Re: [Wikidata] [Wiki-Medicine] [Analytics] Zika

List overview All Threads
Download

newer

older

Wikidata Propbrowse

Three folders about Wikidata

Daniel Mietchen

15 Feb 2016 15 Feb '16

8:27 a.m.

The link to microcephaly has become clearer this week: https://en.wikipedia.org/w/index.php?title=Zika_virus&oldid=704879995#c… states "A complete ZIKV genome sequence [..] was recovered from brain tissue" (of a fetus whose mother had been infected with Zika virus). Given that the mass media are currently all over Zika, simple page view stats are essentially useless for tracking the spread of the disease - the PLOS Computational Biology article that Anthony has linked states "Wikipedia data have a variety of instabilities that need to be understood and compensated for. For example, Wikipedia shares many of the problems of other internet data, such as highly variable interest-driven traffic caused by news reporting and other sources." However, correlating geolocated view stats or searches with external info like http://www.healthmap.org/zika/#timeline might be useful. In addition, if we had some representation of clickstreams for Zika-related articles in languages spoken in affected areas, this could help guide the development of Zika-related content in those languages. Beyond Wikipedia, there is a page on Wikidata to coordinate activities around Zika: https://www.wikidata.org/wiki/Wikidata:WikiProject_Medicine/Zika . Cheers, d. On Mon, Feb 15, 2016 at 4:24 AM, Dan Andreescu <dandreescu(a)wikimedia.org> wrote:

...

On Sun, Feb 14, 2016 at 2:58 PM, Leila Zia <leila(a)wikimedia.org> wrote:

Hey Dan, On Sun, Feb 14, 2016 at 3:02 AM, Dan Andreescu <dandreescu(a)wikimedia.org> wrote:

So, I felt personally compelled in the case of Zika, and the confusing coverage it has seen, to offer to personally help.

Which aspect of the coverage are you referring to as confusing?

Well, so the first reports were that 3500 cases of microcephaly were linked to Zika in Brazil, since October. If you do the math, with Brazil's birth rate of 300,000 per year, 3500 for three months is incredibly high. The number went up to 4400 before it was discredited and the latest I read is that it's down to 404 [1] and there are claims of over-inflation. That same article talks about serious doubts that Zika even has anything to do with microcephaly. In reading around some more about the subject, it seems like a multi-variate analysis gone wrong.

I can run queries, test hypotheses, and help publish data that could back up articles. Privacy of our editors is of course still obviously protected, but that's easier to do in a specific case with human review than in the general case.

I'm up for brainstorming about what we can do and helping. Please keep me in the loop. In general, given that a big chunk of our traffic comes from Google at the moment, it would be great to work with the researchers in Google involved in Google's health related initiatives to produce complementary knowledge to what Google can already tell about Zika (for example, this). I'll reach out to the few people I know to get some more information. Depending on what complementary knowledge we want to produce, working with WikiProject Medicine can be helpful, too.

Cool, yeah, I'm nowhere close to knowledgeable on this, I can data-dog though :) [1] www.cbc.ca/news/health/microcephaly-brazil-zika-reality-1.3442580 _______________________________________________ Wikimedia-Medicine mailing list Wikimedia-Medicine(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikimedia-medicine

Show replies by date

Erik Zachte

15 Feb 15 Feb

1:31 p.m.

New subject: [Analytics] [Wiki-Medicine] Zika

Some observations (maybe stating the obvious): https://tools.wmflabs.org/pageviews/#start=2016-01-16&end=2016-02-14&am… the double peak seems to confirm PV count on wp:en is not correlated much with spread of the disease, but of course wp:es is much more relevant https://tools.wmflabs.org/pageviews/#start=2016-01-16&end=2016-02-14&am… As Wikipedia probably isn't much of a household name in some Spanish countries, absolute PV values per country/region are mostly incomparable, even when normalized for population count in the region, and again for percentage people with internet connections (let alone the latter numbers will be unreliable to start with). A detection of highest relative changes week over week could tell us something. (the oldest week in the comparison should have a minimum PV value, or relative changes from almost zero to negigible with stand out as false remarkables. Of course a decline could indicate saturation in information demand, but not per se in number of people affected. Erik -----Original Message----- From: Analytics [mailto:analytics-bounces@lists.wikimedia.org] On Behalf Of Daniel Mietchen Sent: Monday, February 15, 2016 8:28 To: Wiki Medicine discussion Cc: A mailing list for the Analytics Team at WMF and everybody who has an interest in Wikipedia and analytics.; Discussion list for the Wikidata project. Subject: Re: [Analytics] [Wiki-Medicine] Zika The link to microcephaly has become clearer this week: https://en.wikipedia.org/w/index.php?title=Zika_virus&oldid=704879995#c… states "A complete ZIKV genome sequence [..] was recovered from brain tissue" (of a fetus whose mother had been infected with Zika virus). Given that the mass media are currently all over Zika, simple page view stats are essentially useless for tracking the spread of the disease - the PLOS Computational Biology article that Anthony has linked states "Wikipedia data have a variety of instabilities that need to be understood and compensated for. For example, Wikipedia shares many of the problems of other internet data, such as highly variable interest-driven traffic caused by news reporting and other sources." However, correlating geolocated view stats or searches with external info like http://www.healthmap.org/zika/#timeline might be useful. In addition, if we had some representation of clickstreams for Zika-related articles in languages spoken in affected areas, this could help guide the development of Zika-related content in those languages. Beyond Wikipedia, there is a page on Wikidata to coordinate activities around Zika: https://www.wikidata.org/wiki/Wikidata:WikiProject_Medicine/Zika . Cheers, d. On Mon, Feb 15, 2016 at 4:24 AM, Dan Andreescu <dandreescu(a)wikimedia.org> wrote:

...

On Sun, Feb 14, 2016 at 2:58 PM, Leila Zia <leila(a)wikimedia.org> wrote:

Hey Dan, On Sun, Feb 14, 2016 at 3:02 AM, Dan Andreescu <dandreescu(a)wikimedia.org> wrote:

So, I felt personally compelled in the case of Zika, and the confusing coverage it has seen, to offer to personally help.

Which aspect of the coverage are you referring to as confusing?

_______________________________________________ Analytics mailing list Analytics(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics

Dan Andreescu

8:33 p.m.

New subject: [Wiki-Medicine] [Analytics] Zika

This makes a lot of sense, I'll get started on looking for correlation between that time-line and geolocated interest coming in through the different language wikis. Original Message From: Daniel Mietchen Sent: Monday, February 15, 2016 02:28 To: Wiki Medicine discussion Reply To: Wiki Medicine discussion Cc: A mailing list for the Analytics Team at WMF and everybody who has an interest in Wikipedia and analytics.; Discussion list for the Wikidata project. Subject: Re: [Wiki-Medicine] [Analytics] Zika The link to microcephaly has become clearer this week: https://en.wikipedia.org/w/index.php?title=Zika_virus&oldid=704879995#c… states "A complete ZIKV genome sequence [..] was recovered from brain tissue" (of a fetus whose mother had been infected with Zika virus). Given that the mass media are currently all over Zika, simple page view stats are essentially useless for tracking the spread of the disease - the PLOS Computational Biology article that Anthony has linked states "Wikipedia data have a variety of instabilities that need to be understood and compensated for. For example, Wikipedia shares many of the problems of other internet data, such as highly variable interest-driven traffic caused by news reporting and other sources." However, correlating geolocated view stats or searches with external info like http://www.healthmap.org/zika/#timeline might be useful. In addition, if we had some representation of clickstreams for Zika-related articles in languages spoken in affected areas, this could help guide the development of Zika-related content in those languages. Beyond Wikipedia, there is a page on Wikidata to coordinate activities around Zika: https://www.wikidata.org/wiki/Wikidata:WikiProject_Medicine/Zika . Cheers, d. On Mon, Feb 15, 2016 at 4:24 AM, Dan Andreescu <dandreescu(a)wikimedia.org> wrote:

...

On Sun, Feb 14, 2016 at 2:58 PM, Leila Zia <leila(a)wikimedia.org> wrote:

Hey Dan, On Sun, Feb 14, 2016 at 3:02 AM, Dan Andreescu <dandreescu(a)wikimedia.org> wrote:

So, I felt personally compelled in the case of Zika, and the confusing coverage it has seen, to offer to personally help.

Which aspect of the coverage are you referring to as confusing?

_______________________________________________ Wikimedia-Medicine mailing list Wikimedia-Medicine(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikimedia-medicine

Alex Druk

16 Feb 16 Feb

9:07 a.m.

New subject: [Analytics] [Wiki-Medicine] Zika

My 2¢: http://www.wikipediatrends.com/predictions/Medicine/Zika_virus/ On Mon, Feb 15, 2016 at 8:33 PM, Dan Andreescu <dandreescu(a)wikimedia.org> wrote:

...

On Sun, Feb 14, 2016 at 2:58 PM, Leila Zia <leila(a)wikimedia.org> wrote: > > Hey Dan, > > On Sun, Feb 14, 2016 at 3:02 AM, Dan Andreescu <

dandreescu(a)wikimedia.org>

wrote:

So, I felt personally compelled in the case of Zika, and the confusing coverage it has seen, to offer to personally help.

Which aspect of the coverage are you referring to as confusing?

Well, so the first reports were that 3500 cases of microcephaly were

linked

to Zika in Brazil, since October. If you do the math, with Brazil's birth rate of 300,000 per year, 3500 for three months is incredibly high. The number went up to 4400 before it was discredited and the latest I read is that it's down to 404 [1] and there are claims of over-inflation. That

same

article talks about serious doubts that Zika even has anything to do with microcephaly. In reading around some more about the subject, it seems

a multi-variate analysis gone wrong. > > >> >> I can run queries, test hypotheses, and help publish data that could

back

>> up articles. Privacy of our editors is of course still obviously

protected,

>> but that's easier to do in a specific case with human review than in

the

>> general case. > > > I'm up for brainstorming about what we can do and helping. Please keep

> in the loop. In general, given that a big chunk of our traffic comes

from

> Google at the moment, it would be great to work with the researchers in > Google involved in Google's health related initiatives to produce > complementary knowledge to what Google can already tell about Zika (for > example, this). I'll reach out to the few people I know to get some more > information. > Depending on what complementary knowledge we want to produce, working

with

WikiProject Medicine can be helpful, too.

_______________________________________________ Wikimedia-Medicine mailing list Wikimedia-Medicine(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikimedia-medicine _______________________________________________ Analytics mailing list Analytics(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics

-- Thank you. Alex Druk alex.druk(a)gmail.com (775) 237-8550 Google voice

Dario Taraborelli

18 Feb 18 Feb

5:23 p.m.

New subject: [Analytics] [Wiki-Medicine] Zika

I pointed Reid and his team to this thread. On Tue, Feb 16, 2016 at 12:07 AM, Alex Druk <alex.druk(a)gmail.com> wrote:

...

My 2¢: http://www.wikipediatrends.com/predictions/Medicine/Zika_virus/ On Mon, Feb 15, 2016 at 8:33 PM, Dan Andreescu <dandreescu(a)wikimedia.org> wrote:

On Sun, Feb 14, 2016 at 2:58 PM, Leila Zia <leila(a)wikimedia.org> wrote: > > Hey Dan, > > On Sun, Feb 14, 2016 at 3:02 AM, Dan Andreescu <

dandreescu(a)wikimedia.org>

wrote: > > So, I felt personally compelled in the case of Zika, and the confusing > coverage it has seen, to offer to personally help. Which aspect of the coverage are you referring to as confusing?

Well, so the first reports were that 3500 cases of microcephaly were

linked

to Zika in Brazil, since October. If you do the math, with Brazil's

birth

rate of 300,000 per year, 3500 for three months is incredibly high. The number went up to 4400 before it was discredited and the latest I read

that it's down to 404 [1] and there are claims of over-inflation. That

same

article talks about serious doubts that Zika even has anything to do

with

microcephaly. In reading around some more about the subject, it seems

a multi-variate analysis gone wrong. > > >> >> I can run queries, test hypotheses, and help publish data that could

back

>> up articles. Privacy of our editors is of course still obviously

protected,

>> but that's easier to do in a specific case with human review than in

the

>> general case. > > > I'm up for brainstorming about what we can do and helping. Please keep

> in the loop. In general, given that a big chunk of our traffic comes

from

> information. > Depending on what complementary knowledge we want to produce, working

with

WikiProject Medicine can be helpful, too.

-- Thank you. Alex Druk alex.druk(a)gmail.com (775) 237-8550 Google voice _______________________________________________ Analytics mailing list Analytics(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics

-- *Dario Taraborelli *Head of Research, Wikimedia Foundation wikimediafoundation.org • nitens.org • @readermeter <http://twitter.com/readermeter>

Priedhorsky, Reid

19 Feb 19 Feb

5:30 p.m.

New subject: [Analytics] [Wiki-Medicine] Zika

Dan Andreescu

6:29 p.m.

New subject: [Analytics] [Wiki-Medicine] Zika

Thanks, Reid. When you say there's insufficient data history, do you mean in other sources? Zika was discovered in 1947 and the wiki page for it was built in 2009. We have high quality geolocated data since May 2015. I'm still doing research (I admit the distractions at the foundation have gotten in the way, I apologize for that). I hope to get back to it with renewed force this weekend. On Fri, Feb 19, 2016 at 11:30 AM, Priedhorsky, Reid <reidpr(a)lanl.gov> wrote:

...

We do have more work in progress to extend the 2014 paper, in particular to mosquito-borne diseases in a Spanish-speaking country, though not Zika because there is insufficient data history. I appreciate the pointer. Are there any specific questions folks would like me to address in this thread? Thanks, Reid _______________________________________________ Analytics mailing list Analytics(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics

Priedhorsky, Reid

22 Feb 22 Feb

11:11 p.m.

New subject: [Analytics] [Wiki-Medicine] Zika

Insufficient data in epidemiological sources. Basically, we need fairly decent time series incidence data over a few years in order to train the models; this isn’t available for Zika, just case reports here and there. The expert on our team is Ashlynn Daughton: “[T]here’s been a small amount of surveillance of Zika (http://www.eurosurveillance.org/images/dynamic/ET/V19N02/V19N02.pdf). French polynesia and other islands had an outbreak in 2013 and it sounds like there are *some* reports (pg. 50). There’s also sporadic mentions of imported Zika from travelers from Africa or Asia (e.g. See pg 54). But there hasn’t been anything as systematic, or comprehensive as there is now.” HTH, Reid On Feb 19, 2016, at 10:29 AM, Dan Andreescu <dandreescu@wikimedia.org<mailto:dandreescu@wikimedia.org>> wrote: Thanks, Reid. When you say there's insufficient data history, do you mean in other sources? Zika was discovered in 1947 and the wiki page for it was built in 2009. We have high quality geolocated data since May 2015. I'm still doing research (I admit the distractions at the foundation have gotten in the way, I apologize for that). I hope to get back to it with renewed force this weekend. On Fri, Feb 19, 2016 at 11:30 AM, Priedhorsky, Reid <reidpr@lanl.gov<mailto:reidpr@lanl.gov>> wrote: We do have more work in progress to extend the 2014 paper, in particular to mosquito-borne diseases in a Spanish-speaking country, though not Zika because there is insufficient data history. I appreciate the pointer. Are there any specific questions folks would like me to address in this thread? Thanks, Reid _______________________________________________ Analytics mailing list Analytics@lists.wikimedia.org<mailto:Analytics@lists.wikimedia.org> https://lists.wikimedia.org/mailman/listinfo/analytics _______________________________________________ Analytics mailing list Analytics@lists.wikimedia.org<mailto:Analytics@lists.wikimedia.org> https://lists.wikimedia.org/mailman/listinfo/analytics

Daniel Mietchen

11:23 p.m.

New subject: [Analytics] [Wiki-Medicine] Zika

Dear all, we're currently building a template for Zika-related articles (cf. https://en.wikipedia.org/wiki/Template:Zika ), and I could imagine some of these - e.g. those on the earlier Zika outbreaks - may be useful for training the models. Cheers, Daniel On Mon, Feb 22, 2016 at 11:11 PM, Priedhorsky, Reid <reidpr(a)lanl.gov> wrote:

...

_______________________________________________ Analytics mailing list Analytics(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics _______________________________________________ Analytics mailing list Analytics(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics

Dan Andreescu

23 Feb 23 Feb

9:46 p.m.

New subject: [Wiki-Medicine] [Analytics] Zika

Hm :/ Well, I can't do anything about epidemiological souces, except time travel I guess :) So, if others agree that this is a dead end, I'll downgrade it to a pet project. Daniel, let me know if I can help still. If this particular example isn't good, is there another specific case where our data might help? On Mon, Feb 22, 2016 at 5:23 PM, Daniel Mietchen < daniel.mietchen(a)googlemail.com> wrote:

...

amount

of surveillance of Zika (http://www.eurosurveillance.org/images/dynamic/ET/V19N02/V19N02.pdf). French polynesia and other islands had an outbreak in 2013 and it sounds like there are *some* reports (pg. 50). There’s also sporadic mentions of imported Zika from travelers from Africa or Asia (e.g. See pg 54). But

there

hasn’t been anything as systematic, or comprehensive as there is now.” HTH, Reid On Feb 19, 2016, at 10:29 AM, Dan Andreescu <dandreescu(a)wikimedia.org> wrote: Thanks, Reid. When you say there's insufficient data history, do you

mean

in other sources? Zika was discovered in 1947 and the wiki page for it

was

built in 2009. We have high quality geolocated data since May 2015. I'm still doing research (I admit the distractions at the foundation have gotten in the way, I apologize for that). I hope to get back to it with renewed force this weekend. On Fri, Feb 19, 2016 at 11:30 AM, Priedhorsky, Reid <reidpr(a)lanl.gov>

wrote:

> > We do have more work in progress to extend the 2014 paper, in particular > to mosquito-borne diseases in a Spanish-speaking country, though not

Zika

because there is insufficient data history. I appreciate the pointer. Are there any specific questions folks would like me to address in this thread? Thanks, Reid _______________________________________________ Analytics mailing list Analytics(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics

_______________________________________________ Wikimedia-Medicine mailing list Wikimedia-Medicine(a)lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikimedia-medicine

2979

days inactive

2987

days old

wikidata@lists.wikimedia.org

Manage subscription

9 comments

6 participants

tags (0)

participants (6)

Alex Druk
Dan Andreescu
Daniel Mietchen
Dario Taraborelli
Erik Zachte
Priedhorsky, Reid