Book contents
- Frontmatter
- Dedication
- Contents
- Figures and Tables
- Notes on Contributors
- Acknowledgements
- Foreword
- Introduction
- Part 1 Getting Started
- Part 2 Building Taxonomies
- Part 3 Applications
- Part 4 Business Adoption
- Appendix A Metadata Template to Capture Taxonomy Term Diversity
- Appendix B Semantics – Some Basic Ontological Principles
- Appendix C Metadata Model Template
- Glossary
- Index
Appendix B - Semantics – Some Basic Ontological Principles
Published online by Cambridge University Press: 04 August 2022
- Frontmatter
- Dedication
- Contents
- Figures and Tables
- Notes on Contributors
- Acknowledgements
- Foreword
- Introduction
- Part 1 Getting Started
- Part 2 Building Taxonomies
- Part 3 Applications
- Part 4 Business Adoption
- Appendix A Metadata Template to Capture Taxonomy Term Diversity
- Appendix B Semantics – Some Basic Ontological Principles
- Appendix C Metadata Model Template
- Glossary
- Index
Summary
What is an ontology?
As other comprehensive volumes on this topic are widely available, the following brief explanations are offered with the caveat that this is a book about taxonomy and this discussion of ontologies is in that context.
Useful but incomplete definition 1: An ontology is a taxonomy with more complex and specific relationships between terms
An ontology is a semantic structure of concepts and their relationships. Some ontologies allow only Broader Term–Narrower Term (BT–NT) and Related Term (RT) relationships and are arranged into hierarchies: we call these taxonomies. Therefore, all taxonomies are ontologies, but not all ontologies are taxonomies. This means, of course, that taxonomies are a valid NT of ontologies:
Ontologies
– Taxonomies
Ontologies admit more complex structures than hierarchies and more complex relationships than the now familiar BT–NT and RT thesaural types. An ontology may or may not contain one or more taxonomies.
So, one simple way to introduce ontologies (in the context of a book about taxonomies) is to imagine a taxonomy in which you may specify the type of relationship between terms. To return to my favourite example about ‘Dogs’ and ‘Dog food’: instead of trying to decide whether to make these two terms a BT-NT pair or use an RT to relate them, we may now use any relationship we wish to semantically connect these terms. We might, for example, say that:
Dogs EAT Dog food
Dog food IS EATEN BY Dogs
. . . wherein the second is the inverse of the first, just like BT–NT relationships. We can now reimagine any example using this logic:
Economics HAS SUBFIELD Microeconomics
Microeconomics HAS TOPIC Consumer Demand
This allows us to neatly sidestep some of the problems discussed above (such as mixing topics and subjects) by specifying any relationship we like between any two concepts.
It's important to note that the resulting structure is not a hierarchy, but rather a different and more complex graph.
Useful but incomplete definition An ontology is a bunch of taxonomies tied together
Imagine, to return to an example from publishing, that you are the publisher of some scholarly journals.
- Type
- Chapter
- Information
- TaxonomiesPractical Approaches to Developing and Managing Vocabularies for Digital Information, pp. 233 - 236Publisher: FacetPrint publication year: 2022