Bird Tracking and Language Acquisition Analysis – David Tcheng, Research Scientist for I3, working with David Enstrom, Ornithologist at the State Natural History Survey, used a ground-based microphone array to track bird migration. When we visualized the vocalization locations over time, we could clearly see the thrush migration path and measure species abundance. Working with Dustin Reichard, at Indiana University, we are now testing the hypothesis that song elements are inherited, father to son, in male junco birds.
Automated Pollen Analysis - Tcheng, working with Surangi Punyasena, faculty in Plant Biology, fully scanned a set of 76 microscope slides containing pollen, forming 40TB of image data and then developed a virtual microscope environment for allowing experts to generate ground truth for machine learning. Experts simply circle all the pollen they see on randomly selected images from randomly selected slides. Using 3K of these expert tagged images as training examples, we generated a random forest of decision trees to create a probability field visualization which identifies most pollen grains in a complex environment.
Stakeholder alignment - I3's Mike Haberman has been working with Joel Cutcher-Gershenfeld, Dean of the School of Labor and Employment Relations, and others on a stakeholder alignment project. Every day we rely upon complex socio-technical systems, such as the air, water and ground transportation systems, the power grid, and health care delivery systems, that require cooperation among several different organizations and agencies. Stakeholder misalignment would cause these systems to fall short of their potential and may result in catastrophic failures such as the response to Hurricane Katrina. We are developing tools and methods to advance lateral alignment of stakeholders by enabling transparent data collection for identifying and visualizing behavioral, structural and cultural alignment or misalignment.
SEASR – The Software Environment for the Advancement of Scholarly Research (SEASR) Services Project, led by I3's Loretta Auvil and Boris Capitanu, provides the data synthesis, analytics tools, and supporting infrastructure required to analyze and study digital collections with the use of Meandre. This project has four use cases. Our collaboration with Matthew Jockers from Stanford University led us to integrate into Meandre the Mallet tool for topic modeling to find patterns in a large text collection by identifying words that frequently occur together. Our collaboration with Ted Underwood from UIUC led to spellchecking, OCR error correction and correlation capabilities that were applied to the Google Ngrams dataset with the goal of finding relationships between pairs of ngrams. Our collaboration with Tanya Clement from the University of Texas in Austinled us to integrate the OpenMary tool, which identifies aural features in text and apply further analysis and visualization to help readers examine aural and prosodic patterns in text. Our collaboration with Dan Cohen from George Mason University focused on concept mapping of emotions (Love, Joy, Surprise, Anger, Sadness, and Fear) to capture the sentiment of the 911 dataset.
GroupScope: Instrumenting Research on Interaction Networks in Complex Social Contexts – By applying advanced computing applications to capture, manage, annotate and analyze these massive observational sets of video, audio, and other data, the GroupScope project, led by I3's I-CHASS director M. Scott Poole, will enable breakthrough research into social interaction in large, dynamic groups to be conducted much more quickly and with much higher reliability than was previously possible. It will do this by automating as many functions as possible to the highest degree possible, including managing huge volumes of video, audio, and sensor data, transcription, parsing audio for critical discourse events, annotation and indexing of video streams, and coding interaction.