Abstract : RAND has been investigating the use of semantic data models in object-oriented and knowledge-based simulation under these sponsorships and within RAND's Information Processing Systems Program as part of the Intelligent Databases Project. The research described in this report is one aspect of intelligent databases addressing the derivation of simulation-specific databases from external databases. This report serves two purposes. First, it discusses the data flow gap between (1) external databases acquired from public agencies and (2) datasets derived from these databases to be used as input to specific simulation models. This gap refers to the lack of interoperability between external databases and simulation databases and to the transformations that are necessary to derive compatible datasets. The second goal of this report is to present the computer software system developed to bridge this data flow gap. The report should be of interest to database administrators, simulation developers, and researchers in database management and information processing technology.