5 ms·

Wikipedia stats are painful to browse, but here are some numbers: Al-Qaeda / US Army / ISIS in January 2016 http://stats.grok.se/en/201601/Al-Qaeda http://sta

by steve19 11y ago

Wikipedia stats are painful to browse, but here are some numbers: Al-Qaeda / US Army / ISIS in January 2016 http://stats.grok.se/en/201601/Al-Qaeda http://stats.grok.se/en/201601/Al-Qaeda (~ 3k / day) http://stats.grok.se/en/201601/United%20States%20Army# http://stats.grok.se/en/201601/United%20States%20Army# (~ 2.5k / day) http://stats.grok.se/en/201601/Islamic%20State%20of%20Iraq%20and%20the%20Levant# http://stats.grok.se/en/201601/Islamic%20State%20of%20Iraq%2... ( ~15k / day) Al-Qaeda / US Army / ISIS in May 2013 (before Snowden's revelations & before ISIS/ISIL was known in the West[0]) http://stats.grok.se/en/201601/Al-Qaeda http://stats.grok.se/en/201601/Al-Qaeda (~ 5.8k / day) http://stats.grok.se/en/201601/United%20States%20Army# http://stats.grok.se/en/201601/United%20States%20Army# (~ 3.5k / day) My conclusion, traffic is probably changing due to what is in the news rather than anything else. If the US Army is sent in to invade another country, their wikipedia page will be back on top. ISIS was way higher last year, average about 25k/day in the middle of 2015, but the news media are tiring of them and they are also fading away. That said, I was about to send a completely legit email yesterday that had the keywords "pipe bomb", "suicide vest", "IED", "Al-Qaeda" and "Mali" in it (a friend just lost two of his men in Mali), but I decided to self-censor for the benefit of the recipient rather than myself. [0] ISIS was technically announced at the end of April, but they were not well known and the page was just a stub. It fell below threshold for collecting page view stats. https://en.wikipedia.org/w/index.php?title=Islamic_State_of_Iraq_and_the_Levant&oldid=125623633 https://en.wikipedia.org/w/index.php?title=Islamic_State_of_...

atdt 11y ago
There is a better (and more comprehensive and accurate) API for page view data now: http://blog.wikimedia.org/2015/12/14/pageview-data-easily-accessible/ http://blog.wikimedia.org/2015/12/14/pageview-data-easily-ac... , API specs at https://wikimedia.org/api/rest_v1/?doc#/Pageviews_data https://wikimedia.org/api/rest_v1/?doc#/Pageviews_data
machrider 11y ago
> My conclusion, traffic is probably changing due to what is in the news rather than anything else. I appreciate you throwing some data for a few keywords together in your comment, but you'll excuse me if I lean toward trusting the published paper instead. FWIW, paper is downloadable here (at least, for now): http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2769645 http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2769645
- steve19 11y ago
  Did you read the paper? They did not even track the ISIS pages. They had a small group of pages which they tracked and a control group (see the appendix) They did not compare it to the news cycle, or consider that people are reading different pages about the same subject (AQ in Iraq for example being superseded by ISIS). Sorry if I sound harsh but this study was terribly executed. Doing it properly would require a lot more resources than Wikipedia daily page view dumps. I think the NY Times would be well placed to do a study on this using Wikipedia page views, their page views and their mentions of topics overtime.
  - Baghard 11y ago
    > They did not compare it to the news cycle Such as the April 2013 Boston Marathon bombings. Or the rise of ISIL and Boko Haram in early 2013. I still think the study was well executed, but the instant effect seems to disappear when you remove the 6-7 months before mid june 2013. The long-term effect they claim to observe is more shaky. (I think one can find "statistically significant" groupings of say, Harry Potter articles, that show a similar peak and decline around the Snowden revelations.) Can we find more flaws in this paper?
maho 11y ago
I read your comment half-way, then upvoted it. I mentally prepared a response with further arguments against the size of the chilling effect. Then I read the rest of your reply and saw all the keywords. "Oh no", I thought, "I showed my engagement with this post by upvoting it!" Needless to say, I don't feel like arguing against the chilling effect anymore.

by steve19 11y ago

atdt 11y ago

There is a better (and more comprehensive and accurate) API for page view data now: http://blog.wikimedia.org/2015/12/14/pageview-data-easily-accessible/ http://blog.wikimedia.org/2015/12/14/pageview-data-easily-ac... , API specs at https://wikimedia.org/api/rest_v1/?doc#/Pageviews_data https://wikimedia.org/api/rest_v1/?doc#/Pageviews_data

machrider 11y ago

> My conclusion, traffic is probably changing due to what is in the news rather than anything else. I appreciate you throwing some data for a few keywords together in your comment, but you'll excuse me if I lean toward trusting the published paper instead. FWIW, paper is downloadable here (at least, for now): http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2769645 http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2769645

steve19 11y ago
Did you read the paper? They did not even track the ISIS pages. They had a small group of pages which they tracked and a control group (see the appendix) They did not compare it to the news cycle, or consider that people are reading different pages about the same subject (AQ in Iraq for example being superseded by ISIS). Sorry if I sound harsh but this study was terribly executed. Doing it properly would require a lot more resources than Wikipedia daily page view dumps. I think the NY Times would be well placed to do a study on this using Wikipedia page views, their page views and their mentions of topics overtime.
- Baghard 11y ago
  > They did not compare it to the news cycle Such as the April 2013 Boston Marathon bombings. Or the rise of ISIL and Boko Haram in early 2013. I still think the study was well executed, but the instant effect seems to disappear when you remove the 6-7 months before mid june 2013. The long-term effect they claim to observe is more shaky. (I think one can find "statistically significant" groupings of say, Harry Potter articles, that show a similar peak and decline around the Snowden revelations.) Can we find more flaws in this paper?

maho 11y ago

I read your comment half-way, then upvoted it. I mentally prepared a response with further arguments against the size of the chilling effect. Then I read the rest of your reply and saw all the keywords. "Oh no", I thought, "I showed my engagement with this post by upvoting it!" Needless to say, I don't feel like arguing against the chilling effect anymore.