Creating A Knowledge Base for Supporting Oil Spills Surveillance/Monitoring

摘要

This work deals with the creation of an ontological knowledge base (KB) for supporting oil spills surveillance/monitoring. The methodological framework designed/developed, under the form of an algorithmic procedure, includes 24 activity stages and 7 decision nodes. The respective flow chart can be divided into three conceptually discrete parts: the design of basic ontology and the relevant knowledge exploitation methods, the proof/improvement of functionality by Monte Carlo simulation and laboratory experimentation, and the development/enrichment of the KB by implementation under real incident conditions. An implementation is presented on constructing an oil spill weathering ontology in two steps: in the first, the taxonomic/partonomic (or mereological) function is introduced through the logical operators is-a and part-of to follow the fate of polycyclic aromatic hydrocarbons (PAHs) which constitute one of the most harmful hydrocarbon families in oil spill; in the second, a sample of oil weathering ontology, including two phenomenological levels, a surface and a deeper one. Numerical results, in fuzzy version to count for uncertainty, are also presented, concerning the rate constants of the oil spill constituents in both cases, with and without dispersant addition; its is worthwhile noting that the results referring to oil spill without dispersant exhibit higher fuzziness.