CMS 2005: Content Categorization and Typing
Publishers organize the content inside
their content stores by categorizing and typing their content. This involves the
creation of hierarchical categorization systems, similar to those found on
Yahoo! or a file system, into which content items such as articles and images
are placed. In addition to being placed into a categorical tree, content items
are nominated a type from a library that the publisher maintains, such as
"News" or "Review".
For example, a /Marketing category might
be created, into which are added sub-categories such as /Marketing/Direct Mail
and /Marketing/TV. In a similar way that files are added to folders on a
computer’s disk drives, items such as XML documents, multi-media files and
content bundles are then stored within these hierarchical categories. And, just
as files on a disk drive have types, users can assign nominal types to their
content. For example, inside Direct Mail there might be “News” items containing
news related to direct mail and “Events” items relating to forthcoming
exhibitions and events for marketing professionals.
So that publishers can organize content
in a manner that reflects the real world, which facilitates the provision of
powerful content discovery systems on interactive publications, and the
implementation of sophisticated syndication systems, the System 7 CMS
categorization system enables publishers to create a structure known in
mathematical circles as “directed acyclic graphs” (rather than simple category
trees). Although the term sounds intimidating, it simply means that every
category can have multiple parents, and furthermore, every article can belong
to multiple categories.
For example, a content store containing
articles related to business might contain root categories “Entrepreneurs” and
“Women at work”. The System 7 CMS categorization system would make it possible
to create a single “Female Entrepreneurs” category that is simultaneously a sub-category
of both. This means, for example, that if a pure business magazine only
syndicated “Entrepreneurs”, and a women’s magazine syndicated only “Women at
Work”, both magazines would automatically receive content added to the “Female
Entrepreneurs” category.
The combination of sophisticated
categorization with nominal typing can be used as the basis for a wide range of
content discovery features on interactive publications. For example, a website
section might list only the articles inside a category with the nominal types
of “Features”, “Resource Articles” and “Reviews”. In this case a Web developer
might then add website code that tests whether articles with a nominal type of
“News” have been recently added to the category, and automatically display a
link “Show related News” if they have.
In summary, the provision of directed
acyclic graph categorization functionality together with nominal type
assignment provides huge benefits for the serious publisher. The categorization
of new content is more straightforward, as it need not be artificially
shoehorned into a single category, the management of syndication channels is
greatly simplified and websites can be made more compelling through the
delivery of powerful content discovery features.
next
|