Welcome
Areas of Research
Office of the Vice President
Office of Intellectual Property & Commercialization
Office of Sponsored Projects
Compliance & Review Boards
Research Foundation
Facilities
Forms & Downloads
Funding Opportunities
Policies & Procedures
Proposal Preparation
Resource Links
Routing Proposals
Training & Workshops
Travel FAQs
News Articles
Spotlight Articles
Breeze Newsletter
Events Calendar
Browse Projects
Browse People

Information Fusion, Indexing and Search Algorithms for Active Cyber Modeling and Intelligence

Description :  The rapid advancement in data collection instrumentation and layered network paradigms in cyber space calls for information fusion architectures closely coupled with efficient and accurate data indexing and searching techniques. This necessitates the need for the design and development of algorithms to integrate data from multiple data sources and manage heterogeneous data in a uniform, searchable, and interoperable schema. To achieve this overarching aim, we aim to create concepts for data fusion and information retrieval using algorithms to initialize, populate, integrate, index, mine and eventually search interoperable data space comprising heterogeneous and distributed data sources in cyber modeling schemas. Spatial databases represent, store and manipulate spatial data that represent objects. These databases allow features that represent geometry of objects such as point, lines, areas, surfaces and hyper-volumes in multidimensional space. These databases suffer from the "Curse of Dimensionality", affecting a natural performance degradation of similarity queries as the dimensionality of these datasets increase, thereby limiting the discovered cyber intelligence. One way to overcome the curse of dimensionality in these databases is to develop both efficient and accurate spatio-temporal data structures for indexing. These indexing data structures will enable similarity queries and searches with higher degrees of precision. The project aims to develop algorithms to 1) extract features from selected cyber sources, 2) integrated features from heterogeneous data spaces into a unified information space, 3) analyze the fused information space for discriminatory attributes, and 4) design a spatio-temporal data structure to answer similarity queries from these databases with very low rates of false dismissals and acceptable ranges of false alarms, while allowing multi-fold gains in time performance.
Principal Investigator:  Dua, Sumeet  --  Computer Science
Collaborators:  
Funding Agencies:  DoD/Clarkson
Amount Awarded:  128,584

Start Period:  00/00/0000 End Period:  00/00/0000
Related People
No Affiliated People

October 21st, 2017

Browse Projects

Browse People

Recent Articles

2017 Annual Student Research Symposium...

More News...

Page Tools

Print Page      Email Page

Louisiana Tech University, A Member of the University of Louisiana System
2017 Louisiana Tech University P.O. Box 3178 Ruston, LA 71272