A selection from Stockholm University publication database

• Creating Worlds with Words

  2017. Paul Johannesson, Maria Bergholtz, Owen Eriksson. Conceptual Modeling Perspectives, 169-184

  Chapter

  Conceptual modeling is often viewed as an activity of representing a preexisting world that should be faithfully mirrored in an information system. This view is adequate for modeling physical domains but needs to be revised and extended for social and institutional domains, as these are continuously produced and reproduced through communicative processes. Thereby, conceptual modeling moves beyond analysis and representation in order to cater also for design and creation. Following such a view on conceptual modeling, this paper proposes an ontology for modeling institutional domains. The ontology emphasizes the role of institutional entities in regulating and governing these domains through rules and rights that define allowed and required interactions. Furthermore, the ontology shows how these institutional entities are dependent on and grounded in material entities. Conceptual modelers can benefit from the ontology when modeling institutional domains, as it highlights fundamental notions and distinctions in these domains, e.g., the role of rights, the role of processes in creating institutional facts, and the difference between documents and institutional information. The ontology is illustrated using a case on public consultation management.

  Read more about Creating Worlds with Words

• The case for classes and instances - a response to representing instances

  2019. Owen Eriksson, Paul Johannesson, Maria Bergholtz. European Journal of Information Systems 28 (6), 681-693

  Article

  In “Representing instances: The case for reengineering conceptual modelling grammars”, Lukyanenko et al. (2019) argue that conceptual modelling has been biased towards a focus on knowledge about general phenomena (classes) rather than about specific instances. While we agree that more attention needs to be paid to instances, we critically reflect on their underlying assumptions about instances and classes. Lukyanenko et al. (2019) assume that instances are mainly material things, and also assume that class-based modelling typically requires that class definitions include comprehensive attribute structures that are expected to be stable over time. Based on these assumptions, they conclude that classes are not needed for modelling instances. As an alternative to these assumptions, we suggest that instances can be viewed as language constructs, i.e., as objects that may be anything that is uniquely referred to and identified in human communication. Based on this assumption, we introduce an identity-oriented view of classes, implying that classes are required for modelling objects (instances). We agree with Lukyanenko et al. (2019) that a reengineering of conceptual modelling grammars is required. This reengineering would benefit from approaches such as an identity-oriented view of classes and a class-instance modelling grammar.

  Read more about The case for classes and instances - a response to representing instances

Show all publications by Ann Maria Dorotea Bergholtz at Stockholm University