43rd International Conference on Conceptual Modeling (ER 2024)

Air Quality Index (AQI) is a number aggregated from several air quality sensors deployed in an area. AQI is useful in communicating the air quality to the general public and in making governance decisions to tackle air pollution. However, our ethnographic surveys revealed the existence of a knowledge barrier in interpreting the AQI and data illiteracy in understanding AQI-related charts and trends commonly facilitated by government organizations. This knowledge gap is wider for the marginalized sections of society, who, it turns out, are more exposed to pollution. We use an ontological approach to homogenize the air quality data with social and spatial aspects. The Social Air Quality Index (SAQI) ontology integrates the data from local and central air quality monitoring sensors, meteorological data, and field surveys. This data is converted into a Knowledge Graph, which is used to build an application for civic engagement with the public on pollution to improve community participation of the local institutions and individuals. We evaluated this application through a user survey and received positive feedback. The ontologies, code, and datasets are available under the Apache 2.0 License at https://github.com/kracr/aq-structured-platform.

Time is a fundamental concept in data processing. The growth of social media (SM) data and the rise of the Internet of Things (IoT) underscores the necessity for studying temporal data on the Web. However, accessing realistic temporal data poses significant challenges across data collection, knowledge representation, and real-time provisioning, with no comprehensive solution available yet. To tackle these challenges, we introduce GenACT, a novel data generator rooted in the dynamics of Academic Conference Tweets (ACT), which serves as an ideal domain for eliciting application scenarios spanning temporality, dynamicity, and timeliness. The foundation of GenACT is a domain-specific ontology crafted to conceptualize tweets around an Academic Conference Event (ACE) realistically. The ACE ontology is available in all four OWL 2 profiles. Additionally, RDF instantiation allows for real-time simulation of ongoing academic discussions on Twitter. GenACT stands out for its ability to configure different data segments using SPARQL-based partitioning strategies. This versatility makes it adaptable to various analytical tasks, enabling researchers to focus on specific aspects of the data for their studies. GenACT is designed to seamlessly provide temporal and static data in a streaming format, tailored specifically for applications in studying knowledge graph evolution, temporal reasoning, and stream reasoning. The ontology, code, and documentation are available under the Apache 2.0 License at \url{https://anonymous.4open.science/r/genACT/}.

Board buses at the SEI to go to the Conference Dinner on the Gateway Clipper River Cruise. Boarding takes place from 5:30-5:45 p.m. at the SEI Lobby gathering point. Watch for signage.

Join us for a fun cruise up and down Pittsburgh's three rivers while networking with your fellow registrants. The cruise will start at 6:15 p.m. when the buses arrive from the SEI and end at approximately 9:30 p.m. Then we'll re-board the buses to travel back to the SEI at the end of the evening.

The registration table will be staffed from 8:00 a.m. until sessions end each day. Please stop by to check in and pick up your name badge. If you have any questions, please feel free to stop by and ask our staff. 

Who hasn't dabbled with popular Artificial Intelligence (AI) Large Language Models such as Gemini or ChatGPT, either as part of a project or just to see what it might produce? The popularity and idea that AI can transform human effort is everywhere in our scientific and industrial communities. AI clearly is a growing business driver.

And AI has value for conceptional modelling. For instance, natural language processing can help in identifying the relational data of entities and generative AI and inverse relationship approaches can help create deeply insightful entity relationship (ER) diagrams and models for the systems. To get the usable, reliable AI systems people want, designers and developers need to learn how to repeatably build sound AI systems. With those systems, we will see improvements in areas such as modeling (e.g., digital twins for smart cyber-physical systems), security engineering to defend large-scale systems, and the use of modern software practices (e.g., in developing threat models for DevSecOps).

Building sound AI systems repeatably, we need to pursue these system qualities: robust, secure, scalable, and human centered. Those quality attributes form the basis for forming an AI Engineering discipline. There is much work to do
to create and implement the AI Engineering discipline in areas such as making a commitment to moving R&D in practice to get ahead of the galloping use of AI tools, recognition that we cannot abandon good practices such as practicing cyber hygiene and the use of modern software engineering practices, and the creation of a new concept of the digital-engineering-aware workforce able to interact effectively with AI models.

The BERT model has slow inference speed due to the massive parameters, and it is susceptible to adversarial attacks due to lack of robustness. Knowledge distillation tackles the first issue by compressing the model, while adversarial training resolves the second issue by incorporating disturbances into training. However, existing knowledge distillation methods often struggle to balance between model accuracy and robustness. To address this, we propose a multi-teacher adversarial distillation approach for the BERT model. Our approach employs a clean teacher, providing soft labels to guide the student model in learning the clean samples, and a robust teacher, which employs adversarial training and semi-supervised learning to guide the student model in learning adversarial samples. Additionally, we introduce adaptive loss weights, which enable the student model to focus on the more challenging knowledge during training process. Through experiments with various adversarial attack methods, we validate the effectiveness of our proposed approach. On the typical text classification dataset IMDB, we improve the robust accuracy of the student DistilBERT model from 19.78 to 35.17, while maintaining a high clean accuracy of 93.17.

This paper explores the design and creation of metadata schemas based on the FAIR Data Principles. We provide a clear interpretation of these principles, focusing on how they apply to metadata schemas. Leveraging the OntoUML language, we developed a conceptual model that explains the key components of a FAIR-compliant metadata schema. Through detailed discussion and provision of examples for each model component, this work aims to help metadata designers and curators better understand how to incorporate the FAIR Data principles into their schemas.

Modularization has been extensively investigated for its role in enhancing the comprehension of process models by dividing them into smaller, self-contained and manageable modules. However, existing studies have reported inconclusive results, prompting further exploration into when modularization supports or impedes process comprehension. Among the key factors suggested to influence the effect of modularization is the type of the task at hand. Indeed, the fragmentation of information across several modules caused by modularization can challenge readers' comprehension of process models, especially if the task is not confined to a single module. While the effect of fragmentation has been explored for flow-based tasks, requiring understanding procedural aspects of process models, our work extends to circumstantial tasks, which instead demand comprehending model rules and constraints that rely on the broader context associated with specific process parts. Using eye-tracking, we investigate how the fragmentation of information caused by modularization impacts cognitive integration (i.e., the collection and synthesis of dispersed information) and its subsequent effect on cognitive load and model comprehension. Our findings demonstrate that fragmentation significantly affects cognitive integration in flow-based tasks only. Moreover, our results show that cognitive integration is linked to both cognitive load and readers' comprehension of process models. The outcome of this work provides a detailed model, explaining the task-specific effects of fragmentation. Practically, our findings highlight the need for modeling tools that dynamically adjust their user interface} to reduce cognitive integration demands, thereby enhancing model comprehension.

Big Data is challenging analytical contexts, namely when aligning data and analytical requirements. While the capacity to collect and store new data is expanding rapidly, the pace at which it can be analyzed is developing more slowly. Defining these analytical requirements and selecting the most appropriate visualizations often depends on an in-depth understanding of what users need from the data. To address this problem, this paper proposes an assisted model-driven analytics approach to support visualization, taking domain knowledge and data as input. It allows the user to be guided in the mapping between domain concepts and available data, as well as in the translation of domain questions into analytical tasks that can be supported by useful visualizations for decision support. The approach is supported by a Meta-Model that formalizes concepts needed to answer three fundamental questions, what, why and how. This Meta-Model contextualizes the data, the analytical tasks and the supporting visualizations. The applicability of the proposal is shown through a demonstration case focused on the genome domain. The results highlight how useful visualizations are derived from the specified domain questions.

Over the past decade, extensive research efforts have been dedicated to the extraction of information from textual process descriptions. Despite the remarkable progress witnessed in natural language processing (NLP), information extraction within the Business Process Management domain remains predominantly reliant on rule-based systems and machine learning methodologies. A key reason for this is the data scarcity in the domain, which has so far prevented the successful application of deep learning techniques. However, rapid advancements in generative large language models (LLMs) allow solving many NLP tasks with very high quality, with only a few examples (few-shot), or even just a description of the problem (zero-shot).Therefore, we systematically investigate the potential of LLMs to solve the task of information extraction from textual process descriptions, targeting the detection of process elements, such as activities and actors, and relations between them. Based on a novel prompting strategy, we show that LLMs are able to outperformstate-of-the-art machine learning approaches with absolute performance improvements of up to 8\% $F_1$ score across three different datasets.We evaluate our novel prompting strategy on eight different LLMs, showingit is universally applicable, while also analyzing the impact of certain promptparts on extraction quality. The number of example texts, the specificity of definitions, and the rigour of format instructions are identified as key for improving the accuracy of extracted information. Our code, prompts, and data is available at https://anonymous.4open.science/r/llm-process-generation-2140/README.md and our model generation proof-of-concept at https://anonymous.4open.science/r/pet-to-bpmn-poc-B465/README.md.

By integrating Business Analytics and Process Analytics, organizations can gain a deeper understanding of the relationship between process inefficiencies and business outcomes, leading to improved data-driven decision-making. This integration, however, is often overlooked, with limited methodological guidance for systematically combining these two analytical domains. Motivated by this challenge, this paper proposes a methodological approach for the identification of analytical visualizations that allow the aforementioned integration to be achieved systematically. The proposed approach is tested in a running example, which includes an instantiation of a Data Warehouse system for supporting data integration and analysis for business process analytics.

Event data is a collection of recorded events that capture performed actions and observed states of business processes supported by information systems. It describes the times of event occurrences, event types, event attributes, and process cases of events identified by one or more objects the events relate to. Process mining uses event data to analyze and improve the processes in organizations. These processes are often performed by actors or agents, such as employees, resources, and systems, in different roles within organizations. In this paper, we present Agent System Event Data (ASED), a new type of event data that describes business processes as interactions of agents. ASED provides a new scope for analyzing individual agents involved in multiple processes, interactions of agents, and systems of agents that enact the processes. We formalize ASED as a conceptual data model, discuss its dimensional data modeling aspects, and argue that event data, in general, benefits from dimensional representation. We review existing event data types and discuss the complementary nature of existing models and ASED. Finally, we validate ASED by demonstrating its ability to express existing business process compliance rules, significantly expanding the scope of compliance analysis addressed by existing data models.

As a discipline matures, its research communities often need to reflect on the epistemological underpinnings of their research and discuss how the discipline can, and should, advance. This type of reflection is often found in the natural sciences, where scientists question the nature of their fields, the standards of rigor applied, as well as the research impact. As Systems Analysis and Design (SAND) researchers, we should continuously reflect on and discuss how to conduct SAND research to stay relevant and in synch with ongoing advances in information technologies.
Among more recent changes in the advances in information technologies is the rise of large systems with ill-defined boundaries. Platforms such as Google and Facebook exhibit fuzzy, ill-defined boundaries because they permeate multiple aspects of digital and real-world interactions, making it challenging to delineate where their influence starts and ends. These platforms extend beyond their primary functions—search and social networking—into areas like advertising, data analytics, e-commerce, and even artificial intelligence, creating a vast interconnected ecosystem. Their services often integrate with third-party applications and websites, further blurring the lines of their operational scope. Additionally, their user base spans global demographics, intertwining with various cultural, social, and economic spheres, complicating the identification of distinct boundaries. Accurately modeling these systems and shaping the progression of their designs and their impacts is becoming increasingly complex and difficult.
Traditionally, SAND subscribes to the ontology of individual substances, wherein the reality is made of stand-alone substances, which can be referred to as things, objects, individuals, or entities (Benovsky, 2008; Bunge, 1977; Harman, 2018). These substances commonly have properties or attributes and undergo change, resulting in events and processes. This view is epitomized by Bunge’s ontology that stipulates that the world is “made up of things” that are substantial individuals (Bunge, 1977).
To better understand and design systems with fuzzy boundaries, such as Facebook or Google, we suggest exploring alternative ontological approaches. The traditional ontology of individual substances, while valuable in many contexts, may face limitations when applied to these complex, interconnected systems. Interestingly, similar challenges have been observed in philosophy and natural sciences. Following advances in physics, the claim emerged that “there are no particles, there are only fields” (Hobson, 2013, p. 211; emphasis added). This conclusion was drawn as a result of the accumulation of evidence from diverse scientific disciplines. Unfortunately, unlike many famous ontologies of substance, no established ontology of fields has been developed, with this being a stated goal in modern philosophy (Peuquet et al., 1998). We synthesize the ideas proposed as part of the debate on the ontology of fields and propose foundational concepts of ontology of fields as foundations for SAND scholarship.
The concept of a field is pivotal in both modern science and philosophy, serving as a useful model to understand reality. By adopting the scientific notion of a field, we can broadly conceptualize various aspects of existence. This approach helps to relate the concept of the field to other essential ideas in SAND, such as objects and processes. In this context, a field is defined as any physical or conceptual entity that exhibits different values across space and time, resulting from the oscillations that constitute and sustain these fields.
Fields can be understood by examining the properties at individual points within them. For example, a specific point in the sky modeled as a field may have properties like air temperature, humidity, and chemical composition. Many fields possess multiple properties, which can be represented in a hyperplane— a conceptual space with numerous dimensions corresponding to different property types. The variations in these properties manifest as modeling patterns such as peaks, plateaus, and valleys, where peaks often indicate regions with extreme concentrations of mass, energy, or charge.
The boundaries of fields are typically ambiguous and difficult to measure, often leading to challenges in defining where a field begins or ends. These fuzzy boundaries are a common issue in identifying the limits of natural features like mountains, neighborhoods, or even countries. Determining these boundaries often relies on conventions, which can lead to debates, such as whether Mount Elbrus belongs to Europe or Asia. Fuzzy boundaries can offer explanations for complex phenomena, like the wave-particle duality of light, where particles are viewed as quantized peaks within a field. We have these same types of fuzzy boundaries in our SAND efforts.
By incorporating ideas from physics and geography, ontology of fields offers fresh perspectives on understanding the diverse range of information systems today. The ontology of fields provides a more flexible approach to SAND than was previously possible. This new approach allows for the description of information systems that can be built for a single individual (object) compared to those that are built for many people in (a field). Field-like aspects of information systems may have fuzzy boundaries. Previously, more traditional frameworks (see Siau et al., 2022) are less suitable for the current nature of information systems that are present in today’s society (e.g., intelligent transportation systems).
While the ontology of fields will supplement the existing frameworks, approaches, and ontology of SAND, it does require new ways of thinking about and designing systems. It provides new vocabulary and conceptual tools for understanding and designing modern information systems. Current and future technologies will benefit from the ontology of fields to provide a comprehensive understanding of their composition.
Applying field theory to information systems design encourages a holistic view, considering the entire ecosystem rather than primarily isolated components. Describing systems as fields with characteristics such as peaks and valleys (of mountains) helps in understanding and managing their nature, providing a new lens for various types of engagement.
Many modern applications are inherently field-like, with fuzzy boundaries, multiple stakeholders, and public-private partnerships that blur traditional organizational lines. Viewing these systems through the lens of a field ontology acknowledges their interconnected nature. The ontology of fields is intended to additionally address the need for flexible representations.
This ER Forum submission presents a research endeavor (i.e., using an ontology of fields for SAND). Although still in its early stages, it could potentially develop into a major stream for SAND and conceptual modeling research and perhaps serve as a foundation for impactful research. Future research will expand and apply the ontology of fields to demonstrate its feasibility and effectiveness.

Identity and Access Management (IAM) is a crucial aspect of online interactions, and researchers and practitioners have been exploring this area since the early days of the Internet. Over the years, many digital identity protocols and standards were created. While OAuth 2.0 and OpenID Connect (OIDC) are currently the most widely used and mature protocols, the emerging concept of Self-Sovereign Identity (SSI) and its underlying protocols promise greater user privacy and control. Although much emphasis has been placed on new protocols and implementations, the ontological and formal understanding of digital identity models still needs to be addressed. In this paper, we make three contributions: (i) introduce the Role-Artifact-Function (RAF) framework, which comprises a meta-metamodel and a metamodel; (ii) describe and compare digital identity models through the instantiation of the RAF metamodel; and (iii) discuss and arrive at intriguing conclusions, such as the conceptual equivalence between the well-established family of x509-based protocols and the nascent technologies behind SSI.

Dealing with iterative cycles in machine learning (ML) development presents several challenges because decisions made in one cycle can have effects on subsequent cycles. Goal-oriented conceptual modeling can be used to identify and analyze conflict areas between design decision points by way of refining goals, the alternative tasks that can achieve those goals, and softgoals which those tasks contribute to. Decision-making for ML development involves complex tradeoffs across various iterative cycles, involving conflicts and tensions among business, technical, and Responsible AI goals. Each of these iterative cycles consist of evaluation results that vary within each stage and from stage to stage. This paper investigates the distribution of Responsible AI decisions and tradeoffs along different ML iterative cycles and how they interact with each other. We propose three goal modeling constructs for processes with iterative decision-making that are contingent on observed intermediate results, such as those in ML development: Sensors, Actuators, and Iterative Loops. The goal modeling presented in this paper supports design reasoning about where to place Sensors and Actuators in nested iterative cycles, and to consider which decision criteria are applicable at which level of nesting.

Amid rising global challenges, resilience is attracting increased interest across various disciplines. However, significant ambiguities, vagueness, and variations in its definitions present notable obstacles to interdisciplinary communication and practical application. These semantic issues negatively impact not only the applications that rely on these definitions but also undermine the integrity of the conceptual models built from them, rendering them unable to ensure interoperability and clarity for their intended users. Recognizing the need for a robust and clear definition of resilience, this work addresses the complexity inherent in the concept by performing an ontological analysis using the Unified Foundational Ontology (UFO) and creating a sound core ontology model with OntoUML, UFO’s related conceptual modeling language. This research unfolds the conceptualization of resilience by investigating its fundamental categories and related concepts. We explore essential aspects of resilience, examining whether it preexists disturbances or is developed in response to them. We also identify its relational dependencies, establish how it is actualized, and determine the circumstances under which it becomes perceptible. Following our analysis of the ontological nature of resilience, the paper establishes an ontologically well-founded definition of this concept. To illustrate the practical application of our theoretical findings, we present a specific case in the field of production management.

The advent of generative artificial intelligence, and in particular large language models, has opened up new possibilities for information processing in a multitude of domains. Nevertheless, it is essential to validate their output in order to ensure its validity within the specified context. This is due to their nature as probabilistic models of language, which may lead to the generation of inaccuracies or non-existent facts commonly known as hallucinations. As a solution, we propose a framework and a prompt structure for the validation of the results of generative artificial intelligence in formats that are more human-comprehensible through the use of conceptual models. We denote this as conceptual model augmented generative artificial intelligence (CMAG). We illustrate the approach through application examples in the domains of data management, knowledge graphs and cultural heritage, and software engineering.

Effective conceptual modelling is critical for accurately representing the physical world in digital twin implementations, yet existing methods often lack comprehensive guidance or overlook key aspects like communication modelling. This research proposes the Entity-Relationship Digital Twin (ERDT) modelling method tailored for industrial digital twins. Building on established Entity-Relationship (ER) models, ERDT extends them with constructs for handling historical data, and defining security interfaces and data flows between physical and virtual components. The method provides a systematic process covering digital twin objective setting, physical component selection, conceptual entity definition, interface specification, and communication modelling. It emphasizes principles like encapsulation and separation of concerns to enhance maintainability. Evaluated against ISO 25010 quality criteria, ERDT demonstrates strengths in functional appropriateness, usability, portability, and maintainability while identifying opportunities for further improvement. A logistics case study illustrates applying the method to a real-world industrial scenario. By bridging conceptual modelling with software engineering best practices, ERDT facilitates developing robust, maintainable digital twins.

Typically, four types of human-robot interaction are distinguished: Coexistence, Synchronization, Cooperation, and Collaboration. They differ in the degree of interaction between human and robot and have therefore high impact on safety requirements for a system. Leading to human-robot cooperation and collaboration systems rarely being introduced in industrial practice due to the high risks associated with them. An underlying problem for this, is the lack of understanding for the differences and necessities of these interaction types on a conceptual level. In this paper, we investigate the differences between the four human-robot interaction types with respect to the dependencies between human and robot. Although in all interaction types humans and robots depend on each other, we show that these dependencies are very different in nature. For future use cases, this analysis helps in binding safety risks to concrete properties of the human-robot dependencies.