Fuzzy information management using the roster model

Abstract

The roster model is an array-theoretic alternative to the set-theoretic relational model in handling tabular data. In this paper, the roster model and its operations are extended to incorporate fuzzy data modelling and manipulation with the application of Zadeh's possibility theory. The result is a Fuzzy Roster Model. In the fuzzy roster model, there are two levels of uncertainty being represented the uncertainty in the data itself and the uncertainty in the truth value of the data. Fuzzy projections, selections and joinings in the model are illustrated with the use of examples. The original roster model and the fuzzy roster model are implemented in the Nested Interactive Array Language (NIAL), and are running on a Unix based SUN-3 Workstation.