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

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

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

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

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&gt; wrote: