How Safe Is Your Suburb was an entry in the Mashup Australia contest.
Click here to try How Safe Is Your Suburb
How Safe Is Your Suburb is an easy-to-use interactive web application for the public to gain greater insight into crime statistics in Local Government Areas (suburbs) in New South Wales. The application can be used for informed discussion and policy development by residents, police authorities, and local government. The applicaion shows how statistics can be applied in the everyday life of the community.
How Safe Is Your Suburb embraces the Gov 2.0 philosophy by opening up a static dataset to the public in a useful way. The user can analyse and play with the data, comment on data, and then share their data with others.
For example, the user can choose different ‘reports’, make selections within each report to compare different types of crime over time, and then see which types of crime are more prevalent in their area. They can view an interactive thematic map of crime that provides a spatial visualisation of crime types across LGAs for a given year. They can also identify which suburbs have higher crime rates in total and in per-head of population. (It makes sense that there is more crime in more populous areas). Users can make comments on each visualisation they are working on.
The application mashes up NSW crime data with LGA boundary files and Census data from ABS. Space-Time Research has classified each offence into different categories to enable simpler analysis. More detail could be added to the application at a later date.
The application is built using Space-Time Research’s SuperVIEW product, and is hosted on the Google App Engine. In the spirit of a govhack style competition, our team of three (one database builder, one programmer, one analyst / writer) started working on the application just over 24 hours before it was due.
We would also like to share our experience of mashing up and visualising the data. We have found:
- There is an unexpected spike in road traffic offences in 2001 and 2002 and then no road traffic offences recorded after that. This is seen across most LGAs. Only by visualizing the data in a chart did we see the problem, and would suggest that the data quality be checked with NSW Bureau of Crime Statistics and Research before releasing this data. Perhaps the data should be footnoted.
- We discovered gaps when joining by LGA – our map file, the ABS census data and the NSW Bureau of Crime Statistics and Research data all have slightly different datasets. We don’t know what year the LGAs in the source data were referenced to, and our application currently joins on LGA name rather than LGA id.
- We chose to refer to the spatial areas as ‘suburbs’ to make it easier for the general public to relate to. We are aware that LGAs are different from postcode boundaries and that the general public will not be aware of the difference between the two types of geographic boundary. Most members of the public may not know what an LGA is and we have referenced suburbs with LGA in parenthesis throughout the application.
Ideas for enhancing the application include:
- Enhancing the share functionality by including a share this on twitter, facebook etc application.
- Expanding the application to allow analysis by individual offence types.
- Incorporating other ABS census demographic data, such as population count to calculate offences per head of population, and inclusion of employment, education, age breakdown etc. to see if demographics of an LGA impact crime rates.