The automatic discovery of alarm rules for the validation of microbiological data

In this work, we describe a project, jointly started by University of Bologna and Dianoema S.p.A. in order to build a system which is able to validate microbiological data. Within the project we have experimented data mining techniques in order to automatically discover association rules from microbiological data, and obtain from them alarm rules to be used for data validation. To this purpose, we have exploited the WEKA system and applied it to a database containing data about bacterial antibiograms. Discovered association rules are then transformed into alarm rules, to be used for data validation within an expert system named ESMIS. Among automatically produced alarm rules, we have identified some already considered in ESMIS and suggested by experts according to the NCCLS compendium, and new rules which were not present in that report, but were recommended by interviewed microbiologists.