Challenging but exciting!
Our goal is to make campus safer. It is hard for us to achieve that goal for two reasons. One is that the goal is really too “big”. To make some deliverables in one semester, we must figure out something concrete to do in a short time. We came up with some interesting ideas, such as mobile emergency button and Stingerette reservation system. However, all of them have been realized. We were really frustrated for some time. Is there anything doable to improve campus safety, which can be achieved in one semester? Inpired by crime map on GTPD website, we ended up with a new visualization method to present these informaiton. So, we refined our “big” goal to provide increased awareness about campus safety issues for students to increase their ability to be safe on campus. Things became much easier after we decided the direction to go. Although I didn’t know information visualization at all, I looked at the effects of various visualization methods through the links on the course page of infovis course in Tech. Finally I found “heatmap”, which became the visualization method in our crime map.
The other challenge is that we have no clear stakeholders. We are not tied to a certain organization. Although Tech’s adminstration, GTPD, current and future students may all be concerned about our project, we are directed by none of them. This made the acquirement of the access to crime data extremely difficult. And I am the one which work at the database side. At first, I struggled to use crime data in the excels from GTPD website. It was far more difficult than I expected. For example, some date fields are in “string” format and some are in “float”(which I didn’t remember very clearly..) format, which need to be converted before stored into database. For another instance, a large amount of the addresses are “landmarks”, such as “starbucks”, “Klaus”, which are impossible to be geocoded through Google API. In order to geocode these addresses, I first scaped the mapping of lankmarks and coordinates from gatech campus map website. And then I did some kind of string match to find coordinates for landmark addresses in the excels. We finally can scrape crime data from the dashboard web page made by GTPD in November, although we still have no access to their database.
In summary, the project is very challenging, solution was not well defined at the beginning and we have no clear stakeholders. But finally we made something useful! That is a great experiece for us to overcome these obstacles. Although we don’t achieve much this semester because we spent a lot of time on finding solution and retrieving crime data, the project is very promising, For example, after the integration of cameras on campus, camera feeds can feed the map. What’s more, student outsoucing is also very charming, although many issues need to be considered. For example, several students may input a same crime, and student may not be willing to input crime data or they just don’t have the app in mind when crimes happen. In addition, because of the current achievement, GTPD and APD may be willing to provide crime data.
Thanks to students in C4G class. You gave us very wonderful ideas, such as integration with camera feeds and students outsourcing. Thanks to Prof.Abowd, you kept us always in the right direction. Thanks to Prof.Vempala, I learnt the right attitude to hold through the conversation with you.