Computerized representation of data and knowledge on high-temperature corrosion and corrosion control

Abstract Databases and knowledge bases are tools for the transfer of information from the source to the user of the information, and they are becoming increasingly computerized because both their source and user environments are being computerized. Thus, in the field of corrosion, the challenge is to represent the results of quantitative measurement, metallurgical evaluation and field experience in computer-logical form. The task is not only to computerize, i.e. to make machine-readable a spectrum of information comprising factual data and knowledge ranging from precise mathematical to heuristic categories, but also to generate data documentation for different types of users and applications, i.e. to permit the prediction, analysis, diagnostics and control of corrosion. Contrary to the situation for systems involving mechanical properties, corrosion and corrosion control cannot be structured in factual databases or databanks alone but require an approach combining databanks, knowledge bases and expert systems. The required scope and logics of corrosion metadata to be documented is reviewed in this paper in relation to the organization of corrosion databanks and knowledge bases. Testing and data reporting standards are rather scarce, but some international activities supporting their development are under way. The importance of an international standard terminology is emphasized as an enabling function for database networks.