Trust is essential in our reliance on search engines. But we should understand some of the risks in placing too much trust in search results.
There’s the possibility of bias in what search engines show people based upon the engines’ business practices and operating policies, limitations in indexing and ranking algorithms, and in political and cultural pressures placed upon them.
When I think of conferences like the one to be held next week in Edinburgh, Scotland, during the 15th Annual World Wide Web Conference, I don’t expect to see presentations that are critical of search engines. But, during a workshop on Models of Trust for the Web, there’s a paper being presented that takes a close look at search engine bias, from a couple of researchers at Yuan Ze University in Taiwan.
Position Paper: A Study of Web Search Engine Bias and its Assessment (pdf) by Ing-Xiang Chen and Cheng-Zen Yang
The authors of this paper describe in more detail the three different sources of bias that I mentioned above. How could business practices shape the bias of search engines?
The first source is from the diverse operating policies and the business strategies adopted in each search engine company. As mentioned in , such type of bias is more insidious than advertising. A recent hot piece of news demonstrates this type of bias from the event that Google in China distorts the reality of “Falun Gong” by removing the searched results. In this example, Google agrees to comply with showing in China to guard its business profits.
The next source involves how easily some might find it to manipulate search engine results:
Second, the limitations of crawling, indexing, and ranking techniques may result in search engine bias. An interesting example shows that the phrase “Second Superpower” was once Googlewashed in only six weeks because webloggers spun the alternative meaning to produce sufficient PageRank to flood Google.
The last source listed involves pressures put upon a search engine to conform with the laws and influences of the societies they find themselves within:
Third, the information provided by the search engines may be biased in some countries because of the opposed political standpoints, diverse cultural backgrounds, and different social custom. The blocking and filtering of Google in China and the information filtering on Google in Saudi Arab, Germany, and France are the cases that politics biases the Web search engine.
The paper comes up with some ways to look at, and measure search engine bias. There’s some math, and some graphs in their analysis, but it’s worth struggling through or skipping past to see how the ten different search engines they look at rank in their study.
It’s somewhat interesting that Google happens to be the search engine that they use in their examples of bias because it is also the search engine they determine to have the least amount of bias amongst the ten included.
Some other papers on Search Engine Bias
I enjoyed this paper from Eric Goldman of Marquette University Law School – Search Engine Bias and the Demise of Search Engine Utopianism. I’m not sure that I agree that personalization will lead to less search engine bias, but he presents a thoughtful argument which includes many of the most important documents on the topic in his presentation.
Another paper, in this case a very long undergraduate thesis paper, does an excellent job of presenting issues involved in search engine bias – Through the Google Goggles: Sociopolitical Bias in Search Engine Design. (original link no longer available.)
The author is Alejandro M. Diaz, and he was recognized by the Dean at his school (Stanford) for the paper by being awarded The Firestone Medal for Excellence in Undergraduate Research in 2005, along with a few other undergraduate students.
The author does an excellent job of locating different choices in bias that exist, and while doing so, makes a statement early on that it isn’t a question of “not really whether bias exists, but rather which sort of bias we prefer.”