More like an unhappy disaster, when one realizes that link popularity merely rewards a site or page based on a "measure" that can be artificially manipulated. The notion of "relevance based on link popularity" punishes new sites - especially the smaller ones who do not have the money to generate links quickly. More damning, the entire link popularity scheme effectively means that Google cannot ever effectively or fairly rank news and other "right now" pages or content.
It MAY possibly be true that "popularity" can be a measure of relevance - but many a Libertarion or Free Speech advocate would not agree. But that is not actually "relevant", here, since Google's ranking methods do NOT reflect true popularity.
What Googles's success demonstrates is the poverty of search technology - where relevance is not really based on matching content to the user query, but on the hope that the poor searcher will be sufficiently overwhelmed by a zillion "results", or on persuading the searcher that a bunch of top-of-the-page sites have so many links that they "must be what you want". Jeesh! Give us a break already!
Gush about Google if you must be part of the faddish elitist mob, but it is a *lousy* "technology".
Context-relevant ads are the holy grail - at least for those of us who still prefer an ad-supported Internet to no Internet or one where every site has a "members only" sign on the front page.
The Google "solution" fails so often because the contextual linkage is based on a word, not - sorry for this - a meaningful context. The ad-selection engine is keying into the word the user is using, as opposed to the topical milieu the user has in mind.
All existing media who deploy ads using this basic method are subject to the same high incidence of non-relevance.
Happily, a new engine is in protoytpe that definitively solves this problem, and will enable users to be messaged by ads that are very precisely related to the actual search context.
Not that that will make the users any happier, of course - but if the advertisers are happier, then we have at least that much better chance of seeing this whole wonderful thing survive and grow.
Slashdot Mirror
More like an unhappy disaster, when one realizes that link popularity merely rewards a site or page based on a "measure" that can be artificially manipulated. The notion of "relevance based on link popularity" punishes new sites - especially the smaller ones who do not have the money to generate links quickly. More damning, the entire link popularity scheme effectively means that Google cannot ever effectively or fairly rank news and other "right now" pages or content. It MAY possibly be true that "popularity" can be a measure of relevance - but many a Libertarion or Free Speech advocate would not agree. But that is not actually "relevant", here, since Google's ranking methods do NOT reflect true popularity. What Googles's success demonstrates is the poverty of search technology - where relevance is not really based on matching content to the user query, but on the hope that the poor searcher will be sufficiently overwhelmed by a zillion "results", or on persuading the searcher that a bunch of top-of-the-page sites have so many links that they "must be what you want". Jeesh! Give us a break already! Gush about Google if you must be part of the faddish elitist mob, but it is a *lousy* "technology".
Context-relevant ads are the holy grail - at least for those of us who still prefer an ad-supported Internet to no Internet or one where every site has a "members only" sign on the front page. The Google "solution" fails so often because the contextual linkage is based on a word, not - sorry for this - a meaningful context. The ad-selection engine is keying into the word the user is using, as opposed to the topical milieu the user has in mind. All existing media who deploy ads using this basic method are subject to the same high incidence of non-relevance. Happily, a new engine is in protoytpe that definitively solves this problem, and will enable users to be messaged by ads that are very precisely related to the actual search context. Not that that will make the users any happier, of course - but if the advertisers are happier, then we have at least that much better chance of seeing this whole wonderful thing survive and grow.