| Recent
Articles |
White Paper Business Intellectual Property Overview
Marks are generally an identifier of source and quality. Rights generally accrue
under common law upon adoption and use of a mark (name, symbol, logo, trade dress...
Business
Podcast Marketing Case Study Proves Results
Business Podcast Marketing Case Study Shows How Podcasting Delivers Dramatic Results
for Client. Podcasting has significant business marketing potential...
6
Ways to Leverage Technical Articles
Technology vendors often contribute bylined articles to trade journals. The articles
are great exposure for these companies but they don't come cheap - the trades
rarely pay for these articles but the vendors spend time and resources to assign
pieces, write them...
Another
Way to Narrow Your Margins
With the usual press release-driven fanfare, reserach company eMarketer has released
a new study, "The Business of Blogging." Asking price: US $695. According to eMarketer's
hype, the study covers...
Google
Patent Study
Two months ago, Google released a 63-point patent document outlining how it examines
historical data associated with websites...
|
|
|
12.02.05
Clusty Has No Lust For Personalization
 By
David Utter
Providing personalized search has become a big focus for sites like Yahoo and
Google, but the CEO of Clusty.com's parent company Vivisimo thinks search personalization
is a dead end.
"Search personalization is likely to waste the talents
of top computer scientists," wrote Raul Valdes-Perez in a whitepaper
pointed out by John Battelle.
Valdes-Perez lists five reasons why the quest for more users of personalized search,
one that has Google and Yahoo
promoting search personalization heavily, is a waste of time. We'll summarize
those reason here:
1) People's interests change frequently. 2) Search engines use weak data to personalize
search. 3) Using data from the whole page visited from a search may be misleading
because people visit sites based on a title and a short snippet of text in the
search engine result. 4) In households, frequently shared computers among family
members could skew results. 5) Short queries don't give the search engines enough
to determine one's interests.
Point number three could be the strongest point Valdes-Perez makes, due to the
proliferation of spam blogs and scraper sites in search engine results. Someone
clicking on one based on its entry in a search result page and finding it's a
useless site will exit it quickly. Will personalization take those quick visits
into account?
The argument for point one, where interests change, makes sense from a short-term
perspective. "Seasonal phenomena like elections, the Olympics, sports leagues,
etc. also lead to variable interests: I'll follow the Olympics for the next month
or so, but will pay no attention for another four years," Valdes-Perez wrote.
But the ongoing use of personalization over a long-term should build a more accurate
search experience. That of course depends on the skill of the engineers who create
that experience.
Points two and five are essentially the same. Short one or two word queries make
for weak data. Again, that could be mitigated over long-term use. The fourth point
doesn't apply to personalization schemes like Google's Personalized Home or Yahoo's
My Web 2.0 because both of them require a user to login. Users might forget to
logout from time to time, though, but probably not enough to mess up the results.
About the Author: David Utter is a staff writer
for WebProNews covering technology and
business. Email him here. |
