RiSE - RSPA
oct - dec 2021
A web-based tool to creatively visualise student attendance data to find patterns and determining the cause of attendance issues.
About
RiSE (Research in Society Enhancement) is a group of researchers at De Montfort University, Leicester using Artificial Intelligence to address issues within society, such as crime and school attendance. RSPA (Recommender System for improving upon Pupils Attendance) was the project I was assigned to.My task in the team was to use my skills as a creative to develop a visually interesting way to display attendance data gathered from a school in Milton Keynes. The dataset included information such as student postcode, ethnicity, religion, and attendance values for each individual subject. I was given little guidance or restrictions on what I create and how I create it.
I decided to use JavaScript with the 3D library Three.js. I chose this web-based approach as it would simplify development as opposed to using a more complex system such as Unity, therefore simplifying it's distribution and use. Ease of use was important as it was to be used by school staff who are not expected to have experience with complex data representation software.
To process the data we received from the school, I used Python and the Pandas library for data manipulation and analysis. The formatting of the raw data had to be altered slightly in order for the JS system to read and display it.
The first prototype I created was based on postcode, allowing the user to select a certain subject and view how the attendance of that subject changes based on the student's postcode, displaying the data as pillars on a map. The user could also place pins on the map with representations of certain distances from the pin. Patterns were revealed, such as how proximity to the school changes the attendance of certain subject.
The second prototype was intended to expose how factors relate to and effect each other. Each student was represented by a coloured sphere which are then sorted into boxes of options within a category. The spheres are coloured based on the options of another category, allowing the user to see how much of each colour is in each box and therefore determine patterns in the data. The spheres also represent the overall absence of each student, the larger the sphere, the higher the overall absence. The user can click on each sphere to view all the details of that sphere, allowing them to find who each sphere is representing.