Geography students gain international experience in their BDSS IGERT traineeship

In 2012, Penn State was awarded an Integrative Graduate Education and Research Traineeship (IGERT) grant from the National Science Foundation, which brought together a diverse interdisciplinary team of researchers to create a new training program in the emerging field of social data analytics.

Massive new sources of big data—from human interactions that are increasingly recorded via web, mobile device, and distributed sensors—provide great challenges and opportunities for extracting new scientific, economic, and social value. Big Data Social Science , the collaborative effort of more than 80 faculty from across Penn State, uniquely prepares students to meet those challenges through a new curriculum, training in advanced technologies of data science and analytics, and a series of research rotations.

Geography graduate students Jennifer Smith Mason and Joshua Stevens, two of the seven trainees accepted in the BDSS program’s first year, traveled during summer 2013 to complete research rotations in London and Zurich, respectively.

Mason is a second year Ph.D. candidate with a specialization in GIScience. Her research addresses the complexity of geospatial data uncertainty from socially relevant datasets. “I am studying uncertainty with regard to the data and user, specifically focusing on how to better visualize uncertainty in order to assist users through reasoning and decision-making processes. For example, in a project last spring semester, I worked with another graduate student, Yan Huang from the Department of Media Studies in the College of Communications, comparing different ways to visualize hurricane uncertainty, both its location and magnitude, to assess how it impacts people and their subsequent decisions.”

Mason spent 10 weeks at the giCentre at City University London, where she worked with Jason Dykes and Jo Woods.  The giCentre is engaged in high quality research and education into the role and design of graphical techniques for exploring and analyzing data and disseminating information. Jennifer explains, “Jason Dykes wanted to make sure that I have other tangible training that will be useful and I can take back to Penn State, so I am learning the processing programming language to do all of my visualizations.”

With this new training, Mason worked on a research project that extends her hurricane visualization from the spring, to develop novel and intuitive visualizations to better support hurricane uncertainty. Mason hopes to publish the results of her project. “We will run the evaluation through Amazon Mechanical Turk and hope to, at minimum, develop a poster or presentation for submission at an upcoming conference and at best, end with a project that can be published in a relevant journal. The evaluation will look at the cognitive factors of the visualization to determine how it affects users.”  

Joshua Stevens is a Ph.D. student in geography with an emphasis on geovisualization and visual analytics. Of his research, he says, “I'm primarily interested in the way humans view, interpret, and interact with mapped information—especially within the domains of scientific geovisualization and big data. As a result, my research focuses on bridging the gap between computational systems that are necessarily becoming more sophisticated, and human interfaces that must maintain simplicity and usability. The goal of geovisualization has always been to leverage the processing power of computers with the cognitive abilities of humans. I believe making interactivity salient and intuitive strengthens the human-computer connection and gets us closer to that goal.”

In May, Stevens traveled to Switzerland to work within the Geographic Information Visualization and Analysis (GIVA) unit at the University of Zurich. There he developed an application called GMotions Explorer. Stevens explains, “This tool combines an enormous amount of eye tracking data, video recordings, GPS tracks, and skin conductivity measurements to help researchers understand how users view, reason about, and emotionally respond to visual stimuli. The combination of multimedia, geography, and statistical techniques used by GMotions Explorer requires rethinking traditional interactive methods. By addressing the computational and perceptual challenges of visualizing so much data from so many disparate sources, we hope to uncover new techniques that not only make the system easier to use, but also tell us more about why the techniques work and the cognitive processes that are involved.”

Jenny Mason's NACIS presentation:

Original article: