Jon Udell approaches Cate Gory creation via XPath query. Since he's counting on having used certain terms that he can query on, he's coming close to a Wiki Word approach. Though, since he combines queries on related terms, it's a bit more flexible, kinda like a Wiki Distance approach.

Then he talks about the idea of 3rd-party aggregators offering XPath across feeds. Which would require that query terms work for the varying ways different authors handle things (e.g. he catches book by looking for amazon links, whereas for me it would need bn.com). Which reminds me of Sister Sites models (assuming a large population of WikiLog-s).

User Options Recent Changes Help Page