Content Localization & Gilbane Conference Boston Nov 28, 2006 C. Donner Whats this? Taxonomies.

Content Localization &

Gilbane Conference BostonNov 28, 2006

C. Donner

What’s this?Taxonomies

en_US

Traditional Localization …

Language code Country code

3

Language code: ISO 639-1,

Country code: ISO 3166.

E.g. 'ca_EN' and 'ca_FR'.

AF AFGHANISTAN

AX ÅLAND ISLANDS

AL ALBANIA

DZ ALGERIA

AS AMERICAN SAMOA

AD ANDORRA

AO ANGOLA

…

aa Afar

ab Abkhazian

ae Avestan

af Afrikaans

ak Akan

am Amharic

an Aragonese

ar Arabic

…

The codes

4

en_US

Language = English

Country = United States

Locale.setDefault( new Locale("en", "US") );

5

How Language experts see it …

(Globalization sells better than Localization)

Then, reduce Localization to the act of translating:

Globalization = Internationalization + n x Translation

(the GIRT equation)

6Slide by Pierre Cadieux (i18n.ca) - Reused with permission

7

Something is missing ….

Localization is not only about what language users see, but also about what content users see in different locales.

8

“Give the user the content that he wants,in the format that he understands.”

Localization

“Once he had the Babel Fish in his ear, Arthur understood perfectly.”

9

en_US?en_US?

?

?

10

A Localization Taxonomy

Where is it from?

What region is it for?

What region is it about?

In addition, there is a 4th localization dimension that defines the language.

Note that the dimensions define regions, not countries.

In the reference taxonomy for managed content in multinational organizations, every piece of content can be fully described by three dimensions related to geography:

11

Where is it from?

This can be the name or location of a regional office, for instance.

This information is not very important for the site user in most situations

It can be important internally to control workflow and track content production

Press releases and articles from News feeds are well-known examples of how the source of content is made available to the user/consumer

Still, I don’t know a single example of a website taxonomy that uses this dimension for search and navigation

12

What region is it for?

This dimension describes the target audience(s) for the content

It can be populated automatically based on business rules E.g. “the London office writes for the European markets”

Or manually, based on the judgment of an expert E.g. “this article is universal, therefore I will post it to the

global region”

A site typically uses this dimension to automatically retrieve content based on a user’s profile or selection of a region

13

What region is it about?

Content metadata in its original sense, this dimension discusses what is “in” the content

The most important of the three dimensions, this one is what the users need to conduct searches and retrieve meaningful results

It is also the most difficult one to populate, because the person tagging the content must be a subject matter expert

14

15

Traditional localization techniques deal with countries and languages

Country may be too granular or too coarse for your business objective

Markets do not always align with countries

Examples: EMEA, Asia, Central America

Country is limiting us

16

Region is better

For fine-grained control, let countries roll up into regions, forming a geography dimension

Regions should align with markets, countries can be added to allow for local specifics

17

Going global

The global level node of the geography dimension may not be at the top

Rolling up regions to a global category will create an unusable collection of content

Global can, however, be a separate and independent region, e.g. when there is a corresponding global market

Global content can be used for a global landing page that is transitional for most users

Region 1

Region 2

Region 3

Global

Region 1

Region 2

Region 3 Global

18

The cube

19

Real-world examples …

20

21

One-dimensional Localization

Language only

Everybody gets to see the same content

User can select the language in which the content is presented

Suitable for global organizations with global content

No regional targeting of content possible

Content must be translated into all languages

Using a uniform page layout is unlikely

22

Yahoo.de

Yahoo.fr

Today we plan to show you process, practice, and…

Yahoo.com

Yahoo.de

Yahoo.fr

24

25

Three-dimensional Localization

Content is not shared between Locales

Translations typically not required

User selects a Locale and gets content specific for this Locale

Used on News and Web portals with regional versions

Also used by global organizations with strong regional presences and autonomous regional content production

Today we plan to show you process, practice, and…

A closer look …

Compare: Sports, World, Karikaturen, etc.

27

28

3½-dimensional Localization

Categories differ across Locales

Some similarities across categories can be deceiving, but:Even differences in the order are important

Where is Sports on the German site? In German media, Sports is not considered News

Where is World on the German site? German media often does not distinguish between

National and International politics

29

30

Complex n-dimensional Taxonomy

Target region “Global”

Language “English”

Content category – up to 5 levels (3 in the example) “Indices” – “Equity” – “Italian Indices”

Content type “News & Analysis” – “Index News”

Content region “Europe”

31

microsoft.com

Several years ago, Microsoft had a diverse set of loosely integrated local sites

There were common design cues, but specifically the content was not shared in any way

Imagine the cost of maintaining a local site in every country Microsoft does business in

Today, there is a single design, taxonomy, shared content, translation process, all the works.

32

From the WayBackMachine …. March 2000

33

The Future: Convergence

34

Take-away

Content Localization

Taxonomy Localization

Web CMS Localization

35

Thank you!

Q&A