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 "[https://wikimania2017.wikimedia.org/wiki/Submissions/Brick_by_brick:_Libraries_and_Wikimedia_building_on_each_other 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 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 truly "end user" ready. 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. 

I attended a number of formal sessions as well as informal meetings to discuss Australia's involvement in Wiki Loves Monuments; 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 cultural heritage data (such as used in WLM) being held on Wikidata, so I think we need to be looking at doing this as our goal for WLM 2018.

I was also interested to note the range of activities performed under "Wikipedian in Residence" titles.

Finally, 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.