The Algebra of Hidden Relations as a Means of Modeling Statistical Databases

Abstract

An analogue of relational algebra is considered that is used to model databases when attribute values are not completely visible to the user but only up to a certain equivalence relation on domains. This case is important for so-called statistical databases, when it is necessary to provide and comprehensively analyze the summarized information on data stored in a database, but the user does not have access to personal data in the database. Similarly to the relational approach, this case introduces the concept of a hidden relationship and a system of operations with hidden relationships. Some correlations between these operations are shown and the query language is defined.