Sample of heat map of Barcelona city (using all parameters)
Sample of intensity indicator of zone
Image of Barcelona without many intensity zones (because of the parameters)
Stores for Me
We are a team of two classmates that really like to take parts in Hackathon. Our passion is to visualize data and transform it into useful information. So, when we saw the GFT challenge, we choose it without blinking; we really wanted to use the 10GB of data provided and take advantage of it.
What it does
It is a heat map of transactions made by clients of a Bank (HackUPC '16), data provided by GFT. It consists of a Web application that allows the user to enter an interval of age, average salary and situation; in order to search the best commercial areas that suits their requirements based on previous transactions. So, for example, a mid-aged man that is a father can see where people like him spend their money. It is not only useful for customers, but also for the Bank itself, as it can analyze on which area their clients spend more money based on their features, therefore optimizing aspects such as advertisement, or placing ATMs.
How we built it
We built the system only with our laptops, using Sublime Text and Vi. We use spaces over indentations.
Challenges we ran into
The most difficult challenge was to deal whit the big quantity of data, as it was slow and very demanding. Other challenge was to find a proper visualization for the data.
Accomplishments that we're proud of
We are very proud of the processement speed reached after several optimization processes. Our system can process GBs of data even in less than a second, making the system very fast and usable. This, along the developed map makes the web a very friendly application.
What we learned
We've deepen our knowledge on how to process big quantities of data using MongoDB and NodeJS. Also, we've learnt how to develop heat maps, and how to work in a very stressful ambient.
What's next for Stores for Me!
Next step is to integrate Barcelona Open Data to show data about stores, markets and businesses (for example, to show that data in the heatmap or a list of more popular markets according to customer features). The program is modular so it can be easily extended.