Anna Gerber very nearly was not a computer scientist. In fact, if not for an accident suffered in university, she might be a pianist.
“When I was a kid I used to love computers and programming logo,” Gerber says. “Then for a few years I sort of shied away from them, and I came to UQ to do an Arts degree.
“But I had an injury, so I couldn’t play the piano anymore,” she says. “I had to look for something different and my father suggested computer science; he worked in it and thought computers were the way of the future.
“It’s probably just as well because I absolutely love writing software and IT.”
While her entry into the field was somewhat unorthodox, Gerber is now a highly successful computer scientist, though an extremely humble one. When asked to discuss one of her many accolades, such as winning GovHack, CampJS and a British Library Labs competition, Gerber simply said she doesn’t “really like to think like that”, and instead focus on her work.
Chief amongst her current roles is technical project manager at UQ’s ITEE eResearch Group, where she works with data integration and spatio-statistical analysis at Midja Integrated Knowledge Base for Aboriginal Housing. The project, headed by Carroll Go-Sam from UQ’s Aboriginal Environment Research Centre, is compiling Aboriginal housing data so as to better understand region-specific requirements.
“It’s not without its challenges,” Gerber says. “With indigenous data, because the population is such a small percentage of the broader population and there are obviously issues to do with data privacy, we thought we’d be able to use a lot of the open data sets, like the ABS.
“But it turns out that when you try to drill down to that level of detail with really detailed stats of indigenous population in small, regional areas, that data basically gets anonymised and zeroed out,” she says.
The demographic data comes from a number of sources, notably the Australian Bureau of Statistics, and ranges from housing-specific, such as overcrowding, to public information, such as mobility and employment.
On overcoming problems with data retrieval, Gerber lists the two main approaches as applying directly through the ABS for the raw data, which requires confidentiality and privacy agreements, and creating statistical models by outsourcing specialised data sets.
“ABS applications just mean we have to sign agreements about the use of the data, and obviously it can’t be published,” Gerber says. “So there’s sensitivity about how we store that data and aggregate it in a way that won’t reveal any personal details about people.
“[With statistical models] you’ve got the data at a more aggregated level higher up, you can then try to work out what it would be at the individual community level,” she says. “There are people down in Canberra who do this, and we’ve been trying to use some of the data that they’ve developed because it sort of fills in some of the gaps.”
With a stated goal of “improved housing management, housing design, health and well-being of Indigenous communities and a reduction in crowding and homelessness,” Midja might be Gerber’s project with the furthest social reach, the one with the most potential to utilise data for a broader cause.
However, it is far from her only ongoing venture. Gerber is an ambassador for the Open Knowledge Foundation Australia, worked as lead organiser of this year’s GovHack Brisbane, and runs local meetups for groups such as Nodebots and Open Knowledge Meetup Group. The local tech scene has truly benefited from Gerber’s tireless, if sometimes unacknowledged, coordinating and lending her experience to these groups.
Arguably the biggest complaint with Brisbane’s tech scene, a lack of communication, is not lost on Gerber.
“There are so many awesome things happening in Brisbane but they’re happening in isolation a lot of the time,” Gerber says. “River City Labs, for example, are doing all these amazing things, but I haven’t actually ever been to any of their things; they’re always on at the same time as something else.
“For example, there’s a 3D printing meetup group that has been meeting at River City Labs for months now, and I’ve never been able to make a meetup because it clashes with the Hack the Evening meetup at The Edge, where they build 3D printers,” she says. “So it seems crazy to me that these two groups are doing the same thing on the same night.
“I know that they’ve actually reached out to the Brisbane hackerspace so they are trying to bridge that gap a little, but that’s what I often feel. I go to different meetup groups and I see a completely different group of people at each one, even though there’s obviously a lot of overlap between some of these groups.”
When our conversation turned to open data, Gerber nominated one project as a standout: the development and implementation of an open data annotation model. Gerber’s involvement with the Mellon-funded Open Annotation Collaboration led to the development of the idea at a time when annotations of scholarly communication were traditionally “siloed” and inaccessable from different journal publications.
When a similiar idea for scientific journals was developed by the Annotation Ontology initiative, the groups combined the models and created the W3C Open Annotation Community Group, which rolled out versions of the model in 2013.
“It’s that layer of commentary over the web that was always part of the vision of the web; back when Tim Berners-Lee described it, it was a two-way communication platform, but somehow that part of it just got lost along the way.
Finally, on her success in computer science, Gerber comes once more to music, and stresses the importance of finding people who share your passion.
“It’s like playing an instrument,” she smiles. “You just have to put a lot of time in to get good, and that also means if you love it, that’s really easy.”