Some notes around strategies for yellowpages that may help. I have found the IAWiki to be a useful resource. Related concepts you may wish to visit are expertise mapping, taxonomy and ontology.
The key issues here as I see them are:
* getting all users to buy-in on central domain categories and issues, finding a common language, controlling term variations
* finding an easy way for owners to keep those profiles updated and relevant (motivational issues)
* including relationships and associates, not just domain interests and educational achievements in the mix
* expressing and elicitation of learning desires
Including touch-points e.g. hobbies, professional associations and
informal community membership, locale, recreation activities, photos of
pets certainly helps.
Automated, content driven, user categorization is a rather poor lead
to an individual interest profile construction. Tools that parse a
wider variety of electronic messages and documents (IM, e-mail, forums)
perform better, but are more intrusive. The key to making advances in
this area, in my experience, is to craft an upfront ontology, to have a
blended approach using automatic concept extraction, but incorporate
expert supervision / review of final category allocation.
The most difficult aspect of crafting corporate yellowpages, is
getting buy-in for the skills, competency classification and standard
nomenclature. Topic Maps and XML are useful tools in this area.
More thoughts on yellowpages from my KmWiki [Knowledge-at-work]