Information is of value only if you have a way to find it, at the first place. I had a look at current e-learning repositories or websites with loads of education-related resources. They are generally very poorly structured.

I am a psycholinguist and I have some knowledge of the machine learning techniques. The idea is to combine a system in which the user defines its own keywords (furl and delicious... similar to amazon, the users who choose what you have just selected also choose x and y) and some techniques for the automatic classification of documents by genre and content.

Notes:

• Ontology notes and links on the revolution wiki

Realized so far:

• Education domain (including both standard and computer-based education)
• Experimental Psychology (to capture the content of this wiki)

Models:

• Text Classification Techniques
• Hyperlex (paper in pdf format) (new ways of presenting an interface to query the semantic field)