His and Hers Colors

This visualization by Stephen Von Worley illustrates what names men and women use to describe a particular hue. Each dot represents 2,000 of the most frequently-used color names, sourced by a 5,000,000+ sample from a color survey by XKCD. Color names used by women are in the top half, men in the bottom half. Not only is data showing how the genders perceive color interesting, the results themselves are beautiful. One aspect of the responders that wasn’t evident from the survey was their cultural background, which can affect their perception of color. Indeed, this visualization by Muyueh Lee shows the difference between how the English language and Chinese language describes colors — there are far more English names for colors than Chinese ones.

Humanity’s Cultural History

As a son of immigrant parents, I find the migration patterns of people interesting, often wondering why they migrate to a particular area over another. This visualization was created by Maximilian Schich and others from The University of Texas; they took data from the Freebase and charted the places of birth and deaths over several centuries.  Places where a high concentration of people died provide fascinating insights into the cultural city centers that drew them there. Schich’s team also looked at these results in the context of phrases used in literature at a given time (via Google Ngram Viewer) to infer population and depopulation trends of a key city. This data isn’t as complete as it could be, however. Freebase is a Google-owned catalog of well-known places and people, but there is bias in this source data towards the U.S. and Europe, so the results aren’t globally representative.

The first project I want to introduce is called “Where We Came From and Where We Went“. As the name suggested, it visualized the data of migration from state to state.

It seems people born in the capital tend to choose live in other places, such as Maryland and Virginia.

The portion of Nevada’s residents who was born in Nevada is the least in all states in the US. Most of its residents came from neighboring states.

This work is really easy for me to get information from, which I think is the most important thing to judge a data visualization project — correct me if I was wrong. It’s hard to manage a large amount of data and this work has done a great job.

There are still some points I think should be improved. The allocation of color and height is not explicitly stated. Like shown in Nevada’s chart, the color of “Born in California” and “Born in Utah” is identical. Also, I noticed the bars shown in the graphics are divided into four groups — the native, the states, the regions, and outside the U.S. It won’t come at first glance, and that’s where might make people confused about.

Similar work can be found in the New York Times’s Upshot, like this one — Mapping migration in United States. There is a large possibility that they share the same data. But this one uses graphics instead of charts to represent data.

————————————————————————————

Selfiecity is an interactive website showing data and analyzation about a research into selfies. Although I’m not a selfie person, it’s still of great interest. This project try to figure out the features and trend of selfies in five major cities in the world.

What they do to collect selfies is collecting random photos from Instagram and ask people to label if the photo a selfie or not. I think this data collecting solution is a bit trivial. There are 140,000 images they collected! Maybe it’s a better way to label selfies using some machine learning solutions. Some of the selfie analysis are interesting, like gender distribution and age distribution, but some are not. You can filter selfies using mouth open or closed option, but it is of little importance to me. Maybe it’s better to detect the background in a selfie, to figure out in what conditions people are tend to take a photo of themselves.

————————————————————————————

There are really a lot of good stuff I want to share. Here is the access to these works!

http://gizmodo.com/the-best-data-visualizations-of-2014-1674001488

I want to talk about visualization of expressive movement. To begin, I’d like to define data visualization as taking a complex data set and distilling it into a human-digestible format which enhances our understanding of the data.

Motion Bank

Motion Bank is an ongoing collaboration project involving a group of designers, dancers, and engineers that archives the motion of dance. The visualization techniques output from the collection of three-dimensional data creates a more measurable understanding of dance as a science, something that can be broken down into understandable parts. The purpose for this project is to enable comparisons for learners as well as to understand the range of motion of a dancer and optimize the space around them. Despite this seeming to be their goal I found the most completed aspects of their project to be more of a representation than actual visualization, for there was little to no visible measurement in the final output, which makes it difficult to truly understand. However I found that the process work shown in the website and in parts of videos shows a lot more interesting sets of visualization, making this collaboration a note-worthy project.
I find the website for Motion Bank to be a little confusing, not to mention the language is primarily German. The website was a little confusing to navigate so it is difficult to really get a clear understanding of what’s going on. I did not see any particular documentation to really explain Motion Bank or what its ‘final result’ is, so to speak. It seems to be an on-going process simply exploring methods for archiving dance. That idea alone makes the project an interesting precedent for its playful and experimental qualities, testing numerous techniques for visualizing motion. The idea of using visualization to learn things is not unheard of, so I’d be interested to see the data being used for some other purpose. Like expression. Which leads me to the next project.

More
Link to un-sharable video

E-TRACES by Lesia Trubat. 2014

E-TRACES, memories of dance from Lesia Trubat on Vimeo.

This recent project collects the data of a dancer’s movement and creates something expressive from it. Now I know that the data visualization from this particular project doesn’t collect the amount of data that we will be expected to from our project, but I still think that the purpose of the visualization is something to take note of. The purpose for the visual effects happening here is to work in collaboration WITH the dancer and push the expressionism happening with her motions into a still drawing. Certainly this approach can be pushed to another level, covering more data, and continuously changing itself in response but not necessarily showing measurable data as much as it is expressing…something? Really, I just like the idea of data visualization being something that does not necessarily depict information on a one-to-one level the way a graph or chart would, but instead become its own art form.
I think that the video documentation could be pushed beyond an overlay on the dance itself. What if the actual dance wasn’t seen at all, and its only purpose was to output some visualization? Can that visualization become legible in its own way? Can it be transcribed back into dance if someone learned how to “read” it? I’m interested in this kind of ‘telephone’ game being played and I’d like to see that taken a step farther.

“Humanizing Motion Graphics” by Matt Pyke. 2012

Speaking to my interest in expressive data is Matt Pyke who talks a lot about anthropomorphizing digital graphics so that people can empathize. His work is not data visualization so much as it is visual representation of motion. Though his work does not invoke an understanding of a set of data, it is evocative and exciting. Can data really be this exciting or emotional? Oh yes. Yes it can. Just stick some googly-eyes on it.