Many Sorted Observational Calculi for Multi-Relational Data Mining

Abstract

Observational calculi approved as a tool for study logical properties of association rules. They are defined by modifications of classical predicate calculi. Many sorted observational calculi are introduced as a modification of classical many sorted predicate calculi. It is argued that such defined calculi are suitable for relational data mining. Results on decidability of these calculi are presented. Further research directions are outlined