langreiter.com plain, simple: comments-2005-12-27

LANGREITER.COM plain, simple

START

Dienstag, 27. Dezember 2005

Greg Linden: "[Findory is designed] to show people views from across the spectrum, to give people the information they need to make an informed judgment. For some, that is exactly the problem."

2005-12-27-echoLens

slauti 6905 days AGO
but then, look at wp vs. goog: most (update: i definitely did not mean you) would say they are precisely opposite means of organzing k, i or whatever. one did it by watching and analyzing (unconscious) behaviour, the other by (consciously) organize it under the guidance of old fashioned litaral principles. Most of the time they reflect each other nicely. Nearly every relevant article in wp turns up in the top 3 to ten in goog. most items relevant in goog get an article in wp.

chris 6905 days AGO
I certainly wouldn't say that they are precisely opposite. How is linking on the "macroweb" less conscious than (local) linking within a microweb?

As for k/i organization, that's an interesting question by itself. As far as I am concerned, the only relevant kind of organizational principle Google applies to the information it amasses is (page)ranking(+). Wikipedia articles are embedded in a rich structural framework of categories and systemic links; I think the real value of those remains to be explored as of yet.

The problem Greg seems to address, however, is that more people than he thought expect filtering systems to apply _political_ bias, and in general that people always want more of the same (easy to digest - resolving [complex] conflicting viewpoints is [extremely] hard [cognitive] work!).

In traditional media who take their role seriously, selecting and mixing content (esp. front page content) is no easy affair (I suppose). When consuming information primarily from self-selected ultranarrow sources (as happens with weblogs), are there negative effects? For some who can't and/or don't want to take a walk outside the echo chamber, there sure seem to be, and we should work harder to understand the multitude of biases every filtering system introduces.

slauti 6905 days AGO
Maybe I did not express my view explicitly enough. Allthough am not sure about the + in googles organizational analysis (weighting factor tables from offline analysis e.g.), not from their older papers and not from what I read now, I would have expected political bias from any large-scale system. I would always expect political bias to stem mainly from large-scale agreement processes, more than from anything else.

My guess (educated I hope) is that only societies where important media really are not accessible for all or at least the majority of the people allow for strange compromises between different elite and nonelite groups where you can recognize a dividing gap that makes for large easily recognizable differences. Whoever's got access to the internet, most probably also has access to tv, print and acoustic media.

The digestion problem is large and underestimated by many. And help for it comes in many forms. Some people I know share the perception that biased opinioned media (tv, radio, papers) make large inroads on the intentionally weighed out ones. Now the general intellectual pub-theory would be that information glut and the "emptiness" of relativism drive people there. I'd rather call that a sort of regrouping that comes with the still ongoing and slow decline of the dominance of traditional written language and its inherent structures and by that of nation states, even the US.

And so organisations who's business is the construction of meaning (where GOOG and wp are waxing stars and might not need it) might offer readers tools to accelerate that construction. But only large scale visible information (or rather meaning) can easily be suppressed. So I have no great fears there yet. An explicit slider could be less dangerous than what I feared from personalization only a short time ago.

From the point of view of media ecology it might be more important to expose yourself to different media than to different content. And, I agree, we should work harder to understand multiple bias. And I'm glad you link Greg often.

Please log in (you may want to register first) to post comments!