Mikko Virtanen
Distributing data to different platforms using XML
**************************************************

The ATBC Cancer prevention study is large and complex cohort study with long
history. The path from the raw data to publications is described. Easiest way
would be to standardize the used statistical package and provide methods to
access the data for this package only. In practice the skill level of those
doing the analyses vary and several packages must be supported. Extra care
must be put on ensuring that the versions of the data are same. The XML
offers structured way of combining the definition, documentation and the
revision history of each data item. These definitions are then transformed
using the XSLT to code, meta data (help files) and documentation (HTML and
PDF). The current version defines the data structures in a rather restricted
way, but the extension of the same idea to analysis results will be discussed
briefly.