Building Ontologies with Basic Formal Ontology

Robert Arp, Barry Smith, and Andrew Spear

Detailed Table of Contents

Introduction

Overwhelmed with Information

Obstacles to Accessibility: Human and Technical Idiosyncrasy

The Computer Limitations Problem

Some Implications of Computer Limitations for Information Representation and Management

The Problem of Imprecise Thinking

An Example: The BRIDG Model

Ontology as Part of the Solution

A New Organon for the Information Age

Suggested Further Reading

1 What Is an Ontology?

Introduction

Ontologies Are Representational Artifacts

Artifacts    

Representational Artifacts

Representational Units and Composite Representations  

A Note on “Term”  

Ontology, Terminology, Conceptology

Ontology and Terminology: The Case of ISO         

The Concept Orientation

Philosophical and Historical Background to Conceptualism        

Realism and Ontology    

Accurately Representing Entities in Reality  

Respecting the Use-Mention Distinction      

Ontologies Represent Universals, Defined Classes, and the Relations Between Them       

The Goal of Science Is to Represent General Features of Reality        

Ontological Realism        

Metaphysical Nominalism

Universals and Particulars        

Empty or Potentially Empty General Terms  

Universal vs. Class

Relations in Ontologies   

Basic Relations     

Universal-Universal Relations   

Universal-Particular Relations   

Particular-Particular Relations

Conclusion 

Further Reading on Issues of Epistemological and Ontological Realism 

2 Kinds of Ontologies and the Role of Taxonomies

Philosophical Ontology 

Philosophical Ontology and Taxonomy         

Simple Taxonomies        

Formal vs. Material Ontologies 

Domain Ontology 

Domain Ontology and Taxonomy       

Definition, Taxonomy, Ontology        

Top-Level Ontology        

Semantic Interoperability

Choice of Top-Level Ontology

Application vs. Reference Ontology

Conclusion 

Further Reading on Top-Level and Domain Ontology

Further Reading on Taxonomy and Classification     

3 Principles of Best Practice I: Domain Ontology Design

General Principles of Ontology Design    

1. Realism   

2. Perspectivalism

3. Fallibilism

4. Adequatism

Additional Principles of Ontology Design 

5. The Principle of Reuse

6. The Ontology Design Process Should Balance Utility and Realism   

7. The Ontology Design Process Is Open-Ended   

8. The Principle of Low-Hanging Fruit

Overview of the Domain Ontology Design Process   

Explicitly Determine the Subject Matter of the Domain Ontology        

Domain and Top-Level Ontologies     

Relevance   

Granularity 

The Problem of Nonexistents

Conclusion 

Further Reading on Relevance, Perspectivalism, Granularity, and Adequatism

4 Principles of Best Practice II: Terms, Definitions, and Classification

Principles for Terminology    

Gather and Select Terminology

1. Include in the terminology terms used by scientists

2. Strive to ensure maximal consensus with the scientists’ usage

3. Identify areas of disciplinary overlap where terminological usage is not consistent

4. In terminology construction and ontology design, make use of as many existing resources (terminologies and ontologies) as possible.

Formatting Terminology

5. Use singular nouns.

6. Use lowercase for common nouns.

7. Avoid acronyms.

8. Associate each term in the ontology with a unique alphanumeric identifier.

9. Ensure univocity of terms.

10. Ensure univocity of relational expressions.

11. Avoid mass terms.

12. Distinguish the general from the particular.      

Principles for Definitions

13. Provide all nonroot terms with definitions

14. Use Aristotelian definitions

15. Use essential features in defining terms.

16. Start with the most general terms in your domain.

17. Avoid circularity in defining terms.

18. To ensure the intelligibility of definitions, use simpler terms than the term you are defining.

19. Do not create terms for universals through logical combination.

20. Definitions should be unpackable (Term-definition intersubstitutability)

Principles for Taxonomies

21. Structure every ontology around a backbone is_a hierarchy.

22. Ensure is_a completeness.

23. Ensure asserted single inheritance.

24. Both developers and users of an ontology should respect the open-world assumption.

25. Adhere to the rule of objectivity, which means: describe what exists in reality, not what is known about what exists in reality       

Conclusion     

Further Readings on Definitions and Categorization      

Examples of Critical Reviews

5 Introduction to Basic Formal Ontology I: Continuants

Some Basic Features of BFO

Basic Types of Entity: Continuant and Occurrent

BFO: Continuant  

BFO: Independent Continuant 

BFO: Material Entity       

BFO: Object

BFO: Object Aggregate  

BFO: Fiat Object Part     

Combination Object-Entities    

BFO: Specifically Dependent Continuant     

BFO: Quality        

BFO: Relational Quality  

Relations That Do and Relations That Do Not Have Instances   

BFO: Realizable Entity    

BFO: Role   

BFO: Disposition   

BFO: Function      

BFO: Specifically Dependent Continuant: Summary       

Reciprocal Dependence among Realizable Dependent Continuants     

BFO: Generically Dependent Continuant     

BFO: Immaterial Entity   

BFO: Continuant Fiat Boundary (including Zero-, One-, and Two-Dimensional Continuant Fiat Boundary)   

Boundaries and Granularity     

BFO: Site    

BFO: Spatial Region (including Zero-, One-, Two-, and Three-Dimensional Spatial Regions)

Spatial Regions and Frames of Reference

A BFO Continuant Classification     

Further Reading on Basic Formal Ontology      

Further Reading on Granularity     

Further Reading on Independent Continuants  

Further Reading on Dependent Continuants     

Further Reading on Boundaries, Spatial Regions, and Topology  

6 Introduction to Basic Formal Ontology II: Occurrents

BFO: Process        

BFO: History         

BFO: Process Boundary  

BFO: Spatiotemporal Region    

BFO: Temporal Region    

BFO: Zero-Dimensional Temporal Region     

BFO: One-Dimensional Temporal Region     

An Example of Occurrent Classification       

Classifying Universals with BFO         

Exhaustiveness of BFO Categories    

BFO’s Perspectivalism

BFO’s Perspectivalism in Practice       

Further Reading on Processes and Events       

7 The Ontology of Relations

BFO Relations    

Relations: Formal Properties and Conventions 

Primitive Instance-level Relations  

Universal-Universal Relations in BFO      

Foundational Relation: is_a

Foundational Relations: continuant_part_of and occurrent_part_of

Spatial and Temporal Relations

Spatial Relation: adjacent_to

Temporal Relation: derives_from

Temporal Relation: preceded_by

Participation Relation: has_participant

Some Further Top-Level Relations 

proper_continuant_part_of and proper_occurrent_part_of

has_continuant_part and integral_continuant_part; has_occurrent_part and integral_occurrent_part

Relations and Definitions of Categories   

The All-Some Rule      

Inversion and Reciprocity         

Some Examples of Axioms       

Reflexivity, Symmetry, and Transitivity       

Further Reading on Relations

8 Basic Formal Ontology at Work

The Protégé Ontology Editor and BFO    

The Web Ontology Language (OWL)      

Hypertext Markup Language (HTML) and Extensible Markup Language (XML)       

Resource Description Framework (RDF)       

RDF Schema (RDFS)      

Simple Protocol and RDF Query Language (SPARQL)       

Basic Features of OWL   

OWL vs. Standard Relational Databases      

OWL 2        

Building Ontologies with Basic Formal Ontology

Example: The Ontology for General Medical Science (OGMS)   

Infectious Disease Ontology (IDO)    

Information Artifact Ontology (IAO)

The Emotion Ontology (MFO-EM)      

Facilitation of Interoperability       

Further Reading in OWL, RDFS, and RDF

Appendix on Implementation: Languages, Editors, Reasoners, Browsers, Tools for Reuse

Glossary

Web Links Mentioned in the Text

Notes

Bibliography