|Search Engine Bias|
| 9:36 am on Mar 12, 2002 (gmt 0)|
In an excellent article [searchenginewatch.com], Chris Sherman reports on a independent study aimed at gauging search engine results biases.
|Two computer scientists from the City College of New York have set out to define and measure bias in search engines. |
We've been looking and waiting for such studies to be done for years. Up until now, the search engine industry has been run by cowboy management. There has been no regulation, oversite, or fences that the search engines have had as a guide.
|The authors have a point, namely that the potential for bias in search results is greater today than ever before. On the other hand, no search technology, or for that matter, paper finding tool exists without bias. |
| 1:09 pm on Mar 12, 2002 (gmt 0)|
This article from a Princeton and London School of Economics Professor, first published in 2000 might also be of interest.
Title: "Shaping the Web: Why the Politics of Search Engines Matters"
From the journal: The Information Society, 16(3):1-17, 2000
| 4:11 pm on Mar 12, 2002 (gmt 0)|
I want biased search engines...I want them to all be different, to have different slants on the query
otherwise we may as well just have Google and let the others fade away
the idea that there is a single perfect set of search results for a query is obviously nonsense if you look at it in practical terms...the information is being sought for a purpose...that purpose biases the searcher...in an idea web they can then pick the search engine that has the same bias
there is far too much emphasis placed on the idea that all search engines should be aiming to be the only search engine you will ever need
| 4:33 am on Mar 13, 2002 (gmt 0)|
Good points Eric. Even one, "totally unbiased search engine" to serve the entire web community would be undesirable, though Google might be it one day.
Unfortunately most of the others seem to be biased by who has the biggest pockets. It would be nice to see otehr things skew the results.
| 11:47 am on Mar 13, 2002 (gmt 0)|
My understanding is that audio engineers don't mind using biased audio speakers for mixing and mastering (i.e., speakers which misrepresent certain frequencies, either over- or underemphasizing them). However, they have to know exactly how the speakers are biased, and that takes time to get to know the speakers and suss it out.
The analog to this (no pun intended) is that, yes, it's no problem working with a biased source of info, but it takes time to discern exactly what the biases are. Just part of being an educated person, I guess...