Spring 2016 Workshop Recap: Decision Design & Visual Analytics for Sustainability Applications

The Decision Design and Visual Analytics for Sustainability Applications workshop, run by the Institute for Humanities Research Nexus Lab this past spring, is an excellent example of the kind of transdisciplinary and engaged humanities scholarship for the 21st century that both ASU and the IHR generate and support. Sustainability, or “ecological integrity, human well-being, and social justice for present and future generations,” is a complex goal, one that needs a host of different perspectives as no one field of knowledge can achieve all of those objectives. The needed range of knowledge was on display in the attendees of the workshop: undergraduate students, graduate students, faculty, staff, and community members, in fields ranging from linguistics to industrial engineering, art to ecology. Personally, as a literature scholar focused on the environment, I attended for a variety of reasons: to better understand different ways of looking at environmental data, to begin to work in the digital humanities, and to have interdisciplinary conversations about sustainability decision making. While the material was at times both mentally and technically challenging the workshop space helped us all gain new insights unto our own fields and others.

The workshop was led by Michael Simeone, the director of the Nexus Lab, and was structured around three different modes of analyzing and understanding data: network analysis, geographical information systems, and text analysis.

For the first third of the semester we learned about network theory and modeling. Network theory, at its base, very simple – the researcher identifies the actors in the network and their relationships and then organizes them visually (either by hand or with a computer). For an example, here is a network analysis of the popular Game of Thrones series published in Quartz. The researchers chose the characters in the series as the actors (referred to as nodes in network theory), and then decided to link them together depending on whether or not they interacted (edges in network theory). The computer then counted up and organized the relationships into the network. Ultimately, the network of the particular book confirms that three of the point-of-view characters (Daenerys, Tyrion, and Jon) are centrally important to the social networks of the book. But the data also shows that other characters also hold central importance, such as Robb, who is not a point of view character, and Sansa, who is a point of view character but has no political power, or Robert, who is dead. The network suggests that characters who might not seem influential in the series still move the plot along or alter the plot through their centrality in the social network.

In the workshop we worked with the programs Gephi and ORA to make our on networks. For example, my group made a network model for the Orbit public transportation system in Tempe to map the relationships between the riders, drivers, and Valley Metro who runs the system. I would also note that I am providing a simplified version of network analysis: there is a great deal more to the process than drawing lines between bubbles. I do not claim to be an expert on network analysis now, but thanks to the workshop I better grasp what it is and could begin to investigate the field more with the base of knowledge that I learned from the workshop.

The middle third of the semester was devoted to geographical information systems and mapping. These days were led by Josh MacFayden of the School of Sustainability. Geographical Information Systems, or GIS, is more complicated than being merely a map, as the map itself is able to contain multiple layers of information. For example, Josh showed us a research project where he used mapping tools, in combination with other digital drawing tools, to estimate the amounts of firewood that were being taken from Canadian forests in the nineteenth century. We worked with Google Earth and QGIS to design and manipulate maps with multiple different layers of data. Using GIS to look at data suggested, to me, new ways of understanding data, especially the way that human civilizations and the environment change over time.

The last third of the semester we worked with text analysis tools. These tools read and sort text, usually large bodies of text that would otherwise not be easily readable by a person or persons. Programs like Voyant Tools or Google Ngram Viewer, among many others, allow us to see macro scale pictures of texts and also know where to look at a micro level for further insight. For example, take the 1818 and 1831 editions of Frankenstein. In the 1818 edition the most frequent words (once a series of stop-words like The, a, an etc. are removed) are father, man, time, and life, in that order. In the 1831 edition though the most frequent words in order are man, life, father, and time. This suggests that the two editions might approach the question of fatherhood differently. It is important to note though, as we discussed in the workshop, that this evidence is not enough for a conclusion. Rather we have a lead now that we did not have previously which we could track down at a micro-level, i.e., reading the book with an eye towards fatherhood to see if the texts bear that interpretation out.

At the end of the semester we used text analysis tools in conjunction with network analysis to look through the emails of the Enron corporation right before and after their collapse. One of the tools that I used, in this group project, was Mallet. I sent the emails through the program which sorted them into groups based on shared, commonly occurring words. I noticed an interesting grouping of ‘non-business’ words and examined the top emails in the grouping. There I found quite a few personal emails: people talking about weekend plans, betting on football games, and ultimately trying to setup a get-together to commiserate when the company went under. The emails helped show that while there were major problems at the top, there were also a majority of people just trying to do a good job.

Ultimately, the workshop was not only educational but also fun. While I am by no means an expert on these various tools now, I have the foundations that I can build upon in my future scholarship. The workshop, for me, provided an aperture into the digital humanities. I feel fortunate that the Nexus Lab and Michael hosted this workshop and would encourage members of the ASU community, in addition to the larger Phoenix community, to take advantage of the lab's programming in the future.

Department of English Faculty Associate and recent PhD graduate, Kent Linthicum, is a scholar of American and British literature from 1783-1912, science, and the environment. He recently defended his dissertation titled “Scientific and Cultural Interpretations of Volcanoes, 1766-1901,” which used literary and scientific texts about volcanoes to examine both the popular and intellectual understanding of these geological phenomena.