Wikimania 2017 report - Kerry Raymond
I attended Wikimania 2017 at Montreal on 11-13 August 2017. I was partially self-funded and partially funded by Wikimedia Australia.
I won't do a a "session-by-session" report but rather focus on the big picture issues I observed.
There were many sessions relating in some way to working more closely with GLAM. I delivered a presentation entitled "Brick by brick: Libraries and Wikimedia building on each other" on the work that we have done with State Library of Queensland, which was co-authored with Jacinta Sutton (SLQ). Our presentation was very well-received (we have been asked to repeat it at some other events!) largely because we focussed less on the specifics of what we did but more about what does and doesn't work and why. A number of other presentations went into greater detail about the actual activities but without much reflection on whether the process was working. I noticed that a number of presentations dwelt considerably on "establishing a formal relationship" which sought to commit the GLAM to various contractual obligations (usually in relation to providing access to their collections). As far as I know, WMAU has never established any "formal" relationships with our local GLAMs but access to collections hasn't been an identified problem for us, so our context may be different. Many people were impressed that SLQ had contributed over 1000 citations during 1Lib1Ref this year (a quarter of the world's total), but there seemed some reluctance to take onboard the concerns we raised about the program overall (e.g. the use of Citation Hunt). As much as there is talk about "learning patterns" in the Wikimedia community, there appears a reluctance to grasp that if you want a different outcome, you need to do things differently.
Probably the biggest insights I obtained related to Wikidata and its use for Wikipedia. Too often presentations on Wikidata focus on the ability to query Wikidata directly (e.g. "how many female mayors were there in 2006?") which isn't a compelling argument to why a contributor to Wikipedia should care about Wikidata. However, a number of presentations at Wikimania discussed the automatic generation of infoboxes and other "fact-based" presentations of information (e.g. graphs, tables) which make it clearer that the data can be added once to Wikidata and then extracted for multiple purposes and for many different language Wikipedias, with obvious benefit for consistency.
It is also worth noting that Wikidata can now be accessed from Commons, which is enabling a new initiative called Structured Data on Commons. So I am converted to the benefits of using Wikidata.
However, I would say that the potential of Wikidata is somewhat restricted by the need for more "ordinary user" tools; I don't think that tools like SPARQL are suitable for the average user. Similarly a number of tools for developing infoboxes automatically etc are still "in development". I was also concerned about the way of modelling was being done within Wikidata; it was far from obvious in many cases. So while there is a lot of exciting potential with Wikidata, I think tool development is going to be necessary to create tools that engage existing contributors rather than disenfranchise them.
Wiki Loves Monuments
I attended a number of formal sessions as well as informal meetings to discuss Australia's involvement in Wiki Loves Monuments in 2017 and beyond; I was in constant email conversation with Gnangarra through this process. The conclusion reached was that WMAU would start with just some national heritage lists for our first engagement in WLM and aim to expand to state lists in 2018. It should be noted that there were also a number of discussions about putting cultural heritage data onto Wikidata to better facilitate initiatives like WLM.
As is probably well-known around WMAU, my experience with edit training has shown me that the Visual Editor is very important for engaging new contributors. So I was very interested to hear the presentation about how the Visual Editor has now become the default editor for new and unregistered users on German Wikipedia (and how the sky didn't fall as a result). This is something we urgently need on English Wikipedia if we want to reverse the long-term gradual decline of the numbers of active editors in the English Wikipedian community
Random comments on other Wikimania session
I was also interested to note the range of activities performed under "Wikipedian in Residence" titles.
A number of scheduled "gathering" events (variously called Birds of a Feather, Meetups, etc) generally were not very effective unless the number of people were quite small. Most of the time was taken up with "short introductions", leaving little time for any real discussion of issues.
Finally, attending an event like Wikimania broadens one's horizons. As well as the formal sessions, the informal conversations over coffee etc bring new perspectives to existing issues and raise new issues.
A direct consequence of the talk I presented is that Jacinta Sutton (SLQ) and I invited to participate in the OCLC (a global library cooperative organisation)'s upcoming Wikipedia + Libraries: Better Together series, a 9 week program which seeks to introduce Wikipedia to public librarians in the USA; it has an enrolment of about 300 librarians. We will be presenting the Wikimedia activities at the State Library of Queensland in their live online session on 11 October (which is 3am in the morning for us). Also Jacinta and I will serve as "guides" through the course, which means interacting with the participants through the discussion boards to answer questions and provide advice and support.
In addition, I was also invited to repeat the Wikimania talk at OCLC's Asia Pacific Regional Council Meeting 2017 in Tokyo at the end of November (with funding provided to do so); however after discussions with Pru Mitchell, OCLC agreed to my suggestion that Jacinta Sutton should present instead of me in Tokyo, as, being a librarian event, her views as a librarian would be more relevant to the audience plus, as a librarian, she would benefit far more than me from attending the event overall. We are just awaiting State Library of Queensland's approval of this arrangement.
One of the reasons I agreed to participatein the OCLC program for public libraries is to see how effective that program is, and, if successful, explore whether something similar could be run here in Australia. Currently our Wikipedia training is very much face-to-face and, given our limited resources of people willing to do outreach and the geographic spread of our wide brown land, finding an effective way to deliver training remotely would be a great benefit to us. A search for Australia in the OCLC membership list lists over 100 Australian GLAMs which are members, including the National Library of Australia and a number of state libraries etc, so we potentially have local partners to work with on such an initiative.
Wiki Loves Monuments
To support an ongoing and wider engagement with Wiki Loves Monuments, we need to look at capturing the basic data from the various national and state heritage registers needed for Wiki Loves Monuments. Since this type of informaton is "facts", I don't think there is a copyright issue here so it should be possible to WMAU to webscrape the data from the various heritage register websites (the webscraping task may be easier said than done).
WMAU has already run a session on Wikidata at various cities in Australia, but I believe we need to continue to try to bring the attention of Australian editors to the benefits of Wikidata and how to harness those benefits. In collaboration with Australian editors, we probably need to map out a program of activity around Wikidata. It is likely that the 2016 census will be the first major foray into incorporating Wikidata into Australian Wikipedia articles. In the case of population data, the likely approach is for the template "infobox Australian place" to draw its "pop", "pop_year", and "pop_footnotes" fields from Wikidata unless manually overridden.
Having webscraped the key data from the heritage registers for Wiki Loves Monuments, adding that data to Wikidata is the logical next step.
The Qld state electorate redistribution provides another opportunity to trial replacing the manually-added data for the existing electorates with new data drawn from Wikidata. And there are doubtless many other opportunities of which I am not aware.