Describing and communicating uncertainty within the semantic web

摘要: The Semantic Web relies on carefully structured, well defined data to allow machines communicate and understand one another. In many domains (e.g. geospatial) the being described contains some uncertainty, often due bias, observation error or incomplete knowledge. Meaningful processing of this requires these uncertainties be analysed integrated into process chain. Currently, within there is no standard mechanism for interoperable description exchange uncertain information, which renders automated such information implausible, particularly where must considered captured as it propagates through a sequence. particular we adopt Bayesian perspective focus case inputs / outputs are naturally treated random variables. This paper discusses solution problem in form Uncertainty Markup Language (UncertML). UncertML conceptual model, realised an XML schema, that allows uncertainty quantified variety ways: i.e. realisations, statistics probability distributions. INTAMAP (INTeroperability Automated MAPping) project provides use UncertML. This demonstrates how errors can using wrapped Observations & Measurements (O&M) Observation. An interpolation Processing Service (WPS) uses observations influence improve its prediction outcome. output from WPS may also encoded types, e.g. series marginal Gaussian distributions, set statistics, first three moments, realisations Monte Carlo treatment. Quantifying propagating way results consumed by other services. could part risk management chain decision support system, ultimately paves complex chains Web.