Nepomuk serves as a cross application semantic storage backend. It aims at collecting data from various sources - file indexing, the web, applications, etc, and linking them all together to form a cohesive map of data.
To know more about Nepomuk from a user's point of view, head over to the Nepomuk page on UserBase.
Nepomuk is one of the core projects in KDE. It has a unique position of being an application framework, and interacting with a large number of applications.
If you're interested, come talk to use on IRC - #nepomuk-kde, or join our mailing list - [email protected]