Statistician Clare Taylor, working within the City Technology Platform project for Future City | Glasgow, studied three cities in the USA; from the east to the west coast, and the windy city too. American cities New York, San Francisco, and Chicago are embracing open data. Conducting her research, Clare accessed the 3 USA open data cities’ websites on 19/06/2013. All use software from Socrata to run websites where the cities’ citizens (and anyone else interested) can access a vast amount of open data about city systems and the urban environment.
How we can assess the popularity of a piece of data found on the USA open data websites? One way is to count the views, and another is to count the downloads. It is not an exact means of measuring the popularity since some data sources will have been on the website since the very beginning, but others that appear less popular may be recent additions. However, it does give us some information about what users of these open data websites want to access.
Of course, viewing and download rankings differ but, in general, users across all 3 of the North American cities like information on parking, crime, restaurant safety, and transport – including public transport, associated online ticket buying and taxi registration.
3 Smart Cities: open data pioneers
New York City
The prizes on offer can be as much as $50,000!
The New York City website has an impressive 2160 datasets available. This city is a big supporter of app development; they regularly hold competitions and hackathons. In 2009 NYC BigApps was created, and has helped launch close to 300 apps. The aim is to improve the lives of those who live, work and visit New York City. The prizes on offer can be as much as $50,000. In addition, the NYC BigApps competition has meant that over 1,000 datasets have been made open to developers around the world.
The Chicago website has organised their open data website with types of data available (9 types: including maps, charts, external datasets, APIs, filtered views and more); categories of data (16 categories: including Health & Human Services, Public Safety, Ethics, Community & Economic Development, Education and more); and a wordle – a tool which creates “word clouds” from text that you provide, giving greater prominence to words that appear more frequently – of almost 200 topics. In the wordle, ‘crime’ and ‘police’ jump out as the largest. Then ‘business’ ‘performance metrics’, ‘GIS’ and ‘service delivery’ all appear fairly large. The topics ‘public art’, ‘vaccine’ and ‘snow’ are much smaller. There are many more topics of average size, such as: ‘streets’, ‘trees’, ‘inspections’, ‘lobbyists’ and ‘sustainability’.
We’ll take a closer look the San Francisco website, from the pioneering city on the bay; the backdrop that boasts Silicon Valley, filled with innovation, entrepreneurship and a thriving start-up culture.
Of the 619 “sites”, or sources of data that were to be found online when Clare accessed their system on a summer’s day in June, there are a small number that are the clear favourites.
What are the numbers? Note that these are taken from day one, back in summer 2009 when data.sfgov.org was initially published, through until 19/06/2013, when Clare did her research.
The highest number of views for a single “site” (data source) is 15,863 views: this is for a map of crime incidents over the previous three months. The smallest number of views of a source of data since that dataset was uploaded is 2 views for info from the Fair Political Practices Commission from 1998 to the present.
For downloads, the maximum is 93,952 downloads for a single data source: for off street parking lots and garages. The minimum is 0, with no downloads at all for a total of 86 datasets out of 319.
To unpack this:
With almost half of all views, 46% are looking at 32 out of a possible 619 specific datasets. This represents the top 5%. The remaining 95% are spread across the other 587 datasets.
The top ten data sources views account for one third of all views.
For San Francisco the top three data source sites viewed on their open data platform are:
#1. Crime incidents
#2. Businesses registered
#3. Film locations
The top 5% of downloads of data source sites – again, 32 out of a possible 619 – had almost four-fifths of the total, at 84% of all downloads.
#1. Off-street parking lots and garages
#2. Building footprints
#3. Map: crime incidents
Filling the number 10 slot for both views and downloads is data about city lots. This is potentially quite exciting for Glasgow as it looks at ways datasets could be utilised by users – Glasgow has the more vacant spaces across the city than the rest of Scotland put together.
Check back soon for our blog post about making use of vacant city lots in Glasgow via the international award-winning Stalled Spaces project.