Search

The article describes a technique for evaluation of generalized data management systems. The evaluation and selection of a suitable data management system for an organization can be a frustrating and confusing experience. Done systematically and objectively, however, it can result in substantial savings and relatively smooth growth of the systems and programming effort within the organization. Two prerequisites are mandatory: First, define very clearly your objectives for such a system. And second, clarify your definition of a generalized data management system. At the present time, we are not aware of any one package which satisfactorily performs all of the functions of the mythical system just described. Instead, the packages that are being marketed all perform one or more of these functions. Currently available data management systems can be grouped into the following four general categories: Information Retrieval Systems; Data Base Access Method and Storage Management Systems; Generalized Data Management Systems; Generalized On-line Information Systems. Having determined the objectives and requirements of a generalized data management system for your installation and applications and having formulated and defined the criteria which reflect those objectives, the following approach is suggested for the evaluation and selection of the package or packages to meet particular needs. First, gather information on available data management systems. Second, conduct preliminary evaluation. Third, revise and extend evaluation criteria. Fourth, prepare and mail questionnaire to vendors. Fifth, evaluate and score the alternatives. And finally, contact suppliers and arrange for on-site demonstration and testing of the most promising systems.

The article discusses the selection, grouping, and cross-referencing of terms for inclusion in a coordinate index word list. Most information retrieval systems provide for the needs of the natural sciences and of technology. For many organizations, the provision of an information service is justified, with the depth of indexing geared to the needs of the users. Even for small groups of users, equipment is readily available at reasonable cost. Indexing can be done either by an intelligent human analyzer or, provided that the text to be indexed is available in machine-readable form, by programmed machine operation, the program having been provided by an intelligent human programmer.