Google Blogoscoped

Friday, February 15, 2008

Search Engines in 1996

In October 1996, German Spiegel mentioned web search engines for the first time (I’m using the homemade search tool to dig through their archive). The article shows how much of a need there was for a good search solution. One approach to organizing the information was the manually categorized directory of Yahoo and others. From the article:

The last serious attempts at sorting all the world’s knowledge into publicly accessible categories already failed during the last century. (...) Ironically, at this very age of information overload, the dream of a world encyclopedia is seeing a renaissance. (...)

[Yahoo] employs 50 full-time “surfers.” They browse the web looking for unmapped territory, adding their findings to the collection.

Yahoo already stores over 300,000 addresses in 14 main categories. Like “art,” “society,” or “sports,” which currently branch into 20,000 sub-categories. Step by step, users are supposed to track their search tracket, moving from general to more specific topics.

Attila Boromissza, then head of the German Yahoo division porting the directory for local sites, is quoted saying, “Until today I don’t have a computer of my own. But that’s not what counts here.”
The “impossibility of their task doesn’t scare the Yahoos away,” Spiegel writes, because it solves one of the crucial problems of traditional paper-based systems: every entry can easily be sorted into multiple categories. “This approach sometimes helps tackle the confusing logic of web surfers,” Spiegel says.

Yahoo in October 1996, courtesy of the Wayback machine

Another approach of the time were search engines with automated crawlers, more like the kind we know today. However, result quality was still lacking enormously, as Spiegel reported:

Around ten larger services, as well several smaller organizations, are currently crawling the web using tracking programs – the so-called “spiders.” These little insects patiently crawl from cross reference to cross reference, down to the last branch. When they return, they add their prey to the central archive. (...)

However, this search is best only used for very specific queries. For more general search keywords, the automated archives return a dauntingly large information wave. Search engine “Alta Vista," for instance, replies with 400,000 entries when looking for information on films (using the keyword “movie”).

Note how back then, the focus of discussion was more the quantity of results than the quality of the ranking. It was almost taken for granted that search results were a mere listing of what’s available – the search strategy of the day being to use very specific search queries to manually pull what’s relevant. The Spiegel author went on:

Furthermore, the advice of these robots lack any human logic. For instance, the supposedly German version of search engine Lycos ... serves a Norwegian-language top hit when searching for “film.” Netguide ... spits out a link to a laser graphic company from Mainz university – the best answer according to the machine, listed even before the electronic address of the Berlin film festival.


Blog  |  Forum     more >> Archive | Feed | Google's blogs | About


This site unofficially covers Google™ and more with some rights reserved. Join our forum!