Robert H. Bonczek

A Generalized Decision Support System Using Predicate Calculus and Network Data Base Management

In view of the growing prominence of corporate modeling, an important area of research concerns techniques for facilitating the design and utilization of models. In this paper we show how first-order predicate calculus can be used as a language for formally stating modeling knowledge. Furthermore, knowledge stated in this manner can be subjected to the resolution principle. The result is that application specific modeling knowledge need not be embedded in a computer program. Rather, it can be stored in a data base and utilized as needed by a problem processing system employing resolution techniques. Advantages of a decision support system taking an approach of this sort are considerable modeling flexibility, capacity for automating the model formulation and execution processes, and compatibility with a high-level user interface language.

The DSS development system

As decision support systems become more commonplace, the demand for automatic and semiautomatic DSS development systems increases proportionately. Such systems provide a set of tools that guide the construction of models in response to a users query. This paper describes a set of such tools that provide capabilities for analysis, design, module management, and report and graphics generation.

Developments in Decision Support Systems

Publisher Summary Decision support system (DSS) is a computerized system, which utilizes knowledge about a particular application area to solve ill-structured problems. The number of application areas in which ill-structured problems exist is large, including such diverse applications as financial planning, medical diagnosis, operations management, and market planning. The characteristics of DSS are discussed. A generic framework for understanding the constitution of a decision support system is presented. Knowledge system, language system, and problem-processing system components of a DSS are highlighted. Major streams of research and development in the DSS field are surveyed. The essential components of DSS are also discussed. Research in DSS has concentrated on the effective use of computer technology in aiding a human in the process of making a choice. Though there has been considerable progress in developing concepts and tools for DSS, there is still considerable latitude for further progress. At present, there is software for mainframes to micros that serves as tools for building successful decision support systems.

A transformational grammar-based query processor for access control in a planning system

Providing computer facilities and data availability to larger numbers of users generates increased system vulnerability which is partially offset by software security systems. Much too often these systems are presented as ad hoc additions to the basic data management system. One very important constituent of software security systems is the access control mechanism which may be the last resource available to prohibit unauthorized data retrieval. This paper presents a specification for an access control mechanism. The mechanism is specified in a context for use with the GPLAN decision support system by a theoretical description consistent with the formal definition of GPLANs query language. Incorporation of the mechanism into the language guarantees it will not be an ad hoc addition. Furthermore, it provides a facile introduction of data security dictates into the language processor.

A generalized mapping language for network data structures

This paper addresses itself to the problem of mapping network data structures into network data structures. There are many reasons for studying mapping functions for data bases; some of these are cataloged in 111. The reason for studying network to network maps is simply that complex data structures can naturally be represented by networks. Previous studies, such as [l], have restricted themselves to hierarchical data structures in order to avoid the ambiguity of relationships that arise when dealing with network data structures. When performing network data base retrieval[2], two elements must be specified in order to carry out the retrieval: the required data item names must be given, plus the relationships among the data items. The mapping problem differs from retrieval in that data items and relationships must be determined for both the source and target data bases. Thus in performing a mapping operation, a data base search must be carried out in both the source and target data bases for the appropriate data items. In this paper, a generalized mapping language will be defined. The language will have a linguistic formulation, in order to structurally analyze and easily implement the language constructs. The mapping language will be a high level non procedural language, so that a non-technical user of the data base system would be able to perform complex mapping operations. As a subset of the language, full retrieval capabilities for network data bases are implemented. Finally, the language is compared to other recent research on mapping.

Extensions and corrections for the CODASYL approach to data base management

Abstract In view of the potential contribution of data base management systems as an aid to planning activities, there are certain criteria which such systems must satisfy if the potential is to be realized. This paper suggests several extensions and improvements for the CODASYL approach to data base management. Special attention is devoted to the ease and efficacy of data retrieval and application execution, and to the facility and flexibility allowed with regard to data base creation and maintenance.

Information Transferral within a Distributed Data Base Via a Generalised Mapping Language

This paper investigates some of the problems encountered in the management of a data base that is logically distributed. The term logical distribution is used to connote a situation wherein users in various locales are responsible for creation and maintenance of their own portions of a data base, but a user in one locale can access data maintained by a user in another locale. An example of this situation is drawn from the realm of water quality management. A generalised mapping language is proposed as a mechanism for information transferral within a distributed data base, and a general data structure for supporting the mapping function is illustrated. The presented method accommodates a variety of user views, is independent of whether the data base is geographically distributed or centralised, furnishes a straightforward security mechanism and provides a basis for treating the contingency of uninformed or non-programming users.

REPRESENTATIVE DECISION SUPPORT SYSTEMS

This chapter provides an overview of representative decision support systems, which are developed in or used by the business community. An important trait common to all of these is their integration of models with information-handling capabilities. The significance of the users interface with these systems has become increasingly apparent. With the exception of statistical package for the social sciences (SPSS), the users language allows a user to direct retrieval activities and computational activities. Models themselves access the data bases of these systems, implying that the model is a user of the systems information-collecting abilities. The chapter presents a classification scheme for decision support systems in which humans use some sort of language to direct computations and the computational model, in turn, use a language to obtain needed data from the systems data base. The chapter also discusses the newer usage of models in conjunction with integrated data bases and exemplified by the representative systems.

NEW IDEAS IN DECISION SUPPORT

This chapter provides an overview of decision support systems (DSS). The generic description views a DSS as having three principal components: (1) a language system (LS), (2) a knowledge system (KS), and (3) a problem-processing system (PPS). The chapter examines several computer-based systems that adhere to the generic DSS description. However, most of these systems are not business oriented and few do not lie within the prevalent category E. To solve a stated problem using problem domain knowledge, there must be a PPS that possesses some of the seven abilities involved in decision making. Planner, strips, and mycin have been developed by workers in the artificial intelligence field, which deals with toy problems. Workers in the field typically do not have business backgrounds. On the other hand, designers of systems come from database management-management information systems (DBM–MIS) background. The potential application of artificial intelligence techniques in conjunction with DBM–MIS techniques has yet to be recognized or investigated in the DBM–MIS field. The aim of DSS design is toward a degree of generality that goes beyond such systems as the generalized planning system (GPLAN). The quest for generality is intended as a pedagogical aid, but it also has some practical implications in terms of the economic considerations.

BUILDING A DATA BASE KNOWLEDGE SYSTEM

This chapter focuses on conceptual and operational levels of knowledge representation. Conceptual constructs and operational constructs have been introduced and have been bound together. There is an examination of the various kinds of indirect relationships that can exist, given the conceptual constructs. Depending on which of these indirect relationships appear in a conceptual description of some application, binding produces a schema that strictly conforms to one of the major varieties of logical data structuring. Knowledge of conceptual and operational constructs is only the set of tools that are available for constructing a particular logical structure. Knowledge of the possible varieties of logical structure shows the possible forms that the particular logical structure could assume.