Getting started with DITA requires a multi-step process:

• Understand the topic-based architecture underlying the DITA model.

  Topic-based authoring is the basis of DITA. In order to understand the best approach to DITA, first understand how to work with topics and structured writing. For information about topic-based authoring see http://dita.xml.org/topic-based-authoring

• Create an information model.
  This is the best way to provide your employees with a guideline to follow throughout your DITA adoption and implementation process. An information model includes information such as:

  • Information Type descriptions
  • Content Unit descriptions
  • Metadata and folder/repository structure definitions
  • Naming conventions
  • Document definitions
    
• Read the DITA Architecture and Language Specification to understand how the model works.

  For help creating your information model, you may find the DITA Architecture and Language Specification helpful. You can find both of these at
  Architectural Specification
  Language Specification
  The Language Specification provides you with element and attribute definitions you can use as part of the DITA standard. The DITA Architecture Specification provides you with information about the features DITA provides such as content reuse, conditional processing, and specialization.

• Inventory your own content and transform it into topics.

  In order to fully develop your processes and information model, consider developing a pilot project. Your pilot project should be small enough to not discourage employees and other stakeholders from continuing the project, but also large enough to fully test your content with the DITA model and test your workflow processes.
  It's best to first inventory your content to identify consistencies in structure and information you can minimalize and reuse. Then, start marking up your content using DITA markup to identify special needs to fit your content.
  

• Download the DITA Open Toolkit and processing information from Source Forge (sourceforge.com).

  For processing, you may want to use the DITA Open Toolkit to create PDF, XHTML pages, HTML Help (.chm), Java Help, and Eclipse Help. The DITA Open Toolkit provides processing and rendering techniques to create your output.
  

Have Fun!

For more information, see Michael Priestley's DITA Introduction and Authoring Workshop presentation.