Case-based reasoning (CBR) for multicomponent analysis using sensor arrays: application to water quality evaluation.

This paper reports by the first time the use of a simple, convenient and improved method of artificial intelligence (case-based reasoning, CBR) for multicomponent analysis, involving the data treatment afforded by electrochemical sensor arrays. The suitability of this method has been shown in aqueous solutions containing mixtures of ions of different nature and concentration, being applied in this case to a rapid recognition and evaluation of mineral water samples chosen as an illustrative example. The convenience of such a procedure opens the way to dispose of a versatile tool applicable for obtaining analytical information about ions present in a wide range of other multicomponent systems, and also to ascribe properties inherent to the examined samples.