Tom Raftery has posted his podcast interview with Ross Mayfield at PodLeaders, well worth a listen. Ross, whom I met him briefly at Wikimania 2005, is CEO of SocialText.

To follow up on my question about Semantic Wikis, I think that sometimes there is the misapprehension that anything semantic has to involve some automatic AI-like deduction of metadata from the content by some agent or computer. A big part of the Semantic Web is enabling users to add structured content / annotations to pages (wikis being a good example here!) that can then be used to link things together (see the latter part of my IIA blog post on this). The Wikipedia page about Ross Mayfield links to about 25 pages - but it isn’t possible to get help with even a simple question such as “find me all the organisations that Ross has worked with or for”.

For example, the Semantic MediaWiki system allow people to add structured data into pages, such as typed links and attributes (or relationships and number / text properties). By allowing people to add such extra metadata, the systems can then show related pages (either through common relationships or properties or by embedding search queries in pages). These enhancements are powered by the metadata that the people enter (aided by computers of course, but not too much!)…