Immediate Task List

There are three main areas that need addressing in the immediate future:

1. Develop the web-based prototype
2. Develop a web-based model of sharing of Structures, Views, Modules, Segments, ... (whatever).
3. Develop a basic security model for the web-based system (akin to the Moodle model of protection: lecturer, tutor and student)

But perhaps the most pressing need is to complete the documentation of what already exists ...

Medium Term Task List

Security

There is significant scope to develop a new model of security that takes into account the distributed nature of IVEE. Issues that come to mind that should be explored are:

1. How do you take account of different user group needs in a distributed system environment?
2. Can you avoid the centralized (username/password) model that dominates most systems today?
3. How can one share access information without compromising that access information?
4. ... (others?)

Database

Work is underway on exploring how Jackrabbit can be used to provide database facilities to IVEE. This needs to be integrated with the needs of a distributed environment, and the role of agents (as per the original PIAVEE model) re-searched.

Distribution of Views

The view model (aka structures, modules, ...) is a key concept from the original PIAVEE paradigm. Its definition needs to be refined and formalized, and the following issues addressed:

1. How are views saved?
2. How are views shared and copied around?
3. How is access to a view controlled? (relates to the Security issue above)

Text Mining Task List

The are a number of tasks needing to be completed that reorient text mining through enhanced clustering incorporating textual annotation methods:

1. develop algorithms that allow for intelligent text mining
   • Behrang's basic cluster implementation
   • adaptive decision making through fuzzy methods (?)
   • real-time clustering within text mining (Mohamad's work)
   • dynamic textual annotation as
     • an organizing capacity
     • a basis for subsequent clustering
2. Develop a user-based model of text mining taking this into account
   • The relationship between user behaviour and text mining intention
   • The structure of text mining as a search-and-brows activity
   • The importance of information overload
3. Develop models of interfaces that minimize user sophistication
   • Natural language
   • Speech recognition and speech generation
4. Implement algorithms within the intelliVEE environment
5. Testing of both algorithms and the overall model within educational and commercial environments