'The Filter Bubble' -- the danger of web personalization

Started by littleman, May 20, 2011, 05:55:39 AM

Previous topic - Next topic

littleman

This was an excellent conversation on the social problems caused by personalized search results and media exposure.  In the interview Eli Pariser explains the corporate motivation in pushing personalized results and how it is at odds with the public good.

Worth the listen:
http://www.kqed.org/a/forum/R201105191000

inbound

Indeed, personalisation is a double edged sword. If a service knows what I like, and only delivers me things it thinks I'll be interested in, what likelihood is there of me discovering something new (and a bit or a lot different) through that service.

Furthermore, being constantly fed the same viewpoint is dangerous(which is what could happen if meaning is successfully extracted from the sources on offer); we can see at all sorts of examples where people choose to read views which are similar to their own (newspapers are a great example of this) - the last thing we need is for the web (a great source of conflicting views) to be divided up so that people believe that they are right because all they see agrees with them. Such a setting is a breeding ground for extremism (and I'm not just talking about religion) - an example is the peak oil vs abiotic oil argument; either side emerses itself in its own literature and can develop a fierce hatred of the "opposition" (although I have to agree that those that think oil is abiotic are nuts...)

But...

This does remind us that a good way to repeat traffic is to be on the extreme edge of an argument, or at least consistent in reporting a (relatively) commonly held viewpoint that is under-represented in the mainstream media.

dogboy

Personally, I think 'bubble' or 'framing' issues like these are sometimes the worst types of problems to initially identify. But once 'the bubble' snaps into focus, the resolution usually does as well.

I worked on an art print site where you would rate images and compare your results, to other Facebook friends. As a result, everyone ended up rating all the same images, and ignoring the majority of everything else out there:) It took a little while to figure this out.