I was originally intending to work with specific G.I.S. crime data. My source was Bryan Kinney, a Phd candidate in the School of Criminology at SFU. Bryan had originally taken this course a couple years ago, and was understanding of my queries. He directed me towards a few suggestions, including the resources at the Crime Prevention Analysis Laboratory or CPAL lab.
Unfortunately, after weeks of emailing and asking, I was turned down by the lab, and I was told that the information was highly classified and needed special permission of the Police Department. Seeing as time was drawing near, I decided to forego that option, and use data that was on the network drive instead. Most of the data used was found S:\GIS Data\City of Vancouver. This data was in ArcView and when I was deciding which of the three variables to choose, I overlayed the themes and chose three of the most appropriate subjects.
After choosing my three files, I proceeded to use the FME translator to convert the files to IDRISI, this was a relatively simple procedure that thankfully did not cause me any problems. The most glaring problem I found in data collection was the various sources that I tried to go to for specific crime data. Other than my criminology contacts, I did extensive research from internet sources. I found that I was quickly able to locate promising websites. But after downloading the data, I realised that a lot of the data out there is quite specific and tailored specially for the project the author was working on. There were many files that were incomplete... ie Arcview .shp files without its accompanying counterparts like .dbf, .shx and .prj files.
I believe
that it is not easy to find usable crime data, its just so specialised
a type of data, that it would not be cost-efficient for the government
to make such data available to the public, but merely for internal purposes.
This part of the project was certainly the most frustrating, as I was not
able to fulfill my original intentions, and had to compromise with canned
data from the network, that prevented me from doing a real hard hitting
analysis. I do believe that I've learnt how to use G.I.S. towards
detecting crime patterns, with the right data in the future, I'm sure the
further analyses are limitless. So while the availibility of data
in this project greatly hinders its original purpose, the learning of the
tool has certainly not been forsaken.