Mefisto: a functional model for statistical entities

Abstract

There have been numerous proposals aimed at correcting the deficiency in existing database models in order to manipulate statistical data. The manipulation of these data, such as statistical tables that are widely used in many statistical database application areas, is examined. A functional model Mefisto, which is based on a data structure called statistical entity and on a set of operations capable of manipulating this data structure, is proposed. The characteristics that an aggregate data model has are discussed and a brief survey of the main proposals in literature is made. The operators that allow statistical entities to be manipulated both from the descriptive and from the summary data point of view are discussed, and some examples are given. Each operator can be seen as a family of operators, and each is able to automatically compute the summary values of the statistical entity obtained by its application. A brief discussion regarding the limitations of the relational model for this type of data and a comparison with other proposals are presented. The advantages of the Mefisto model over those proposals are illustrated. It is shown that it is possible to define user friendly query languages based on the Mefisto model.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>