A newly published study shows that despite a high degree of interest in web personalization, most search engine web sites offer few options that can be tailored to individual needs.
The ideal search engine, at least according to Google's chief technology officer Craig Silverstein, would resemble the computer on Star Trek, always responding to queries for information that precisely matched the needs of the user. Today's search engines are far from that ideal, according to the authors of Search Engine Personalization: An Exploratory Study.
The team of Pennsylvania State University researchers looked at 60 sites listed on the SearchEngines.com web site. They asked three questions:
How many search engine web sites offer personalization features?
What features can be personalized?
How accessible are the personalization features?
Only 13% of the 60 sites examined offered some level of personalization. And most of those personalization features were related to email, business and financial information, searching of a reference tool, such as yellow pages, entertainment listings, sports, and news headlines.
Why the lack of personalization features? The authors note that though personalization seems to be a simple concept, it's difficult to implement. By its very nature, personalization means different things to different people, and there's no one size fits all solution.
Accessibility of features is another important consideration. In many cases, personalization features are available, but they are buried deep within a site and difficult for users to locate.
All of these factors lead to a relative paucity of personalization features on search engine sites.
The study is rich in detailed analysis and graphs of key findings. It's a fascinating read, though with a couple of cavils.
First, the study was conducted in the first two weeks of May, 2001, so the findings are more than two years out of date -- an eternity in the world of search engines. This isn't a problem for some of the major players, such as Yahoo, Google and MSN, where personalization features haven't changed much over the past few years.
But the study includes many sites that are long gone, such as Direct Hit, nbci.com and Northern Light, and others that have changed substantially since 2001, such as HotBot and Excite.
Another quibble lies with the definition of "search engine" used by the study. While many of the acknowledged search engine leaders are included, the study also includes "portal" sites that tend to be aggregators of content, like AOL, MSN and so on, where the search function is incidental.
There are also non-crawler based directories like Yahoo and ODP, and sites that really aren't search engines at all, like Blink and Octopus, and many marginal search guides like aeiwi.com and efind.com.
While all of these sites admittedly offer help for navigating the web to one degree or another, referring to them all as "search engines" seems something of a stretch, and dilutes the usefulness of the conclusions. But limiting the study to major search engines wasn't the point of the research, according to Amanda Spink, one of the study's authors.
"I think the definition used in the study was fairly broad," she wrote in an email. "It was really an exploratory study of personalization in relation to the Web and search engines." The original analysis was done by students who have since graduated, but Spink says she will probably continue the work once she gets settled in at her new position at the University of Pittsburgh.
Minor quibbles aside, Search Engine Personalization offers an interesting, systematic analysis of the state of personalization in the world of web search. It's a useful examination of an important feature set of search engines, and how personalization may -- or may not -- help our efforts as web searchers.
About the Authors of Search Engine Personalization
Amanda Spink and C. Lee Giles are from the School of Information Sciences and Technology at Pennsylvania State University. Yashmeet Khopkar, Prital Shah, C. Lee Giles, and Sandip Debnath are from Pennsylvania State Univesity's Department of Computer Science and Engineering.
NOTE: Article links often change. In case of a bad link, use the publication's search facility, which most have, and search for the headline.