Information Typing Architecture (DITA) is an XML-based architecture used to organize, author, and deliver technical information. It is a standard approach to creating structure for online and onscreen information. DITA can be used for online help, web sites, and procedures, saving companies time and money, and increasing productivity when creating content for these and other purposes.

DITA was developed in 2000 by IBM as a way for various user-assistance teams to standardize the structure of information. Since then, DITA was released as an open-source tool, managed by the Organization for the Advancement of Structured Information Standards (OASIS), and is gaining popularity among technical communications professionals. OASIS is a not-for-profit, international consortium that drives the development, convergence, and adoption of e-business standards.

This architecture consists of a set of design principles for creating information-typed topic modules and for using that content in various ways such as online help. The rewards of DITA require some thought about what your content is, how you categorize it, and what you do with it.

The DITA solution has three parts:

—Authors need to rethink how to write content to thoroughly separate form from content. Rather than using a linear structure on nonlinear information, a writer should write in chunks or topics of information. Each chunk is a discrete entity and deals with a specific subject, signified by its title. It has no sentence or paragraph at the start to tie it backward and no sentence or paragraph at the end to tie it forward. Writing in chunks allows the writer to create certain types of topics: task information, concept information, or reference information. Each of these types of topics should contain specialized types of information with their own distinguishing features.

—Architects need to rethink how to classify and design information. DITA is an XML document type definition (DTD) that expresses many design principles. A DTD is a file that defines the allowable markup and markup rules for a particular markup language. Writers can concentrate on topic types rather than entire document architectures. The rules within a single topic are a lot easier to analyze than the rules that might apply across an entire book or web. Also, you can create new topic types relative to existing topic types. So, instead of defining the entire markup you’ll need and all the rules that apply, you will only have to define the new required elements and rules.

—Programmers need to rethink how to create transforms and processes, to allow content-specific information to be exchanged, and to make it easier to create and maintain specialized processes. To get the reuse and compatibility your specialized types require, you need to create processes around topic types instead of around document types and create new processes by extending existing ones and referencing general logic. As with content chunking and DTD design, reuse by reference is key. For processes, it means that if a fix or enhancement is applied to the original process, your new processes automatically pick up the change.

For more information regarding DITA, refer to http://www-128.ibm.com/developerworks/library/x-dita1/