What is it about?
The paper introduces a process for designing ontologies that will be used to set up knowledge bases. Thus, the process is driven by validation and verification metrics to ensure that the ontology aims are reached in terms of quality criteria for design quality and end-user usability, learnability and completeness. The paper explains this process in a real case study.
Featured Image
Why is it important?
Ontologies might be used for eliminating terms ambiguity, to create taxonomies with semantics or for setting knowledge bases that will be used to perform reasoning in case studies. The process in this paper is intended for the latter, and thus, it will help in building ontologies to create knowledge bases (KBs) that will be used by information systems. These KBs might also be used for encoding business rules as knowledge inference rules, then the designing quality is important for preventing unexpected situations in reasoning. The ontology validation is also important from the end-user perspective since they will be the KBs users. Then, the process proposed in the paper is closely and iterative driven by verification and validation of the ontology against competency questions and metrics.
Perspectives
We defined this process proposed in the paper because we had specific needs when defining ontologies in several research projects for different application domains (smart grids, health care, anti-money laundering). The relevant and well-known processes, which are important and inspiring, did not have enough activities or the quality control that we required. For instance, our process iteratively uses metrics to improve the ontology and KB until reaching the end-user requirements through metrics for validation and verification. These metrics are proposed to be defined based on the end-user competency questions. Also, our process proposes a cycle and sub-cycles to control the quality according to the project specifications; thus, the process is adaptable to the project (e.g. fewer quality requirements, fewer process activities). The included case study will help to guide the usage of our process. We expect this paper can help other researchers and practitioners.
Angelina Espinoza-Limon
University College Cork National University of Ireland
Read the Original
This page is a summary of: A validation & verification driven ontology: An iterative process, Applied Ontology, July 2021, IOS Press,
DOI: 10.3233/ao-210251.
You can read the full text:
Contributors
The following have contributed to this page
