Presentation Abstracts

Loretta Auvil
D2K and Data Analytics

In order to facilitate our research activities, the Automated Learning Group has developed the D2K application environment for data mining. D2K (Data to Knowledge) is a rapid, flexible data mining and machine learning system that integrates analytical data mining methods for prediction, discovery, and deviation detection, with data and information visualization tools. It offers a visual programming environment that allows users to connect programming modules together to build data mining applications and supplies a core set of modules, application templates, and a standard API for software component development. This talk will highlight data modeling aspects of D2K and touch briefly on recent work on stream mining and text mining. Some examples of D2K's applications include Larry Schook with discovering genetic susceptibility to infectious diseases, Harris Lewin with the development of Evolution Highway for comparative genomics, John Unsworth with text analysis of literary data, Tanya Gallagher with data modeling and text analysis of Social Security Administration data, and the Office of Naval Research with large-scale streaming text analysis.

Noshir Contractor
From Disasters to WoW: Enabling Communities with Cyberinfrastructure

Advances in digital technologies invite consideration of organizing within communities as a process that is accomplished by global, flexible, adaptive, and ad hoc networks that can be created, maintained, dissolved, and reconstituted with remarkable alacrity.

This presentation will outline the major socio-technical challenges for the successful development, deployment, and growth of social and knowledge networking tools as part of the cyberinfrastructure to support communities engaged in a wide range of activities, such as public health, disaster response, environmental engineering, economic resilience, educational diversity, and gaming. In particular it will underscore the importance of a principled methodology to empower the social and knowledge networks of the individuals and organizations that comprise these communities.

Stephen Eubank
Cyberinfrastructure for Epidemiological Simulations

High-fidelity, individual-based simulations of the spread of infectious disease through a population create virtual public health laboratories. Such simulations are based on the interactions among millions to hundreds of millions of individuals. Because individual-based models require both large-scale computing and detailed estimation of interaction patterns, researchers have long had to be content with aggregate or collective models. Although aggrgate models may reproduce the main features of outbreaks, by their very nature they cannot provide answers to important public health questions such as the efficacy of targeted control measures. Recently, however, the information and computing resources necessary to build useful individual-based models have become available. This talk will describe how these cyberinfrastructure requirements arise in epidemiology, how they can be satisfied, and why it's worth the effort.

Download the presentation: .ppt | .pdf

David Freedman
Use of International Travelers as Sentinels for Global Infectious Disease Events

GeoSentinel is a global travel/tropical diseases surveillance network supported by the International Society of Travel Medicine and the U.S. Centers for Disease Control. Tens of millions of people from industrialized nations travel to the developing world each year. Travel/tropical medicine clinics are ideally situated to effectively detect emerging infections of potential global impact at their point of entry into domestic populations and to track ongoing trends in travel-related morbidity. Returning travelers seen at relatively few sentinel sites provide a sample of disease agents in over 230 countries. Real-time data is captured at the clinical point of service, avoiding the delays that are inherent in laboratory-based surveillance.

Download the presentation: .ppt | .pdf

Uriel Kitron
Surveillance of Zoonotic Diseases

All components of a zoonotic disease transmission system must be included in. an effective surveillance program. These include the pathogen, reservoir hosts (domestic and wildlife), arthropod vectors (in case of vector-borne zoonoses), human infection and disease, and environmental and climate/weather factors associated with transmission.

A successful surveillance system will include tools for collection of all these data, their storage, integration, analysis and output. Collaborative arrangements between various institutions and administrative units are often necessary. Examples from research on West Nile virus and Lyme disease in the Midwestern US and Chagas disease in Argentina, where epidemiological, entomological, demographic, and environmental data are considered and geographic information system (GIS) and spatial analysis applied.

Download the presentation: .ppt | .pdf

Martin Kulldorff
Space-Time Scan Statistics for the Early Detection of Disease Outbreaks

The ability to detect disease outbreaks early is important in order to minimize morbidity and mortality through timely implementation of disease prevention and control measures. Many state and local health departments are setting up systematic syndromic surveillance systems with daily analyzes of hospital emergency department visits, ambulance dispatches, primary care physician visits, nurses hotline calls, pharmacy sales and other electronic health data records.

The space-time scan statistic can be used for the early detection of disease outbreaks. Critically, it makes minimal assumptions about the time, geographical location or geographical size of the outbreak. Adjustments can made for natural spatial and temporal variation in the data due to for example geographically uneven population density, geographical differences in data coverage or day-of-week differences in health care utilization. Scan statistics are in general computer intensive methods, and efficient computer algorithms and computing resources are vital for geographically large data sets. The method is illustrated using daily analyses of hospital emergency department visits in New York City and ambulatory care visits in Boston.

Jim Myers
Cyberenvironments for Multi-disciplinary Research

NCSA works with a number of scientific and engineering communities to provide them with an integrated set of tools and services -- cyberenvironments tailored to their specific disciplines -- that allow these communities to realize the full potential of the national cyberinfrastructure in their research, development, and teaching activities. Over the past two years, NCSA has developed the concept of Cyberenvironments to describe a conceptual architecture for these systems that focuses on the goal of increasing end-to-end, systems-level scientific productivity and capacity. Cyberenvironments combine interaction with local and shared instruments and sensor networks, data stores, computational capabilities, and analysis and visualization services within a secure framework for the management of complex projects, automation of processes, and collaboration with colleagues both near and far. This talk will provide further motivation and definition of Cyberenvironments and highlight capabilities in current projects that demonstrate their potential for supporting Infectious Disease Surveillance, Modeling, and Response efforts.

Gilbert L. Rochon
Satellite Remote Sensing for Infectious Disease Mapping and Vector Habitat Monitoring

The current state of the science and anticipated satellite launches are reviewed with respect to deployment of archival and real-time remotely sensed earth observing data, in support of infectious disease mapping and the monitoring of habitat for disease vectors, intermediate hosts and reservoirs for an array of diseases endemic in specific regions within Africa, Asia and Latin America. More broadly, information technology, ranging from high performance computing to appropriate intermediate information technology's role in analyzing and distributing spatial, temporal and spectral data is discussed. Given the synergistic relationship between malnutrition and infection, identified by Scrimshaw, Gordon & Taylor, the role of remote sensing in supporting famine early warning is also addressed. Moreover, since poverty has been etiologically associated with increased vulnerability, the resources available for poverty mapping are explored. Finally, the national and international facilities for real-time remote sensing research support are described, including groundstations, existing network infrastructure, software options and data accessibility.

Download the presentation: .ppt | .pdf