Content uploaded by Artem Polyvyanyy
Author content
All content in this area was uploaded by Artem Polyvyanyy on Jan 29, 2017
Content may be subject to copyright.
A preview of the PDF is not available
Structuring process models
Abstract and Figures
One can fairly adopt the ideas of Donald E. Knuth to conclude that process modeling is both a science and an art. Process modeling does have an aesthetic sense. Similar to composing an opera or writing a novel, process modeling is carried out by humans who undergo creative practices when engineering a process model. Therefore, the very same process can be modeled in a myriad number of ways. Once modeled, processes can be analyzed by employing scientific methods. Usually, process models are formalized as directed graphs, with nodes representing tasks and decisions, and directed arcs describing temporal constraints between the nodes. Common process definition languages, such as Business Process Model and Notation (BPMN) and Event-driven Process Chain (EPC) allow process analysts to define models with arbitrary complex topologies. The absence of structural constraints supports creativity and productivity, as there is no need to force ideas into a limited amount of available structural patterns. Nevertheless, it is often preferable that models follow certain structural rules. A well-known structural property of process models is (well-)structuredness. A process model is (well-)structured if and only if every node with multiple outgoing arcs (a split) has a corresponding node with multiple incoming arcs (a join), and vice versa, such that the set of nodes between the split and the join induces a single-entry-single-exit (SESE) region; otherwise the process model is unstructured. The motivations for well-structured process models are manifold: (i) Well-structured process models are easier to layout for visual representation as their formalizations are planar graphs. (ii) Well-structured process models are easier to comprehend by humans. (iii) Well-structured process models tend to have fewer errors than unstructured ones and it is less probable to introduce new errors when modifying a well-structured process model. (iv) Well-structured process models are better suited for analysis with many existing formal techniques applicable only for well-structured process models. (v) Well-structured process models are better suited for efficient execution and optimization, e.g., when discovering independent regions of a process model that can be executed concurrently. Consequently, there are process modeling languages that encourage well-structured modeling, e.g., Business Process Execution Language (BPEL) and ADEPT. However, the well-structured process modeling implies some limitations: (i) There exist processes that cannot be formalized as well-structured process models. (ii) There exist processes that when formalized as well-structured process models require a considerable duplication of modeling constructs. Rather than expecting well-structured modeling from start, we advocate for the absence of structural constraints when modeling. Afterwards, automated methods can suggest, upon request and whenever possible, alternative formalizations that are "better" structured, preferably well-structured. In this thesis, we study the problem of automatically transforming process models into equivalent well-structured models. The developed transformations are performed under a strong notion of behavioral equivalence which preserves concurrency. The findings are implemented in a tool, which is publicly available.
Figures - uploaded by Artem Polyvyanyy
Author content
All figure content in this area was uploaded by Artem Polyvyanyy
Content may be subject to copyright.
... As a process model can describe infinitely many instances, it is challenging to specify a query that addresses all process instances of the model over the finite graph of this model. Secondly, infinitely many structurally different process models can describe the same behavioral relations [15,16]. Consequently, it is challenging to specify a query that accounts for all possible structural patterns of models that can describe the relations. ...
... To exemplify PQL, we use an example process repository consisting of ten process models sourced from the SAP R/3 reference model [5] and Polyvyanyy's Ph.D. thesis [15] and depicted in BPMN in Fig. 8. For simplicity, the models in Fig. 8 use alphabet letters as abstract task labels. ...
... Models 1 to 5 are acyclic, while models 6 to 10 contain cycles. Models 1, 2, 6, and 10 are well-structured, where a model is well-structured if and only if every node with multiple outgoing arcs (a split) has a corresponding node with multiple incoming arcs (a join), and vice versa, such that the set of nodes between the split and the join induces a Single-Entry-Single-Exit component [63] [15]. Models 3,4,5,7,8, and 9 are unstructured. ...
... Unstructuredness and loops make the execution and analysis of process models difficult [2,3,8,16,17,20,22,23,30,31]. Polyvyanyy summarizes that wellstructured process models are more comprehensible for humans, are more likely to contain fewer errors, and, therefore, improve their quality [20]. ...
... Unstructuredness and loops make the execution and analysis of process models difficult [2,3,8,16,17,20,22,23,30,31]. Polyvyanyy summarizes that wellstructured process models are more comprehensible for humans, are more likely to contain fewer errors, and, therefore, improve their quality [20]. Arbitrary loops ⋆⋆ The third author's work was done during his PhD program at Inje University. ...
... Arbitrary loops ⋆⋆ The third author's work was done during his PhD program at Inje University. D R A F T tend to increase the probability of errors in process models [17] and prevent the structuring of process models or at least increase the effort [20]. The (Refined) Process Structure Tree (RPST) describes a hierarchy of single entry and single exit (SESE) structures [2,23,30,31] that is often used to find independent structures in unstructured process models [31] and to speed up analysis [6,11], but do not help to solve the problem of unstructured and cyclic process components [3]. ...
Business process models are usually described in a visual notation and reflect actual processes in systems. As a result, process models often are unstructured and cyclic. Unfortunately, unstructured and cyclic models are difficult to analyze and execute as research shows. Unstructuredness could be overcome using the existing studies, however, the analysis of cyclic models is still an open research problem. For this reason, this paper presents a decomposition of cyclic process models into sets of acyclic models. Together with a simple execution semantics for the acyclic models, the semantics of the decomposed model coincides with the original model if soundness is assumed. The decomposition can be achieved in a quadratic runtime complexity and gives the possibility to apply many existing analysis methodologies for acyclic process models. A short evaluation shows the feasibility of the approach.
... An entire process model is sound if each of its fragments is sound. The drawback of this method is its incompleteness regarding soundness [47], as inherently unstructured fragments (known as rigids) [34] In summary, most approaches fail by too restricted OR-join semantics, high computational complexity, or incompleteness [27,15,50,38,47]. The approaches by using symbolic execution of Favre and Völzer [22] as well as by searching activation edges and intersection points by Prinz and Amme [39] are ecient and provide good diagnostic information, but are restricted to acyclic process models. ...
... This will involve deriving a new execution semantics after loop decomposition to build event logs and establish theoretical foundations for more ecient proofs. Additionally, a loop-completed process structure tree will be derived to investigate loops and structures in unstructured regions [34] ...
... Such conditions usually address both the class of systems and the class of logs for which rediscoverability can be assured. For example, the Inductive Mining algorithm guarantees rediscoverability for the class of systems that are captured as block-structured process models [18] without duplicate actions and in which it is not possible to start a loop with an activity the same loop can also end with [16]. In contrast to rediscoverability guarantees, we study the problem of measuring how well a discovered model describes an unknown original system. ...
Process mining extracts value from the traces recorded in the event logs of IT-systems, with process discovery the task of inferring a process model for a log emitted by some unknown system. Generalization is one of the quality criteria applied to process models to quantify how well the model describes future executions of the system. Generalization is also perhaps the least understood of those criteria, with that lack primarily a consequence of it measuring properties over the entire future behavior of the system when the only available sample of behavior is that provided by the log. In this paper, we apply a bootstrap approach from computational statistics, allowing us to define an estimator of the model’s generalization based on the log it was discovered from. We show that standard process mining assumptions lead to a consistent estimator that makes fewer errors as the quality of the log increases. Experiments confirm the ability of the approach to support industry-scale data-driven systems engineering.
KeywordsProcess miningGeneralizationBootstrappingConsistent estimator
... To rediscover the true process, it imposes two requirements on an input event log: it should contain all directly-follows relations present in the true process, and the true process should be block-structured [20]. A similar argument holds for the Inductive Miner [15]. ...
A process discovery algorithm aims to construct a process model that represents the real-world process stored in event data well; it is precise, generalizes the data correctly, and is simple. At the same time, it is reasonable to expect that better quality input event data should lead to constructed process models of better quality. However, existing process discovery algorithms omit the discussion of this relationship between the inputs and outputs and, as it turns out, often do not guarantee it. We demonstrate the latter claim using several quality measures for event data and discovered process models. Consequently, this paper requests for more rigor in the design of process discovery algorithms, including properties that relate the qualities of the inputs and outputs of these algorithms. We present four incremental maturity stages for process discovery algorithms, along with concrete guidelines for formulating relevant properties and experimental validation. We then use these stages to review several state of the art process discovery algorithms to confirm the need to reflect on how we perform algorithmic process discovery.
... Such conditions usually address both the class of systems and the class of logs for which rediscoverability can be assured. For example, the Inductive Mining algorithm guarantees rediscoverability for the class of systems that are captured as block-structured process models [18] without duplicate actions and in which it is not possible to start a loop with an activity the same loop can also end with [16]. In contrast to rediscoverability guarantees, we study the problem of measuring how well a discovered model describes an unknown original system. ...
Process mining extracts value from the traces recorded in the event logs of IT-systems, with process discovery the task of inferring a process model for a log emitted by some unknown system. Generalization is one of the quality criteria applied to process models to quantify how well the model describes future executions of the system. Generalization is also perhaps the least understood of those criteria, with that lack primarily a consequence of it measuring properties over the entire future behavior of the system when the only available sample of behavior is that provided by the log. In this paper, we apply a bootstrap approach from computational statistics, allowing us to define an estimator of the model's generalization based on the log it was discovered from. We show that standard process mining assumptions lead to a consistent estimator that makes fewer errors as the quality of the log increases. Experiments confirm the ability of the approach to support industry-scale data-driven systems engineering.
... Among others, the following projects are based on jBPT. bpstruct 7 is a collection of techniques proposed in [10] for transforming unstructured process models into well-structured ones. BPM Academic Initiative 8 is an academic version of a professional business process management tool. ...
This paper presents the idea of a compendium of process technologies, i.e., a concise but comprehensive collection of techniques for process model analysis that support research on the design, execution, and evaluation of processes. The idea originated from observations on the evolution of process-related research disciplines. Based on these observations , we derive design goals for a compendium. Then, we present the jBPT library, which addresses these goals by means of an implementation of common analysis techniques in an open source codebase.
Robotic Process Automation (RPA) is an emerging automation technology that creates software (SW) robots to partially or fully automate rule-based and repetitive tasks (aka routines) previously performed by human users in their applications’ user interfaces (UIs). Successful usage of RPA requires strong support by skilled human experts, from the detection of the routines to be automated to the development of the executable scripts required to enact SW robots. In this paper, we discuss how process mining can be leveraged to minimize the manual and time-consuming steps required for the creation of SW robots, enabling new levels of automation and support for RPA. We first present a reference data model that can be used for a standardized specification of UI logs recording the interactions between workers and SW applications to enable interoperability among different tools. Then, we introduce a pipeline of processing steps that enable us to (1) semi-automatically discover the anatomy of a routine directly from the UI logs, and (2) automatically develop executable scripts for performing SW robots at run-time. We show how this pipeline can be effectively enacted by researchers/practitioners through the SmartRPA tool.KeywordsRobotic Process AutomationProcess miningUser Interface (UI) logsReference data model for UI logsSegmentationAutomated generation of SW robots from UI LogsSmartRPA
We present a query language, called QuBPAL, for retrieving knowledge from repositories of business processes represented in the BPAL (Business Process Abstract Language) framework. BPAL combines in a single modeling framework the procedural and the ontological aspects of business processes. This is done by providing a uniform, logic-based representation of both the workflow, with its associated procedural semantics, and the domain knowledge that captures the meaning of the entities participating in the process. This uniform representation is achieved by using Logic Programming (LP) as an intermediate language, to which we map BPMN models and OWL/RDF definitions. QuBPAL queries allow combining structural, behavioral, and domain-related knowledge and hence enable reasoning about the process from all these perspectives.
Synthesis of optimal business processes has practical applications in: manufacturing, scheduling, process mining, agent planning, and parallel computing. Block-structured models, in particular process trees, have certain advantages compared with other approaches regarding correctness and robustness. In this work we propose algorithms for the automated synthesis of optimal block-structured processes and then we perform a sound analysis of their correctness and complexity.
This paper addresses the problem of transforming a process model with an arbitrary topology into an equivalent well-structured process model. While this problem has received significant attention, there is still no full characterization of the class of unstructured process models that can be transformed into well-structured ones, nor an automated method to structure any process model that belongs to this class. This paper fills this gap in the context of acyclic process models. The paper defines a necessary and sufficient condition for an unstructured process model to have an equivalent structured model under fully concurrent bisimulation, as well as a complete structuring method.
This article addresses the problem of estimating the Quality of Service (QoS) of a composite service given the QoS of the services participating in the composition. Previous solutions to this problem impose restrictions on the topology of the orchestration models, limiting their applicability to well-structured orchestration models for example. This article lifts these restrictions by proposing a method for aggregate QoS computation that deals with more general types of unstructured orchestration models. The applicability and scalability of the proposed method are validated using a collection of models from industrial practice.