Translating RDF schemas and Ontologies

Position paper for SWAD-E / CEN ISSS MMI-DC workshop "metadata for a multilingual world".

Charles McCathieNevile, W3C

This short paper outlines some f the issues that arise in translation of the human-readable parts of RDF vocabularies

Introduction

RDF, and in particular the RDF Vocabulary Definition Language (RDF Schema), provides mechanisms for providing human-readable labels and comments on individual terms in a vocabulary. This is done with the label and comment properties, whose typical content is a literal. By using xml:lang, it is possible to define labels in mutliple languages.

Use Cases

There are several obvious use cases

Locallised user interfaces

Tools can usee the label information to present properties and classes through a short, easily readable label

Assisting translation

A translator who is not familiar with english might use other languages to assist in translation. For example, a Mongolian translator may use the Russian and Chinese versions of label and comment to understand how to translate the definitions into Mongolian.

Context-sensitive help

A tool can use the comment information to provide more detailed explanation about the meaning of a particular property or class on the request of the user.

Authority

It is possible for anyone to provide a translated label. There is therefore a problem in knowing which translation to trust. More particularly, automated interfaces may be used to search for information available on the Web.

Is there an appropriate mechanism for asserting that one translation is better than another?