|0. Project Management|
|1. Requirement Analysis|
|6. Testing and Validation|
|7. Information Dissemination and Exploitation|
The previous work packages are not strictly sequential.
WP1 should be reasonably short; apart from a few topics requiring a deeper analysis (Tasks 1.3-5), this phase is essentially meant to rapidly reach a good level of inter-operability among the different sites.
Most part of the work is based on the possibility to have at our disposal, and as soon as possible, large collections of documents encoded with semantic markup. One strategy is the import of material (e.g. journal articles) written in LaTeX. The development of a suitable LaTeX based authoring tool (Task 4.4) will need to begin immediately, as an appropriate semantic encoding in LaTeX has to be developed first. The delivery of the first prototype of the authoring tool is scheduled for month 18.
A more rapid way to get meaningful repositories of fully structured mathematical knowledge is by exporting them from the available libraries of Logical Frameworks and Proof Assistants (Task 2.1). The intelligence contained in the exported XML files should reflect the requirements defined in the previous work package, requiring a deep analysis of the markup model. After six months from the beginning of the project we plan to have a first prototype of the Exportation Module and a first draft of the Document Type Descriptor for thelow, logical level. This is our first Milestone.
At this point we may start the study of the intermediate format of the information, and the implementation of the stylesheets performing the transformation (tasks T2.2-3). This part of the work is expected to be essentially completed after one year (second Milestone). Since a strong feedback is expectedwith presentational issues, we plan to begin the development of presentational stylesheets around month 9. In turn, the need of rapidly have at our disposal good presentational engines suggests to begin their development as soon as possible (task 4.1).
In parallel with these transformation issues, we shall start the study and classification of metadata, and their concrete modelling.
Summing up, at the end of first year we plan to have:
During the first half of second year we shall perform, in parallel, four major activities:
The second half of the second year is devoted to the completion of the previous tasks, and to their integration inside a single, compound application. Around this time we shall also start a detailed validation of the application, according to three pilot applications:
The last six months are mainly devoted to testing, debugging, validation, dissemination of results and exploitation plans.