Hybrid Data Visualization
“Finding a way to convey both informativity and aesthetic value”
Data has become a hot issue today with the expeditious development of informative technology. This trend has impacted the inspiration and motivation of artists and designers to the extent that many creators have created a bunch of data-driven projects, ranging from informative data visualization to visual art. With this phenomenon, the most observable fact is that conveying both informativity and aesthetic value is not easy when creators deal with data visualization. Take informative data visualization for instance; aesthetic value is often ignored in this form of creation because visual elements, such as color, form, and movement are strictly controlled for the clear and precise transmission of information. Meanwhile, in terms of data art, informativity is often neglected in order to amplify aesthetic impression of data. For this reason, I have explored ways to satisfy both standpoints by being participating in several different data visualization projects. And, I have believed that there exist hybrid data visualization as an ultimate solution that brings together both aesthetics and informativity. Based on this, I will delve more closely into the intricacy and reality of my belief, reviewing a general knowledge of data visualization and analyzing representative examples.
Data is a series of collection or record about a fact, such as numbers, words, measurements, and observations, even descriptions of things. Also, it can be defined as the plural of the singular piece of data, called “datum”. This is only recorded facts and thus, has not any message and narrative par se. Even so, this can be transformed into a form of information or a specific insight by being analyzed and interpreted by a creator.
And, data-driven works can briefly be divided into infographic and data visualization based on the means of creation, the amount of data, and aesthetic value. For one, infographics are manually drawn, using graphic tools like Adobe Illustrator, Photoshop, and so forth. Also, in general, these are aesthetically rich due to strong visual contents, while these are relatively data-poor because each piece of information must be manually encoded. By contrast, being rendered with the help of computerized methods usually algorithmically draws data visualizations. This product feature results in the fact that this form of works is easy to regenerate with different data. By its repercussion, the same form may be repurposed to represent different datasets with similar dimensions or characteristics. Furthermore, this is often aesthetically barren since data is not decorated, whereas, this is relatively data-rich because large volumes of data are welcome and viable, in contrast to infographic. 
INFORMATIVE VS. AETHETIC VS. HYBRID DATA VISUALIZATION
Based on this fundamental definition of data and data-related works, I will look more into informative, aesthetic, and hybrid data visualization. For a start, the focal point of informative data visualization educates readers by providing specific information.  For this reason, it has the most highest informativity compare to other forms of data visualization. Thus, informative data visualization is used in the projects that reveal scientific, cultural, and social research. For instance, from old time, scientists have recorded sunspots. In the case of Galileo Galilei who is an astronomer, he made a series of sunspot observations by drawing the patterns of sunspots’ trace in 1600s. Since then, sunspots-related data visualization has been made by an official science research organization like NASA, being analyzed the impact of the scientific phenomenon. This research and visualized data enable scientists to unveil many secrets of the Sun and the Solar System, which lets people to understand more about the space. Meanwhile, main media organizations, like New York Times has allotted their budget and efforts to visually organize what they report to the public because visualized data allows readers to rapidly understand massive information by enabling them to cognize patterns and meanings of data. Take ‘President Map’ for instance, this site shows the result of 2012 US election, providing several different dimensions that readers can analyze data. On a glance, it is evident that the approval rate of Obama supersedes that of Romney with 332 and 206 respectively. Also, the given site illustrates that the South and East citizens, despite the fact that the Central citizens defend Romney, mainly support Obama. When readers click left buttons, the site allow them to explore the data along with different dimensions, such as States, Counties, Size of lead, and so forth. By this process, the data visualization enables readers to newly learn facts and knowledge.
On a closer look about the informativity of the project, the most apparent part is the form of map. This metaphor intuitively implies about what the data visualization presents, arousing readers’ curiosity about the concept of project. Also, we should look into its color-set. The main colors comprise of two complementary colors including less saturated primary blue and red. These colors guide readers to be aware of the fact that two candidates are in a competitive relation at glance. What is more, the contrast between the colors obviously demonstrates two candidates’ competitive relation over the all visualized maps thus, helping people clearly to compare the given data. In terms of typography, well-labeled text information is conducive to thoroughly understand the meaning of information because this presents even very detail information, which cannot be clearly conveyed with visual elements. Together with the typography system, its interaction system also induces users’ further absorbing of information. A bubble-window that pops up by a users’ mouseover action is a great affordance, so much so, that users successively click selected buttons to read more information.
Based on this elucidation, we can see that metaphor, color, typography, and interaction are closely associated with the informativity of the data visualization. Though we should consider the specific characteristic and goal of projects, these elements are a prerequisite of maintaining informativity not only in data visualization works, but also other design projects. Personally speaking, figuring out these features would become a good guidance for creators when they deliberate which elements in a data visualization piece should be maintained or changed in order to make their works more aesthetically.
Meanwhile, aesthetic data visualization often entails unidirectional encoding of information, meaning that the reader may not be able to decode the visual presentation to understand the underlying information.  This form of data visualization often focuses on showing visual, audio and any other forms of patterns which are transformed from dataset; thus these can give audience emotional impression. Take ‘Fiber Optic Tapestry — 50 Different Minds’ for instance, this work is a computer controlled light system that displays Twitter streams into a woven fiber-optic tapestry. This shows a color pattern that lively transform according the Twitter Tweets of color words. By this point, it allows audience to meditate the color change while looking the art piece. In addition, some works reveal their aesthetic impression with the power of media and space. In the case of ‘Unnumbered sparks’ created by Janet Echelman and Aaron Kobin allow people to engage in the part of artwork, inducing them to paint on the surface with the use of their smartphones. To elaborate, the system assembles the drawing data that visitors send to the project server. Thereafter it lively projects the images on the interactive sculpture in the sky. By doing so, this sculptural data visualization not only represents the beauty of data, but also creates an ambient environment. Also, ‘Moveable Type’ created by tech wizards Ben Rubin and Mark Hansen is another representative example of aesthetic data visualization. This is a sculptural art pieces installed in the lobby of the New York Times building that turns two opposing walls into large information dashboards. The content of the dashboard constantly changes. A live feed from The New York Times provides snippets of text to the displays. Additional input comes from nytimes.com website visitor stats and comments as well as content from the newspaper’s archives.  The most interesting part of the creation is sound. A minimally designed sound effect is rhythmically played along with the change of typography. Interestingly a serious of audio effects sounds like music, allowing visitors to meditate the generated sound together with the movement of typography in 560 screens.
I believe that the form of hybrid data visualization can be defined based on the above elucidation. I would think that the epicenter of hybrid data visualization is the form of informative data visualization. Also, hybrid data-driven works include more aesthetic features compare to information-centered projects. For this reason, not only the project conveys information, it also represents the beauty of data. One of the observable examples is “wind map” created by artists Fernanda Viégas and Martin Wattenberg. On a glance, though it seems like a data art due to the beautifully illustrated wind pattern on the world map, it obviously includes several different informative elements. Like the previously referred ‘President Map’ project, the USA map intuitionally shows users the fact that the site is about USA related issue. Also, the legends in the left side of the website implies the fact that this project demonstrates a serious of wind speed. When users zoom in the screen by clicks, the site provides newly detected wind patterns of a selected area. With this interaction, bubble windows which are activated by users’ mouse over events constantly provides its area-specific information, including wind direction, location, and so forth. In supplement, we need to look through the project, named ‘Cascades’. This is a data visualization tool for tracking twitter activity around New York Times content. And, the members of New York Times data lab, including Jer Thorp who is a data artist, create this project. In an article, he explains the project, saying that cascade works by showing us something that we’ve never seen before — the underlying “architecture” of sharing systems within social networks. Not only can we see that a piece of content has been shared from person A to person B, we can see, in really granular detail, all of the activity that happened in between.  With this concept, this project allows users to explore the information with two main modes and 3 different view options under the three-dimensional environment. For this reason, well-designed interaction and information architecture provide users to cognize a new knowledge, while at the same time the three-dimensional graphic environment and quantity based visual representation create the aesthetic value of the project at large.
Considering the elucidation of three styles of data visualization, we can deduce that a hybrid data-driven project needs to have a concrete information structure, as well as a well-formed visual system that reflects its data narrative and metaphor. I believe that adjusting a balance between the freedom of visualization and informativity is the core of hybrid data visualization. In other words, a means of stringently empathizing informativity will adversely affect the creation of aesthetic value, while over-expressed visual imperils the present of informativity, adversely affecting readability and legibility and users’ interaction flow in a project. In addition, typography is that we should deliberately consider while creating hybrid data visualization since texts have an ultimate power which clearly convey the concept of project. Because of this reason, it is inevitable that a serious of text should be designed when the project should reveal the detailed and sophisticated meaning of data.
Despite the above research and analysis, I suppose that there might be not an objective solution that gratifies both informativity and aesthetic impression in data-driven projects because each person has a different concept about these terms. For this reason, we cannot also jump to a conclusion that a certain work must only be in one category. Nonetheless, I believe that this consideration and inquiry might be good starting points to develop an unique method of manipulating data as a designer and artist in contrast to data analysts and developers who mainly focus on informativity. Based upon this expectation, I will study more closely into the intricacies of this question and other related subjects further.
1) Designing Data Visualization / Noah Iiinsky & Julie Steele / O’REILLY/ P8~P11
2) Designing Data Visualization / Noah Iiinsky & Julie Steele / O’REILLY/ P8~P11
3) Designing Data Visualization / Noah Iiinsky & Julie Steele / O’REILLY/ P8~P11
4) Moveable Type / http://art-nerd.com/newyork/moveable-type/
5) Cascade / http://www.chrism.com/visualize-tweets-cascade/
‘President Map’ http://elections.nytimes.com/2012/results/president
‘Fiber Optic Tapestry — 50 Different Minds’ http://ligoranoreese.net/fiber-optic-tapestry/
‘Wind Map’ http://hint.fm/wind/gallery/mar-21.js.html
‘Unnumbered sparks’ http://www.unnumberedsparks.com/
‘Moveable Type’ http://art-nerd.com/newyork/moveable-type/