Table of Contents

Introduction

This document presents a catalog of patterns to publish Multilingual Linked Data and identify some issues that should be taken into account.

The patterns are classified according to the common tasks that have to be performed to publish linked data:

Naming

This task refers to the process of URI design and dataset description from a multilingual point of view.

Dereferencing

This task describes patterns to handle dereferencing in a multilingual environment. For example, should we return the same or different representation depending on the language?

Labeling

When publishing linked data, there are a number of reasons to label different resources. In a multilingual setting, how can we provide labels for different languages?

Longer descriptions

We consider that resource labels are different from longer textual information. Not all textual information attached to resources are labels and in fact, longer descriptions, like comments, or even book chapters, can be encoded in literals and contain textual information in different languages.

Linking

In a multilingual setting, it is possible to have different resources representing the same thing in different languages. How can we link those resources?

Reuse

Linked data is all about reuse of data. When multilingual data is linked to vocabularies, is it better to have multilingual vocabularies or to localize existing ones?

Catalog

More information

• MLOD Patterns paper
• Slides of talk: Multilingual Linked Open Data Patterns (W3c Multilingual Web Workshop, Rome)
• Slides of talk: Best practices for Multilingual Linked Open Data (MLODE 2012)
• Slides of talk: Best practices for Multilingual Linked Open Data (W3c Workshop on Multilingual Web )
• Video: Video on Best practices for Multilingual Linked Open Data (W3c Workshop on Multilingual Web )

Acknowledgments