The ethics of conceptual, ontological, semantic and knowledge modeling
Robert J. Rovetto
DOI: https://doi.org/10.1007/s00146-022-01563-3
2023-04-14
AI & Society
Abstract:The ethics of artificial intelligence (AI) is a research topic with both theoretical and practical significance. However, the ethical and moral aspects of conceptual, ontological, semantic, and knowledge modeling, more specifically, and which are sometimes found in AI applications, is not being given sufficient attention. I argue that it should. Whether considering using or developing these meaning-focused models, there are ethical aspects. This paper offers a preliminary outline about this potentially new research field, discussing: some questions and areas of concern, their significance; scenarios of ethical import, and initial ethical guidelines. A focus area, potential case study, and lens through which to see the ethical significance at hand, is ontologies, with emphasis on the most abstract ontologies (so-called generic, foundational, top-level or upper ontologies). This would aim to show the ethical import of highly abstract (often philosophical) ontological choices, assumptions, and their effects on data, developers, users, and society. In particular, I stress the need for critical examination of such models, their content, their intended meanings, assumptions, ideologies, as well as the development process. Both the ethics of, and ethics in, these activities are, therefore, of interest. Topics in the ethics of AI that are relevant include: transparency, bias, data-sharing, privacy, control over data, etc. This paper thereby contributes to the ethics of AI, as well as the ethics of information, of knowledge managment and relevant topics by developing and discussing the ethics of ontology (and similar systems). Future development will refine and add to the ideas expressed in this paper.