Data Science and Information Fusion
As the world continues to become more networked, availability and access to a wide variety of data can often overwhelm people and obscure the information they are seeking. CUBRC’s scientists and software developers sit at the forefront of data science and multisource information fusion research, development, testing, and evaluation in an ever expanding data environment. Our work encompasses a class of technologies that turn massive amounts of data into actionable information for decision makers in government and commercial applications. CUBRC’s data extraction, alignment, analysis and systems optimization technologies provide powerful tools for use by government, business and industry to effectively utilize the data they have available to them; discover unseen relationships; make more informed decisions; and better understand their customers.
CUBRC's Data Science and Information Fusion Group consists of skilled researchers, computer scientists, engineers, and subject matter experts who specialize in the development of software and hardware systems supporting:
CUBRC/UB Center For Multi-Source Information Fusion
The CUBRC/University at Buffalo Center for Multisource Information Fusion (CMIF) was jointly established in 1997 as the first Information Fusion center of its kind. CMIF focuses on performing research, and the development of data alignment and analysis algorithms that efficiently and effectively help people answer questions and make decisions in “big data” environments. Paired with CUBRC’s software engineering and development capabilities, CMIF delivers fully functional and supported software systems in both Service Oriented Architecture (SOA) and “cloud” based information data storage and analysis environments. CMIF’s educational mission supports the creation of the next generation of data scientists through formal education programs, internships and collaboration opportunities across a wide variety of application domains.
Work in the center is performed from two separate CMIF facilities. The first is located at the University at Buffalo, where basic research is conducted in an unclassified, academic environment. A second, secure facility performs applied through advanced research and systems software development supporting long-term technology transition into productized or fielded systems.
Common Core Ontologies for Data Integration
The Data Science and Information Fusion Group’s work in ontologies started in 2008. Since 2010, our participation in IARPA’s Knowledge, Discovery and Dissemination program focused our work on the development of the Common Core Ontologies (CCO). The CCO provide a base vocabulary that is to be extended to a fuller suite of ontologies that serve as the unified semantics for the content of all data sources within an enterprise. To accomplish this extension, the CCO is accompanied by a rule-guided method. The combination of the CCO and this methodology allows different groups within an enterprise to develop ontologies and still arrive at semantically interoperable models.
Rapid Ontology Development
The ten modular but interrelated ontologies that compose the CCO are not a monolithic, global structure that attempts to describe every element of every specialized domain. Rather, the ontologies provide a starting set of general, commonly used terms. The entities of interest to data communities within the enterprise not covered by these common terms are those that need to be integrated by developing extension ontologies...
Transportation and Healthcare Applications
Atlas & Database of Air Medical Services (ADAMS)
In conjunction with the Association of Air Medical Services (AAMS) and the air medical industry, CUBRC has established a national database containing air medical rotor and fixed wing services, including main and satellite base locations, communication centers and receiving hospitals. This data resource, called the Atlas & Database of Air Medical Services (ADAMS), is implemented in a Geographic Information System (GIS) on the web. The need for such a data resource is driven in part, by newly-emerging, Automatic Crash Notification (ACN) technologies which are changing both the content and manner in which emergency messages from car crashes are received and routed.
CUBRC specializes in providing advanced tools to help people that are overwhelmed by data to discover relationships and information from which they can rapidly make critical decisions. CUBRC’s healthcare informatics team has collaborated with leading physicians, providers, insurers, and medical device manufacturers to develop innovative predictive analytic methods designed to improve patient outcomes while managing cost. We specialize in combining clinical, demographic, and environmental information into a single picture to help our customers build on their strengths, respond to patient needs, predict intervention efficacy, and improve outcomes.
Transportation Research Informatics Platform (TRIP)
The Transportation Research Informatics Platform (TRIP) is an informatics based system designed to handle massive amounts of transportation data, provide researchers an efficient way to interact with this data, and allow for the straightforward use of analytical tools to assess the data. TRIP is an ongoing project funded under the Federal Highway Administration's Exploratory Advanced Research Program.
Transportation researchers and practitioners have access to unprecedented amounts of data but lack the tools to easily store, manipulate, and analyze this data. TRIP provides tools for researchers enabling them to conduct 'Big Data' analytics in an efficient way...