DataTerminology.comNEW
An ontological principle specifying when two entities are the same and under what criteria their identity persists.
DESCRIPTION
Identity conditions determine when entities are identical and how they endure through change. They ground matching, deduplication, and linking in data systems. Philosophical principles like Leibniz’s Law inform practical identity modeling. In RDF/OWL, owl:sameAs asserts identity across datasets, requiring caution to avoid over-linking. Clear identity criteria improve data quality, governance, and reliable reasoning.
ALTERNATIVE DEFINITIONS
Wikipedia
Identity is the relation each thing bears only to itself.
Defines identity as a reflexive, unique relation.
Stanford Encyclopedia of Philosophy
Identity is a relation that everything bears to itself and to no other thing.
Highlights reflexivity and exclusivity in identity.
W3C
OWL 2 provides built-in support for asserting sameness using owl:sameAs to indicate that two IRIs refer to the same individual.
Connects philosophical identity to practical data linking with owl:sameAs.
RELATED TERMS
Leibniz’s Law Individuation Equivalence SameAs Universal Particular Knowledge Representation Description Logic RDF OWL
COMPETING TERMS
Indiscernibility Non-Identity
RELATIONS
- Narrower: Account
KNOWLEDGE AREAS
Semantic & Knowledge Modeling Data Modeling Data Management
TARGET AUDIENCE
Data Modeler Data Architect Information Analyst Data Steward MLOps Engineer
EXTERNAL SOURCES
- Wikipedia - Identity (philosophy)
- W3C - OWL identity and owl:sameAs
- Stanford Encyclopedia of Philosophy - Philosophical foundations of identity
- Wikipedia - Overview of owl:sameAs usage
TOP QUESTIONS
What is identity in ontology and data modeling?
It specifies criteria under which entities are the same, guiding linking, deduplication, and consistent referencing.
Category: Definition
How does owl:sameAs relate to identity?
It asserts that two IRIs denote the same individual. Use cautiously to avoid unintended equivalence propagation.
Category: Application
What is the biggest misunderstanding about identity?
Treating owl:sameAs as a loose similarity; it denotes strict identity, not approximate match.
Category: Misunderstanding
Why is identity important for governance?
Clear identity rules prevent duplication, ensure accurate reporting, and support lineage and compliance.
Category: Relevancy
What risks arise from sloppy identity modeling?
Over-linking corrupts datasets and inference; under-linking fragments knowledge and impairs analytics.
Category: Risk
How do you define identity criteria?
Use domain-specific invariants, keys, and persistence conditions informed by realist ontology and business rules.
Category: Requirements
Where can I learn more about identity?
Read SEP’s identity entry and W3C OWL guidance on equality, then explore data matching best practices.
Category: Learning
Is identity always absolute?
Philosophically, yes; operationally, linking may use probabilistic matching but must not be conflated with owl:sameAs.
Category: Opinion
What’s a practical example of identity criteria?
A machine’s serial number and manufacturer identifiers may uniquely determine identity across systems.
Category: Example