Abstract
Knowledge can be represented in different formats. At present, knowledge representation as ontologies is the mainstream, while much less is heard about frames. In order to understand the state-of-the art of knowledge frames application, the authors overviewed recent research work in IEEE Xplore Digital Library, SpringerLink, ScienceDirect, and ACM Digital Library from 2000 till 2020. The overview touched such aspects as knowledge acquisition techniques, constituent parts of frames, the actual state of technologies used, capabilities of frames integration with other formats, and limitations. The results showed that native limitations of knowledge frames lead to creation of hybrid knowledge bases. However, hybrid systems also have known issues in performance and inconsistency of knowledge due to a conflict between paradigms. Moreover, a large part of technologies mentioned is not supported nowadays. The results can be useful for researchers who investigate whether to use knowledge frames for managing knowledge.
TopBackground
Software development lacks rigor separation of concerns in specifications. Model Driven Architecture (MDA) (Miller & Mukerji, 2001) proposed by the Object Management Group (OMG) introduces three concerns (or viewpoints) and related models, namely, a computation independent model (CIM), a platform independent model (PIM) and a platform specific model (PSM). MDA can be defined as “an approach to system development and interoperability that uses models to express and direct the course of understanding, requirements elicitation, design, construction, deployment, operation, maintenance, and modification” (OMG, 2010).
The first model that can be used for problem domain analysis is a CIM. This model allows specifying a domain knowledge expressed as business rules and processes, data vocabulary and requirements to the system and to the software (Miller & Mukerji, 2001). Therefore, usually the representation formats are unstructured (informal) or semi-structured (semi-formal) text and graphical representations, e.g., BPMN (Yassine Rhazali, Hadi, & Mouloudi, 2016), Data Flow Diagrams (Kardoš & Drozdová, 2010), UML diagrams (e.g. use case and activity diagrams), user stories, business rules (Essebaa & Chantit, 2016), etc. Advantages of formal and structured formats is that they can be used for automated processing and inferring. Therefore, the question on what a format to use in order to specify and automatedly process enough domain facts and rules is still open.
Key Terms in this Chapter
Topological Functioning Model: A formal mathematical model for specification of systems functional characteristics and causal relations among them based on the system theory and algebraic topology.
Knowledge Frame: A format for knowledge representation in computer science. It joins both declarative and procedural knowledge. Knowledge Frames support closed-world paradigm.
Problem Domain: An environment (or a field of expertise) that need to be investigated in order to improve or enhance its characteristics.
Model Transformation: A transformation process when a model is simplified, refined or otherwise changed according to mapping rules.
Solution Domain: A set of elements that can be applied in order to improve or enhance characteristics of a problem domain.
Computation Independent Model: One of three models in Model Driven Architecture that can be applied to specify business rules, business processes, data vocabulary and requirements to the system or software.
Ontology: A format for declarative knowledge representation in computer science. It supports open-world paradigm.
Knowledge: A fact that describes a phenomenon in a domain.