gusl: (Default)
[personal profile] gusl
I think it is well-known that the news media copies stories from each other (presumably with enough changes to not seem too blatant*, but nevertheless the typical lack of citations would suffice to be hung for plagiarism in any academic field). This process can be described as a DAG in which the nodes are stories and the arrows represent copying, and only some nodes are observed (printed). If you abstract over the rewordings, a diagram displaying merges and splits might not look that different from a version control graph.

Query: What aspects of these networks are public knowledge, and to what extent can we trace the history of an individual news item?

Have people developed tools to help this science of "news forensics"? My hope is that by figuring out who the original sources are and what they actually said, we can cancel out the effect of the replicator dynamics (what you see is what sells), and thus get more objective information.



* - If this were a single (rather than a developing) piece of news, and if everyone were honest, one would expect that the non-blatant copying means that the news gets phrased more and more awkwardly. In reality, I suspect they rephrase things without regard for the facts (and often towards sensationalism).

(no subject)

Date: 2010-01-03 12:12 pm (UTC)
From: [identity profile] selfishgene.livejournal.com
The biggest problem with the news is not the inaccuracy of what they report. The real issue is the selection of what events to report in the first place. Many people can detect bias/error in what is said but it is more difficult to notice what was not said. Dramatic is not the same as important.

(no subject)

Date: 2010-01-03 05:20 pm (UTC)
From: [identity profile] bhudson.livejournal.com
I saw someone who published a graph of autocopying (repeatedly publishing a paper with near-identical text). They gave a talk oriented towards graph drawing, at Google Pittsburgh in late 2006 or early 2007. So the techniques are there, and would be trivial to implement from within google which has trawled all the news sites for you already.

If it hasn't been done already, it would be a fascinating summer-project length piece of work. You'd need a journalism prof to get the embarrassing results into that community, and perhaps a stats prof or just a mentor at google with the knowledge you already have to get the math and programming right.

(no subject)

Date: 2010-01-03 05:21 pm (UTC)
From: [identity profile] bhudson.livejournal.com
Oh, another really awesome one: half of CNN stories about NASA are manglings of the NASA press release. As in, there'll be a paragraph that makes no sense whatsoever; that's the one written by CNN. The rest is verbatim from NASA and makes perfect sense.

(no subject)

Date: 2010-01-04 06:30 am (UTC)
From: [identity profile] easwaran.livejournal.com
Is it clear that media copy stories from each other? I thought that there were some organizations (UPI, AFP, AP, etc.) that have syndicated stories that run quite widely (sometimes with slight modifications), but otherwise stories are "original". That is, the different outlets each have their own reporter who writes her own story. The biggest thing is that they will tend to make sure they have a story on any topic that any other news outlet has a story on. But the story itself will be original, in whatever sense the first story was original. As [livejournal.com profile] bhudson points out, they're often recycled press releases, and combinations of pool reports with only a little truly original writing.

(no subject)

Date: 2010-01-04 01:34 pm (UTC)
From: [identity profile] gustavolacerda.livejournal.com
maybe I worded it too strongly.


<< but otherwise stories are "original". That is, the different outlets each have their own reporter who writes her own story. >>

but do they do the research independently? when you see the same thing written in two independent outlets, do you have independent confirmation that it really happened that way?

(no subject)

Date: 2010-01-04 05:34 pm (UTC)
From: [identity profile] easwaran.livejournal.com
One standard practice for certain types of story is that only one news source conducts the interview or gathers quotes or whatever, and "the pool" all share it. But I think that's more standard for routine stories from the White House and such, rather than for more "breaking news" type stories. Of course, they're not fully independent, because the reporters tend to be attending to the same things at the same scenes, and talking to many of the same people. But I don't think they literally use the other news source's story as their source, except when they explicitly acknowledge it, like when the National Enquirer broke the story about Jesse Jackson's illegitimate child.

February 2020

S M T W T F S
      1
2345678
9101112131415
16171819202122
23242526272829

Most Popular Tags

Style Credit

Expand Cut Tags

No cut tags