Developing SIEV-nets from non-Bayesian Resources

Abstract

George Mason University has developed a scalable, information-based, real-time methodology for heterogeneous sensor management in support of maritime domain awareness. We propose to continue the development of an enabling technology for this research by investigating the possibility of developing a methodology for converting existing non-Bayesian connectivity data into a Bayes net formulation which is the underlying format of the SIEV-net. The technical approach is to locate an open-source example of non-Bayesian, directed connectivity data among social entities which is already formatted and used by a commercial application. The data will be evaluated to determine if there are parameters associated with these graph edges which indicate the strength of the causal relationship. The theoretical approach will be to develop mathematics for converting these causal linkage strengths to conditional probabilities while maintaining the causal relationships as are required for Bayes nets. The anticipated outcome of this fundamental research is a better understanding of the process by which subject matter expert data can be converted to a structural, causal model with the associated quantitative probabilities derived from connectivity strength parameters. The anticipated outcome is a methodology by which existing social network data can be reformulated into a Bayes net for use in the information based sensor management (IBSM) paradigm and the valuing of collection alternatives. The principle investigator for this proposed research is Dr. Kenneth J. Hintz, Ph.D. who will perform the research at George Mason University, Fairfax, VA. The anticipated public benefit of this research is to enable effective resource allocation and collection planning for situation information gathering in the event of an unplanned disaster. This abstract is publicly releasable.

Document Details

Document Type
DoD Grant Award
Publication Date
May 17, 2018
Source ID
N002441810004

Entities

People

  • Kenneth Hintz

Organizations

  • George Mason University
  • United States Navy

Tags

Fields of Study

  • Computer science

Readers

  • Distributed Systems and Data Platform Development
  • Economics
  • Regression Analysis.

Technology Areas

  • AI & ML
  • AI & ML - Bayesian Inference