Expert Systems

Computers are indispensible tools for processing and retrieval of the increasing amounts of chemical data. But, especially in spectroscopy, where the knowledge and experience of experts play a major role in the handling of spectroscopic information. Conventional computers will never be able to recognize, associate or interpret information in such an effective way as a human brain. However, they can assist the spectroscopist by reducing and organizing the huge amounts of information that are available. This is what expert systems are designed to do. Before we start describing the application of expert systems in vibrational spectroscopy some terms should be defined. Expert systems are computer programs that aid an expert in making decisions about a certain problem. An expert system typically operates with rules that are evaluated to predict a result for a certain input. For the generation of rules a prior knowledge about the correlation between query and output data is necessary. Such a correlation can be evaluated by artificial intelligence (AI) methods in an inductive learning process by analyzing experimental data. Knowledge-based systems can generally be defined as computer systems that have knowledge in the domain of problem solution. Human experts rely on experience, as well as on knowledge. This is the reason why problem solving behavior cannot be performed by using simple algorithms. Experience can be regarded as a specialized kind of knowledge created by a complex interaction of rules and decisions. Instead of representing knowledge in a static way, rulebased systems represent knowledge in terms of rules that lead to conclusions. A simple rule-based system consists