The Filter Bubble
Limiting the opinion formation or reducing the mass of information?
Already used by the big internet companies for years, it is still unclear whether the filter bubble is isolating peoples views or saving their time. However, the fact that this phenomenon is often not known by those who are affected by it daily could be a problem.
The “Filter Bubble” is a term coined by Eli Pariser (http://www.thefilterbubble.com/) and describes the process of websites using algorithms to select which topics could be relevant for an individual user. Previously stored information about the user induces the choice of topics shown to him next. Location, search history and past click behavior uncover the main interests of the user and allow server to get a hint about what their clients might wish to see in the future.
And, by selecting the issues in that manner, which subjects could be neglected. Thereby contrasting views can disappear and isolate the user in his own, small “Filter Bubble”. This may have negative implications for civic discourse, according to Eliser. Besides, it can reinforce confirmation bias, the tendency of humans to favor information that supports their opinions.
The two most famous corporations using these algorithms are Facebook and Google. Every click of the user is stored and can lead to specific reactions in the future. Eliser describes an example of two people searching for BP with Google and receiving two different topics. One got financial issues, the other got news about the Deepwater Horizon catastrophe. Another example should have happened on facebook: By repeatedly ignoring the updates of a liberal friend, facebook simply removed everything similar from the users site.
Nevertheless, there are contrasting views as well.
Several people like Jacob Weisberg, Paul Boutin or Jonathan Zittrain question Elisers opinion about the extent and/or consequences of the “Filter Bubble”. They tried small experiments like telling some people to search for a specific word and compare their results. They were nearly identical which led to the conclusion that the effect of the “Filter Bubble” is overestimated by Eliser.
Besides, other critics mention that the mass of information today makes filter techniques inevitable. There is no alternative. Paul Resnick states personalization is not bad in principle. It is necessary and because of the possible great influence, the people in the background have a responsibility towards the society.
A study from The Wharton School, University of Pennsylvania comes to the conclusion that personalization does not fragment the online population but could even help the user to widen their interests.
In conclusion, it remains unclear, whether the “Filter Bubble” has a big influence or not. On the other hand, it is obvious that filter are necessary to handle the fast growing information. This applies to the internet in particular.
There exists a similar phenomenon called the relevance paradox. This is about searching information and while doing so deciding which is important and thereby possibly neglecting the really relevant sources. See also: http://en.wikipedia.org/wiki/Relevance_paradox for further information.