Studying information visualization

What an information visualization profile could look like

After being admitted for the computer science masters programme at the FHNW, I was asked to define a profile I would like to work towards. From the start it had been clear that I wanted to have information visualization at the center of my studies. But what should an information visualization curriculum look like? To my surprise, I didn't find a university which offered a complete information/data visualization programme. It's usually just a single course mostly in design or computer science programmes.1 So I took on the challenge and composed my own curriculum. It's naturally based on the courses that are available to me at the FHNW. But maybe it's interesting to someone else, looking for some input, like me.

Update (6.10.2017)

Here is the list of classes that I took which seemed relevant:

  • Information Visualization (obviously)
  • Predictive modelling (different statistical modelling techniques)
  • Data processing with Python
  • Image Processing
  • Machine Learning

What I am missing most in this list are more communication-related things like writing, visual communication, etc.

Update (5.11.2018)

It is definitely interesting to have an overview of the research in Information Visualization. Probably the fastest way is to just read Tamara Munzners book Visualization Analysis and Design where she summarizes much of the most relevant research in the area. After that it's easy to dive directly into the research papers that sound interesting. A good starting point is everything that's published at IEEE Vis. There are a lot of great ideas in research papers. Researchers are usually not that interested in putting them into practice because they get paid for ideas and not products. Most of them will probably be very happy to hear that their research has been used by a practicioner.

When doing information visualization, it is essential to master three types of tools throughly:

  1. A statistical computing environment (like R) to bring any data into the shape you need it to be in. In my case it's pandas and Jupyter Notebooks.
  2. A rather high-level visualization tool (like Tableau) to explore the data and find interesting patternstories For me it's Vega-Lite
  3. A low-level visualization tool (like D3) to build custom visualizations that present the data in the form most suitable for the story.

Don't worry to much about the precise tools. The most important thing is to know, that there are three types of tools with very different goals but you will need all of them to be quick and good in creating visualizations.