A Knowledge-Based System for Earthquake Risk Inventories

Techniques for seismic risk assessment have improved rapidly in the past decade. Much more is known about how different types of structures respond to an earthquake. To make assessments of seismic risk, this earthquake hazard information must be combined with an inventory of buildings by structure type. The cost of obtaining such inventory information presents a key stumbling block to implementing risk assessment techniques for large urban areas. Structure inventories based on field surveys are too expensive and become outdated as soon as they are completed. No single secondary data source contains the detailed structural information required by today's risk assessment techniques. A knowledge-based system is described that makes inferences about structure types based on secondary data. Existing land use, structure age, size, number of stories, and improvement are used to classify buildings by structure type. The knowledge-based system uses a set of rules to make inferences about structure type based on these characteristics and thus creates new data not available from secondary sources.