DITA (Darwin Information Typing Architecture) is an XML content model. XML stands for eXtensible Markup Language, which means there are many different “flavors” of XML, such as DocBook, S1000D, DITA, and more.
What defines DITA specifically as an XML content model is that it is based on categorizing topics into information types, meaning each document will be constrained to a limited subset of the content model. The main topic types are “task”, “concept”, and “reference”. The idea behind this is to make authoring of content more consistent. For example, instructional content is always written using the task model so that certain elements follow others, like steps after prerequisites. It can however be argued that it is hard to define content so rigidly, and that such topic typing can be over-restrictive.
The model does allow for specialization of topic types, to create your own DITA topic types in order to reduce the limitations and extend it to fit your needs. However, specialization can introduce another level of complexity in creating that information architecture and maintaining it.
Topic typing can be done quite easily in other content models such as DocBook, but it would then typically be an option to enforce it rather than being built into the model. The DITA open standard is developed and maintained by the OASIS DITA Technical Committee.