United States of Autocomplete also via Flowing Data
Deeply amusing visualization of autocompletion of search terms by the Google search engine.
Most data manipulation/visualization focuses on some data set.
What if those techniques were turned inward?
That is to say what if data manipulation/visualization choices were visualized? Either in terms of the choices they make or the processes they use?
I suspect that is a largely unexplored area of visualization and one where given the differences in terminology, topic map could be helpful.
Questions:
- What data manipulation/visualization choices would you suggest for visualization? What about them do you think could be visualized? (3-5 pages, citations)
- How should we go about designing a visualization for #1? What is it that we wish to illustrate? (3-5 pages, citations)
- How would we compare that visualization to other visualizations? In topic map terms, what are the subjects and how do we identify them? (3-5 pages, citations)
PS: One additional thought. What if we had a data set and compared two techniques, manipulation or visualization, on the basis of how they treated the data? What data was lost or not included by one technique but was by another? And does it vary by the nature of the data set?
Could be useful for topic map engineers trying to decide on the best tools for a particular job. I am sure that sort of information is generally available in data mining books or in stories about data mining exercises, but being able to visualize the same could be quite useful.