Concepts for the identification of Semantic Equivalence
In the scope of Industry 4.0, one challenge is the semantic interoperability between different applications. While various metamodels have been or are being developed for syntactic interoperability, there is a lack of concepts to ensure semantic interoperability. Many metamodels only specify the possible modelling classes and the semantics is outsourced to information models, e.g., in the form of submodels for asset administration shells or in companion specifications for OPC UA.
In many cases, the same information in the information models is modelled with different modelling elements. In order to be able to work with this information in an automated way, concepts for the identification of semantic equivalence shall be researched. Examples for possible solutions to this problem could be the identification by hand with the definition of integration rules (as shown in a similar fashion in prior dissertations at this chair about the topic of semantic interoperability), or algorithmic identification, using the semantic annotations in ontologies with inference- and reasoning-algorithms. A third solution type may be approaches using artificial intelligence.