#3 Duplicate Listing Checking
Add the ablity to determine if two listings are likely
to be duplicates and flag them for the administrator to
join them. This will in the long term be useful when
directories are spread out over several servers and we
have cross-directory syndication. Even with out that
often the same information gets added twice.
Ways to determine duplicate entries.
Check geographic location. If two entries are in the
same city, state, and country then they are much more
likely to be duplicates. Once we start comparing
listings within a city we can trying to run some
analisis on the the organization description and meta
information to determine similarity.
A good running code example is at:
http://laughingmeme.org/archives/000547.html
Even if we don't remove the entries for duplicates,
having an automatically generated 'similar
organizations' link would be cool. It could be
suplimented by a friendsters style 'this other
organization is my friend' type designation. Then when
looking at Indymedia St.Louis you could see they fellow
traveler organizations which share the CAMP building
they co-own.
