I really liked Manovich’s overview of ‘big social data’, and was glad to have read it first of his three readings for the week, to use as a guide. In particular, I was fascinated by two of his paradigms—between ‘surface’ and ‘deep’ data and, in context of information visualization, between description and prediction in explaining what information visualization is (and/or is supposed to do).
What does data visualization do? Is it just descriptive (and if so, how does including a description that is visual improve our comprehension of the thing in question, as compared to or in addition to a description that is purely textual?)? Or predictive, a counterpart to inferential statistics? I’ve always put things into pictures to understand them, and I am crazy about Scott McCloud’s high recommendable book on comics [Understanding Comics], which shows (in comic form) how pictures and text can go together to explain or evoke things like space and time better than just pictures or text alone. So I’m a big fan of this new infographics/data visualization trend.
But I wanted to discuss Manovich’s discussion of sampling and behavior data, though class starts in 5 minutes. To be continued..