{"51196":{"#nid":"51196","#data":{"type":"news","title":"Georgia Tech to Analyze Massive Data Sets Using Visual Analytics","body":[{"value":"\u003Ch3\u003E$3 million award will build a foundation for emerging research field\u003C\/h3\u003E\n\u003Cp\u003E\n\u003C\/p\u003E\n\u003Cp\u003EATLANTA (August 6, 2008)\u2014Enormous amounts of data are being generated\u003Cbr \/\u003E\nin health care, computational biology, homeland security and other\u003Cbr \/\u003E\nareas, but analyzing these massive and unstructured data sets has\u003Cbr \/\u003E\nproven cumbersome and difficult. An emerging research field known as\u003Cbr \/\u003E\ndata and visual analytics is helping sift through such mountains of\u003Cbr \/\u003E\ninformation to find and put together individual pieces of a picture.\u003C\/p\u003E\n\u003Cp\u003EThe Georgia Institute of Technology has received a five-year grant to lead and coordinate a new initiative that will develop foundational research in massive data analysis and visual analytics. A research team headed by Haesun Park, a professor and associate chair in the Computational Science and Engineering Division of the Georgia Tech College of Computing, will investigate ways to improve the visual analytics of massive data sets through machine learning, numerical algorithms and optimization, computational statistics, and information visualization.\u003C\/p\u003E\n\u003Cp\u003E\u201cDeveloping new and improved mathematical and computational methodologies will further enable systems developers, intelligence analysts, biologists and health care workers to implement new methods to \u2018detect the expected and discover the unexpected\u2019 among massive data sets,\u201d Park explained.\u003C\/p\u003E\n\u003Cp\u003EThe $3 million joint National Science Foundation and Department of Homeland Security grant establishes Georgia Tech as the lead academic research institution for all national Foundations of Data and Visual Analytics (FODAVA) research efforts. Seven other FODAVA Partnership Awards will be announced later this year, all working in conjunction with eleven Georgia Tech investigators to advance the field.\u003C\/p\u003E\n\u003Cp\u003EOver the next five years, the Georgia Tech-led research team will work to establish FODAVA as a distinct research field and build a community of top-quality researchers that will collaborate on research workshops and conferences, industry engagement and technology transfer.\u003C\/p\u003E\n\u003Cp\u003E\u201cFODAVA seeks to put an improved science base under one portion of the problem \u2013 how can we transform large, complex data sets into reduced computational models or mathematical formalisms that retain the information content while better supporting the human in extracting critical information from the data,\u201d said Lawrence Rosenblum, program director for graphics and visualization at the National Science Foundation. \u201cScientific advances here are critical to future advances in the science of data and visual analytics that will keep us safe and provide technological and commercial advances that benefit mankind.\u201d\u003C\/p\u003E\n\u003Cp\u003EGeorgia Tech\u2019s expertise in advanced computer-based analysis, probability and statistics, numerical algorithms and optimization, machine learning, and human-computer interaction techniques provides a strong foundation to lead this new initiative.\u003C\/p\u003E\n\u003Cp\u003EPark specializes in using numerical linear algebra and optimization techniques to develop computer-based algorithms that dramatically reduce the dimension and number of data points in massive data sets. Dimension reduction is essential for efficient processing of high-dimension data sets while removing the noise in the data.\u003C\/p\u003E\n\u003Cp\u003EPark is especially interested in developing methods for dimension reduction that exploit prior knowledge in the data sets \u2013 such as clustered structures and non-negativity. This process is important because it leads to more accurate classification and prediction results.\u003C\/p\u003E\n\u003Cp\u003EAlexander Gray, an assistant professor in the Computational Science and Engineering Division of the College of Computing, has experience developing efficient algorithms that allow statistical and machine learning methods to be applied to massive datasets. He employs ideas from computational geometry and computational physics to statistical computations.\u003C\/p\u003E\n\u003Cp\u003E\u201cReducing the computation time for an analysis from hours to seconds makes all the difference, since data analysis is inherently an iterative and interactive process,\u201d explained Gray, also a principal investigator on the project.\u003C\/p\u003E\n\u003Cp\u003ELarge data sets may also include multiple objects of high dimensionality, such as images, that must be analyzed based on a relatively small number of samples. The mathematical analysis of problems like these requires expertise in statistics and probability methods, which Georgia Tech School of Mathematics professor and principal investigator Vladimir Koltchinskii will contribute to the new initiative.\u003C\/p\u003E\n\u003Cp\u003EOnce massive amounts of data are collected and processed, relevant information must be pulled from it and presented using visual and interactive means. John Stasko, a principal investigator on this project and professor in the School of Interactive Computing, conducts research in the field of visual analytics.\u003C\/p\u003E\n\u003Cp\u003EHe heads a team that developed Jigsaw, a visual analytics system that helps analysts better assess, analyze and make sense of large document collections. The system provides multiple coordinated views to show connections between entities extracted from a document collection.\u003C\/p\u003E\n\u003Cp\u003E\u201cJigsaw essentially acts as a visual index of the document collection \u2013 helping analysts identify particular documents to read and examine next,\u201d explained Stasko, whose team won the university division of the 2007 Visual Analytics Science and Technology contest using Jigsaw.\u003C\/p\u003E\n\u003Cp\u003EStasko also serves as Georgia Tech\u2019s director in the Department of Homeland Security-sponsored SouthEast Regional Visualization and Analytics Center (SRVAC), a regional center created in 2006 to perform research in visual analytics. SRVAC is a partnership between the Georgia Tech and the University of North Carolina Charlotte, and is one of five national university centers connected to the National Visualization and Analytics Center located at Pacific Northwest National Laboratory.\u003C\/p\u003E\n\u003Cp\u003EAll of the steps involved in massive data analysis and visual analytics \u2013 data collection, processing, analysis and visualization \u2013 require optimization. Renato Monteiro, a professor in the H. Milton Stewart School of Industrial and Systems Engineering and principal investigator, specializes in this research field.\u003C\/p\u003E\n\u003Cp\u003E\u201cThis new center provides me the opportunity to apply optimization techniques to new and unique problems and applications that I haven\u2019t studied in the past,\u201d said Monteiro.\u003C\/p\u003E\n\u003Cp\u003EFrom law enforcement and intelligence gathering to electronic heath records and computational biology, the accurate and timely analysis of massive amounts of information is critical to deeper understanding and effective decision making.\u003C\/p\u003E\n\u003Cp\u003E\u201cCollaborations across Georgia Tech\u2019s computing, engineering and mathematics disciplines aim to develop better scientific and foundational methods to help practitioners in many different lines of work analyze and interactively explore large data sets more efficiently and effectively,\u201d Park added. \u003C\/p\u003E\n\u003Cp\u003EFor more information, contact:\u003Cbr \/\u003EStefany Wilson\u003Cbr \/\u003EGeorgia Tech College of Computing\u003Cbr \/\u003E404.894.7253\u003Cbr \/\u003E\u003Ca href=\u0022mailto:stefany@cc.gatech.edu\u0022\u003Estefany@cc.gatech.edu\u003C\/a\u003E\u003C\/p\u003E\n\u003Cp\u003ETechnical Contact: Haesun Park (404-385-2170); E-mail: (\u003Ca href=\u0022mailto:hpark@cc.gatech.edu\u0022\u003Ehpark@cc.gatech.edu\u003C\/a\u003E)\u003C\/p\u003E","summary":null,"format":"limited_html"}],"field_subtitle":"","field_summary":[{"value":"\u003Cp\u003EATLANTA (August 6, 2008)\u2014Enormous amounts of data are being generated in health care, computational biology, homeland security and other areas, but analyzing these massive and unstructured data sets has proven cumbersome and difficult. An emerging research field known as data and visual analytics is helping sift through such mountains of information to find and put together individual pieces of a picture. Source: Office of Communications\u003Cbr \/\u003E\u003C\/p\u003E","format":"limited_html"}],"field_summary_sentence":"","uid":"27154","created_gmt":"2010-02-09 21:40:51","changed_gmt":"2016-10-08 03:04:33","author":"Louise Russo","boilerplate_text":"","field_publication":"","field_article_url":"","dateline":{"date":"2008-08-06T00:00:00-04:00","iso_date":"2008-08-06T00:00:00-04:00","tz":"America\/New_York"},"extras":[],"groups":[{"id":"47223","name":"College of Computing"}],"categories":[],"keywords":[{"id":"8432","name":"massive_datasets"}],"core_research_areas":[],"news_room_topics":[],"event_categories":[],"invited_audience":[],"affiliations":[],"classification":[],"areas_of_expertise":[],"news_and_recent_appearances":[],"phone":[],"contact":[],"email":[],"slides":[],"orientation":[],"userdata":""}}}