Introducing λ-Specialization into the Fuzzy EER Model

Abstract

The entity–relationship (ER) model and its extensions (EER) have been widely used as conceptual models in database design and business modeling. The traditional ER model consists of three basic concepts—entity, relationship, and attribute. An ER model M can be denoted as M = (E, R, A), in which E, R, A are sets of entity types, relationship types, and attributes, respectively. An entity type E in an ER model is an element of E. In the traditional EER model, the constraints would affect database maintenance when updating data. With total specialization, an entity must be inserted into at least one subclass if it is inserted into the superclass. An entity must be deleted from the superclass or inserted into another subclass if it is deleted from a subclass and does not belong to any other subclass. In the setting of λ-specialization, where λ can be regarded as the strength of participation of entities in the specialization concerned, a change in the entity membership between nonzero values would cause a change in the degree of specialization. Hence the updating strategy would be different.