URREF Reliability Versus Credibility in Information Fusion.

摘要

For many operational information fusion systems, both reliability and credibility are evaluation criteria for collected information. The Uncertainty Representation and Reasoning Evaluation Framework (URREF) is a comprehensive ontology that represents measures of uncertainty. URREF supports standards such as the NATO Standardization Agreement (STANAG) 2511, which incorporates categories of reliability and credibility. Reliability has traditionally been assessed for physical machines to support failure analysis. Source reliability of a human can also be assessed. Credibility is associated with a machine process or human assessment of collected evidence for information content. Other related constructs for URREF are data relevance and completeness. In this paper, we seek to develop a mathematical relation of weight of evidence using credibility and reliability as criteria for characterizing uncertainty in information fusion systems.