TS ISO 12620

TS ISO 12620 is a standard concerned with how data categories are defined, selected, and managed for language resources, providing a common framework for terminology, lexicography, corpus-based resources, machine translation, and related content types. It sets out principles for applying a Data Category Registry and addresses the management of data categories, including their representation in exchange formats.

What a data category registry is meant to organise

In general technical terms, a data category is a specific property or value used in a language resource, such as a grammatical feature of a term, a field in a dictionary entry, or metadata attached to a translation unit. A registry makes these categories identifiable, distinguishable, and reusable in a consistent way. This helps different systems and resource types refer to the same concept with less ambiguity and greater comparability.

Interoperability across language-resource systems

The standard is framed broadly rather than being limited to one linguistic discipline. By covering terminology, lexicographical, collection-based, machine-translation and similar resource types, it reflects the fact that structured data categories are important both for content description and for exchange between systems. Clear data-category definitions help reduce divergent interpretation across tools and support more disciplined resource management.

Its focus is on creating, selecting, and managing data categories, not on providing a list of the contents of individual language resources. As such, it is useful for anyone working with language technologies and terminology management who needs conceptual consistency, precise data definition, and orderly registry practice.