A panel entitled "Search by another name: New ideas in Search" didn't yield much in the way of new ideas or Web 2.0 breakthroughs. The panel highlighted search companies that are in the shadows of Google/Yahoo/MSN with more specialized engines.
Michael Tanne introduced Wink, an early stage search engine based on user tagging from its own site as well as tags from sites tags on del.icio.us and Flickr. It a more democratic engine, derived from atomic individual interactions of users with the Web and with search, Tanne said. Wink's "tag rank" concept is analgous to what Google's PageRank was in the late 1990s--a way to improve the accuracy and freshness of search results, he added. For now, Wink is providing both its own tag-based results and Google's for queries. However, without a open standard for tagging information, it will be difficult to get a big enough data set to deliver quality results. Tanne said a group effort, called TagCamp, is tackling that problem as well as creating a tag taxomony that could match related keywords. Interesting idea...long way to go.
From left: Bob Wyman (PubSub), Paul Forster (Indeed.com), Michael Tanne (Wink), Rahul Lahiri (Ask Jeeves)
Bob Wyman of PubSub talked about an open source effort to get structured information on the Web, a la semantic Web. By tagging parts of a blog post, for example, such as titles, names, rankings, etc., you get both machine and human readable structured information. "It's about making the Web easy for the computer to use," said Wyman quoting Amazon's Jeff Bezos. Definitely a good idea. More to come on this topic...
Update on structured info on the Web: The open source organization is called Structured Blogging, and it will initially be supported by WordPress, MoveableType, Drupal, and Reger.com blogging platforms. PubSub is working with Marc Canter's Broadband Mechanics to build PubSub's Structured Blogging prototype.