Question Crime

Team Name: 
Question Crime


The project was driven out of a desire to raise awareness of crime statistics data and help the public take practical measures to protect them from being a victim of crime.  Looking through the media it was clear that many types of crime are increasing and this is creating fear in our communities.  This is therefore a pressing issue.  We could see in the data that a lot of this crime was the result of careless or poor informed individual choices.  So our hope was to create a tool to raise awareness and provide people with useful tips and information to help them stay safe, as well as correct some common misconceptions.

Our initial step was heavy analysis of crime data, not just in isolation but joining it with a host of publicly available data.  This included census data (ABS), poverty data (NATSEM), accessibility data (AURIN), gaming statistics, charity data and so on.  We created a tableau project to analysis this, not just in 2-dimensions but with multi-variable analysis.  We saw in this analysis key trends, such as the detrimental impact inequality has on house and car crime, along with the high overall crime rates (both offender and victim) associated with low income and education.  This analysis gave us insight into the solution required and to whom we should target it.

After discussions with Brad, the Crime Statistics Agency representative, we decided the best solution was to build a mobile app.  It would be focused on three specific areas which impact people, these being the threats of personal assault, car theft and house burglary.  The app would have two facets.  The first is a “here and now” solution.  Based on the user’s location and time (both time of day and season) the app will provided relevant information and tips.  It is built on a simple traffic light concept to communicate risk levels in a simple way without alarming people.  It would provide useful tips and information which the user needs at that immediate moment.  The second facet is educational and for this we decided to produce a quiz game.  This is targeted at younger people and school children and is written in such a way as to educate in a fun way.  The questions are chosen analytically to ensure both them and the answers are relevant and informative.  The app is data light as we believe a lot of our users have limited data literacy or even interest.  Yet, they want to be sure that the information we provide is indeed built on robust insight.  The app also goes further to provide a deeper level of data to those who are interested.

From a technical perspective, the app itself was built on java script, ruby on rails, html and css.  It was designed meticulously with a well thought wire frame. The goal is to be a simple as possible with as few barriers to entry for the user as possible.  A first version of the app was successful implemented.  The app is built in a flexible way to allow new data to be loaded which is relevant to the current circumstance, such as rising car crime or terrorist threats. It app itself worked well, although the answers to some of the questions from our developers was less than desirable!

We believe the app is a very effective ‘mash up’ of a host of open source data in a simple yet impactful way.  Although it is not a focused ‘get home safe’ app, it provides to insight and education to help users to just that, get home safe.  There are many great journalistic stories in the analysis, such as the relationship between inequality and car and home theft.

Next steps are to do further testing on the target audience to fine tune its use.  It has capacity to grow and become a usable repository of crime prevention information, such as maps of police stations and other safety spots.  Likewise we can build links to more resources. There will also be increased data exploration and interaction - an area we hope we can educated our users more on.  We are very keen to expand to domestic and family violence as this is a very pressing issue.

Overall, we are pleased with the outcome and hope that the app will help people in Victoria feel a lot safer in their communities and make life more difficult for those who may steal or cause harm.

Datasets Used: 
Crime Statistics Agency: This was the basis of our study and the website has a host of information. The issue we faced was that the data was heavily aggregated but that just meant we needed to be more creative in our analysis. ***** Victoria Police: The crime stats were supplemented with data from Victoria Police, which was particularly useful for quiz questions. ***** Victoria Police Station Locations: ***** ABS Census Data: A lot of data was used from the census, which allowed us to compare crime with a host of variables, including ethnicity, income, age ranges and dwelling types, as well as gathering basic information such as population and dwelling counts. We also used the data to build our own measures, such as inequality within LGAs. ***** AURIN: The vampire data was also very useful for comparing various types of accessibility with various types of crime. ***** Data Vic: This was a great repository of all sorts of measures we can join to the crime data at LGA – such as the gaming data. ***** Charities Data: We did not have time to map this, but relevant charities will be mapped along with the police data. ***** NSW Crime Data: Gave us lots of ideas! ***** NATSEM: The poverty data we scraped from here was invaluable.

Local Event Location: