Computer scientist, Lawrence Berkeley Laboratory, University of California, Berkeley, CA
Pages: 16 Published: Jan 1989
Material properties data present numerous challenges that current data base management technology does not address. Major problems include data complexity, naming conventions, measurement units, summary levels of abstraction, and data sparseness. Networks and gateways to multiple material information systems compound these challenges and add others.
Experience from projects such as the Materials Information for Science and Technology (MIST) system prototype suggests several general approaches that address these problems, including modular system architecture; a three-tiered approach to naming materials, properties, and variables; an active data thesaurus; material and variable class hierarchies; structured representation of tables and graphs; and existence tables to indicate data availability by the cross product of materials, properties, and independent variables for any given data base. It also suggests that we need to go beyond the simple relational model to solve our problems and implement solutions.
information systems, information retrieval, thesauri, data description, tables (data), units of measurement, distributed data bases, electronic publishing, dictionaries, directories, handbooks, nomenclature
Paper ID: STP18622S