Relational Database Systems: Analysis and Comparison

ACM SIGPLAN Notices, 1987

OOPSIA

ACM Transactions on Database Systems, 1980

TAXIS, a language for the design of interactive information systems (e.g., credit card verification, student-course registration, and airline reservations) is described. TAXIS offers (relational) database management facilities, a means of specifying semantic integrity constraints, and an exception-handling mechanism, integrated into a single language through the concepts of class, property, and the IS-A (generalization) relationship. A description of the main constructs of TAXIS is included and their usefulness illustrated with examples.

Cybernetics and Systems Analysis, 1996

The present article discusses the foundations of the relational approach to database models, which goes back to Codd's work [1-6] (see also surveys [i-9] and monographs [10-16]). The state of the art in this area of research is characterized by a large number of scattered results. Further progress is constrained by three missing components: a unified methodological and conceptual base, formalization of the main paradigms, and an understanding of the interrelationships between the corresponding formalizations. The next stage in the development of relational databases thus should involve a shift of emphasis from extensional to intensional approach to the subject. Certain steps in this direction are presented in our paper, whose goal can be summarized as follows: use the principles of programmology developed in the framework of composition programming [ 17-20] to provide adequate formalizations of the main paradigms for the construction of a substantive theory of relational databases while remaining on a sufficiently high level of abstraction. Database models, and in particular relational models, have at least two major aspects: the information aspect associated with formalization of data structures, and the manipulation aspect, associated with formalization of operations on these structures. Joint formalization of these aspects reduces to the construction of a corresponding algebra of data structures. Relation is the key concept of the relational approach to databases, and it is therefore the first that must be formalized. Like any natural intuitive concept, the database relation potentially admits a variety of different formalizations. The first formalization of database relation, following Codd, was the classical view of a finite-place logical relation as a subset of the Cartesian product of sets taken in a certain order (in other words, a relation is a set of ordered n-tuples of fixed length, n = I, 2 ....). This formalization was in fact responsible for the initial success and popularity of the relational approach to databases. Indeed, informal concepts could be formalized and analyzed using a well-developed formal apparatus (theory of logical relations, fragments of languages of first-order predicate logic, etc.). Yet even this traditionally established formalization of the concept of database relation is not free from weaknesses, which lead to a number of irrelevancies. These irrelevancies include the following: first, the information content of a database relation is usually independent of the order of the components, whereas for logical relations this order is essential (thus, for instance, it is not always meaningful to consider the inverse of a given database relation, by analogy with the concept of symmetry); second, the use of standard names l, 2 .... to access the components of a database relation (tuples in this case) is quite burdensome and leads to certain complications, such as the appearance in the metalanguage, but not in the object language, of arbitrary ("mnemonic") names denoting the standard names; moreover, with the exception of set-theoretical operations, the only natural many-place operation on database relations under this formalization is the so-called Cartesian product (not to be confused with the eponymous set-theoretical operatioi,), and all other natural operations (for instance, 0-join) are introduced as parametric operations on the standard names of the components. Also note the following typical twist: the operation of multiplication of (binary) logical relations, which is the core of relational algebra (see, for instance, [21-23]), does not play the same role for the general case of database relations, but it is of course simulated by the 0-join with ordinary equality acting as the parameter 0. No less important is the fact that the common database management systems (DBMS) are not strictly relational in the sense of the above formalization, because they admit arbitrary component names (see, e.g., [24]).

ACM SIGMOD Record, 1993

This paper very briefly summarizes the features and technologies implemented in the IBM relational DBMS products. The topics covered include record and index management, concurrency control and recovery methods, commit protocols, query optimization and execution techniques, high availability and support for parallelism and distributed data. Some indications of likely future product directions are also given.

1991

Object-oriented data models are receiving wide attention since they provide expressive abstraction mechanisms to model naturally and directly both structural and behavioral aspects of complex databases applications. In an objectoriented data model, a database is modeled in terms of objects grouped in classes, organized into subclasses hierarchies. Moreover, associations between entities are modeled by defining properties of objects whose value is the related object. However this way of modeling associations has several limitations which make the description of some aspects of associations unnatural. To overcome these limitations an object-relationship data model is proposed which supports both the mechanisms of an objectoriented data model and a separate mechanism to model explicitly associations and to express declaratively common constraints on them. Constructs to support this model for a statically and strongly typed object-oriented database programming language are defined.

2003

In this paper, we describe an environment developed to support a rich learning experience in which practice and theory in relational databases are better integrated, enabling students from various backgrounds to appreciate the significance of relational theory and the logical flaws in SQL. Our lightweight open source software includes the aspects of a commercial database system that are most relevant to teaching relational databases and can be run on several platforms.

Databases based on relational, object-oriented, and object-relational models represent significant advances in database technologies. In the context of general-purpose database management systems, the fundamentals of database models are examined. A historical perspective on the evolution of major database models is provided. The principal concepts underlying relational, object-oriented, and object-relational database models are presented with examples. Finally, a brief view of database federation issues is introduced and served as the foundation for computerized databases are essential and inseparable components of most of today’s information systems and data base technology.

Springer eBooks, 1983

In this paper, we present a comprehensive approach for privacy preserving access control based on the notion of purpose. Purpose information associated with a given data element specifies the intended use of the data element, and our model allows multiple purposes to be associated with each data element. A key feature of our model is that it also supports explicit prohibitions, thus allowing privacy officers to specify that some data should not be used for certain purposes. Another important issue addressed in this paper is the granularity of data labeling, that is, the units of data with which purposes can be associated. We address this issue in the context of relational databases and propose four different labeling schemes, each providing a different granularity. In the paper we also propose an approach to representing purpose information, which results in very low storage overhead, and we exploit query modification techniques to support data access control based on purpose information.

Proceedings of the 6th ACM SIGACT-SIGPLAN symposium on Principles of programming languages - POPL '79, 1979

... Such updates are called transactions [ES2] . Most of the work concerned with the correct-ness of multi-user database systems [BE1,ES2,LA2, PA1,ST1,TH1] assumes that only transactions access ... The familiar operations [cH1,co1,HE1,LAII --retrieve, insert, delete, update-...

ACM SIGMOD Record, 1990