There was a small amount of (presumably unintentional) teasing, which prompted cries of:
Even within Spatial Vision, excitement was high.
Then finally, G-NAF was released to a fervent audience and hailed a saviour.
So now that it has been public for over a month, what is G-NAF actually like to work with?
Early ideas focussed on the aspects of G-NAF that make it unique. One of the touted features of G-NAF is its aliases, which are features that reference a primary address, but differ in some other part of the address, such as number, level, street name or locality. The majority of these aliases represent surface differences – such as referring to a single street number rather than a range (think ’55 Smith St’ rather than ’53-57 Smith St’), level duplication, and so on.
What we thought might be very interesting was the idea of using this data to identify cases of “vanity” addresses, where people intentionally misrepresent their address to appear more… affluent? Cool? Prestigious? Who knows, but wouldn’t it be curious to see the places that people most- and least-wanted to be associated with?
We thought so, but despite the capacity for G-NAF to support this kind of observation, and as unlikely as it seems, it just looks like people in Victoria aren’t vain enough to produce the amount of data required to say anything interesting. The most aliased- locality was Bairnsdale (a rural centre in East Gippsland) with 27 instances, but of the 1800-some localities with aliases, only the top 13 had instances of aliasing over 10, and just over 1100 have a single address that would rather be there than somewhere else, including our predicted frontrunner, Toorak.
At this point we decided to change our approach. One of the beautiful things about open data – especially governmentally created and maintained open data – is that it gives the public a peak under the hood. It gives the man on the street access to the raw facts about the country. But what can the man on the street do with millions of rows in dozens of tables about addresses? Honestly, probably not much.
Something sprang to mind here (probably because many of us at SV have been working on their print maps for the last 2 months): the Australian Electoral Commission (AEC) have just redefined the federal election boundaries in WA and NSW, and NSW in particular had a few divisions removed or significantly modified. But who knows what the effects of these changes on upcoming elections could be, right?
Well, man on the street, what if you had the tools to interrogate what effect the boundary redistributions have had on the number of addresses within the division? Enter “Addressing the changes in NSW” – a simple web map that lets users view and explore their new divisions, and form their own ideas about what the redistributions mean. Danielle and I developed the tool using CartoDB so the map can be easily and freely distributed in the spirit of open data.
At the end of the day, G-NAF is a complex but fascinating dataset, with bags of potential to people who aren’t afraid to get their hands dirty. We learnt this through our project at Innovate@Locate16, and even though when I say we “got our hands dirty”, it is probably more accurate to say that our arms were in up to the elbow, and we had to go home and change because we got G-NAF all over our clean work clothes, I still can’t wait to dive back in!
2016 Graduate Cadet, Spatial Vision