Enhancing CMMN: Conceptual Development of a Notational Variant for Case Management Modeling

Abstract

The Case Management Model and Notation (CMMN) supports the modeling of dynamic, semi-structured, and knowledge-intensive processes, but its adoption remains limited due to conceptual and visual shortcomings. Using a Design Science Research Method (DSRM), this study introduces a notational variant of CMMN, termed CMMN+, comprising three structural and visual enhancements: explicit representation of activation logic, enriched data entity modeling through semantically grounded metadata and structured role assignments based on the RACI framework. Cognitive effectiveness is analytically evaluated using the nine principles of the Physics of Notations (PoN). The analysis demonstrates clear improvements in semiotic clarity, semantic transparency and perceptual discriminability, confirming enhanced interpretability of the proposed notational variant. As expected, trade-offs arise with respect to graphic economy, while principles such as cognitive fit require subsequent empirical validation. CMMN+ constitutes a conceptually and technically grounded notational advancement in case management modeling by systematically aligning language design with cognitive-effectiveness theory. The presented results establish a strong foundation for integrating more intuitive and semantically rich modeling support into practice.

1. Introduction

Business processes are at the core of organizational efficiency and value creation [1]. As businesses increasingly move toward dynamic and flexible operations, traditional structured process modeling techniques, such as Business Process Model and Notation (BPMN) [2], often struggle to capture the complexity and variability inherent in unstructured or semi-structured processes [3]. In response to this challenge, the Case Management Model and Notation (CMMN) [4] was introduced as a complementary standard, offering greater adaptability for processes that require case-by-case decision-making. However, despite its potential, CMMN has not achieved widespread adoption in practice [5]. Concerns regarding its practical usefulness have also been raised in the practitioner-oriented literature; for example, in [6] authors note limited adoption of CMMN.

Despite this adaptability, CMMN exhibits several conceptual and practical weaknesses that limit its effective use in real-world modeling. In particular, users experience difficulties in interpreting activation logic, the notation provides only minimal support for representing and managing data, and the definition of roles and responsibilities remains insufficiently explicit. Together, these issues directly contribute to its limited adoption in practice. Addressing these shortcomings forms the core problem that this study seeks to address.

These shortcomings contribute directly to the perceived complexity of CMMN. As a result, one of the key reasons behind its limited use is the lack of intuitive visual cues to clearly understand activation logic, manage data, and define roles within the process [7]. By contrast, BPMN benefits from a more structured approach and clearer visual representation [1], supported by extensive industry-driven standardization efforts [8]. As a result, CMMN is often perceived as less accessible, particularly by novice users and by modelers unfamiliar with its flexible nature [9]. This discrepancy highlights the fundamental need to enhance the cognitive effectiveness of CMMN in fostering its broader acceptance in various organizational contexts [7].

To address these challenges, this study introduces structural and visual improvements to the CMMN. The aim is to mitigate the identified shortcomings by introducing more intuitive visual representations of activation logic, strengthening data support structures, and integrating a clear role assignment mechanism [7]. By focusing on cognitive effectiveness, according to Physics of Notations (PoN) theory [10], the proposed modifications seek to make CMMN models more approachable, thereby improving its conceptual clarity in business environments that require flexible process management.

The development of the proposed notational variant follows the Design Science Research Method (DSRM) [11], which supports a rigorous and iterative approach to innovation. The improvements were informed by a systematic literature review (SLR) [12] that identified specific gaps in the current use of CMMN [7]. This methodological foundation ensures that the extensions are conceptually well-grounded and systematically derived.

This paper contributes a conceptual notational variant that addresses key structural and visual shortcomings of CMMN modeling. The proposed enhancements—covering activation logic, data support, and role specification—aim to strengthen the cognitive effectiveness of the notation, thereby laying a foundation for broader applicability in dynamic business environments.

In light of these shortcomings, this study seeks to determine whether the proposed structural and visual enhancements improve the cognitive effectiveness of CMMN.

It is important to emphasize that CMMN+ is proposed as a CMMN-based notational variant rather than a standards-compliant extension. The contribution focuses on refining the visual and representational layer of CMMN, while preserving its original execution semantics and metamodel.

The remainder of this paper is structured as follows. Section 2 presents the theoretical background. Section 3 reviews related work and identifies research gaps. Section 4 outlines the methodological framework, followed by Section 5, which summarizes the key shortcomings of CMMN. Section 6 introduces the proposed structural and visual enhancements of the notational variant, and Section 7 presents their serialization and interchange mechanisms. Section 8 combines a scenario-based demonstration with an analytical evaluation. Section 9 discusses theoretical and practical implications, and Section 10 concludes the paper.

2. Theoretical Background

This section presents the theoretical foundations underlying the proposed notational variant. It introduces the framework for assessing cognitive effectiveness and summarizes the core concepts of Case Management and CMMN that form the domain context of this study.

2.1. Cognitive Effectiveness and Physics of Notations (PoN)

The cognitive effectiveness of visual modeling languages plays a central role in determining how accurately, efficiently, and reliably users can interpret and apply conceptual models. A rigorous theoretical foundation for the analysis and design of visual notations is provided by the Physics of Notations (PoN), proposed by Moody [10]. PoN defines cognitive effectiveness as the speed, ease, and accuracy with which information can be processed by the human mind and positions it as the primary criterion for evaluating visual notations.

PoN is grounded in theories from cognitive psychology, visual perception, semiotics, and communication. It explicitly distinguishes between the semantic representation of a modeling language (what concepts are represented) and its visual or concrete syntax (how these concepts are graphically encoded), emphasizing that visual representation constitutes an independent and critical design dimension rather than a purely esthetic concern. The theory formulates nine principles for cognitively effective visual notation design: semiotic clarity, perceptual discriminability, semantic transparency, complexity management, cognitive integration, visual expressiveness, dual coding, graphic economy, and cognitive fit. These principles address both the encoding of information into visual form and its decoding by the human perceptual and cognitive systems [10].

PoN can be applied analytically, to identify cognitive deficiencies in existing notations, as well as prescriptively, to guide the systematic design of new or extended notational systems. Its applicability and robustness have been demonstrated across a wide range of conceptual and process modeling languages (e.g., [13,14,15]). In the context of this study, PoN provides the primary theoretical foundation for the design and evaluation of the proposed CMMN-based notational variant.

2.2. Case Management and CMMN

Case Management was introduced in 2004 by van der Aalst, Weske, and Grünbauer as a new paradigm for business process support, initially referred to as Case Handling [3]. The concept was later further developed under the term Adaptive Case Management (ACM) [16,17]. It emerged in response to the limitations of traditional Business Process Management (BPM), which was shown to be too restrictive and insufficiently flexible for dynamic and knowledge-intensive environments [3,16]. In this context, ACM is referenced only as part of the historical and conceptual background leading to the development of CMMN.

The central concept of Case Management is the case, representing a real-world situation that evolves over time. Unlike traditional workflow management systems with strictly predefined execution paths, Case Management treats activities as flexible “chunks of work” that can be selected, executed, and adapted by knowledge workers [3]. Based on their understanding of the entire case context, knowledge workers decide which activities are relevant and when they should be performed, which constitutes one of the essential principles of Case Management [3].

The Case Management Model and Notation (CMMN) was standardized by the Object Management Group (OMG) [4] as a declarative modeling language for the specification and execution of case-based processes. It was conceived as a complementary notation to BPMN [2], specifically targeting flexible, weakly structured, and knowledge-intensive processes that cannot be fully captured using imperative control-flow semantics [18].

CMMN defines a set of core modeling constructs, including cases, tasks, stages, milestones, event listeners, sentries, and case file items, which together specify the structure, behavior, and data context of a case instance [4,7]. In contrast to workflow-oriented notations, the execution of case elements in CMMN is governed by declarative rules rather than by a fixed sequence of activities [18].

Despite its formal definition and conceptual expressiveness, the practical uptake of CMMN has remained limited. In the practitioner-oriented literature, concerns have been raised regarding its applicability in real-world settings. Freund and Ruecker report that due to limited adoption and perceived lack of practical usefulness, CMMN was omitted from the next edition of this book [6,19].

3. Related Work

This section builds on the findings of our previous study—a systematic literature review on CMMN [7], which analyzed both the theoretical foundations and practical applications of the notation. That study identified three recurring shortcomings of the standard CMMN: (i) the lack of an explicit and intuitive representation of activation logic, (ii) limited support for structured and semantically rich data modeling, and (iii) insufficient mechanisms for specifying roles and responsibilities. These shortcomings have also been recognized in the broader body of CMMN-related research, where various authors have proposed partial enhancements targeting specific aspects of the notation.

The following overview selectively highlights the most relevant notational enhancements and evaluation studies in order to position the proposed CMMN enhancements within the current state of the art.

Within this body of work, several previous studies have identified key shortcomings of CMMN and proposed various enhancements to improve its practical applicability. A notable contribution was presented in [20], who introduced the concept of entity life cycles to address the challenge of unclear task inputs, outputs, and performer assignments. They proposed linking life cycle phases with the states of informational entities and integrating a Unified Modeling Language (UML)-based information model to better represent the process structure and improve task allocation transparency. The authors demonstrated the application of their extension using an example related to academic conference submissions, illustrating its potential to improve process flexibility and stakeholder communication.

A related study [21] tackles the problem of task assignment in CMMN. Given that CMMN supports discretionary task execution based on the case context, it lacks explicit mechanisms for assigning responsibilities. The authors addressed this limitation by proposing a model that formalizes task–performer relationships, leveraging the structure of case information and role definitions to enable automated task recipient selection and visualized data flows between tasks.

In parallel with structural proposals, several studies have adopted an empirical approach to evaluate the usability of CMMN. For instance, Ref. [22] compared flexible BPMN with CMMN in real-world modeling scenarios and reported that CMMN is perceived as less accessible, particularly by users unfamiliar with declarative paradigms. These findings highlight the need for clearer semantics and more intuitive visual constructs for CMMN.

Whereas most existing work focuses directly on CMMN, insights can also be drawn from research on enhancing other modeling notations. For instance, Ref. [23] proposed a BPMN extension that introduced explicit role modeling into the standard notation, allowing for the representation of multiple actor types—human, software, and device—within a unified framework. Their RBPMN approach improves expressiveness in socio-technical systems and supports context-dependent variability in business process behavior.

Additional studies, including [22,24,25,26,27,28,29,30,31,32,33,34], contribute to partial improvements, such as refined data structuring or integrated process flow logic. However, these studies tend to focus on isolated aspects of the notation and often lack architectural coherence. While they offer valuable insights, most proposed enhancements remain localized in scope, and do not provide a systematic approach to strengthening the overall conceptual foundations of CMMN.

Beyond academic research, developments in CMMN are also shaped by ongoing standardization efforts of the OMG. Since its initial release in 2014 [4], CMMN has undergone minor revisions, but it has not achieved the same level of maturity or adoption as BPMN. Discussions within OMG working groups indicate that the standard is still evolving and that several of the identified shortcomings, such as limited data support, insufficient role specification, and unclear activation semantics, remain open challenges [32]. This highlights both the relevance and timeliness of research contributions that seek to conceptually strengthen the notation and provide a foundation for more coherent future developments.

Overall, existing research provides valuable partial improvements of CMMN modeling, but lacks a coherent, theory-driven approach to systematically strengthening the notation as a whole. To address this gap, this study follows DSRM for the development and evaluation of the proposed CMMN-based notational variant.

4. Methodological Framework

The proposed improvements to CMMN modeling were developed using the Design Science Research Method (DSRM). DSRM is a rigorous research approach specifically designed to address complex real-world problems through the systematic creation and evaluation of innovative artifacts. It provides a structured process for identifying problems, defining objectives, designing and developing solutions, demonstrating their use, evaluating their effectiveness, and communicating results. This makes DSRM particularly suitable for studies that focus on the development of modeling artifacts, methods, or frameworks in response to limitations observed in existing systems [11,35].

According to [11], DSRM consists of six main activities: (1) problem identification and motivation, (2) definition of the objectives of a solution, (3) design and development of the artifact, (4) demonstration of its use, (5) evaluation of the artifact, and (6) communication of the results. These activities are not strictly linear but can be carried out iteratively, allowing continuous refinement of the artifact and alignment with both practical requirements and theoretical contributions.

The choice of DSRM as the methodological framework for this study is grounded in its structured and iterative nature, which enables a systematic development process that balances scientific rigor with practical relevance. The methodology supports traceability between identified problems, design objectives, developed artifacts, and evaluation outcomes, which is essential for ensuring the validity and transparency of design-oriented research [11].

In this study, the DSRM process was applied iteratively through three design cycles, as illustrated in Figure 1. Each iteration followed the same methodological pattern, consisting of (1) the identification of a specific design problem, (2) the formulation of corresponding design objectives, (3) the design and development of a CMMN-based notational variant addressing these objectives, and (4) the demonstration and conceptual evaluation of the resulting artifact. The iterations were organized in a cumulative manner, such that each subsequent cycle builds on the results of the preceding one.

The evaluation in this study is primarily conceptual and analytical in nature and focuses on assessing the potential of the proposed notational enhancements to improve the clarity and usability of the CMMN-based models. The communication phase of DSRM is realized through this article, which documents the methodological process, the developed notational variant, and its conceptual evaluation for both the academic and practitioner communities.

5. Identified Shortcomings

The identification of key shortcomings of CMMN is based on a thorough examination of existing research and practical applications. These shortcomings were identified through an SLR, which was previously conducted and published as a separate study [7]. The insights from this review serve as a foundation for the development of the notational enhancements presented in this paper. The three primary shortcomings (S1, S2, and S3) are summarized below.

S1

Poor understanding of activation logic. One of the most commonly cited shortcomings is the lack of an intuitive and structured representation of activation logic. Unlike BPMN, which inherently supports explicit visual representations of process flow relationships [2], the flexible nature of CMMN makes it difficult to depict the progression of tasks as explicit relational structures, especially for novice users. This gap in visual representation results in lower cognitive effectiveness, making the CMMN-based models less suitable for business process modeling.

S2

Weak data support. Another major shortcoming is the limited support for representing and managing data within CMMN-based models. While the CMMN is inherently designed for flexibility, this focus on dynamic task management often overlooks the need for structured data handling, making it challenging to integrate data-driven decision-making processes into case models. This limitation reduces the practical applicability of CMMN-based modeling in data-intensive scenarios.

S3

Insufficient role definition. The existing CMMN modeling framework lacks a clear and systematic way to define roles and responsibilities within CMMN-based case models. This ambiguity can lead to inconsistencies in role assignment, particularly in complex cases involving multiple stakeholders. The absence of an integrated role assignment mechanism further complicates the adoption of CMMN-based modeling approaches in professional settings where role clarity is crucial.

Addressing these shortcomings is essential for enhancing the cognitive effectiveness of CMMN-based modeling. By systematically tackling these issues, the proposed improvements aim to make CMMN-based models more intuitive and practically applicable, particularly in dynamic business environments where adaptability and clarity are of primary importance.

6. Development of Structural and Visual Enhancements

To address the identified shortcomings in CMMN-based modeling, we developed a series of structural and visual enhancements aimed at increasing the cognitive effectiveness of CMMN-based models. These enhancements have been designed to bridge the gap between CMMN’s theoretical potential and its practical application. By focusing on more explicit activation logic, structured data support, and clear role definition, the proposed enhancements aim to make CMMN-based models more intuitive and accessible for both novice and experienced users.

In this section, we present the conceptual foundation of each enhancement, followed by a detailed description of the implementation process. We then demonstrate how these improvements address the specific challenges identified earlier, with examples illustrating the practical benefits of the proposed notational variant.

6.1. Visual Enhancement of Activation Logic Representation (R1)

One of the primary challenges identified in CMMN-based modeling is the lack of a clear and intuitive representation of activation logic, which makes it difficult for users to understand how tasks are triggered and how they progress within a case model. Unlike BPMN, which inherently supports structured workflows with explicit sequencing, the CMMN emphasizes flexibility and dynamic task execution based on declarative conditions. However, this flexibility often leads to confusion, particularly among novice users who expect a more structured and visually guided representation of task relationships and activation conditions [7,20,22,27].

An analysis of the existing CMMN metamodel segments revealed that the behavior of tasks and stages is governed by Sentry elements, which rely on event triggers (OnPart) and conditions (IfPart). These elements are linked to entry and exit criteria, graphically represented as diamond-shaped symbols attached to plan items [4]. Whereas this structure enables declarative modeling, its visual representation in CMMN-based models can be ambiguous for users unfamiliar with the underlying semantics.

To address this issue, we developed a visual enhancement (R1) that introduces explicit graphical elements to improve the representation of activation logic within CMMN-based models. It incorporates directional connectors that visually suggest the conditions under which tasks may become active. By doing so, we aim to reduce ambiguity and increase the interpretability of CMMN-based case models without altering their underlying declarative semantics.

The design of this enhancement is based on the PoN theory [10], particularly focusing on perceptual discriminability and semantic transparency. By using arrow styles, the proposed visual encoding enables users to immediately grasp the activation dependencies between activities without needing to analyze the entire model in detail.

An illustrative example of the enhanced activation logic representation is shown in Figure 2. The comparison between a standard CMMN model and the proposed CMMN-based notational variant (CMMN+) is intended to improve clarity and facilitate user comprehension. This improvement is particularly beneficial in case management scenarios, where the dynamic nature of processes requires flexible yet transparent visualization of task activation conditions.

R1 uses a single graphical element to represent both entry and exit criteria. This choice does not merge their semantics but treats them as two manifestations of the same underlying activation logic, with the distinction encoded relationally through connector direction and attachment points rather than separate symbol shapes. From a visual notation design perspective, this supports semiotic clarity by avoiding unnecessary symbol proliferation, while direction-based differentiation preserves semantic transparency and improves perceptual discriminability by making triggering and completion conditions directly observable from the model structure.

In summary, the visual enhancement (R1) aims to improve the cognitive effectiveness of CMMN-based models by providing a clearer and more intuitive representation of activation logic. By supporting users in understanding task activation conditions, this enhancement has the potential to facilitate better comprehension and decision-making in environments characterized by unstructured processes.

6.2. Strengthening Data Entity Support (R2)

A second shortcoming in CMMN-based modeling lies in weak data support. Whereas the CMMN specification includes the CaseFileItem elements to represent information relevant to a case, these elements are often treated superficially in practice and lack the structure and expressiveness necessary for complex, data-driven scenarios [27,28]. As a result, modelers face difficulties in effectively capturing data dependencies and relationships, which are essential in domains such as healthcare, legal casework, or insurance processing.

To establish a foundation for this enhancement, we analyzed the CaseFile package in the CMMN metamodel, which defines how data are structured and managed within CMMN-based case models. Central to this structure are the classes CaseFileItem and CaseFileItemDefinition. The former represents individual data items used during case execution, while the latter defines the data types to which these items refer. Each CaseFileItem instance is linked to exactly one definition, allowing for multiple instances to share a common semantic structure. The metamodel also defines several supporting attributes—such as multiplicity, sourceRef, targetRef, and definitionRef—that govern relationships between data instances and their roles in case execution.

Building on this foundation, we introduced structural enhancement R2 as part of the proposed notational variant, focusing on augmenting data representation at the metamodel level. This enhancement builds on the existing data constructs defined in the CaseFile package [4] and introduces a more explicit structure for defining data entities within CMMN-based models. By incorporating additional metadata, the Dublin Core Metadata Element Set (DCMES) [36] was employed to ensure consistency and semantic clarity. At the metamodel level, R2 introduces a set of additional, non-standard attributes for the CaseFileItem and CaseFileItemDefinition classes (see Figure 3), enabling a more precise and semantically grounded representation of data within the proposed notational variant. These include resourceState, which captures the current life cycle status of a data item using a compact, domain-agnostic enumeration; access, which defines access rights; and status, which indicates predefined conditions or classifications. In addition, optional metadata attributes, such as locationRef, author, version, and description, support improved data traceability and integration with external systems. Together, these metamodel-level augmentations provide a foundation for richer and more interpretable data representations in CMMN-based case models, particularly in data-intensive case management environments.

To complement the structural enhancements introduced as part of R2, a set of visual elements was defined to support the explicit representation of data semantics in CMMN-based models. These enhancements include dedicated icons for key metadata values such as resource state (e.g., draft, approved), access level (e.g., read-only, editable), and data type (e.g., document, folder, schema). The icons are assigned to instances of the CaseFileItem and CaseFileItemDefinition classes, as detailed in

Table 1. Selected CaseFileItem attributes with values and icons, as introduced in the proposed notational variant. Source: Adapted from CMMN 1.1 Specifications [4], with modifications by the authors.

Attribute	Value and Annotation	Icon
multiplicity MultiplicityEnum	ZeroOrOne (0..1)
	ZeroOrMore (0..*)
	ExactlyOne (1)
	OneOrMore (1..*)
	Unspecified
	Unknown
resourceState ResourceStateEnum	Draft
	InReview
	InApproval
	Approved
	Published
	Unpublished
	Archived
	Canceled
	Unspecified
	Unknown
access AccessEnum	Private_ReadOnly
	Private_Editable
	Public_ReadOnly
	Public_Editable
	Unspecified
	Unknown

and

Table 2. Selected CaseFileItemDefinition attributes with values and icons, as introduced in the proposed notational variant. Source: Adapted from CMMN 1.1 Specifications [4], with modifications by the authors.

Attribute	Value and Annotation	Icon
definitionType URI	Folder in CMIS
	Document in CMIS
	Relationship in CMIS
	XML-Schema
	Unspecified
	Unknown
status StatusEnum	Active	–
	Deprecated	–
	Archived	–
	Unspecified	–
	Unknown	–

, respectively. Each icon is accompanied by a corresponding annotation, reinforcing its semantic meaning and supporting the principle of dual coding as part of cognitive effectiveness. The annotation displays the value defined for the corresponding metadata attribute (e.g., Draft, ZeroOrOne (0..1)), thereby making the semantic content accessible in both graphical and textual form. Although the annotations are not explicitly shown in the tables, the listed values correspond to their textual representation in the notation. Designed in line with cognitive effectiveness principles—particularly perceptual discriminability and semantic transparency—these visual encodings are expected to support intuitive interpretation of the operational role and status of each data element. A visual example of this representation is shown in Figure 4.

In CMMN-based notational variant (CMMN+), data objects are not intrinsically classified as input or output artifacts. Instead, whether a data object is consumed or produced by a task is determined by the modeling context and the semantics of the task–data association, rather than by an explicit object-level attribute, which is outside the scope of the present enhancement.

Figure 5 illustrates a real-world example modeled using the proposed CMMN-based notational variant (CMMN+). The diagram depicts a case for patient data processing in a healthcare context, where tasks interact with semantically enriched data objects. Each CaseFileItem is annotated with metadata attributes—such as access rights, resource state, and data type—represented through intuitive visual symbols. This explicit visual encoding is designed to enhance the interpretability of data-related aspects in the model and to clarify how data are produced, consumed, and controlled throughout the process.

In addition to the structural and visual improvements, enhancement R2 is supported by a formal semantic specification that precisely defines the behavior of metadata-enriched data entities in models expressed using the proposed CMMN-based notational variant (CMMN+). A complete formal semantic specification of enhancement R2 is available online (https://github.com/matejabule/CMMNplus-formalSemantics, accessed on 12 December 2025) [37]. This specification formalizes key aspects such as access constraints, resource states, and multiplicity using logic-based rules that complement the visual notation without altering the underlying execution semantics of CMMN. Two representative fragments are presented below to illustrate the underlying semantics and provide insight into the modeling rationale.

The first example focuses on the newly introduced attribute resourceState, which defines the current life-cycle state of a data object. In the formal model, this attribute is represented as a function (Equation (1))

$$resourceState:F\to ResourceStateEnum,$$

(1)

where F is the set of all CaseFileItem instances. The function returns one of the predefined values, with Unspecified as the default.

The enumeration ResourceStateEnum includes the following values: Draft, InReview, InApproval, Approved, Published, Unpublished, Archived, Canceled, Unspecified, Unknown.

The second example focuses on the attribute access, which determines visibility and editability permissions for a data object. Formally, this attribute is defined as (Equation (2))

$$access:F\to AccessEnum,$$

(2)

where F is the set of CaseFileItem instances. The function returns one of the predefined values, with Unspecified as the default.

The enumeration AccessEnum includes the following values: Private_ReadOnly, Private_Editable, Public_ReadOnly, Public_Editable, Unspecified, Unknown.

The following rule defines access conditions for different user roles depending on the value of the access attribute. Let CFI be a CaseFileItem, U be a user, R denote the set of roles, and let hasRole $\subseteq U\times R$ denote the role-assignment relation. Access to CFI is permitted if one of the following holds (Equation (3)):

$$\begin{array}{cc}\hfill CFI.access& =Public\_ReadOnly\Rightarrow U\mathit{may}\mathit{read}CFI\hfill \\ \hfill CFI.access& =Public\_Editable\Rightarrow U\mathit{may}\mathit{read}\mathit{and}\mathit{write}CFI\hfill \\ \hfill CFI.access& =Private\_ReadOnly\wedge \exists r\in R:(U,r)\in hasRole\wedge r\in CFI.roles\hfill \\ \hfill & \Rightarrow U\mathit{may}\mathit{read}CFI\hfill \\ \hfill CFI.access& =Private\_Editable\wedge \exists r\in R:(U,r)\in hasRole\wedge r\in CFI.roles\hfill \\ \hfill & \Rightarrow U\mathit{may}\mathit{read}\mathit{and}\mathit{write}CFI\hfill \end{array}$$

(3)

This specification supports a consistent interpretation of access rights and supports consistent reasoning about user–data interactions based on assigned permissions.

In summary, enhancement R2 strengthens data-centric modeling in CMMN-based models expressed using the proposed notational variant (CMMN+) by enriching visual elements with semantically meaningful metadata and grounding them in a formal rule-based specification. Whereas this enhancement focuses on clarifying how data are structured and interpreted, the following enhancement (R3) addresses another aspect of case execution: the definition of roles and responsibilities.

6.3. Role Definition Enhancement (R3)

The third shortcoming identified in CMMN-based modeling is the insufficient support for defining roles and responsibilities in a clear and structured manner. Whereas the CMMN includes elements where roles are required, such as human tasks, it lacks a systematic way to assign, differentiate, and communicate responsibilities among multiple stakeholders involved in a case [20,27]. In practice, this can lead to ambiguity, overlapping responsibilities, and reduced accountability—especially in collaborative, multi-actor environments.

To establish a foundation for this enhancement, we analyzed how the Role class is used across different packages of the CMMN metamodel. In the Case package, roles are modeled as integral parts of a case and are associated with specific plan elements via composition relationships. Within the EventListener package, the UserEventListener class is directly linked to a Role, allowing certain user roles to trigger events in the case execution. Similarly, the PlanningTable package connects roles to DiscretionaryItems, enabling knowledge workers in specific roles to dynamically adapt the course of action by selecting optional elements from planning tables. Finally, in the Task package, particularly in the HumanTask class, roles are associated with tasks that require user execution. Each HumanTask can reference a role, specifying who is responsible for performing the task [4]. These structural insights provided a foundation for introducing a more explicit and semantically grounded approach to responsibility assignment within models, expressed using the proposed CMMN-based notational variant (CMMN+).

The structural enhancement introduced in R3 addresses the limitations in role modeling practices within CMMN-based models by enabling more precise and semantically rich responsibility assignments. At its core, the proposed notational variant introduces an explicit association construct, denoted as Responsibility (see Figure 6), which mediates the relationship between the Role concept and the elements it is associated with—namely HumanTask, UserEventListener, and TableItem. This construct includes five Boolean attributes: responsible, accountable, support, consulted, and informed, reflecting the widely adopted RACI (or RASCI—although the structural model supports the full RASCI responsibility matrix (Responsible, Accountable, Support, Consulted, Informed), we use the more commonly known term RACI throughout this paper for clarity and ease of reading) responsibility matrix [38].

The use of an association construct allows for the annotation of responsibilities directly on the relationship, rather than requiring the creation of new subclasses or auxiliary structures. This addresses a key limitation observed in CMMN-based role modeling practices, where roles are typically represented in a generic manner and lack the capacity to distinguish between different types of involvement in task execution [4].

This structural refinement lays the foundation for explicit and traceable responsibility definitions and is designed to support both formal interpretation and visual representation. It also replaces the previously informal practice of using textual annotations to indicate responsibilities, thereby improving the precision and maintainability of CMMN-based models expressed using the proposed notational variant.

It is important to note that this enhancement is not intended to collapse responsibility information into a single graphical shape within the CMMN diagram. Instead, the responsibility matrix provides a dedicated, model-integrated view that complements the case diagram. Unlike an external spreadsheet, the matrix is formally linked to the underlying CMMN-based notational variant (CMMN+) and its elements, and is therefore subject to the same consistency rules and validation mechanisms as the rest of the model. This design preserves the visual clarity of the case diagram while enabling an explicit and analyzable representation of organizational responsibilities at the notational level.

To support the practical use of this structural enhancement, a corresponding visual notation improvement was introduced as part of the proposed notational variant. Each type of RACI responsibility is represented using standardized abbreviations (R, A, S, C, I) placed within a responsibility matrix. In this matrix, rows correspond to case plan elements (e.g., tasks), while columns denote roles. This tabular layout complements the graphical model by explicitly specifying which responsibilities are associated with each role for a given activity. This tabular view, illustrated in Figure 7, provides a compact and structured overview of stakeholder involvement across activities. The matrix is directly linked to graphical elements in models expressed using the proposed CMMN-based notational variant (CMMN+), ensuring consistency between textual responsibility assignments and their corresponding visual elements. Designed in line with cognitive effectiveness principles, this combined representation enhances semantic transparency and minimizes interpretative effort. It enables users to quickly understand role assignments in terms of responsibility, accountability, support, consultation, and information for a given task, thereby improving model interpretability, governance compliance, and team coordination in collaborative case environments.

In addition to its visual integration, enhancement R3 is grounded in a formal semantic specification that defines how roles and responsibilities are modeled and interpreted in models expressed using the proposed CMMN-based notational variant (CMMN+). A complete formalization is available online (https://github.com/matejabule/CMMNplus-formalSemantics, accessed on 12 December 2025) [37], detailing the constraints and logic behind role assignments, task involvement, and the consistent application of the RACI matrix. This formal layer complements the underlying execution semantics of CMMN by ensuring unambiguous interpretation of role-related metadata and supporting model validation, especially in collaborative scenarios where a clear delineation of responsibilities is critical. To illustrate the semantic underpinnings of this enhancement, one representative rule is presented below.

In the following, we define an accountability uniqueness constraint that can be used to express the RACI principle of single-point accountability.

Let T be the set of tasks, U the set of users (actors), and $RACI=\{R,A,S,C,I\}$ the set of role types corresponding to Responsible, Accountable, Supporting, Consulted, and Informed. Define the relation (Equation (4))

$$\mathit{assignedAs}\subseteq T\times U\times RACI$$

(4)

such that $(t,u,r)\in \mathit{assignedAs}$ denotes that user u is assigned to task t with role type r. The following constraint represents an optional modeling assumption that can be checked over models expressed using the proposed notation. For a task t ∈ T, the accountability uniqueness constraint holds if and only if (Equation (5))

$$\begin{array}{cc}\hfill & \forall t\in T:\hfill \\ \hfill & \exists !u\in U\mathrm{such}\mathrm{that}(t,u,A)\in assignedAs,\hfill \\ \hfill & \forall r\in \{R,S,C,I\}\exists u_1,\dots ,u_k(k\ge 1)\mathrm{such}\mathrm{that}(t,u_i,r)\in assignedAs.\hfill \end{array}$$

(5)

The second part of the constraint (of Equation (5)) requires that, for each task, at least one user is assigned to each of the remaining RACI roles (Responsible, Supporting, Consulted, and Informed). This requirement reflects a modeling assumption aimed at ensuring explicit responsibility and communication coverage for every task. By enforcing the presence of these role assignments, the model avoids underspecified task responsibilities and enables systematic analysis and consistency checking of role distributions within the case model. While in practice some of these roles may be optional, their explicit inclusion in the formal model makes implicit organizational assumptions explicit and analyzable.

This constraint ensures the uniqueness of the Accountable role per task, reflecting the organizational principle that responsibility for decision-making must be clearly assigned. It also supports consistency checks across models expressed using the proposed notational variant, thereby reinforcing the clarity and accountability objectives of enhancement R3.

To summarize, role definition enhancement R3 equips CMMN-based models expressed using the proposed notational variant with a clear and semantically rich mechanism for modeling roles and responsibilities. By incorporating RACI semantics into the metamodel-level constructs and the visual notation used by the variant, the enhancement promotes transparency, accountability, and role clarity—key requirements in collaborative, real-world case management scenarios. This enhancement aligns with established organizational practices and has the potential to strengthen the support for structured coordination and effective decision-making in knowledge-intensive environments when using CMMN-based modeling approaches.

7. Interchange and Serialization of the Proposed Notational Variant

To support the practical use of the proposed notational variant, it is essential to provide an interchange format that enables model exchange and potential tool integration. While the conceptual and notational enhancements described in the previous section address shortcomings observed in CMMN-based modeling, their adoption in practice requires a clear mapping to an XML-based representation. For this purpose, a dedicated XML Schema Definition (XSD) has been developed and is available in the accompanying repository (https://github.com/matejabule/CMMNplus-serialization, accessed on 12 December 2025) under the namespace cmmnplus [39]. The schema leverages the extension mechanisms provided by the official CMMN 1.1 XML Schema by means of schema import and type extension, without modifying any original CMMN declarations. This ensures compatibility with existing CMMN tooling, whereby models can be parsed even if the additional elements introduced by the notational variant are not fully supported. This section presents the structure of the schema, outlines the mapping between the conceptual notational enhancements and their XML representations, and provides examples demonstrating how models expressed using the proposed notational variant can be serialized in a tool-independent manner.

7.1. Schema Overview

The XML schema provides two complementary schema-level extension mechanisms that support the structural and notational enhancements introduced in Section 6, namely R2 and R3. These mechanisms are realized using the standard XML Schema extension facilities defined by the CMMN specification and operate entirely within a separate namespace (cmmnplus), thereby preserving compatibility with existing CMMN models and tools. First, attribute-level metadata and semantic qualifiers are introduced for existing CMMN elements, specifically:

cmmn:CaseFileItem;

cmmn:CaseFileItemDefinition.

This is achieved through extended complex types defined in the cmmnplus namespace:

cmmnplus:tCaseFileItemPlus;

cmmnplus:tCaseFileItemDefinitionPlus;

which are activated using the standard xsi:type mechanism. For CaseFileItem, the extended type introduces additional, non-standard metadata attributes, including:

cmmnplus:resourceState (ResourceStateEnum);

cmmnplus:locationRef (xs:anyURI);

cmmnplus:access (AccessEnum);

cmmnplus:author;

cmmnplus:version.

Similarly, the extended type for CaseFileItemDefinition provides additional descriptive metadata:

cmmnplus:description;

cmmnplus:status (StatusEnum);

cmmnplus:version.

Second, responsibility assignments corresponding to enhancement R3 are represented using dedicated elements defined in the

cmmnplus:respContainer;

cmmnplus:responsibility.

These elements can be embedded into selected CMMN elements—namely HumanTask, UserEventListener, and TableItem—via extended types:

cmmnplus:tHumanTaskPlus;

cmmnplus:tUserEventListenerPlus;

cmmnplus:tTableItemPlus.

The proposed CMMN-based notational variant follows an explicit opt-in principle. As long as the CMMN+ variant is not used, models remain fully valid with respect to standard CMMN. When the notational variant is applied, instances are interpreted according to the constraints defined for CMMN+, including the provision of all required attributes.

7.2. Mapping of Concepts to XML

Table 3. Mapping of R2-related data modeling elements of the proposed CMMN-based notational variant associated with CaseFileItem (CFI) to XML schema elements and attributes. Source: Adapted from CMMN 1.1 Specifications [4], with modifications by the authors.

Metamodel	XSD Element/Attribute
CSI resourceState	<cmmn:caseFileItem xsi:type="cmmnplus:tCaseFileItemPlus" cmmnplus:resourceState={Draft, InReview, InApproval, Approved, Published, Unpublished, Archived, Canceled, Unspecified, Unknown}>
CSI locationRef	<cmmn:caseFileItem xsi:type="cmmnplus:tCaseFileItemPlus" cmmnplus:locationRef="anyURI">
CSI access	<cmmn:caseFileItem xsi:type="cmmnplus:tCaseFileItemPlus" cmmnplus:access={Private_ReadOnly, Private_Editable, Public_ReadOnly, Public_Editable, Unspecified, Unknown}>
CSI author	<cmmn:caseFileItem xsi:type="cmmnplus:tCaseFileItemPlus" cmmnplus:author="string">
CSI version	<cmmn:caseFileItem xsi:type="cmmnplus:tCaseFileItemPlus" cmmnplus:version="string">

,

Table 4. Mapping of R2-related data modeling elements of the proposed CMMN-based notational variant associated with CaseFileItemDefinition (CFID) to XML schema elements and attributes. Source: Authors’ own work.

Metamodel	XSD Element/Attribute
CSID description	<cmmn:caseFileItemDefinition xsi:type="cmmnplus:tCaseFileItemDefinitionPlus" cmmnplus:description="string">
CSID status	<cmmn:caseFileItemDefinition xsi:type="cmmnplus:tCaseFileItemDefinitionPlus" cmmnplus:status=Active, Deprecated, Archived, Unspecified, Unknown>
CSID version	<cmmn:caseFileItemDefinition xsi:type="cmmnplus:tCaseFileItemDefinitionPlus" cmmnplus:version="string">

and

Table 5. Mapping of R3-related role and responsibility modeling elements to XML schema elements and attributes in the proposed notational variant. Source: Authors’ own work.

Metamodel	XSD Element/Attribute
Role assignment	<cmmnplus:respContainer>   <cmmnplus:responsibility     cmmnplus:responsible="true"     cmmnplus:accountable="false"     cmmnplus:support="false"     cmmnplus:consulted="false"     cmmnplus:informed="false"/> </cmmnplus:respContainer>

summarize the mapping between the conceptual elements of the proposed CMMN-based notational variant and their corresponding XML representations defined in the cmmnplus XSD. This overview illustrates how the conceptual and notational enhancements introduced in Section 6 are reflected at the serialization level and clarifies the opt-in nature of the approach—namely, that standard CMMN instances remain fully valid unless constructs of the notational variant are used.

7.3. XML-Based Technical Validation

To demonstrate practical usage, the following XML snippets illustrate how the conceptual and notational enhancements introduced by the proposed CMMN-based notational variant are represented at the instance level through XML-based serialization constructs.

• <cmmn:caseFileItem
  •  id="cfi1"
  •  definitionRef="cfd1"
  •  xsi:type="cmmnplus:tCaseFileItemPlus"
  •  cmmnplus:resourceState="InReview"
  •  cmmnplus:access="Public_ReadOnly"
  •  cmmnplus:author="John"
  •  cmmnplus:version="1.0.1"
  •  cmmnplus:locationRef="https://example.org/docs/123"/>

• <cmmn:caseFileItemDefinition
  •  id="cfd1"
  •  name="DocumentType"
  •  xsi:type="cmmnplus:tCaseFileItemDefinitionPlus"
  •  cmmnplus:status="Active"
  •  cmmnplus:version="1.0.0"
  •  cmmnplus:description="Definition of a lab document type."/>

• <cmmn:humanTask id="ht1" name="Review document"
•          xsi:type="cmmnplus:tHumanTaskPlus">
•  <cmmnplus:respContainer name="RACI for Review">
•   <cmmnplus:responsibility
•     cmmnplus:responsible="true"
•     cmmnplus:accountable="false"
•     cmmnplus:consulted="false"
•     cmmnplus:informed="false"/>
•  </cmmnplus:respContainer>
• </cmmn:humanTask>

In summary, the proposed XML schema enables the consistent serialization of models expressed using the proposed CMMN-based notational variant, while preserving the validity of standard CMMN constructs. By providing a well-defined serialization format for the additional notational elements, the approach supports technical interoperability and lays the groundwork for future tool support without imposing requirements on existing CMMN-compliant tools. Having established a concrete and standards-aligned interchange representation, the following section illustrates how the notational enhancements can be applied in a realistic case scenario.

8. Demonstration and Evaluation

To illustrate the practical applicability of the models expressed using the proposed CMMN-based notational variant (CMMN+) and to assess their potential cognitive effectiveness, this section integrates a scenario-based demonstration with an analytical evaluation. Together, these perspectives provide a structured view of how the proposed notational enhancements function in practice and how they align with established principles of visual notation design.

8.1. Scenario-Based Demonstration

To demonstrate the practical implications of the proposed notational enhancements, a healthcare scenario addressing patient management tasks such as diagnostics, treatment planning, and follow-up care is used. This domain provides a representative example of a knowledge-intensive, data-rich, and collaboratively executed case process—an environment in which limitations of CMMN-based modeling practices are frequently observed.

The illustrated scenario models a diagnostic stage in which a patient undergoes a series of medical examinations and assessments. Within this stage, the execution of diagnostic and organ function tests is determined by the current case context, and their outcomes condition the assessment of the severity of the patient’s condition. These tasks update key data objects within the case model, which in turn influence clinical decision-making and coordinate the involvement of different medical roles.

In order to explicitly illustrate the contribution of the proposed notational enhancements, the scenario is first modeled using the standard CMMN and subsequently using a model expressed with the proposed CMMN-based notational variant (CMMN+). This side-by-side demonstration enables a direct visual and conceptual comparison between the baseline representation and the enhanced model, serving as a qualitative illustration of the potential benefits introduced by the notational variant.

Figure 8 presents the baseline representation of this scenario modeled using the standard CMMN. In this representation, task activation is governed implicitly by entry and exit criteria, which are not visually emphasized. Case data are represented through CaseFileItems elements without additional semantic information, and roles are treated generically without explicit responsibility differentiation. As a result, the interpretation of task dependencies, data usage, and actor responsibilities relies largely on the reader’s familiarity with CMMN semantics rather than on explicit visual cues in the model.

Figure 9 shows the same healthcare scenario modeled using a model expressed with the proposed CMMN-based notational variant (CMMN+). The activation logic between tasks—such as “Diagnostic tests”, “Organ function tests”, and “Assessment of the severity of the condition”—is made visually explicit through directional connectors introduced in R1, thereby reducing ambiguity in the visual interpretation of Sentry-based activation conditions. Each data object, including “Medical Records”, “Laboratory results”, and “Diagnostic results”, is annotated with metadata specifying access level, resource state, and data type, as introduced in R2. In addition, responsibilities for each activity are defined using a RACI-based representation (R3), assigning roles such as “Medical Staff”, “Professional Staff”, and “Doctor” according to their involvement in the corresponding tasks. Together, these elements illustrate how the proposed notational variant enables a more explicit, semantically enriched, and operationally transparent representation of diagnostic case behavior when applied to CMMN-based models.

Building on this, the integrated example demonstrates how the combined use of all three notational enhancements supports a cohesive modeling approach that simultaneously addresses activation logic, data semantics, and role allocation. The scenario serves primarily as an illustrative demonstration and highlights the potential of models expressed using the proposed CMMN-based notational variant (CMMN+) to better accommodate the needs of complex, collaborative, and information-intensive case management settings.

A direct comparison between the baseline and the enhanced models shows that, while the standard CMMN representation requires implicit reasoning about task activation, data usage, and actor responsibilities, these aspects are made explicit in models expressed using the proposed CMMN-based notational variant (CMMN+). By jointly addressing activation logic, data semantics, and role allocation, the proposed notational enhancements support a more transparent and interpretable representation of the healthcare case. The scenario thus establishes a concrete basis for the analytical evaluation presented in the following subsection.

8.2. Analytical Evaluation

Building on the scenario-based demonstration, this subsection evaluates the cognitive effectiveness of models expressed using the proposed CMMN-based notational variant (CMMN+) by applying the principles of the PoN [10]. This evaluation is conducted as an analytical, expert-based assessment and does not involve an empirical user study or any form of participant-based evaluation. It draws on Moody’s framework, which provides a basis for assessing the cognitive quality of visual modeling languages. Several studies have already applied PoN to assess or improve the cognitive quality of visual notations in practice—for example, in the analysis of the WebML notation [40], in empirical investigations of BPMN diagram comprehension [41], in experimental evaluations of high-level process diagrams [42], and in studies examining cognitive effectiveness across conceptual modeling languages [43].

In addition, several studies have operationalized PoN using structured evaluation procedures and quantitative metrics. These include analyses of BPMN’s notation design [14] and evaluations of UCM and related modeling languages [44], as well as empirical validation studies demonstrating improvements in cognitive effectiveness following PoN-informed notation redesigns [45].

Before conducting the PoN-based analysis, it is important to emphasize that the standard CMMN models and the corresponding models expressed using the proposed CMMN-based notational variant (CMMN+) are informationally equivalent, as they encode the same underlying case semantics. The proposed notational enhancements do not introduce additional functional expressiveness, but rather reorganize and enrich the visual encoding of existing information. Consequently, any observed differences in cognitive effectiveness can be attributed to changes in visual representation rather than to differences in the underlying informational content.

Against this background, the analysis aims to determine whether the notational enhancements introduced in Section 6 have the potential to improve the cognitive effectiveness of models expressed using the proposed CMMN-based notational variant. The evaluation focuses in particular on two principles—semiotic clarity and semantic transparency—which are examined in depth for each individual enhancement (R1, R2, R3). The remaining principles are assessed holistically in order the overall cognitive alignment of the notational variant with PoN principles (see

Table 6. Analytical evaluation of the proposed CMMN-based notational variant (CMMN+). Source: Authors’ own work.

Principle PoN	Evaluation Summary
Semiotic clarity	(Enhancement level—R1, R2, R3) The proposed CMMN-based notational variant affects semiotic clarity in different ways. R1 replaces the visual representation of EntryCriterion and ExitCriterion without altering their underlying semantics and therefore does not affect the one-to-one mapping between concepts and symbols. R3 improves semiotic clarity by introducing an explicit graphical symbol for the existing Role concept, thereby resolving a missing-symbol violation present in standard CMMN. In contrast, R2 introduces additional icons to CaseFileItem that do not correspond to separate semantic constructs. While this constitutes redundant mapping from a strict semiotic perspective, these markers function as informative visual cues that support interpretability despite formally reducing one-to-one correspondence.
Perceptual discriminability	The analysis of the proposed CMMN-based notational variant indicates an improved level of perceptual discriminability among its symbols compared to standard CMMN (0.98 vs. 0.95). The enhanced shapes and clearer visual cues introduced by the notational enhancements further minimize perceptual confusion and support faster, more accurate symbol recognition.
Semantic transparency	The proposed CMMN-based notational variant exhibits a higher level of semantic transparency compared to standard CMMN (0.53 vs. 0.38). The redesigned and extended symbols convey their intended meaning more directly, resulting in a higher proportion of immediately and intuitively interpretable constructs.
Complexity management	Standard CMMN and the proposed CMMN-based notational variant provide a comparable level of complexity management, as both rely on the same core mechanisms of modularity, hierarchy, and abstraction. While the notational variant introduces clearer and more explicit modeling constructs, these enhancements do not fundamentally extend the structural means available for managing complexity. Consequently, observed improvements are modest and primarily related to increased clarity rather than to enhanced complexity-handling capability.
Cognitive integration	The proposed CMMN-based notational variant exhibits a modest but meaningful improvement in cognitive integration across process, data, organizational, and rule-related perspectives compared to standard CMMN (0.29 vs. 0.26). The introduced notational enhancements provide more explicit and semantically clearer relationships between modeling concepts, supporting a higher degree of qualitative cohesion within the notation.
Visual expressiveness	Standard CMMN and the proposed CMMN-based notational variant exhibit a comparable level of visual expressiveness, as both rely on essentially the same set of visual variables, including shape, line style, and iconography. While the notational variant introduces additional symbols, these enhancements do not extend the underlying palette of graphical dimensions and therefore do not substantially increase the expressive capacity of the notation. Consequently, the level of visual expressiveness remains broadly aligned with that of standard CMMN.
Dual coding	Standard CMMN exhibits a moderate level of dual coding (0.69), reflecting a relatively consistent combination of textual and graphical cues across most elements. The proposed CMMN-based notational variant reaches a comparable but slightly lower value (0.47), as the notational enhancements extend the symbol set without substantially altering naming conventions. This indicates that the added constructs do not increase the degree of textual–graphical integration. Consequently, both notations rely primarily on visual forms for conveying meaning, with limited systematic use of textual labels.
Graphic economy	Standard CMMN exhibits a relatively high level of graphic economy at the level of its core notation, relying on a compact set of symbols to express its basic semantics. With the introduction of the proposed CMMN-based notational variant, the number of graphical constructs is moderately increased in order to externalize information that is only implicitly or informally represented in standard CMMN (e.g., through textual annotations or comments). Under the assumption of informational equivalence between standard CMMN and models expressed using the notational variant, this change does not introduce additional information but instead redistributes existing semantics into explicit, first-class visual constructs. Consequently, the observed reduction in graphic economy reflects a deliberate trade-off between visual vocabulary size and semantic explicitness rather than a loss of notational efficiency.
Cognitive fit	This principle depends on empirical validation with representative users and concrete task contexts and therefore cannot be meaningfully assessed through analytical evaluation alone.

).

The evaluation was conducted as an expert-based qualitative analysis. No empirical data were collected, and no participants were involved; instead, the assessment relies on the systematic application of PoN principles to the visual characteristics of the proposed notational variant. Each of the nine PoN principles was applied as an analytical criterion to assess the potential cognitive effectiveness of models expressed using the proposed CMMN-based notational variant. The analysis combined a detailed symbol-by-symbol assessment (for principles such as perceptual discriminability and semantic transparency) with a conceptual and structural examination of the metamodel-level constructs introduced by the notational enhancements (for principles such as semiotic clarity and complexity management). Where possible, results were supported with quantitative indicators derived from systematic rating procedures. For principles that inherently require empirical user validation, such as cognitive fit, only an analytical justification was provided, and no formal evaluation was performed. The analysis is summarized in Table 6, where the reported values represent qualitative, ordinal judgments resulting from this analytical assessment and are intended to support comparative interpretation rather than to report measured empirical effects.

The detailed analytical procedures underlying the evaluation—including rating matrices, symbol-to-concept mappings, and calculation steps for individual PoN principles—are available upon request. Overall, the findings indicate that, from a theoretical and analytical perspective, the proposed CMMN-based notational variant (CMMN+) has the potential to improve several dimensions of cognitive effectiveness while remaining aligned with the design foundations of CMMN. At the same time, these theoretically grounded advantages call for complementary empirical validation to assess their impact in real modeling practice. The following section discusses these implications in depth, reflecting on the strengths, limitations, and practical relevance of the proposed notational variant.

9. Discussion

The proposed CMMN-based notational variant addresses the identified shortcomings of standard CMMN through targeted structural and visual enhancements. By reorganizing and enriching the visual representation of existing case semantics, the notational variant has the potential to improve cognitive effectiveness and conceptual clarity when modeling dynamic, knowledge-intensive processes. The design of these enhancements is grounded in the principles of cognitive effectiveness as defined by the PoN theory [10], aiming to ensure that the visual syntax better supports the understanding and interpretation of case behavior. This section discusses the broader implications of the proposed notational variant, contextualizing its conceptual contributions, language-level implications, analytical evaluation outcomes, methodological limitations, and directions for further development.

9.1. Conceptual Positioning

Although this study focuses on the development of a CMMN-based notational variant, it is important to position CMMN+ relative to other established process and case modeling approaches. Procedural notations such as BPMN provide a flow-oriented representation that is highly effective for structured and predictable processes, but they offer limited support for flexible, data-driven casework. In contrast, declarative modeling approaches emphasize constraint-based execution semantics rather than predefined control flows, which can be well-suited for rule-intensive or highly variable scenarios, but may be less intuitive for practitioners who rely on explicit visual cues to reason about case progression.

From a broader business process modeling perspective, BPMN, CMMN, and DMN address complementary aspects of organizational behavior. BPMN primarily targets structured, repeatable, and well-defined process flows, whereas CMMN supports less structured, knowledge-intensive, and situation-driven case work. DMN, in turn, provides a formalism for modeling decision logic that can be integrated with both procedural and case-based processes. Together, these standards form a methodological ecosystem for capturing structured processes, unstructured or semi-structured case work, and decision-making logic within a unified modeling framework.

The proposed CMMN-based notational variant occupies a conceptual space between these paradigms. It retains the flexibility inherent to case-based modeling while systematically addressing several limitations of standard CMMN, in particular with respect to activation logic, data semantics, and the explicit specification of responsibilities. In this way, the notational variant complements rather than replaces existing procedural modeling notations, strengthening CMMN’s role as a semi-structured approach for dynamic, knowledge-intensive environments where both flexibility and visual clarity are essential. Beyond this conceptual positioning, the proposed notational variant introduces systematic changes at the level of visual and notational design.

9.2. Language-Level Implications of the Proposed Notational Variant

The proposed CMMN-based notational variant has important implications for how language constructs are used and interpreted at the modeling level. The three notational enhancements (R1, R2, R3) affect the visual articulation of execution semantics, data-related information, and organizational responsibilities within CMMN models by making several aspects that were previously implicit or informally documented more explicit and systematically represented. These changes do not introduce new execution semantics but reorganize the visual and structural representation of existing concepts to support clearer interpretation and analysis.

With R1, task activation semantics—which in standard CMMN are expressed implicitly through EntryCriteria and ExitCriteria elements—are made visually explicit through directional connectors. This notational enhancement does not introduce new modeling entities nor modify the underlying execution semantics, but elevates activation relationships to first-class visual constructs. As a result, enabling conditions become directly observable in the diagram without increasing the functional expressiveness of the notation.

A more substantial impact on language use and interpretation is achieved through R2, which semantically enriches existing first-class data entities at the notational level. In standard CMMN, CaseFileItems elements and their definitions are formally part of the metamodel but remain semantically under-specified in their visual representation. The R2 notational enhancement augments these elements with explicit metadata capturing resource state, access rights, and data type. This enhancement does not introduce new data concepts; rather, it reorganizes and exposes existing structural information in a way that makes data objects more semantically transparent. As a result, data dependencies, life-cycle states, and access constraints become more cognitively accessible and formally analyzable within the notation itself.

The most substantial impact on organizational modeling is achieved through R3. While roles are conceptually present in standard CMMN, they are not supported as explicit graphical constructs and are typically documented only through informal annotations. The R3 notational enhancement introduces an explicit responsibility association with integrated RACI semantics, making role involvement visually and structurally explicit within CMMN-based models. Responsibilities are represented as first-class notational constructs with dedicated visual encodings and corresponding serialization support. This enables systematic validation of responsibility assignments, supports automated consistency checks at the model level, and strengthens the organizational dimension of case models, particularly in collaborative and multi-actor settings.

The validation of responsibility assignments is performed at the modeling level by tools that interpret models expressed using the proposed CMMN-based notational variant. Since roles and responsibilities are made explicit through dedicated notational constructs and corresponding serialization support, validation rules can be applied during model creation or analysis to detect missing, conflicting, or inconsistent assignments (e.g., violations of accountability uniqueness or incomplete RACI coverage). While the present work focuses on the design and analytical evaluation of the notational variant, such validation can be realized using standard model validation mechanisms provided by CMMN-compliant modeling environments.

Taken together, these notational enhancements systematically transform the way CMMN models articulate execution logic, data-related information, and organizational responsibilities. Aspects that are partly implicit or informally documented in standard CMMN are made explicit and systematically represented at the modeling level, supporting clearer interpretation and analysis. This establishes a solid foundation for tool support, automated consistency checks, and organizational accountability, while preserving the underlying case semantics of the original CMMN.

9.3. Interpretation of Evaluation Results

The analytical evaluation provides several insights into the cognitive implications of the proposed notational enhancements. The most substantial improvements emerge in the areas of semiotic clarity, semantic transparency, and perceptual discriminability, where redesigned symbols and additional visual cues reduce ambiguity and support more immediate interpretation. These effects are consistent with prior studies demonstrating that even relatively small refinements in visual notation design can significantly enhance the comprehensibility of process models.

At the same time, the evaluation shows that not all PoN principles are equally affected. Improvements in cognitive integration and complexity management are observed but modest, reflecting that the notational enhancements refine the visual articulation of existing modeling concepts rather than fundamentally restructuring CMMN’s underlying modeling mechanisms. Similarly, graphic economy is reduced due to the introduction of additional icons, reflecting a trade-off between visual vocabulary size and semantic expressiveness.

9.4. Limitations of the Analytical Evaluation

While the analytical evaluation provides valuable initial insights into the cognitive effectiveness of models expressed using the proposed CMMN-based notational variant (CMMN+), it also entails several limitations that should be acknowledged. PoN-based assessments support systematic reasoning about visual syntax, but they cannot capture how users actually interpret, comprehend, or construct models in real modeling situations. As a result, important dimensions such as comprehension time, error rates, modeling performance, or user preferences remain unexamined.

Additionally, some PoN principles—most notably cognitive fit—cannot be meaningfully assessed without empirical studies involving representative user groups and concrete task scenarios. The present evaluation therefore remains limited to what can be inferred from expert judgment, conceptual inspection, and symbol-level analysis. Although such analytical assessments are a well-established first step in notation design, they should be complemented by empirical validation once tool support and stable prototypes are available.

9.5. Trade-Offs and Remaining Limitations of the Proposed Notational Variant

Although the proposed notational enhancements improve several aspects of cognitive effectiveness, they also entail trade-offs that merit careful consideration. The additional icons introduced in R2 increase the semantic transparency of CaseFileItems representations, but they do so at the cost of a more redundant symbol–concept mapping. This trade-off reflects a common tension in visual notation design, where greater semantic explicitness can reduce strict semiotic clarity while still supporting improved interpretability in practice modeling contexts.

A related effect is observed with respect to graphic economy. Models expressed using the proposed CMMN-based notational variant introduces a slightly larger visual vocabulary to support the notational enhancements, which increases the number of graphical constructs that modelers must recognize. While the additional symbols remain conceptually coherent and visually consistent, they nonetheless expand the overall palette of graphical elements required for modeling. Importantly, this increase in visual vocabulary does not represent additional informational content, as standard CMMN models and models expressed using the notational variant remain informationally equivalent; rather, it results from a deliberate redistribution of existing semantics into more a more explicit visual form.

Improvements in complexity management and cognitive integration remain relatively modest. This outcome is expected, as the proposed notational enhancements refine rather than fundamentally restructure the underlying modeling mechanisms of standard CMMN. The notation continues to rely on modularity, decomposition, and declarative activation logic, and the enhancements primarily clarify the visual articulation of these mechanisms rather fundamentally. Consequently, certain structural limitations inherent to CMMN persist, particularly in scenarios requiring highly declarative, highly procedural, or multi-layered perspectives.

These trade-offs do not diminish the value of the proposed notational enhancements; rather, they highlight the design choices involved in balancing expressiveness, clarity, and notational economy. Explicitly acknowledging these limitations provides useful guidance for future refinements and empirical investigation.

9.6. Directions for Future Research

Building on the analytical evaluation conducted in this study, several directions emerge for future research. A natural next step is the development of prototype tooling that supports models expressed using the proposed CMMN-based notational variant and enables users to create, manipulate, and interpret such models in realistic scenarios. Tool-supported environments would also enable controlled experiments to empirically assess key performance indicators such as comprehension accuracy, modeling efficiency, and error rates.

In addition, the serialization mechanism introduced in Section 7 provides an important technical foundation for tool integration. By defining explicit interchange formats for models expressed using the proposed CMMN-based notational variant, it ensures that the notational enhancements are not only conceptually specified but also technically representable in a standards-compliant manner. Future work can build on this specification when developing modeling environments and automated components capable of supporting the creation, exchange, and analysis of such models. While a fully featured tool prototype is outside the scope of the present work, the inclusion of an explicit serialization provides a concrete foundation for future tool support.

Empirical studies involving practitioners, domain experts, and novice modelers are essential for evaluating principles that cannot be meaningfully assessed analytically, particularly cognitive fit and the practical implications of reduced graphic economy. Such studies would also help identify which notational enhancements provide the greatest benefits across different organizational contexts, thereby informing further refinement of the notational variant.

Another promising direction for future work is the empirical evaluation of models expressed using the proposed CMMN-based notational variant in comparison with alternative modeling approaches in practical settings, particularly in scenarios where case-based representations are considered alongside more procedural process models. Such studies could provide insights into the conditions under which enhanced case-based modeling offers measurable advantages, for example, in terms of flexibility, comprehensibility, or support for responsibility-aware modeling, compared to more rigid process-oriented representations.

Finally, future research will explore how the proposed CMMN-based notational variant can be further enriched to cover additional perspectives—such as decision logic or data quality—or how it can be integrated with complementary frameworks for business process analysis. These explorations further strengthen the expressive power and practical utility of CMMN+ in dynamic, knowledge-intensive environments.

10. Conclusions

This study introduces CMMN+, a CMMN-based notational variant that addresses key conceptual and visual shortcomings observed in standard CMMN modeling practice. Developed through the Design Science Research Method, the proposed notational enhancements strengthen the representation of activation logic, enrich the modeling of case data, and introduce explicit responsibility assignments. Together, these enhancements improve the cognitive effectiveness and conceptual clarity of case models, thereby supporting their use in dynamic, knowledge-intensive environments.

The analytical evaluation grounded in the PoN theory demonstrates that models expressed using the proposed CMMN-based notational variant deliver improvements along several cognitive dimensions—most notably semiotic clarity, semantic transparency, and perceptual discriminability—while maintaining compatibility with the design foundations of standard CMMN. These results establish CMMN+ as a coherent and theoretically grounded refinement of the original notation.

Despite these promising findings, the evaluation presented in this work is analytical in nature and does not capture how users perform, interpret, or construct models in real modeling situations. Future research will therefore focus on implementing tool-supported prototypes and conducting controlled experiments or field studies to assess model comprehension, modeling performance, and practical applicability across diverse domains. Further work should also examine the scalability of such models in complex, multi-stakeholder settings and investigate how the proposed notational enhancements interact with existing procedural modeling approaches.

The results of this research will also inform the evolution of case management modeling standards. The proposed CMMN-based notational variant provide a theoretically grounded example of how cognitive principles can be incorporated into visual language design and offer useful guidance for organizations such as the OMG when considering refinements to CMMN or related notations aimed at improving clarity, expressiveness, and conceptual precision.