Lima and Sinclair both emphasize the importance of the viewer being able to interpret the data in their own way and not be led to a specific conclusion by the creator of the visualization. This means that everyone should be able to draw their own conclusions and even find make a new discovery in the data. I believe the purpose of DuBois’ work is to take common and well-known datasets, display them in artistic form that allows the viewer to easily understand the data, and develop their own understanding based on what they see.

To achieve this, DuBois implements his own technological hardware and software. He enjoys writing his own scripts to scrape data from the Internet, and developing a beautiful physical rendition of the information that allows it to be intuitively understood. One example of this was a network graph that DuBois generated of all of his email communication. He used his own software to generate the visualization, and then took it one step further by hand drawing the connections and individually writing each name. His email self-portrait effectively demonstrates his ability to bridge the physical and digital worlds.

My favorite work of DuBois is “A More Perfect Union”, which he created by first scraping data from 21 dating sites and downloading 19 million online dating profiles of Americans. He then applied algorithms to the data that found the most unique word for each city, by analyzing the biographical data from each profile. He then used this data to generate very aesthetically pleasing maps that plotted this data on a road atlas of the United States. I enjoyed this piece because DuBois found a unique way of displaying differences between different cities in the US. Most people, including myself, would not have considered scraping all of this data and then filtering it the way he did.

Assignment 1

Listen To Wikipedia

Listen To Wikipedia

Although most people use Wikipedia on a daily basis, it is difficult to appreciate the scale of the web site. Listen To Wikipedia, a real-time audio-visualization of Wikipedia edits, helps effectively conceptualize that Wikipedia is one of the largest information libraries on the Internet. This application monitors all edits being made to Wikipedia by displaying circles for each edit made on the site. The size of each circle corresponds to the size of the edit and the colors represent the type of user making the change. Since Wikipedia edits are comprised of both addition and removal of text, Listen to Wikipedia complements the display of circles by playing bells sounds when additions are made, and strings sounds when subtractions are made. Listen To Wikipedia is interesting because it becomes easy to understand how dynamic Wikipedia is when seeing and listening to this project. Stéfan Sinclair explains that  interactive visualizations allow for the presentation of data that is sequential and iterative. As opposed to a static visualization, which would have provided a high level, generalized view of the types of edits made to Wikipedia, this project provides a live data stream of edits. This not only ensures that the viewer will never see the same information twice, but also gives them the ability to see exactly what pages are being updated. The high frequency of sounds and circles appearing on the screen gives the viewer a better level of understanding and information as compared to what we would see on a static graph.



As Warren Weaver explained in his article entitled “Science and Complexity” (1948), our modern society is dealing with “Problems of Organized Complexity.” We have access to large amounts of data, and need to find the ways of linking them in useful ways. This issue of finding the right data is prominent in the research community. When looking for research papers, it is often hard to find exactly what you are looking for. Paperscape eliminates this issue by visually mapping out research papers. This network visualization helps bring order to the complexity of finding research data by grouping and coloring different categories. Each paper is represented by dot, and is placed in a cluster of similarly colored dots, where each cluster covers the same topic. The size of each dot shows how often the paper has been cited and the dots are linked by similarities in citations. This dynamic visualization makes it much easier to explore research topics, compared to conventional search mechanisms. The versatility of this interface allows the visualization to be used in many ways. Rather than sifting through lists of papers, Paperscape makes it possible to find an interesting topic, see the most popular papers in that field, and find closely linked papers, all in one interface. As a result, it can be useful for non-experts to find gain a basic understanding of a topic by looking at the most cited papers for that topic (represented by the larger dots). Alternatively, experts in a field are able to find less common papers related to a certain subtopic by looking at smaller dots surrounding papers they are already familiar with. Experts can also find related subtopics to ones they are interested in by looking at nearby clusters of dots. This workflow provides a more intuitive and useful means of discovering new information for all researchers.

Bad Data Visualizations

In everyday media such as magazines and newspapers, we are constantly bombarded with attractive graphics that attempt to catch our attention while displaying useful data and statistics. Unfortunately, even the best looking graphics can have major flaws. Below, we’ll explore some example graphics and try to find where they went wrong.

Bad Visualization Ex1


The intention of the map in the above graphic is to visualize the statistic, “29% Millennials (Born after 1980).” Firstly, it is difficult to understand what this statistic means. It could use some additional context or explanation. Moving to the map of the image, the bottom of the bottom of the map is filled in proportionally to statistic, but there is no reason to use a US map since the location where the map is filled in does not relate to the statistic at all. This map misleads people into thinking the that millennials live in the south, and therefore poorly implements the usage of self-representing images.


Bad Visualization Ex2


At first glance, this chart is confusing because the percentages do not appear to match the relative sizes of the chart segments. For example, when seeing a percentage such as 93%, you would expect its corresponding segment to fill 93% of the chart. This graphic could have been better implemented using a different type of visualization for the statistics. Additionally, the graphic should have provided additional data or further explanation to help the viewer understand the importance of social media marketing, as was expected from the title.