Valdes-Peres concludes (and points I tend to agree with):
If not search personalization, then what? Many companies, including my own, are placing bets on a display of search results that goes beyond simple ranked lists. The idea is to analyze the search results, show users the variety of themes therein, and let them explore their interests at that moment, which only they know. The best personalization is done by persons themselves.
In other words, let the content on each open-web page speak for itself. I’m still not sur that the masses will begin tagging all of their web page content.
Of course, in the specialty database arena, dynamic clustering can be enhanced with the metadata (aka cataloging info) associated with each record often applied from controlled vocabularies and professional catalogers. Vivisimo’s ClusterMed shows this in action with the ability to cluster records in different ways based ot the metadata associated with each PubMed record
John’s Searchblog post includes several comments agreeing with Raul’s ideas ((to my somewhat surprise). One of those comments includes a link (from me) to another excellent and brief paper by Vivisimo titled, “Needed: A More Selective Ignorance (PDF).”
Finally, one of Vivisimo/Clusty’s biggest challenges is what’s known a s the dreaded learning curve. Their software can be powerful and useful but to some degree (not much) you have train people how to take full advantage of what’s available. In an age where it’s “type two words and that’s your answer”, Vivisimo’s challenges are as much about developing great technology as they are about training. I can speak from experience that once people learn and understand Clusty, they love it. Another challenge? What happens if/when other companies debut clustering techonology. Sure, it’s possible that Clusty might be the best in class but trying to keep-up with the pr juggernauts that others possess will also be a challenge Clusty must face.
See Also: FirstGov Switches to Vivisimo & MSN