One of the original use cases for topic maps was the creation and merging of traditional indexes for printed materials. The range of uses for topic maps has expanded to include archives and libraries, image collections, navigation of European administrative nomenclature and software development. And the list continues to grow with each topic maps oriented conference. But what of the original use case with indexes?

The recently approved Ecma standard, Office Open XML (OOXML) is 6,036 pages long in five separate parts. How useful is it really to be able to search for w:p and to find it occurs 775 times in more than 6,000 pages? Granted that a useful index could have been constructed for this standard, but it wasn’t.

This presentation addresses three separate but closely related issues:

• How to extract topics from OOXML markup.
• How to extract associations from OOXML markup despite the lack of customary XML structures.
• Creation of a topic map based index for the OOXML standard.

Some of the material will be specific to the OOXML standard per se but for the most part the techniques to be demonstrated are generally applicable, assuming proper document analysis, to any OOXM