Browse Prior Art Database

Semantic Integrity Mechanism for A Shared Database

IP.com Disclosure Number: IPCOM000089368D
Original Publication Date: 1977-Oct-01
Included in the Prior Art Database: 2005-Mar-05
Document File: 3 page(s) / 17K

Publishing Venue

IBM

Related People

Fernandez, EB: AUTHOR [+2]

Abstract

Means to preserve the semantic integrity of information are a basic requirement for a shared database. A semantic integrity system prevents some types of inconsistencies introduced by errors of users on their application programs. By enforcing semantic restrictions on the information, it is possible to insure that the contents of the database are at least plausible, if not correct, and that no inconsistencies exist between related information.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 39% of the total text.

Page 1 of 3

Semantic Integrity Mechanism for A Shared Database

Means to preserve the semantic integrity of information are a basic requirement for a shared database. A semantic integrity system prevents some types of inconsistencies introduced by errors of users on their application programs. By enforcing semantic restrictions on the information, it is possible to insure that the contents of the database are at least plausible, if not correct, and that no inconsistencies exist between related information.

An integrity system consists of a set of rules that include assertions about the contents of the database (expressed in some suitable language), a validation mechanism that checks changes to the database for compliance with the integrity assertions, and an enforcement mechanism that performs some predefined actions upon detecting that one or more of the integrity assertions have been violated.

An integrity rule is the 5-tuple (o,t,p,c,e), where o is the data object to which the rule applies, t is an access type which indicates for what type of database access the rule will be invoked, p is an assertion stating a semantic constraint which must be true for an occurrence of the object o, c is a predicate expressing a condition which must be true in order for p to apply to o, and e is an enforcement type that specifies the action that will be taken by the system if p is not true. See [1] and below for examples of integrity rules.

The validation mechanism can be understood better in the context of a specific database (which is representative of a broad class). This database consists of a set of base relations (or Database Structures), which are presented to the end users through application-oriented views. These views, or templates, are basically joins and/or projections of the base relations, constructed for specific purposes. The database is accessed through an extended high-level language which, by referring to predefined templates, allows the user programs to manipulate database elements as any other program variable [2]. The organization of this database lends itself very well to apply to it an authorization model [3], which, due to the fact that the extended language makes program data intentions manifest, can be enforced partially at compile time.

There are in this system four kinds of data objects: templates, database (DB) structures, fields, and types. The DB structures constitute the shared system view of the database, and they are composed of one or more fields. Types, which carry attributes, serve as prototypes for fields. A template includes a defining expression, which indicates which DB structures participate in its construction and how they are combined. An application program views the database as a set of templates.

Four basic roles for users are contemplated in this system: 1. Database administrators (DBA) define and maintain the shared DB structures and their integrity and authorization rules. 2. Database designers build templ...