Good Data Workshop

6 min read November 22, 2017 at 9:08pm on Data and Future Wise

I flew to Brisbane on Wednesday to attend the the "Pathways to Good Data" workshop at my old stomping ground, QUT. It was organised by the team of Angela Daly (twitter), Kate Devitt (Twitter) and Monique Mann (twitter), and the Future Wise team put in a strong attendance with both Kate and Justin being there as well as me. 

Sadly, I missed the introduction (thanks to a Jetstar fault on my flight down and a long delay), but arrived for a discussion of what makes "good" vs "bad" data. There was a range of opinions, from the inevitable "it's what you do with it that's the problem" to the extreme data nihilist "all data is bad'. I'm a bit split on this, I think that some data is great, some can be put to multiple uses and some is unequivocally bad. 

The first content session was talking about Indigenous Data Sovereignty, with a number of perspectives (2 Indigenous Australian, and 1 Maōri). This has been a bit of a hot topic of late with a recent workshop addressing IDS issues in more detail. Good examples from overseas (New Zealand, Canada) were shared, and probably unsurprisingly, the news that Australia has a way to go. There's a free to download monograph from ANU and a decent recent article in Crikey [$] for those interested in reading more about the concept.

After afternoon tea was an interesting session where someone from the community presented an idea on the use of data, and an academic with some sort of relevant interest presented an opinion. 

First up was an industry presentation from Extensia - who, as far as I could tell - are promoting their own layer of electronic medical record, which takes the primary records from clinicians, sorts and filters them in a way which gives the patient more choice and then passes the filtrate to myHealthRecord. It sounds like an almost irredeemably bad idea, and no thanks to Kate for pulling faces at me from the other side of the room. The academic discussant seemed like a nice guy, but there was very little in the way of content to engage with in his discussion.

Up next was Ellen Broad (twitter) from the Open Data Institute - "we're not an institute and we're not really about Open source". Her talk was excellent, and it was great to hear from people with a strong commitment to data use and transparency issues.

The Queensland Community Wi-Fi initiative (it should be but no link because I can't confirm their website right now- it appears to have both a SSL error and a permissions error on their server :-/ ) sounds like a cross between the movement and pirate radio. They want to stick WiFi repeaters at various locations around Brisbane and create a distributed open WiFI network which avoids the major carriers. There were a couple of unresolved issues for me - it wasn't clear where they are getting the backhaul, and it also seems likely they would be considered a service provider under the data retention laws. The guy speaking was a 20ish uni student, so I don't think anyone (except Justin) wanted to rain on his parade too much.

Finally, Justin Warren (@jpwarren) spoke on his transformation from less-engaged-citizen through rowdy whinger to online activist for #notmydebt to FOI warrior and member of the board of Electronic Frontiers Australia. It was a great talk for a few reasons.

One, I'd not yet met Justin in meatspace, and it was nice to press the actual flesh. Second, he was almost pathologically enthusiastic, and it was great to see the room respond to his enthusiasm (it also helped that there were lots of people who believed in #notmydebt in the room). Justin's story also paralleled my own - I joined EFA in response to the #NoCleanFeed campaign back in the day, and Future Wise now exists because of the data retention laws. So you can be reassured at least, that terrible government policy at least has one benefit - creating activists.

The academic discussant was A/Prof Kate Galloway from Bond Uni - fellow Future Wise member, and a rowdy whinger on the internet if ever there was one. Sitting at Kate's table was a guy from the Tax Office doing an academic secondment to QUT who looked progressively more and more uncomfortable through Justin's talk and then would have disappeared into a chasm in the floor if one had opened up while Kate was railing at the Government's renegotiation of its civil contract. He asked Justin a question at the conclusion of the polemics and promptly got shot down in flames. It was magnificent.

The final session was a rapid-fire discussion of some of the work going on in the area, which largely ended up being a request for funding of interesting projects and probably didn't add much to the day. The team are editing an open-access book on data issues which is seeking contributions, and I'm considering both possible topics and my ridiculous level of overcommitment.

After the session proper had finished, we moved to the Thoughtworks Brisbane office for a discussion of machine learning by Charlie Brooking from TW, and then a chat with Asher Wolf.


The data science Venn

The Data science venn - From @CharlieBrooking, photo by me
(elapsed time between tweeting it and a "#notallhackers" reply: 5min)

Charlie's talk was an excellent, practical demo of (social) bias in datasets, and how machine learning depends on being "fed" data to make the algorithms work. It also highlighted the dangers of "deidentifying" data and assuming that everything will be fine, because the (made up) dataset Charlie used showed that removing "tribe" and university still had significant clustering of the tribe characteristic by postcode, which you wouldn't automatically think of as being a sensitive dataset.

The second speaker was Asher Wolf - Asher's formal bio is "award-winning journalist and CryptoParty founder", but she is universally known for being a force of nature on Twitter, talking about how she ended up moving in to her current role. I won't break down the details of her talk, but the key message was about connecting with people, and Asher's belief you can achieve more by bringing the right people together than by any well-placed letter or phone-call (although they're important, too).

It's worth acknowledging how important Asher has been to the digital rights scene in Australia - not only by her raising awareness of important issues and scouring twitter in a way that is clearly superhuman, but for her role in linking up people in the space online and in person. I'm happy to have her as an internet friend, and again, it was great to meet her in person.

It was a long day, but I enjoyed it, and I'm looking forward to more work in this area... added to my list of "things I don't really have time for but are very important".


Kate, Trent and Justin - Future Wise
Kate, Justin and I - Future Wise rocking the front row (Kate's iPhone)

Photo credit: Statistics on a computer screen - by Carlos Muzak via Unsplash - CC0