Introduction


In the first phase of the project, along with initiating a community process, we are working on a catalog of requirements in coordination with public institutions, interested economical enterprises and citizens.

While working on our project, we were lucky to discover OpenOffice.org because the OOo file format seems to fulfill most of the requirements we have found until now.

What is missing from our point of view, is a more sophisticated, hierarchical meta data model which would enable document management systems, workflow management systems, archiving systems etc. of any vendor to automatically identify and process documents and application specific content like address fields, order numbers, etc..

Regarding the necessity of semantic structures, we share the opinion of the people of Linux Information Systems AG who came to similar results in their study already discussed on OpenOffice.org.

Implementing a meta data model would lead to a new paradigm for handling information contained in electronic documents inside organisations and in several situations of communication, such like B2B, B2C, G2C (government to citizen), G2G (government to government), etc.

Based on this technology a public institution, for example, may supply application forms for "something" (templates) which might be downloaded by a citizen, filled out and send back.
The returned document could be identified automatically as an application form for "something" by a workflow management system and an appropriate workflow might be started.

Even on the desktop computer of the citizen, a new associative paradigm of storing documents would be introduced:
The only thing the user would have to do to store the document would be to press the "store button", without further need to select directories or to specify file names.

By the office product (OOo) the document would be saved as what the one itself claims to be: an application form for "something", filled out by "somone" on "date", sent to "someone" etc.
Stored documents could be retrieved later on by specifying these same attributes.


Proposal


Extending the OOo file format.
We propose to implement the ability of to store semantically structured, arbitrary content specified by XML schemas inside the OOo file format.
The underlying meta data model should be hierarchical in an object oriented manner. That means that starting from a basic document which contains only rudimentary information, such like document class, author, creation date, version of the file format etc., more complex document classes like application forms, delivery notes, invoices etc. could be derived.
A document inside this hierarchical structure would inherit all attributes of its predecessors.

In order to show our intention clearly, we emphasize that the specification of the content is not part of our proposal. Our focus lies on the ability of the OOo file format to contain arbitrary content semantically described by corresponding XML schemas.
Information about application specific content could be taken from other specifications like EDIFACT, ebXML, openTRANS, DOMEA etc..

See www.1dok.org/eng/ for an illustration of the 1dok.org proposal for a meta data model.

Storing documents
Storing documents could be done in an associative fashion as described inside the introduction.

Editing forms
OOo might offer an XML schema driven toolbox of specialized input/editing controls.

If, for example OOo found an element type definition for an ISBN inside a given XML schema, it might offer a corresponding (specialized) text input field for ISBN's inside the toolbox for form controls.