Generalized data management systems and scientific information - Archive ouverte HAL Accéder directement au contenu
Rapport (Rapport De Recherche) Année : 1978

Generalized data management systems and scientific information

Systèmes de gestion de bases de données et information scientifique et technique

Résumé

This report is aimed at people with scientific background, such as physicists, chemists, biologists, etc. as well as management personnel. Its purpose is to stimulate scientists of all disciplines to consider the advantages of using a generalized data management system (GDMS) for storage, manipulation and retrieval of the data they collect and often need to share. The report should also be of interest to managers and programmers who need to make decisions on the management of scientific (numeric or nonnumeric) data. Another goal of this report is to expose the features that a GDMS should have which are specifically necessary to support scientific data, such as data types and special manipulation functions. It is hoped that the way the report is organized, starting with basic concepts and the terminology of GDMS, then discussing the requirements of GDMS for scientific data, and describing case studies, will be of value to people who have not been exposed to GDMS before. At the same time, the reader more familiar with GDMS can benefit from guidance on available systems, from discussion of other users’ experience, and capabilities in GDMS helpful in handling scientific data. More specifically, the report will: a. Give a clear introduction to what GDMS are, why they may be useful, and what computing hardware is needed. b. Include a list of the capabilities required in a Generalized Database Management System for handling scientific data. Presented at a time when considerable effort is being invested in GDMS software development, such an inventory may be in time to influence the specifications of this third generation of Data Management Systems. c. Compare possible alternatives: do-it-yourself software, APL, file management systems. d. Show by case studies of a variety of existing and potential GDMS applications to scientific data (in different fields, with more or less numerical content) what is involved in GDMS use, and what advantages may result. e. Survey the direction of development work in GDMS: hardware development trends, software development trends, distributed Database systems, and database conversion. A GDMS is a system that provides generalized tools for the purpose of defining a database structure, for loading the data, for modification of the data, and for organizing the database for efficient retrieval and formatted output. A data management system is ’’generalized" when it provides a user-oriented language for the different functions, so that it is possible to define any new database, its internal organization, and to retrieve and modify the data without the need to develop special purpose software (program) for each new database. The main purposes of a GDMS are quoted from a recent survey [1]: • to make an integrated collection of data available to a wide variety of users; • to provide for quality and integrity of the data; • to ensure retention of privacy through security measures within the system; and • to allow centralized control of the database, which is necessary for efficient data administration. From the user point of view GDMS should provide: • data independence, i.e. that application software does not need to be modified when data or data structures are changed • languages and facilities to perform the spectrum of data management functions: data definition, data entry and updating, conditional search and data retrieval, and data output and report generation. These facilities could be available in on-line or batch mode depending on the needs of the application. • the representation and access of both numerical and literal data. The above points will be discussed in more detail in the next section of the report. The advantages of using a GDMS are numerous, but for scientists and other users who are not computer specialists (and who cannot afford the time to become expert programmers) the main advantage is the immediate availability of a system for database handling. If their data handling requirements are relatively simple, these may perhaps be satisfied without further programming by use of the query language/report writer facilities of a suitable GDMS. Where user requirements are more complicated, applications programs in a high-level ’host language’ (such as COBOL or FORTRAN) may be linked to the database by embedding GDMS Data Manipulation Language statements in those programs. The available programming effort can be concentrated on the specific problems in hand. The generality of GDMS can sometimes degrade the efficiency of an application compared to a specially developed program, but this consideration is usually outweighed by the savings in software development time and cost, by the availability of the data for numerous applications, and by facilities providing integrity and security and easy modification and manipulation of the data. Database management, like the use of highlevel programming languages instead of assembler, is the latest in a long series of compromises in which increased user convenience is traded against computer efficiency. Although initially these trade-offs are expensive, hardware and software have historically tended to evolve in a way which reduces the cost of convenience. Recognizing that scientific data may have different characteristics from ’’commercial” data, it is the aim of this study to explore the differences and point out systems that have features amenable to the handling of scientific data. Therefore, the people involved in the study were selected both from the scientific community and the computer field. In particular, cases where GDMS were used for the handling of scientific data are described. The report introduces first the spectrum of GDMS approaches, techniques and terminology. This is done at a level which describes the functionality of GDMS without going into the details of how it is achieved. Then a survey of (mostly commercially) available systems is given, together with comparative features. This is followed by a discussion of requirements and capabilities needed to deal with scientific data and a description of several data management systems designed specifically to handle scientific data. In the next sections several example scientific applications that use GDMS successfully are described, followed by examples of potential applications now under consideration. Finally, a discussion of future trends in the development of GDM S and related areas is given so that the reader may consider their possible effect on his environment. In the conclusion section (part of the epilogue) we summarize the viewpoints of participants relative to when should G D M S be used and what to expect in using them.
Fichier principal
Vignette du fichier
OCDE_26.pdf (8.98 Mo) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-01512694 , version 1 (24-04-2017)

Identifiants

  • HAL Id : hal-01512694 , version 1

Citer

Organisation de Coopération Et de Développement Économiques (ocde). Generalized data management systems and scientific information. [Research Report] Organisation de coopération et de développement économiques (OCDE). 1978, 347 p. ⟨hal-01512694⟩

Collections

LARA
103 Consultations
539 Téléchargements

Partager

Gmail Mastodon Facebook X LinkedIn More