Overview
ISO/IEC 13250-5:2015 - "Information technology - Topic Maps - Part 5: Reference model" defines the Topic Maps Reference Model (TMRM). TMRM is an abstract, ontologically neutral reference model for subject maps that specifies a formal basis for representing subjects, minimal access/navigation functionality, merging behavior, and a constraint framework for interpreting maps. It is intended as a concise conceptual foundation to enable interoperability, information retrieval, and mapping between different subject-centric data models.
Key Topics
- Subjects and subject proxies: A subject is represented by a single proxy (a set of properties). Properties are key/value pairs where keys are labels and values come from a value set (labels, numbers, strings).
- Proxy and map structure: A subject map is a finite set of proxies. Labels and values are handled by defined sets (L and V), and proxies are identified by a partial injective labeling function.
- Navigation (primitive path language P): TMRM defines primitive navigation operators (e.g., local keys x#, remote keys x", local values x!k, and reverse navigation v k) that form a minimal path language for information retrieval from maps. These primitives underpin higher‑level query and constraint languages.
- Ontological commitments: TMRM intentionally minimizes ontological assumptions. It defines two core predicates - ako (subclass of) and isa (instance of) - with specified reflexivity/transitivity properties but leaves subject identification and proxy derivation methods open to implementers.
- Merging semantics: Merging of proxies is modeled as a partial function ./ that may depend on an environment E. The model requires merging to be commutative, associative and idempotent, and describes how equivalence classes of proxies produce merged views of maps.
- Constraints: TMRM specifies requirements for constraint languages (C), including an operator applying constraints to maps and a satisfaction relation (m |= c). It does not mandate specific constraint formalisms, but provides the formal foundation for languages such as TMCL.
Applications / Who uses this standard
- Knowledge engineers and architects designing subject-centric metadata and Topic Maps implementations
- Developers of information retrieval, semantic search, knowledge graph and ontology integration tools
- Systems performing data integration, entity resolution and canonicalization across heterogeneous metadata sources
- Digital libraries, archives, and enterprise content management systems requiring unified views of subject-centric data
Practical uses include defining interoperable Topic Maps implementations, specifying query/navigation APIs, formalizing merging/identity rules, and building constraint-driven validation and transformation tools.
Related Standards