2 in Table 5); iv) [0.1, 0.2, 0. 3,2] for the minimum number of samples required to split an internal node (float values indicate a percentage of the training data); v) [1,10,16] for the minimum number of samples required to consider a node a leaf node; vi) the number of the most informative features to use in the feature selection phase, i.e., 50%, 30% and the square root of the total number of initial features (after ranking them by using the mutual information score). ...
Preprint
Full-text available
Prescriptive Process Monitoring systems recommend, during the execution of a business process, interventions that, if followed, prevent a negative outcome of the process. Such interventions have to be reliable, that is, they have to guarantee the achievement of the desired outcome or performance, and they have to be flexible, that is, they have to avoid overturning the normal process execution or forcing the execution of a given activity. Most of the existing Prescriptive Process Monitoring solutions, however, while performing well in terms of recommendation reliability, provide the users with very specific (sequences of) activities that have to be executed without caring about the feasibility of these recommendations. In order to face this issue, we propose a new Outcome-Oriented Prescriptive Process Monitoring system recommending temporal relations between activities that have to be guaranteed during the process execution in order to achieve a desired outcome. This softens the mandatory execution of an activity at a given point in time, thus leaving more freedom to the user in deciding the interventions to put in place. Our approach defines these temporal relations with Linear Temporal Logic over finite traces patterns that are used as features to describe the historical process data recorded in an event log by the information systems supporting the execution of the process. Such encoded log is used to train a Machine Learning classifier to learn a mapping between the temporal patterns and the outcome of a process execution. The classifier is then queried at runtime to return as recommendations the most salient temporal patterns to be satisfied to maximize the likelihood of a certain outcome for an input ongoing process execution. The proposed system is assessed using a pool of 22 real-life event logs that have already been used as a benchmark in the Process Mining community.
... Process mining techniques [1] are used to analyze the behavior and performance of a variety of processes. A mandatory input of the techniques is an event log composed of cases, wherein each case is a sequence of events. ...
... We implemented a prototype tool for EC-SA-RM. 1 Using this tool, we conducted three experiments to evaluate the accuracy of our approach, and compared the results with EC-SA [9] and EC-SA-Data [7] as a baseline. ...
Conference Paper
Process mining analyzes business processes’ behavior and performance using event logs. An essential requirement is that events are grouped in cases representing the execution of process instances. However, logs extracted from different systems or non-process-aware information systems do not map events with unique case identifiers (case IDs). In such settings, the event log needs to be pre-processed to group events into cases – an operation known as event correlation. Existing techniques for correlating events work with different assumptions: some assume the generating processes are acyclic, others require extra domain knowledge such as the relation between the events and event attributes, or heuristic information about the activities’ execution time behavior. However, the domain knowledge is not always available or easy to acquire, compromising the quality of the correlated event log. In this paper, we propose a new technique called EC-SA-RM, which correlates the events using a simulated annealing technique and iteratively learns the domain knowledge as a set of association rules. The technique requires a sequence of timestamped events (i.e., the log without case IDs) and a process model describing the underlying business process. At each iteration of the simulated annealing, a possible correlated log is generated. Then, EC-SA-RM uses this correlated log to learn a set of association rules that represent the relationship between the events and the changing behavior over the events’ attributes in an understandable way. These rules enrich the input and improve the event correlation process for the next iteration. EC-SA-RM returns an event log in which events are grouped in cases and a set of association rules that explain the correlation over the events. We evaluate our approach using four real-life datasets.
... For now, we consider each activity name as "atomic" -later in this chapter we will see that activities themselves can have some "structure" themselves. A trace σ ∈ Σ * is a finite sequence of activities 1 . It describes that this sequence of activities had been observed at some point in the past. ...
... This hierarchical structure is formalized in the XES-standard [2,6,7]. See also other formalizations of event logs [1] Exercise 13. Provide the cases and traces of the structured event log of Table 2 for case identifier delivery. ...
Preprint
Full-text available
Event data is the basis for all process mining analysis. Most process mining techniques assume their input to be an event log. However, event data is rarely recorded in an event log format, but has to be extracted from raw data. Event log extraction itself is an act of modeling as the analyst has to consciously choose which features of the raw data are used for describing which behavior of which entities. Being aware of these choices and subtle but important differences in concepts such as trace, case, activity, event, table, and log is crucial for mastering advanced process mining analyses. This text provides fundamental concepts and formalizations and discusses design decisions in event log extraction from a raw event table and for event log pre-processing. It is intended as study material for an advanced lecture in a process mining course.
... Analyzing these data provides great opportunities for operational improvements, for example, reduced cycle times and increased conformity with reference process models. Therefore, process mining [17] comprises data-driven techniques to analyze event data to gain insights into the underlying processes; for example, automatically discovered process models, conformance statistics, and performance analysis information. Since service-oriented computing is concerned with orchestrating services to form dynamic business processes [6], process mining can provide valuable insights into the actual execution of processes within organizations [16]. ...
... Most process mining techniques [17] define process executions, termed traces, as a sequence, i.e., a strict total order, of executed activities. In reality, however, processes can exhibit parallel behavior, i.e., several branches of a process are executed simultaneously. ...
Preprint
Full-text available
Event logs, as viewed in process mining, contain event data describing the execution of operational processes. Most process mining techniques take an event log as input and generate insights about the underlying process by analyzing the data provided. Consequently, handling large volumes of event data is essential to apply process mining successfully. Traditionally, individual process executions are considered sequentially ordered process activities. However, process executions are increasingly viewed as partially ordered activities to more accurately reflect process behavior observed in reality, such as simultaneous execution of activities. Process executions comprising partially ordered activities may contain more complex activity patterns than sequence-based process executions. This paper presents a novel query language to call up process executions from event logs containing partially ordered activities. The query language allows users to specify complex ordering relations over activities, i.e., control flow constraints. Evaluating a query for a given log returns process executions satisfying the specified constraints. We demonstrate the implementation of the query language in a process mining tool and evaluate its performance on real-life event logs.
... Analyzing these data provides great opportunities for operational improvements, for example, reduced cycle times and increased conformity with reference process models. Therefore, process mining [17] comprises data-driven techniques to analyze event data to gain insights into the underlying processes; for example, automatically discovered process models, conformance statistics, and performance analysis information. Since service-oriented computing is concerned with orchestrating services to form dynamic business processes [6], process mining can provide valuable insights into the actual execution of processes within organizations [16]. ...
... Most process mining techniques [17] define process executions, termed traces, as a sequence, i.e., a strict total order, of executed activities. In reality, however, processes can exhibit parallel behavior, i.e., several branches of a process are executed simultaneously. ...
Chapter
Full-text available
Event logs, as viewed in process mining, contain event data describing the execution of operational processes. Most process mining techniques take an event log as input and generate insights about the underlying process by analyzing the data provided. Consequently, handling large volumes of event data is essential to apply process mining successfully. Traditionally, individual process executions are considered sequentially ordered process activities. However, process executions are increasingly viewed as partially ordered activities to more accurately reflect process behavior observed in reality, such as simultaneous execution of activities. Process executions comprising partially ordered activities may contain more complex activity patterns than sequence-based process executions. This paper presents a novel query language to call up process executions from event logs containing partially ordered activities. The query language allows users to specify complex ordering relations over activities, i.e., control flow constraints. Evaluating a query for a given log returns process executions satisfying the specified constraints. We demonstrate the implementation of the query language in a process mining tool and evaluate its performance on real-life event logs.
... [1,2]. "Modern" or online Process Mining considers additional activities as detect, predict and recommend [3]. ...
... Over the last decade, the industry has embraced Robotic Process Automation (RPA) as a new level of process automation aimed at mimicking the behavior of humans interacting with user interfaces (UI) rather than orchestrating sequences of API calls (van der Aalst, 2016). It is recognized that a successful RPA adoption goes beyond simple cost savings but also contributes to improved agility and quality (Asatiani and Penttinen, 2016;Capgemini, 2017;Lacity and Willcocks, 2015). ...
... Furthermore, some aspects of the actual process may not be taken into account. Process-mining-based solutions [4] have been presented as one method for reducing the impact of these issues. Process mining is a set of process management Zahra Mohammadnazari is with Kingston University, United Kingdom (email: z.mohammadnazari@kingston.ac.uk). ...
Article
Full-text available
Corporations have always prioritized efforts to examine and improve processes. Various metrics, such as the cost and time required to implement the process and can be specified in this regard. Process improvement can be defined as an improvement of these indicators. This is accomplished by looking at prospective adjustments to the current executive process model or the resources allotted to it. Research has been conducted in this paper to the improve the procurement process and aims to explore assessment prospects in the project using a combination of process mining and simulation (benefiting from Play-In and Play-Out methodologies). To run the simulation, we will need to complete the control flow diagram, institution settings, resource settings, and activity settings. The process of mining event logs yields the process control flow. However, both the entry of institutions and the distribution of resources must be modeled. The rate of admission of institutions and the distribution of time for the implementation of activities will be determined in the next step.
... Process mining techniques can identify bottlenecks, anticipate problems in process execution, record business rule violations, recommend countermeasures, provide information for decision-making, and streamline the process. In this way, the process discovery task through an algorithm can automatically produce a business process model from the behavior observed in the event log without requiring a-priori information [4,5]. The conformance checking task verifies the relationship between the behavior observed in the event log and the behavior modeled in the business process, making it possible to identify possible deviations [1,6]. ...
Article
Full-text available
Process mining is a novel alternative that uses event logs to discover, monitor, and improve real business processes through knowledge extraction. Event logs are a prerequisite for any process mining technique. The extraction of event data and event log building is a complex and time-intensive process, with human participation at several stages of the procedure. In this paper, we propose a framework to semi-automatically build an event log based on the XES standard from relational databases. The framework comprises the stages of requirements identification, event log construction, and event log evaluation. In the first stage, the data is interpreted to identify the relationship between the columns and business process activities, then the business process entities are defined. In the second stage, the hierarchical structure of the event log is specified. Likewise, a formal rule set is defined to allow mapping the database columns with the attributes specified in the event log structure, enabling the extraction of attributes. This task is implemented through a correlation method at the case, event, and activity levels, to automatic event log generation. In the third stage, we validate the event log through statistical analysis and business process discovery. The former allows determining the complexity of the event log built using the metrics of the average time of cases and average time of the number of events. The latter evaluates the business process models discovered through precision, coverage, and generalization metrics. The proposed approach was evaluated using an autonomous Internet of Things (IoT) air quality monitoring system’s database, reaching acceptable values of 1.0 in the precision and coverage metrics and between 0.980 and 0.991 in the generalization metric.
... As a set of process-focused data science techniques, process mining is widely used in various business environments [1]. A core value proposition of process mining is an increased understanding of the actual execution of processes and leveraging that knowledge for process improvements. ...
Conference Paper
Full-text available
On their trajectory through educational university systems, students leave a trace of event data. The analysis of that event data with a process lens poses a set of domain-specific challenges that is addressed in the field of Educational Process Mining (EPM). Despite the vast potential for understanding the progress of students and improving the quality of study programs through process mining, a case study based on an established process mining methodology is still missing. In this paper, we address this gap by applying the state-of-the-art process mining project methodology (PM2) in an EPM case study with a focus on student trajectory analysis at a German university. We found that process mining can create actionable items to improve the quality of university education. We also point out domain-specific challenges, like handling reoccurring exams (retaken after failing) for future research in EPM. Finally, we observe insights of some value in our case.
... It is indispensable for organizations to continuously monitor their operational problems and take proactive actions to mitigate risks and improve performances [1]. Constraint monitoring aims at detecting violations of constraints (i.e., operational problems) in business processes of an organization by analyzing event data recorded by information systems [7]. ...
Preprint
Full-text available
Constraint monitoring aims to monitor the violation of constraints in business processes, e.g., an invoice should be cleared within 48 hours after the corresponding goods receipt, by analyzing event data. Existing techniques for constraint monitoring assume that a single case notion exists in a business process, e.g., a patient in a healthcare process, and each event is associated with the case notion. However, in reality, business processes are object-centric, i.e., multiple case notions (objects) exist, and an event may be associated with multiple objects. For instance, an Order-To-Cash (O2C) process involves order, item, delivery, etc., and they interact when executing an event, e.g., packing multiple items together for a delivery. The existing techniques produce misleading insights when applied to such object-centric business processes. In this work, we propose an approach to monitoring constraints in object-centric business processes. To this end, we introduce Object-Centric Constraint Graphs (OCCGs) to represent constraints that consider the interaction of objects. Next, we evaluate the constraints represented by OCCGs by analyzing Object-Centric Event Logs (OCELs) that store the interaction of different objects in events. We have implemented a web application to support the proposed approach and conducted two case studies using a real-life SAP ERP system.
... Engineering digital twins is being studied also at ISoLA [55,56,82], including some initial, recent attempts in the railway domain [77]. This requires the use of a variety of techniques from formal methods, in particular probabilistic (and statistical) model checking to deal with the inherent probabilistic nature of the (machine-learned) model, game theory (for instance for controller synthesis), and automata or model learning, but also specific techniques from data or process mining [1]. Railway system models are characterised by the need to deal with real-time aspects and a degree of uncertainty. ...
Chapter
In 2020, the EU launched its sustainable and smart mobility strategy, outlining how it plans to have a 90% reduction in transport emission by 2050. Central to achieving this goal will be the improvement of rail technology, with many new data-driven visionary systems being proposed. AI will be the enabling technology for many of those systems. However, safety and security guarantees will be key for wide-spread acceptance and uptake by Industry and Society. Therefore, suitable verification and validation techniques are needed. In this article, we argue how formal methods research can contribute to the development of modern Railway systems—which may or may not make use of AI techniques—and present several research problems and techniques worth to be further considered.
... Put differently, each entity can be considered as a 'generating process' that emits events related to it, and we can distinguish different sources. This is a common in many real-world datasets, specifically to process mining [23] (events have a 'case identifier'), retailer data Shoppers selected for training. The grey shoppers did not have a visit in the three weeks from which the journey is extracted. ...
Preprint
Full-text available
Predicting the behaviour of shoppers provides valuable information for retailers, such as the expected spend of a shopper or the total turnover of a supermarket. The ability to make predictions on an individual level is useful, as it allows supermarkets to accurately perform targeted marketing. However, given the expected number of shoppers and their diverse behaviours, making accurate predictions on an individual level is difficult. This problem does not only arise in shopper behaviour, but also in various business processes, such as predicting when an invoice will be paid. In this paper we present CAPiES, a framework that focuses on this trade-off in an online setting. By making predictions on a larger number of entities at a time, we improve the predictive accuracy but at the potential cost of usefulness since we can say less about the individual entities. CAPiES is developed in an online setting, where we continuously update the prediction model and make new predictions over time. We show the existence of the trade-off in an experimental evaluation in two real-world scenarios: a supermarket with over 160 000 shoppers and a paint factory with over 171 000 invoices.
... Process mining provides techniques to extract insights from event data recorded by information systems, including process discovery, conformance checking, and performance analysis [1]. Especially performance analysis provides techniques to analyze the performance of a business process, e.g., bottlenecks, using process models as representations of the process [6]. ...
Chapter
Full-text available
Performance analysis in process mining aims to provide insights on the performance of a business process by using a process model as a formal representation of the process. Existing techniques for performance analysis assume that a single case notion exists in a business process (e.g., a patient in healthcare process). However, in reality, different objects might interact (e.g., order, delivery, and invoice in an O2C process). In such a setting, traditional techniques may yield misleading or even incorrect insights on performance metrics such as waiting time. More importantly, by considering the interaction between objects, we can define object-centric performance metrics such as synchronization time, pooling time, and lagging time. In this work, we propose a novel approach to performance analysis considering multiple case notions by using object-centric Petri nets as formal representations of business processes. The proposed approach correctly computes existing performance metrics, while supporting the derivation of newly-introduced object-centric performance metrics. We have implemented the approach as a web application and conducted a case study based on a real-life loan application process.KeywordsPerformance analysisObject-centric process miningObject-centric Petri netActionable insightsProcess improvement
... Once business processes are supported by information systems, it is often possible to extract event data on the execution of individual activities as part of a specific case [1]. Such data, which are commonly captured in the form of event logs, provides a valuable starting point for operational monitoring, analysis, and improvement. ...
Article
Event logs recorded during the execution of business processes provide a valuable starting point for operational monitoring, analysis, and improvement. Specifically, measures that quantify any deviation between the recorded operations and organizational goals enable the identification of operational issues. The data to compute such process-specific measures, commonly referred to as process performance indicators (PPIs), may contain personal data of individuals, though, which implies an inevitable risk of privacy intrusion that must be addressed. In this article, we target the privacy-aware computation of process performance indicators. To this end, we adopt tree-based definitions of PPIs according to the well-established PPINOT meta-model. For such a PPI, we design data release mechanisms for the functions in a PPI tree. Using a probabilistic formulation of the expected result of a privatized PPI, we further show how to determine the combination of release mechanisms that inflicts the least loss in utility. Moreover, given a set of PPIs, we provide an algorithmic framework to manage an inherent trade-off: Privatization may strive for maximal utility of each single PPI or for maximal reuse of privatized functions among all PPIs to use a privacy budget most effectively. Results from experiments with synthetic as well as real-world data indicate the general feasibility of privacy-aware PPIs and shed light on the trade-offs once a set of them is considered.
... Rights reserved. [1][2][3]. Im Gegensatz zum Process Mining stellt Task Mining einen neueren Ansatz dar, der eine detailliertere Betrachtungsweise auf die Prozesse verfolgt. Der Fokus liegt hierbei auf den einzelnen ausgeführten Arbeitsschritten entlang der Prozesse und ist demnach aufgrund der feineren Granularität eher auf der Workflowebene angesiedelt [4]. ...
Article
Full-text available
Zusammenfassung Task Mining ist ein Ansatz, mit dem Tätigkeiten entlang der Prozesse aus der Sicht des Arbeitsplatzes betrachtet werden. Durch die enge Verbindung zu Prozessen und Tätigkeiten erfährt Task Mining eine zunehmende Beliebtheit im Geschäftsprozess- und Workflowmanagement. Hierfür werden Ereignisdaten erfasst, die Interaktionen zwischen dem Benutzer und der Softwarelandschaft dokumentieren. Die Auswertung dieser Daten erfolgt mithilfe der Techniken des Data Minings, allerdings liegt der Fokus auf der Prozessebene. Letztendlich lassen sich so die realen Ist-Ausführungen von Prozessen durch die Anwender auf höchster Detailebene in Prozessmodelle zu überführen. Der primäre Einsatzzweck von Task Mining ist derzeit im Bereich der Robotic Process Automationen (RPA) zu verordnen, verlagert sich aber zunehmend in die Ergänzung von Process Mining. Dieser Artikel bietet eine Einführung in das Themenfeld Task Mining, einschließlich dessen Durchführung, und grenzt den Begriff Task Mining vom Process Mining ab. Zudem wird abschließend eine Übersicht der bestehenden Softwarelösungen geboten.
... Educational Process Mining (EPM) [4,27] is a sub-field of PM [28], using various, commonly known PM techniques in the educational context, e.g. higher education. While we focus on CMS data, most work in EPM has been done using Learning Management Systems (LMS) data with similar aims. ...
Preprint
Full-text available
This paper presents an approach of using methods of process mining and rule-based artificial intelligence to analyze and understand study paths of students based on campus management system data and study program models. Process mining techniques are used to characterize successful study paths, as well as to detect and visualize deviations from expected plans. These insights are combined with recommendations and requirements of the corresponding study programs extracted from examination regulations. Here, event calculus and answer set programming are used to provide models of the study programs which support planning and conformance checking while providing feedback on possible study plan violations. In its combination, process mining and rule-based artificial intelligence are used to support study planning and monitoring by deriving rules and recommendations for guiding students to more suitable study paths with higher success rates. Two applications will be implemented, one for students and one for study program designers.
Article
Full-text available
In this paper, we respond to Grover and Lyytinen (2022). We agree with them that the advent of the digital age is calling for a reconsideration of the role of theory and theorizing. We also think their proposal does not go far enough. The time is ripe to question the role of theory in our field more fundamentally. We propose to instead focus on establishing IS research as a platform through which we can collect, organize, and provide access to digital trace data from various sources to analyze contemporary socio-technical phenomena. We believe that such a move allows us to more fully unleash the unique socio-technical competences of our field in the digital age.
Article
Full-text available
In spite of recent advances in process mining, making this new technology accessible to non-technical users remains a challenge. Process maps and dashboards still seem to frighten many line of business professionals. In order to democratize this technology, we propose a natural language querying interface that allows non-technical users to retrieve relevant information and insights about their processes by simply asking questions in plain English. In this work we propose a reference architecture to support questions in natural language and provide the right answers by integrating to existing process mining tools. We combine classic natural language processing techniques (such as entity recognition and semantic parsing) with an abstract logical representation for process mining queries. We also provide a compilation of real natural language questions and an implementation of the architecture that interfaces to an existing commercial tool: Everflow. We also introduce a taxonomy for process mining related questions, and use that as a background grid to analyze the performance of this experiment. Finally, we point to potential future work opportunities in this field.
Chapter
Decision mining enables discovery of decision rules guiding the control flow in processes. Existing decision mining techniques deal with different kinds of decision rules, e.g., overlapping rules, or including data elements, for example, time series data. Though online process mining and monitoring are gaining traction, online decision mining algorithms are still missing. Decision rules can be, similarly to process models, subject to change during runtime due to, for example, changing regulations or customer requirements. In order to address these runtime challenges, this paper proposes an approach that i) discovers decision rules during runtime and ii) continuously monitors and adapts discovered rules to reflect changes. Furthermore, the concept of a decision rule history is proposed, enabling (manual) identification of change patterns. The feasibility and the applicability of the approach is evaluated based on three synthetic datasets, BPIC12, BPIC20 and sepsis data set.KeywordsOnline decision miningDecision rule evolutionDecision rule monitoringProcess-aware information systems
Conference Paper
Several organizations started to adopt the strategy to have data scientist professionals to help them identify the true value of the cost vs benefits and since data science is on fire for a quite long time, adopting the same comes in handy for several factors. i.e., investment, net profit, etc. At the same time, the clarity is missing around what data science is and this tends to introduce the concept of unambiguous theory or assumptions.In this research, we are aiming to pin down the fact of what data science is and how it can lead to crucial decision-making using a data-driven approach. We firmly believe that the accurate limitations of data science scope cannot be defined due to its vast possibilities. If the business wants to align themselves using the data science by understanding the true potential, the following needs to be underlined: 1) understand the importance of related concepts using the relationship approach, 2) identify how fundamental principles of data science can be useful for the different business use case, 3) Identify what the data science can offer depending on our requirement. In this research, we offer a decision-making solution for businesses using the underlying data science principles.
Chapter
Norms have been widely proposed as a way of coordinating and controlling the activities of agents in a multi-agent system (MAS). A norm specifies the behaviour an agent should follow in order to achieve the objective of the MAS. However, designing norms to achieve a particular system objective can be difficult, particularly when there is no direct link between the language in which the system objective is stated and the language in which the norms can be expressed. In this paper, we consider the problem of synthesising a norm from traces of agent behaviour, where each trace is labelled with whether the behaviour satisfies the system objective. We show that the norm synthesis problem and several related problems are NP-complete.
Chapter
Cyber-physical manufacturing systems with industry 4.0 technologies have the ability to generate real-time data on the behavior of the system in each of its components, so predictions can be generated from this data. This article presents a method for the development of a predictive model where process mining techniques and data mining algorithms are combined. Through the discovery techniques of process mining, a descriptive analysis of the system is carried out to subsequently develop a predictive model with predictive data mining algorithms that provide information on the time remaining for a product that is in process to be completed. This prediction allows decision makers to reconfigure the manufacturing system variables and its schedule to optimize its performance. The method was applied in a production system that is currently installed in the Computer Integration Manufacturing Lab at Pontificia Universidad Javeriana.
Article
Learning, or addressing a gap in one's knowledge, is an important motivator behind information‐seeking activities. The Search as Learning research community advocates that information search systems should be reconfigured to become educational platforms to foster learning and sensemaking. Modern search systems have yet to adapt to support this function. An important step to foster learning during search is to identify behavioral patterns that distinguish searchers gaining more vs. less knowledge during search. Previous efforts have primarily studied searchers in the short term, typically during a single lab session. Researchers are concerned over this ephemeral approach, as learning is not fleeting, and takes place over time. We propose an exploratory longitudinal study to analyze the long‐term searching behaviour of students enrolled in a university course, over the span of a semester. Our research aims are to identify if and how students’ searching behaviour changes over time, as they gain new knowledge on a subject; and how processes like motivation, metacognition, self‐regulation, and other individual differences moderate their “searching as learning” behaviour. Findings from this exploratory longitudinal study will help to build improved information search systems that foster human learning and sensemaking.
Chapter
Learning design platforms aim to facilitate teachers with their learning design processes. Existing studies have mostly focused on developing design features for learning design platforms yet the number of studies that investigate the collective behaviour of educators when using such tools is scarce. To this end, this study proposes the use of data analytics techniques, namely, process mining, to analyse the behaviour of teachers as they engage in using an online learning design platform as part of a teacher professional development course. The findings of the study shed light on teachers’ collective behaviour and motivational aspects related to their participation within an online learning design community.KeywordsLearning designLearning communitiesCommunity of educatorsProcess discovery
Chapter
Full-text available
Traditional process mining techniques take event data as input where each event is associated with exactly one object. An object represents the instantiation of a process. Object-centric event data contain events associated with multiple objects expressing the interaction of multiple processes. As traditional process mining techniques assume events associated with exactly one object, these techniques cannot be applied to object-centric event data. To use traditional process mining techniques, object-centric event data are flattened by removing all object references but one. The flattening process is lossy, leading to inaccurate features extracted from flattened data. Furthermore, the graph-like structure of object-centric event data is lost when flattening. In this paper, we introduce a general framework for extracting and encoding features from object-centric event data. We calculate features natively on the object-centric event data, leading to accurate measures. Furthermore, we provide three encodings for these features: tabular, sequential, and graph-based. While tabular and sequential encodings have been heavily used in process mining, the graph-based encoding is a new technique preserving the structure of the object-centric event data. We provide six use cases: a visualization and a prediction use case for each of the three encodings. We use explainable AI in the prediction use cases to show the utility of both the object-centric features and the structure of the sequential and graph-based encoding for a predictive model.
Preprint
Full-text available
Historically asymmetric availability or ownership of Data Asset, Information Asset, Knowledge Asset and Wisdom Asset has essentially contributed to the revelation of cognitive understanding, development and planning of various economic society in human civilization processing. Various models and theories of Economics lay their exchange and transaction foundation on the asymmetry of the availability and the asymmetry of the demands of commercial goods or services. However from the DIKWP Capital materialization and DIKWP Governance perspective, with the rapidly development of the modern information technology and widely progressing digital communication facilities, Asymmetric
Article
Designing healthcare facilities and their processes is a complex task which influences the quality and efficiency of healthcare services. The ongoing demand for healthcare services and cost burdens necessitate the application of analytical methods to enhance the overall service efficiency in hospitals. However, the variability in healthcare processes makes it highly complicated to accomplish this aim. This study addresses the complexity in the patient transport service process at a German hospital, and proposes a method based on process mining to obtain a holistic approach to recognise bottlenecks and main reasons for delays and resulting high costs associated with idle resources. To this aim, the event log data from the patient transport software system is collected and processed to discover the sequences and the timeline of the activities for the different cases of the transport process. The comparison between the actual and planned processes from the data set of the year 2020 shows that, for example, around 36% of the cases were 10 or more minutes delayed. To find delay issues in the process flow and their root causes the data traces of certain routes are intensively assessed. Additionally, the compliance with the predefined Key Performance Indicators concerning travel time and delay thresholds for individual cases was investigated. The efficiency of assignment of the transport requests to the transportation staff are also evaluated which gives useful understanding regarding staffing potential improvements. The research shows that process mining is an efficient method to provide comprehensive knowledge through process models that serve as Interactive Process Indicators and to extract significant transport pathways. It also suggests a more efficient patient transport concept and provides the decision makers with useful managerial insights to come up with efficient patient-centred analysis of transportation services through data from supporting information systems.
Chapter
In modern business processes, the amount of data collected has increased substantially in recent years. Because this data can potentially yield valuable insights, automated knowledge extraction based on process mining has been proposed, among other techniques, to provide users with intuitive access to the information contained therein. At present, the majority of technologies aim to reconstruct explicit business process models. These are directly interpretable but limited concerning the integration of diverse and real-valued information sources. On the other hand, Machine Learning (ML) benefits from the vast amount of data available and can deal with high-dimensional sources, yet it has rarely been applied to being used in processes. In this contribution, we evaluate the capability of modern Transformer architectures as well as more classical ML technologies of modeling process regularities, as can be quantitatively evaluated by their prediction capability. In addition, we demonstrate the capability of attentional properties and feature relevance determination by highlighting features that are crucial to the processes’ predictive abilities. We demonstrate the efficacy of our approach using five benchmark datasets and show that the ML models are capable of predicting critical outcomes and that the attention mechanisms or XAI components offer new insights into the underlying processes.KeywordsMachine learningProcess miningTransformerXAI
Chapter
Full-text available
Business processes are at the core of a well-functioning organization. They reflect how information should flow and what actions should be taken to achieve business goals. The discipline of Business Process Management (BPM) focuses on managing and improving business processes within an organization and is often partially adopted during an audit. This chapter discusses how various process analyses can support both internal and external auditors. To ensure that process analysis leads to correct insights, process mining can be applied. Process mining is a collective name for all data-driven process analysis techniques. The insights generated from a process mining analysis provide a good basis for improving business processes in terms of efficiency and risk. This broad view of processes ensures that the generated insights are relevant to both the internal and external auditor.
Article
Runtime monitoring is a central operational decision support task in business process management. It helps process executors to check on-the-fly whether a running process instance satisfies business constraints of interest, providing an immediate feedback when deviations occur. We study runtime monitoring of properties expressed in ltl f , a variant of the classical ltl (Linear-time Temporal Logic) that is interpreted over finite traces, and in its extension ldl f , a powerful logic obtained by combining ltl f with regular expressions. We show that ldl f is able to declaratively express, in the logic itself, not only the constraints to be monitored, but also the de facto standard rv -LTL monitors. On the one hand, this enables us to directly employ the standard characterization of ldl f based on finite-state automata to monitor constraints in a fine-grained way. On the other hand, it provides the basis for declaratively expressing sophisticated metaconstraints that predicate on the monitoring state of other constraints, and to check them by relying on standard logical services instead of ad hoc algorithms. We then report on how this approach has been effectively implemented using Java to manipulate ldl f formulae and their corresponding monitors, and the RuM rule mining suite as underlying infrastructure.
Article
Full-text available
The field of process mining focuses on distilling knowledge of the (historical) execution of a process based on the operational event data generated and stored during its execution. Most existing process mining techniques assume that the event data describe activity executions as degenerate time intervals, i.e., intervals of the form [t, t], yielding a strict total order on the observed activity instances. However, for various practical use cases, e.g., the logging of activity executions with a nonzero duration and uncertainty on the correctness of the recorded timestamps of the activity executions, assuming a partial order on the observed activity instances is more appropriate. Using partial orders to represent process executions, i.e., based on recorded event data, allows for new classes of process mining algorithms, i.e., aware of parallelism and robust to uncertainty. Yet, interestingly, only a limited number of studies consider using intermediate data abstractions that explicitly assume a partial order over a collection of observed activity instances. Considering recent developments in process mining, e.g., the prevalence of high-quality event data and techniques for event data abstraction, the need for algorithms designed to handle partially ordered event data is expected to grow in the upcoming years. Therefore, this paper presents a survey of process mining techniques that explicitly use partial orders to represent recorded process behavior. We performed a keyword search, followed by a snowball sampling strategy, yielding 68 relevant articles in the field. We observe a recent uptake in works covering partial-order-based process mining, e.g., due to the current trend of process mining based on uncertain event data. Furthermore, we outline promising novel research directions for the use of partial orders in the context of process mining algorithms.
Conference Paper
Surgical center scheduling is challenging to schedule physical spaces to reduce costs and increase productivity. This study presents a framework called PM4SOS (Process Mining for Simulation, optimization, and Scheduling) that facilitates the generation of an operating room schedule in an automated way and integrates, reducing cognitive overload and waste of time. This framework allows the evaluation of restrictions to get the best performance of surgical schedules. PM4SOS combines process mining with data from event logs, generation of the automated simulation model, and case-based reasoning (CBR) to analyze management indicators, allowing significant gains in optimization of the scheduling of surgical centers. The results with PM4SOS help decision-making in hospital environments to better use physical resources and human resources, such as reducing waiting time, optimizing surgery execution time, and resource capacity management.
Article
Full-text available
Zusammenfassung Data-Science-Projekte sind typischerweise interdisziplinär, adressieren vielfältige Problemstellungen aus unterschiedlichen Domänen und sind häufig durch heterogene Projektmerkmale geprägt. Bestrebungen in Richtung einer einheitlichen Charakterisierung von Data-Science-Projekten sind insbesondere dann relevant, wenn über deren Durchführung entschieden werden soll – beispielsweise anhand von Kriterien wie Ressourcenbedarf, Datenverfügbarkeit oder potenziellen Risiken. Nach bestem Wissen der Autoren fehlt es jedoch in Wissenschaft und Praxis bisher an einschlägigen Ansätzen. Mit diesem Artikel wird ein erster Schritt auf dem Weg hin zu einem Ansatz für eine einheitliche Charakterisierung von Data-Science-Projekten gegangen, indem ein morphologischer Kasten vorgeschlagen wird, der im Rahmen einer dreischrittigen Analyse auf Basis eines Fragenkataloges abgeleitet wurde. Er umfasst sieben Dimensionen mit 32 Dimensionsausprägungen und wird anhand einer Fallstudie aus dem Gebiet der Predictive Maintenance illustriert. Der morphologische Kasten bietet theoretische und praktische Anwendungspotenziale für den strukturierten Vergleich von Data-Science-Projekten und die Definition von Projektportfolios, erhebt jedoch keinen Anspruch auf Vollständigkeit. Er ist somit als Vorschlag und Anstoß zum Einstieg in einen weiterführenden Diskurs anzusehen.
Chapter
Event logs are used for a plethora of process analytics and mining techniques. A class of these mining activities is conformance (compliance) checking. The goal is to identify the violation of such patterns, i.e., anti-patterns. Several approaches have been proposed to tackle this analysis task. These approaches have been based on different data models and storage technologies of the event log including relational databases, graph databases, and proprietary formats. Graph-based encoding of event logs is a promising direction that turns several process analytic tasks into queries on the underlying graph. Compliance checking is one class of such analysis tasks. In this paper, we argue that encoding log data as graphs alone is not enough to guarantee efficient processing of queries on this data. Efficiency is important due to the interactive nature of compliance checking. Thus, anti-pattern detection would benefit from sub-linear scanning of the data. Moreover, as more data are added, e.g., new batches of logs arrive, the data size should grow sub-linearly to optimize both the space of storage and time for querying. We propose two encoding methods using graph representations, realized in Neo4J & SQL Graph Database, and show the benefits of these encoding on a special class of queries, namely timed ordered anti-patterns. Compared to several baseline encoding, our experiments show up to 5x speed up in the querying time as well as a 3x reduction in the graph size.
Chapter
Resource allocation to execute business processes is increasingly crucial for organizations. As the cost of executing process tasks relies on several dynamic factors, optimizing resource allocation can be addressed as a sequential decision process. Process mining can aid this optimization with the use of data from the event log, which records historical data related to the corresponding business process executions. Probabilistic approaches are relevant to solve process mining issues, especially when applied to the usually unstructured and noisy real-world business processes. We present an approach in which the problem of resource allocation in a business process is modeled as a Markovian decision process and batch reinforcement learning algorithm is applied to get a resource allocation policy that minimizes the cycle time. With batch reinforcement learning algorithms, the knowledge underlying the event log data is used both during policy learning procedures and to model the environment. Resource allocation is performed considering the task to be executed and the resources’ current workload. The results with both Fitted Q-Iteration and Neural Fitted Q-Iteration batch reinforcement learning algorithms demonstrate that this approach enables a resource allocation more adherent to the business interests. Per the evaluation we performed on data of a real-world business process, if our approach had been used, up to 37.2% of the time spent to execute all the tasks could have been avoided compared to what is represented in the historical data at the event log.KeywordsReinforcement learningProcess miningResource allocationBusiness processes
Article
Process descriptions are the backbones for creating products and delivering services automatically. Computing the alignments between process descriptions (such as process models) and process behavior is one of the fundamental tasks to lead to better processes and services. The reason is that the computed results can be directly used in checking compliance, diagnosing deviations, and analyzing bottlenecks for processes. Although various alignment techniques have been proposed in recent years, their performance is still challenged by large logs and models. In this work, we introduce an efficient approach to accelerate the computation of alignments. Specifically, we focus on the computation of optimal alignments, and try to improve the performance of the state-of-the-art A∗-based method through Petri net decomposition. We present the details of our designs and also show that our approach can be easily implemented in a distributed environment using the Spark platform. Using datasets with large event logs and process models, we experimentally demonstrate that our approach can indeed accelerate current A∗-based implementations in general.
Article
Full-text available
Anomalous behavior detection in business processes inspects abnormal situations, such as errors and missing values in system execution records, to facilitate safe system operation. Since anomaly information hinders the insightful investigation of event logs, many approaches have contributed to anomaly detection in either the business process domain or the data mining domain. However, most of them ignore the impact brought by the interaction between activities and their related attributes. Based on this, a method is constructed to integrate the consistency degree of multi-perspective log features and use it in an isolation forest model for anomaly detection. First, a reference model is captured from the event logs using process discovery. After that, the similarity between behaviors is analyzed based on the neighborhood distance between the logs and the reference model, and the data flow similarity is measured based on the matching relationship of the process activity attributes. Then, the integration consistency measure is constructed. Based on this, the composite log feature vectors are produced by combining the activity sequences and attribute sequences in the event logs and are fed to the isolation forest model for training. Subsequently, anomaly scores are calculated and anomalous behavior is determined based on different threshold-setting strategies. Finally, the proposed algorithm is implemented using the Scikit-learn framework and evaluated in real logs regarding anomalous behavior recognition rate and model quality improvement. The experimental results show that the algorithm can detect abnormal behaviors in event logs and improve the model quality.
Thesis
Full-text available
This research works on improving the customer support process of an Education as a Service provider with a data-driven approach. A literature review on process mining to improve customer support and on frequently asked questions in customer support is used as a method to answer the first research question. The first research question is about identifying suitable methods to analyze the customer support. Quantitative and qualitative data analysis were chosen as methods to answer the second research question about finding the most frequent incoming customer requests of the customer support. As the data analysis revealed that most tickets are labeled with the correct categories, the most frequent customer requests could be defined by their assigned categories. Additionally, other patterns in the data, such as who processes specific customer requests, were found. The process mining tool Disco with its integrated fuzzy mining algorithm analyzed the event log of the most frequent customer requests in more detail. The fuzzy miner based on frequency and time performance metric was able to identify the most important activities, the most dominant process flows, and the main bottlenecks in the process of handling incoming customer requests. Before answering the third research question about formulating suggestions for improvement for the customer support process, expert interviews were conducted to identify the main current problems. The findings of the data analysis and process mining then led to several suggestions for improvement of how the identified problems of the customer support can be diminished
Chapter
Full-text available
Although our capabilities to store and process data have been increasing exponentially since the 1960s, suddenly many organizations realize that survival is not possible without exploiting available data intelligently. Out of the blue, “Big Data” has become a topic in board-level discussions. The abundance of data will change many jobs across all industries. Moreover, also scientific research is becoming more data-driven. Therefore, we reflect on the emerging data science discipline. Just like computer science emerged as a new discipline from mathematics when computers became abundantly available, we now see the birth of data science as a new discipline driven by the torrents of data available today. We believe that the data scientist will be the engineer of the future. Therefore, Eindhoven University of Technology (TU/e) established the Data Science Center Eindhoven (DSC/e). This article discusses the data science discipline and motivates its importance.
Article
Full-text available
Conformance checking techniques compare observed behavior (i.e., event logs) with modeled behavior for a variety of reasons. For example, discrepancies between a normative process model and recorded behavior may point to fraud or inefficiencies. The resulting diagnostics can be used for auditing and compliance management. Conformance checking can also be used to judge a process model automatically discovered from an event log. Models discovered using different process discovery techniques need to be compared objectively. These examples illustrate just a few of the many use cases for aligning observed and modeled behavior. Thus far, most conformance checking techniques focused on replay fitness, i.e., the ability to reproduce the event log. However, it is easy to construct models that allow for lots of behavior (including the observed behavior) without being precise. In this paper, we propose a method to measure precision of process models, given their event logs by first aligning the logs to the models. This way, the measurement is not sensitive to non-fitting executions and more accurate values can be obtained for non-fitting logs. Furthermore, we introduce several variants of the technique to deal better with incomplete logs and reduce possible bias due to behavioral property of process models. The approach has been implemented in the ProM 6 framework and tested against both artificial and real-life cases. Experiments show that the approach is robust to noise and applicable to handle logs and models of real-life complexity.
Article
Full-text available
Recently, Nüttgens and Rump proposed a formal semantics for Event driven Process Chains (EPCs), which should be fully compliant with the informal semantics of EPCs. But, their semantics has a severe flaw. This flaw reveals that there is a fundamental problem with the informal semantics of EPCs. Here, we pin-point the cause of this problem, we show that there is no sound formal semantics for EPCs that is fully compliant with the informal semantics, and we discuss some consequences.