Distributed Problem Solving in Seismic Event Analysis: Final Report

摘要

The accuracy of analyses for seismic events visible to more than one monitoring station can be improved by combining data from several stations and ''triangulating'' analyses. This sort of analysis is a natural candidate for distributed problem-solving approaches. Data from different stations must be related in both time and space, raising problems of reasoning from multiple perspectives toward a common interpretation. Differences in local geography and monitoring conditions will raise problems of reasoning using different knowledge and local constraints. Just which problem aspects should be the provinces of which local analyzers (the issues of problem decomposition and task allocation) are uncertain. We proposed to investigate how emerging research on to these questions may be applied to seismic analysis in particular. (ERA citation 12:029473)