CI Year in Review 2015


Climate change and agriculture. The environment, infrastructure, and activity of cities. The damage of electric shock. Police relations with the public. The balance of innovation and conservatism across the history of science. These subjects would seem to have little overlap, but in 2015, they were all subjects of research at the Computation Institute, with researchers finding new and brilliant ways of using data and computation to study a broad range of important topics. Below, you'll find our annual summary of the year's activity at the CI, and we look forward to bringing you more exciting news from the cutting edge of computation in 2016.

As artist-in-residence at Microsoft Research, CI fellow Jason Salavon had a rare opportunity to further experiment with the intersection of art and technology, including sentiment analysis of business meetings and the identification of religious figures in photos of nonhuman objects.
This month’s edition of Inside the Discovery Cloud brought together CI Senior Fellows Juan de Pablo and Michael Wilde to discuss the young, promising field of molecular engineering, and how the modeling and simulation at its heart can be accomplished using parallel computing.
A series of videos recapped the 2014 Data Science for Social Good summer fellowship, where teams worked with organizations such as the World Bank Group, the Office of the President of Mexico, and the Chicago Department of Public Health on data-driven solutions with real world impact.
The inaugural Discovery Engines workshop focused on pandas, a data analysis module for the Python programming language, as taught by Knowledge Lab researcher (and former DSSG fellow) Misha Teplitskiy.
The same computational tools used to make movie magic are put to use by CI senior fellow Gary An in the fight against sepsis and inflammation, as this article from Student Science described.
A new API expanded the utility of metagenomics platform MG-RAST, used by biologists around the world to study the genetics of the microbial world.
A healthy investment from the University of Chicago Innovation Fund gave CI Senior Fellow Michael Wilde’s new business a welcome boost in its mission to bring parallel computing to industrial and engineering businesses.
Many cities, including Chicago, are working to become more energy efficient. But in order to set goals and measure improvements, a city needs a benchmark. Researchers from the Urban Center for Computation and Data helped the City of Chicago analyze statistics from the early years of its building efficiency program, assessing the energy demands of skyscrapers.
Electric shocks are a unique form of injury, creating damage through exposure to both powerful electrical fields and high temperature. To study these effects upon the human body and find better treatments, University of Chicago Medicine surgeon Raphael Lee worked with CI staff to build a new model of electrocution on the Beagle biomedical supercomputer.
The Array of Things urban sensing project headlined February’s Inside the Discovery Cloud, with complementary talks from UrbanCCD director Charlie Catlett and School of the Art Institute of Chicago designer Douglas Pancoast, while the Discovery Engines workshop focused on climate science you can do at home with RDCEP’s Michael Glotter and David Kelly. [video]
A visualization from Knowledge Lab researcher Cody Braun illustrated the differences between the “bad internet” of Phishing and scam sites and the “safe internet” where we’d all prefer to stay.
CI Director Ian Foster added a new award to his shelf: the IEEE Award for Excellence in Scalable Computing. The honor saluted his critical role in the creation of grid computing for physicists and earth scientists up through recent achievements in cloud computing for science with Globus.

To create a “new vocabulary” for the visual communication of complex data, CI faculty and fellow Gordon Kindlmann introduced Algebraic Visualization, which creates mathematically-driven rules for the presentation of graphs, maps, and other visualizations.
Network analysis methods have made billions of dollars for companies such as Google and Amazon, and Bill Shi from Knowledge Lab provided a useful introduction for March’s Discovery Engines. Meanwhile, Inside the Discovery Cloud looked at computational biology, highlighting the study and simulation of microbial communities with talks from CI fellow Christopher Henry and Argonne’s Jack Gilbert.
As part of the University of Chicago’s commitment to global outreach, the CI’s Robert Gardner, Paul Dave and Dinanath Sulakhe traveled to India for a workshop on Accelerating Collaboration and Science Through Connective Computation.
The spring finalists for the UChicago Innovation Fund had a familiar flavor, as 4 of the 6 nominees included CI researchers. From this crop, only the Array of Things project would reach the funding stage, receiving $150,000 (and an additional matching sum from Argonne) to develop the first nodes for the urban sensing project.
Globusworld’s 2015 edition once again brought together an international crowd of researchers and computational experts to discuss exciting new solutions for research data management. In this year’s keynote, CI Director Ian Foster announced expanded Globus publication services, making research data easier to publish, share, and discover.
Globus also made headlines with the announcement of Wrangler at the Texas Advanced Computing Center, a data-intensive resource that will use Globus services to rapidly transport research data in and out of the system.
Big computation and “big data” was the focus of this month’s Inside the Discovery Cloud, with a tag-team talk from CI senior fellows Andrew Chien and Pete Beckman about how the future of computer science will disprove much of the accepted wisdom of the field’s past.
Writers in science and other fields are often told to keep articles short and concise, densely packed with information. But a study from the lab of CI faculty and fellow Stefano Allesina found that longer and more flowery abstracts are more likely to be cited by other scientists, highlighting flaws in modern scientific publishing and search engines.
As health care begins to see the potential of data analytics and computation to revolutionize the field, administrators and entrepreneurs alike are looking for early signs of success to build upon. In a panel at Chicago’s MATTER work space in the Merchandise Mart, CI Director Ian Foster, Data Science for Social Good Director Rayid Ghani, and former CI Deputy Director Jonathan Silverstein participated in a panel on advances in detecting lead poisoning, predicting disease, and getting the most from cancer data sources.
Another promising confluence of data and medicine is the use of text mining to extract new knowledge from the millions of biomedical journal articles already published. The final Inside the Discovery Cloud of 2014-15 focused on these methods, with talks from Knowledge Lab director James Evans and Ishanu Chattopadhyay of the Conte Center for Computational Neuropsychiatric Genomics.
The occupation of “data scientist” is so new that few people, even within the field, know how to properly define the job or the best path to becoming one. 2014 DSSG fellow Carl Shan and collaborators set out to dispel some of this fog, interviewing data scientists from several companies and schools to produce a pay-what-you-like ebook, The Data Science Handbook.
The 2015 edition of the Data Science for Social Good summer fellowship kicked off in Chicago’s River North neighborhood, bringing 42 students in statistics, computer science, and social sciences to work in teams with nonprofits and government organizations. Projects this year included efforts to predict adverse police incidents, improve graduation rates and college persistence, and detect urban blight in Cincinnati.

The entire world of science would seem to be immeasurably vast, impossible for man or computer to fully comprehend and categorize. But an analysis from the Knowledge Lab’s Bill Shi, Jacob Foster, and James Evans found that a “hypergraph” network of scientific findings connected by diseases, chemicals, methods, and authors was surprisingly compact -- and perhaps capable of steering future scientists to the most useful experiments.
One less-discussed consequence of climate change is the forecasted effect on the world’s food supply, as rising temperatures and changing precipitation patterns affect farmers and their crops. In a book published by the United Nations Food and Agriculture Organization, CI and RDCEP researcher Joshua Elliott wrote two chapters on how computer modeling can help governments prepare for these potentially catastrophic changes.
For a class on how technology has shaped law and culture, taking it online in MOOC form only makes good sense. This month, CI Senior Fellow Randy Picker debuted his self-paced Coursera course, “Internet Giants: The Law and Economics of Media Platforms,” which has so far attracted thousands of students.
MIT Technology Review examined work from Knowledge Lab’s Eamon Duede, Misha Teplitskiy, and Grace Lu on how the collaborative online encyclopedia Wikipedia cites scientific journals, detecting a bias towards open access journals over their firewalled peers.
Distinguished Fellow is one of the highest honorary titles that an Argonne scientist can receive, and this year all three new fellows shared a common link in the CI. Charles “Chick” Macal, Branko Ruscic, and Barry Smith, experts in agent-based modeling, computational chemistry, and mathematics, spoke about the influence of the CI on their groundbreaking work.
Chameleon, a new cloud computing testbed led by CI senior fellow Kate Keahey, officially entered its production phase, allowing researchers around the country to test out new cloud architectures and applications for tomorrow’s computational research projects and industries.
A new report warning about the dangers of food insecurity related to climate change -- including research from RDCEP’s Joshua Elliott on more frequent “once in a century” droughts -- was released, bringing coverage from The Guardian, Science, and the BBC.
The 2015 edition of Data Science for Social Good wrapped up with their annual Data Fest, where each of the 12 teams presented brief talks on their summer projects. You can watch the highlights of the event, or full video of each of the twelve talks.
A $3.1 million National Science Foundation grant gave Array of Things the funding boost it needed to roll out 500 sensors in Chicago over the next three years. The project will collect information on the city’s environment, infrastructure, and activity, releasing all data publicly to work towards a cleaner and more livable city.
The diseases that most hurt the global population are not always those that attract the most research dollars. To help identify and fill these gaps, a team including CI Senior Fellows Andrey Rzhetsky and James Evans created a new tool called the Research Opportunity Index (ROI), published in Nature Biotechnology.
There’s cloud computing, and then there’s *cloud computing*. Read about how the CI’s Virendra Ghate and Chris Schwartz rode research flights from California to Hawaii, collecting data that will improve computer modeling of cloud patterns in the Pacific Ocean.
A DSSG project on predicting police officers at risk of adverse interactions with the public was acknowledged by the White House as part of their Police Data Initiative. The team collaborated with the Charlotte-Mecklenburg Police Department to build a new early intervention system that identifies officers in need of more training and suggests operational changes to avoid conflicts.
The American power grid has remained largely unchanged for decades, despite new energy sources and “smart grid” technologies coming to fruition. In a talk at the CI, Argonne’s Jianhui Wang talked about how computer simulations can help the energy industry prepare for the “smart grid” and its reinvention of how electricity is distributed and consumed.
Globus opened up its new data publication tools to its user base, providing a new way for researchers to share large, complete datasets that underlie scientific findings. The service will also allow for new ways of discovering data, facilitating collaboration and accelerating discoveries.

Ever wanted to visit the Large Hadron Collider at CERN, but couldn’t afford the trip to Switzerland? CI researcher Ilija Vukotic created a virtual reality tour of the LHC’s ATLAS detector, available for use with the Oculus Rift head-mounted device.
Another Rzhetsky-Evans collaboration examined millions of journal articles to model the historic battle between safe and risky research, finding that science has grown more conservative over time and doesn’t follow the most efficient strategies for making discoveries.
Perhaps CI fellows Shashi Aithal and Stefan Wild knew something when they named their computational model for automobile engine testing ACCOLADES. At this year’s Supercomputing conference, the project took home the HPC Innovation Excellence Award.

After a successful first phase combining climate and crop models and comparing their results against historical data, the Global Gridded Crop Model Intercomparison (GGCMI) Project, co-led by RDCEP’s Joshua Elliott, entered its second phase, further developing tools to help policymakers prepare for climate change’s effect on agriculture. 

Written By: