Factwhat

Factwhat is a term used in information science to describe a standardized, machine-readable representation of a factual claim about the real world. In this context, a factwhat unit encodes a proposition in a structured form that can be stored, queried, and compared across systems.

A factwhat statement typically comprises a subject, a predicate, and a value, and may include optional metadata such as provenance, timestamp, and confidence. The aim is interoperability with existing knowledge representations, enabling integration with knowledge graphs, databases, and automated fact-checking workflows. The format is intended to be domain-agnostic and minimally ambiguous.

Although factwhat is presented here as a coined term for illustration, the concept aligns with triple-based models such as RDF. The emphasis on explicit metadata supports traceability and auditability of claims, while acknowledging that standardization and parsing burdens remain points of discussion among researchers.

Examples: a factwhat statement could express the height of the Eiffel Tower as subject 'Eiffel Tower', predicate 'height', value '324 meters', source 'Official Monument Records', date '1889'. Another example: subject 'Earth', predicate 'orbits', value 'Moon', source 'Satellite Observations', date 'Date uncertain'.

Applications of the idea behind factwhat appear in research on fact extraction, data provenance, and automated verification. In practice, the term is not widely standardized and is used chiefly as a schematic concept to illustrate how factual claims might be structured for machine processing.

See also knowledge graph, data provenance, fact-checking, semantic web.