Keywords: Artificial intelligence, Cloud platform, Conceptual modeling, Knowledge graph, Model transformation, Model-driven engineering
Astract: Semantic processing of conceptual models is a focus of research for several years, bridging the disciplines of knowledge-based systems, conceptual modeling, and model-driven software engineering. With Knowledge Graphs, this research area gained momentum. In this paper, we introduce CM2KGcloud, a generic and extensible Web-based platform for transforming conceptual models into Knowledge Graphs. The platform can work on models created by state-of-the-art metamodeling platforms (e.g., EMF, Papyrus, ADOxx) and transforms models created with them into standardized Knowledge Graph representations like GraphML, RDF, and OWL. CM2KGcloud can be used as a service and can be integrated into software systems by its exposed API.

Keywords: blended modeling, GLSP, Langium, model server
Astract: Combining textual and graphical modeling i.e., representing textual models in the form of diagrams, has been a popular topic ever since in the field of model engineering. Most often modeling tools only provide users the possibility to create models either in textual form or in the form of a diagram, and the users have to decide upon initial creation of the model whether they would like to use a textual or a graphical model editor. So far, blended modeling tools combining both approaches have only been developed based on traditional frameworks e.g., Xtext and EMF. The next generation frameworks Langium and the Graphical Language Server Platform (GLSP) promote new opportunities such as increased modularity in architecture and deployment options, more flexibility in user interface design, web-based and cloud-friendly development possibilities, while eliminating the dependency to Java.This thesis aims to revisit and explore the topic of combining textual and graphical modeling with the next-generation frameworks Langium and GLSP. A concept for blended textual-graphical modeling based on these frameworks is developed, which utilizes a model service to jointly manage the textual and graphical editor’s underlying modification model. The concept considers that the graphical and textual editor must operate on the same model, simultaneous updates must be possible between the two editors and non-semantic information of the model must be maintained during updates of the model. The concept is realized as an artifact based on the Workflow language. An existing GLSP framework for the Workflow language providing graphical modeling is extended by a Langium language server providing textual modeling, and a model server handling model access, provision and updates between the textual and graphical editors.To evaluate the developed concepts and artifacts, the implemented solution concepts are instantiated by two UML use cases of the bigUML modeling tool: the package diagram and the class diagram. These two use cases are evaluated against the conceptualized requirements of blended textual-graphical modeling.

Keywords: Langium, GLSP, Web Modeling, Model-driven Engineering, Model Management, Metamodeling, Generator
Astract: In recent years, a shift from feature-rich Integrated Development Environments (IDE) to lightweight web clients could be observed. In order to be able to make that shift, for textual editors, the Language Server Protocol (LSP) has been created, while for graphical editors, an enhancement of the LSP has been introduced in the Graphical Language Server Platform (GLSP). However, the model management for graphical editors has still been handled by heavy-weight Java model servers. This includes both the creation of metamodels and the runtime handling of models.This thesis aims to make the next step toward web-based graphical model editors and shift the model management to a TypeScript-only technology stack. For this, the functionalities of the next-generation language framework Langium are explored and extended by a model server API. This enables model-oriented clients to access the Abstract Syntax Tree (AST), which is created by Langium and holds the current state of a model. Furthermore, a new TypeScript native grammar language is conceptualized to provide a TypeScript native solution to define metamodels.To combine the model server API with the TypeScript-based grammar language, a generator is created that sets up the entire model management component. Typically, in Java-based environments, the metamodel for the model management component is created using the EMF Ecore metamodel. Therefore, to ease the transition from Ecore to the TypeScript-based grammar language, a mechanism to create the TypeScript-based grammar definition from the Ecore metamodel is added to the implementation of the generator.The work in this thesis is evaluated in two parts: First, the TypeScript-based grammar language is evaluated by comparing it with the widely used Ecore metamodel. Second, two state-of-the-art modeling tools, which utilize GLSP, are rebuilt to evaluate the generator by creating the metamodel and model management using the generator’s capabilities.

Keywords: Model-Driven Engineering
Astract: Model-driven engineering (MDE) and Artificial Intelligence (AI) have gained momentum in recent years, and the fusion of techniques and tools in the two domains paves the way for several applications. Such integrations—which we call MDE Intelligence—are bidirectional, i.e., MDE activities can benefit from the integration of AI ideas and, in return, AI can benefit from the automation and subject-matter-expert integration offered by MDE. The 5th edition of the Workshop on Artificial Intelligence and Model-driven Engineering (MDE Intelligence), held in conjunction with the IEEE/ACM 26th International Conference on Model-Driven Engineering Languages and Systems (MODELS 2023), follows up on the success of the previous four editions, and provides a forum to discuss, study, and explore the opportunities offered and the challenges raised by integrating AI and MDE.

Keywords: Data set, Enterprise architecture, Enterprise modeling, FAIR, Machine learning
Astract: The conceptual modeling community and its subdivisions of enterprise modeling are increasingly investigating the potentials of applying artificial intelligence, in particular machine learning (ML), to tasks like model creation, model analysis, and model processing. A prerequisite—and currently a limiting factor for the community—to conduct research involving ML is the scarcity of openly available models of adequate quality and quantity. With the paper at hand, we aim to tackle this limitation by introducing an EA ModelSet, i.e., a curated and FAIR repository of enterprise architecture models that can be used by the community. We report on our efforts in building this data set and elaborate on the possibilities of conducting ML-based modeling research with it. We hope this paper sparks a community effort toward the development of a FAIR, large model set that enables ML research with conceptual models.