Smajevic, M. (2022). A generic conceptual model to graph transformation framework and its application for enterprise architecture analysis [Diploma Thesis, Technische Universität Wien]. reposiTUm. https://doi.org/10.34726/hss.2022.83010
E194 - Institut für Information Systems Engineering
-
Date (published):
2022
-
Number of Pages:
130
-
Keywords:
conceptual modeling; model-driven engineering; model transformation; graph-based analysis; knowledge graphs
en
Abstract:
Core assets of enterprise architecture (EA) are its models. They encapsulate enterprise structure and IT knowledge by providing a holistic view of an organization that can drive decision-making. For making good decisions, it is necessary to have an excellent overview and to know how to analyze the model to gain information that different business stakeholders need. From the perspective of history, EA has evolved a lot, and the models have become larger and more complex. Models’ change in size and complexity requires other techniques to analyze the model since the analysis of models in their raw form can exceed human capabilities by far. Therefore current approaches focus on studying partial views over the model, thus providing a partial valuation. This thesis proposes a slightly different approach to cope with EA analysis. The method is based on transforming conceptual models into graphs and analyzing the model in a graph-based manner by using knowledge and possibilities from graph theory. To put this into the real world, we have developed a generic framework and implemented it in the prototype platform, which we call CM2KG. The platform allows analysts to upload different models created in industry-standard tools and transform them into a standardized graph format. From there on, it becomes possible for them to visualize and interact with the model using multiple techniques, query the whole model, and define custom functions that will provide valuable information. We discuss some basic examples for each part to scratch the surface of what is possible. We also put one existing application of EA analysis into the perspective of this approach by implementing queries that are used to identify smells in EA models.