From 5th to 11th February, the whole semester visited Berlin's creative offices in order to get familiar with interactive agencies such as art+com or kling klang klong and their field of work. We met many inspiring people at these offices who gave us a lot of insight in their creative processes. Especially for me, art+com and kling klang klong were highlights of the trip, as their work is one of the related works and their projects are the most inspiring ones for my spatial installation.
From 12th to 16th of February, I was attending the Z-module «Video als Medium der künstlerischen Entdeckung». I used the very limited time left for a meeting with Benjamnin Wiederkehr who is the head of the Zurich-based agency Interactive Things. Benjamin gave me a lot of insight in the agency's creative process, how they approach a new data set and how they finally end up with a final visualization. It is mainly an extended double diamond, adjusted to the very specific needs of a data visualization process. It starts with exploration of the data set, followed by desktop research and intense exchange with the data provider. The next step is visualizing the data through plotting it in a quick way, a data set prototype so to say. One string of the process is evaluating these plots on a data-basis; the other string is creating an interface and interaction principles with example data. These results are brought together and are tested with users. This is a very long-term iterative phase that lasts until a final visualization is created.
While talking, he was asking a lot of critical questions which helped me immensely to get a more clear picture of what I want to do and where I situate my work. He also suggested to visit other swiss design agencies who are dealing with data visualization and recommended me two books, which deal with strategies of how to set up and structure a data visualization process.
My main goal of this short phase was, according to the advice of Benjamin, to get to know my data set really profoundly. Therefore, I intensely studied the data set and its given variables and got in contact with Noah Bubenhofer again to clarify potential questions.
Furthermore, I was getting in touch with the topic of 3d-visualization in d3.js in order to generate a 3d scatterplot, which will be my prototype and is necessary to derive a 3-dimensional spatial concept from it. There are several ways to approach this, but nearly all deal with other open source libraries, which are using DOM commands and can be directly implemented into the d3.js synthax. Possible libraries are for instance Three.js, x3dom and the d3-3d plug-in from Stefan Nielke, which is specially designed to work with d3.js, whereas the other tools are independent java-based 3d libraries. Since the plug-in is not working properly in my case, I decided to use x3dom. It has a fairly easy understandable sythax and an extensive online documentation with some tutorials, easing the beginning.
To start off, I explored scatterplots that have been realized with these tools and found good examples, which I want to apply onto my data set. The first example shows one possible way and a demo of a scatterplot in a related way Noah Bubenhofer realized it. I also found a nice online tool to explore and quickly review graphical data. This enables users to understand connections easily by quickly changing variables. The next one uses x3dom and d3.js in combination to create a 3d visualization and this simple version of a scatterplot as well.
The next step is to create a 3d prototype of a scatterplot in the way Noah Bubenhofer made it, using the last two examples as a base, and then to plot the different variables in different combinations. This will allow me to have a basis on which I can implement the findings of the user testings and surveys.
