Physicists Find Users Uninterested After 36 Hours

SuperGrads writes "Statistical physicists working in the US and Hungary have found that the number of people reading a particular news story on the web decreases with time by a power law rather than exponentially as was previously thought. The finding has implications for the study of information flow in social networks, marketing and web design."

Comments

In related news

[lecithin]

Users Find Physicists Uninteresting After 3.6 seconds.

Re:In related news

[JPribe]

Wow. It didn't even take that long.

But seriously, I wonder if this will change ad placement for revenue models? If an ad gets a click on a story older than 36 hours, is it worth more? Hmm, I smell a patent in the works, too.

Re:In related news

[The_Wilschon]

Well, I Am A Physicist, and I certainly find YOU uninteresting!

Old news

[Percent Man]

Are we still talking about this?

However ...

[blowdart]

The story will get posted again on slashdot 37 hours later.

maybe because it's not "news" anymore?

[holden caufield]

I'm wondering if these same researches tried to define what their subjects defined as "news"? If something was newsworthy, I'm guessing they likely found out about it over time. Maybe the people didn't read it because they were informed from other sources?

Sounds like a bit of a flawed evaluation to me.

Re:maybe because it's not "news" anymore?

[truthsearch]

It might also be relevant that this study was done only on a Hungarian news site. It's possible there would be different results in other countries due to cultural differences and the number of available news sources.

Re:maybe because it's not "news" anymore?

[The_Wilschon]

The "news" in this story is not that people become disinterested in a story, but that the rate at which they become disinterested is quite different from what was expected.

Furthermore, the study was not done by taking people and finding out how quickly they became disinterested in one story or another. A quick glance at the summary informs us that the subject of the study was the number of people reading a news story (more likely downloading the story) at a given time. That this number decreases with time is obvious. However, it was expected that the decrease would follow an exponential curve, whereas the experiment showed a power law curve instead.

Re:maybe because it's not "news" anymore?

[yfnET]

To prove the point, they actually did such a reversal in the case of telephone-queue waiting times. Traditionally, these have been assumed to follow a Poisson distribution, but some recent research suggests they actually follow a power law. Analysing the participants’ responses suggests that a power law, indeed, it is.

——

Science & Technology / Psychology

Bayes rules
Jan 5th 2006
From The Economist print edition

A once-neglected statistical technique may help to explain how the mind works

IMAGE

SCIENCE, being a human activity, is not immune to fashion. For example, one of the first mathematicians to study the subject of probability theory was an English clergyman called Thomas Bayes, who was born in 1702 and died in 1761. His ideas about the prediction of future events from one or two examples were popular for a while, and have never been fundamentally challenged. But they were eventually overwhelmed by those of the “frequentist” school, which developed the methods based on sampling from a large population that now dominate the field and are used to predict things as diverse as the outcomes of elections and preferences for chocolate bars.

Recently, however, Bayes’s ideas have made a comeback among computer scientists trying to design software with human-like intelligence. Bayesian reasoning now lies at the heart of leading internet search engines and automated “help wizards”. That has prompted some psychologists to ask if the human brain itself might be a Bayesian-reasoning machine. They suggest that the Bayesian capacity to draw strong inferences from sparse data could be crucial to the way the mind perceives the world, plans actions, comprehends and learns language, reasons from correlation to causation, and even understands the goals and beliefs of other minds.

These researchers have conducted laboratory experiments that convince them they are on the right track, but only recently have they begun to look at whether the brain copes with everyday judgments in the real world in a Bayesian manner. In research to be published later this year in Psychological Science, Thomas Griffiths of Brown University in Rhode Island and Joshua Tenenbaum of the Massachusetts Institute of Technology put the idea of a Bayesian brain to a quotidian test. They found that it passes with flying colours.

Prior assumptions
The key to successful Bayesian reasoning is not in having an extensive, unbiased sample, which is the eternal worry of frequentists, but rather in having an appropriate “prior”, as it is known to the cognoscenti. This prior is an assumption about the way the world works—in essence, a hypothesis about reality—that can be expressed as a mathematical probability distribution of the frequency with which events of a particular magnitude happen.

The best known of these probability distributions is the “normal”, or Gaussian distribution. This has a curve similar to the cross-section of a bell, with events of middling magnitude being common, and those of small and large magnitude rare, so it is sometimes known by a third name, the bell-curve distribution. But there are also the Poisson distribution, the Erlang distribution, the power-law distribution and many even weirder ones that are not the consequence of simple mathematical equations (or, at least, of equations that mathematicians regard as simple).

With the correct prior, even a single piece of data can be used to make meaningful Bayesian predictions. By contrast frequentists, though they deal with the same probability distributions as Bayesians, make fewer prior assumptions about the distribution that applies in any particular situation. Frequentism is thus a more robust approach, but one that is not well suited to

Re:First post

[Luctius]

Please reread your own sig.

Impulse function

[courtarro]

Users losing interest in this particular news story follow an impulse function.

Re:Impulse function

[m0nstr42]

This is "interesting"? It should just be "funny"!

I mean, EVERYONE knows what an impulse function is, right?

Right?

It's so very lonely here.

P.S. It's not really a function. It's a distribution, measure, functional, possibly some other things, but not a function.

Yes, very lonely.

Another massive triumph for statistical physicists

[Rob T Firefly]

Actually, nobody cares about this sort of thing, and these so-called "statistical physicists" would all be cleaning gutters for a living right now.. except the guy from HR is too terrified to go downstairs and fire them. The last time he tried, they somehow irrevocably proved to him that not only was it statistically impossible that he had arrived to give them their pink slips, but they also proved his trousers, eyebrows, and cat out of existence with nothing more than a slide rule and a whiteboard.

Possible other causes?

[Todd Knarr]

One has to wonder how the site's story policy affects the drop-off. That is, is the drop-off because users are uninterested or not reading, or is it because after that time the story drops off the main pages and becomes hard to find to read?

BREAKING NEWS

[27,000]

PHYSICISTS REPORT ARTICLES NOT ON FRONT PAGE READ LESS

ALSO NOTE THAT SITES HAVE FINITE NUMBERS OF USERS

And nothing about 'uninterested users'. This implies that, well, a reader is not likely to read an article more than once. Shocking, much unlike the answer to the question who is funding these people?

Methodological issues?

[gilroy]

I've read the linked article but not the actual Phys Rev paper, so I'm likely blowing smoke but...

• The "news cycle" is 24 hours, due to historical roots in daily newspapers (augmented by the evening news, etc.) Assume for the moment that people stay interested in a news story. After a day, if the story is ongoing, the original article is likely to be replaced by an update. Real-life example: Over the weekend, the NY Times Science section had these stories in a row: "Shuttle astronauts complete spacewalk", "shuttle astronauts inspect tiles", "shuttle Discovery meets space station", "shuttle Discovery set for launch". (paraphrased) Clearly, the first story in the list is the most recent and, were I looking for news on the Discovery, I'd probably click that one. Even if I really liked the Times' coverage of the rendez-vous, I'm not likely to read that article again if a new one has been posted. Does that mean I've "lost interest" in the shuttle?
• The results seem drawn from traffic at a particular Hungarian portal and might not have any generalized relevance.
• Ease of navigation seems important but not addressed. If stories "fall off" the homepage after 36 hours, it would make it look like people were less interested. (Or, really, the fact that some stories are highlit on the front page makes it look like people are more interested than they really are.)

Heh.

[wfberg]

Good luck in explaining the spike in traffic 3 full days after the article was posted.

Suckers!

Re:Who Knew?!!

[refriedchicken]

"36 Hours is the exact age of a story before it drops from the bottom of slashdot's Main page."

And is recycled back to the the top.

Exponent? Power?

[ch-chuck]

Color me ignorant, but I thought exponentials and powers were the same thing?
Or are they talking about natural exp -vs- a higher order power, like 4 or 5?

Re:Exponent? Power?

[DaoudaW]

The difference is whether the independent variable is the base or the exponent. A power function is something like f(x)=x^(.5) whereas an exponential function could be f(x)= (.5)^x.

We Just Killed Their Theory

[Anonymous Coward]

I can see these physicists really scratching their heads when the article gets slashdotted 72 hours after the published date (July 7).

Summary misstates article

[Nurf]

If you read the article, it says the distribution of half-lives of stories decreases as a power law, not that hit rates on stories decrease as a power law.

Half lives are a measurement of exponential decay. Individual stories still decrease in hits exponentially over time. If you look at lots of stories, the decays are distributed according to a power law.

The article directly contradicts the Slasdot summary.

Hits on stories do decrease exponentially.

I am stunned that I am the only one so far who seems to have picked up on this. Did anyone actually read the article, or did they just read into it what they were told they would see?

Duh.

[StikyPad]

Current tags: boring, slownewsday, yawn, uninteresting and duh.

Put another way: Slashdot -- Now with 20% real nerds!