A logic approach for exceptions and anomalies in association rules

Association rules have been used for obtaining information hidden in a database. Recent researches have pointed out that simple associations are insu cient for representing the diverse kinds of knowledge collected in a database. The use of exceptions and anomalies deal with a di erent type of knowledge sometimes more useful than simple associations. Moreover ex- ceptions and anomalies provide a more comprehensive understanding of the information provided by a database. This work intends to go deeper in the logic model studied in [5]. In the model, association rules can be viewed as general relations between two or more attributes quanti ed by means of a convenient quanti er. Using this formulation we establish the true semantics of the distinct kinds of knowledge we can nd in the database hidden in the four folds of the contingency table. The model is also useful for providing some measures for assessing the validity of those kinds of rulesPeer Reviewe