News Entropy

#1

Post by Naa4 » Sat, Jun 12, 2021, 11:52 AM UTC

Re: News Entropy

#2

Post by whatshisface » Sat, Jun 12, 2021, 4:29 PM UTC

I guess this is rigorous confirmation of the slow news day = talk about anything, crisis = talk about Michael Jackson's death for six months model of news broadcasting.

Re: News Entropy

#3

Post by crsn » Sat, Jun 12, 2021, 4:52 PM UTC

I’m working on a commercial side project along these lines — if anyone is interested in helping or collaborating, shoot me an email!

Re: News Entropy

#4

Post by gtt » Sat, Jun 12, 2021, 5:22 PM UTC

crsn wrote: post #3

I’m working on a commercial side project along these lines — if anyone is interested in helping or collaborating, shoot me an email!

When I've tried to predict social unrest or market volatility from the news, I've found, than news actually trailing important events and are useless for any sorts of prediction.

Re: News Entropy

#5

Post by qorrect » Sat, Jun 12, 2021, 5:27 PM UTC

I was a bit bummed I thought this would be more inline with Shannons entropy, or a measure of how much an article reveals about the topic.

Re: News Entropy

#6

Post by Macuyiko » Sat, Jun 12, 2021, 5:57 PM UTC

qorrect wrote: post #5

I was a bit bummed I thought this would be more inline with Shannons entropy, or a measure of how much an article reveals about the topic.

They use Shannon's entropy in the paper, no?

Re: News Entropy

#7

Post by fogof » Sat, Jun 12, 2021, 6:23 PM UTC

Macuyiko wrote: post #6

qorrect wrote: post #5

I was a bit bummed I thought this would be more inline with Shannons entropy, or a measure of how much an article reveals about the topic.

They use Shannon's entropy in the paper, no?

I would summarize it as the entropy over distribution of topics over the whole news spectrum, rather than a the entropy of a particular piece of news.

Re: News Entropy

#8

Post by webmaven » Sun, Jun 13, 2021, 3:56 AM UTC

whatshisface wrote: post #2

I guess this is rigorous confirmation of the slow news day = talk about anything, crisis = talk about Michael Jackson's death for six months model of news broadcasting.

> I guess this is rigorous confirmation of the slow news day = talk about anything, crisis = talk about Michael Jackson's death for six months model of news broadcasting.

Well, yes, but it is also implying "slow news is good news".

Re: News Entropy

#9

Post by webmaven » Sun, Jun 13, 2021, 3:59 AM UTC

gtt wrote: post #4

crsn wrote: post #3

I’m working on a commercial side project along these lines — if anyone is interested in helping or collaborating, shoot me an email!

When I've tried to predict social unrest or market volatility from the news, I've found, than news actually trailing important events and are useless for any sorts of prediction.

> When I've tried to predict social unrest or market volatility from the news, I've found, than news actually trailing important events and are useless for any sorts of prediction.

It might be useful though to be able to distinguish actual social unrest from the astroturf variety.

Re: News Entropy

#10

Post by crsn » Sun, Jun 13, 2021, 6:53 AM UTC

gtt wrote: post #4

crsn wrote: post #3

I’m working on a commercial side project along these lines — if anyone is interested in helping or collaborating, shoot me an email!

When I've tried to predict social unrest or market volatility from the news, I've found, than news actually trailing important events and are useless for any sorts of prediction.

Luckily, that’s not my application area if these ideas right now. :) But it might indeed fail for some of the same structural regularities!