Background

Ontologies and conceptual data models are used by intelligent systems to manage data more efficiently by a number of applications [1]. One of the applications is being applied to the Semantic Web, considered as the next generation of the World-Wide Web [2], where data is envisioned as a linked web. The idea is to express knowledge as an ontology or conceptual data model.   How the actual data of a domain is linked to the domain’s conceptual representation is a consideration that a theoretical architecture called KnowID has taken into account. Its purpose is to carry out a ‘knowledge-to-data’ pipeline so that we can have a graphical interface representing conceptual data as an EER, popular for mapping over relational databases, as a map of inter-related concepts and expressive relationships and query over them to obtain the actual data we are looking for. One of the important steps in realizing this is ensuring the model’s expressivity which requires making implicit knowledge explicit. This process was the core focus of the materialization of deductions component of the system where inferred facts needs to be amended to the model.   One of the most prominent existing tools that carried inference-based ontology edits was an earlier version of Protégé [9]. However, Protégé’s original functionality that to materialized inferences is widely inaccessible and not tailored to suit EER features. Another prominent implementation is the OWL API [7], a high-level API implemented in Java that allows users to load, manipulate and query ontologies [7]. However, object-oriented programming approaches like Owlready [8] are easier to use to modify an ontology than APIs like the OWL API [7]. Owlready provides functions to access and edit entities and constraints. It is a python module designed to use object-oriented programming principals to support OWL 2 ontologies where entities can be accessed like objects in programming languages. Moreover, an important feature of another tool called ICOM [6] brings about an important service. Like the other implementations discussed, it is a tool capable of making certain edits based on inferences from its DL reasoner, albeit not EER specific. More importantly, editing, the user can then decide if they want to keep the changes.