Asad Khailany

Advanced structured COBOL programming

There is a large market demand for COBOL programmers. Can a single introductory course in COBOL programming provide the students with sufficient professional expertise to qualify them for these job openings? If not, is it feasible to offer a second course in COBOL programming? In this paper the technical capabilities of those who have completed an introductory COBOL programming course and the skills and technical qualifications which are desired by employers of entry level COBOL programmers are analyzed. It is concluded that a single introductory COBOL programming course is not sufficient to produce COBOL programmes with the qualifications and skills expected by employers. Therefore, an advanced structured COBOL programming course is proposed. The goals and the course outline including texts and references of such an advanced course are presented. Several samples of integrated information systems which have been completed by student teams in Advanced Structured COBOL programming will be available for the audience.

Alternative teaching strategy for an introductory computer language course

Many consider that the best strategy to teach an introductory computer language course is to start programming very early in the semester. This paper discusses the drawbacks of such a strategy and presents an alternative way to teach such a course. Based on experiments the alternative strategy appears to be more effective. The advantages and outline of the alternative strategy are discussed. Furthermore, the course outline to teach a FORTRAN language based on the alternative strategy is presented.

Information systems curriculum (abstract): where we should be going?

The recommendations of the ACM Curriculum Committee on Information Systems indicate an attempt to keep abreast of both curricular changes in academia and job skill demands of the computing profession. As tie needs of both changed, new recommendations were made. The latest modell published in 1981, was the culmination of a process to update the ACM 1973 recommendations for undergraduate programs. It has been more than a decade however, since the publication of the 1981 model curriculum. Much has changed over the last dozen years in the job skills demands of undergraduate students and the curricular offerings of universities. The makeup of information systems has changed quite dramatically, with an increasing use of microcomputers and advances in telecommunications as evidence. As part of the process to keep the information systems curriculum up-to-date, this panel reviews the recommendations of 1981. The purpose of this panel is to address the following issues: (1) changes in the information systems profession which demand a response in curricular change, (2) the adequacy of the 1981 model in meeting the needs of the IS graduate in the 90s, and (3) specific recommendations for an Information System Curriculum for the 90s. 1 Jay F. Nunarnaker, Jr., J. Daniel Couger, and Gordon B. Davis, “Information Systems Curriculum Recommendations for the 80s: Undergraduate and Graduate Programs, A Report of the ACM Curriculum Cornnrittee on Information Systems”, ACM Cwricufa Recommen&tions for Information Systems, Volume 11, Association for Computing Machinery, 1983 2 J. Daniel Couger, “Curricrrhrm Recommendations for Undergraduate programs in Information Systems”, Association for Computing Machinery, 1973

Extended object oriented model to design relational databases (abstract)

The recent trend in designing relational databases systems has been toward using object oriented model in high level. The next step is to convert the high level model into the relational diagrams from which relational are extracted. Finally the relations are examined to make it sure that they are fully normalized. In the current Object Oriented model a number of important concepts of the relational data bases such as generalization, subset hierarchy, the relation of a record type to itself..etc either can not be represented or it hard to represent them. In this presentation the object oriented model is extended to have a better representation of the concepts such as optional or mandatory relationship, aggregation objects, association objects, generalization, subset hierarchy and others.

A data model for software representation (abstract)

As software systems evolve, they grow larger and more complex. Thus, they become difficult to maintain. This is true, because making a change to a software system requires an understanding of how the part being changed fits into the system and of the potential ripple effect the change may have on the system. Understanding the software system is a time consuming activity. This is due to the fact that a software system is an abstract complex object. It may consist of many parts which are interrelated in many different ways forming a complex structure. Moreover, if the source code is the only reliable document that is available about a software system, understanding the software system will be more time consuming. The code of a software system show very little of their internal state and does not reveal the structure of the software system for the maintainer. Two of the factors that help the understanding of the programs are: 1) the ability to see the information relevant to what we want to understand and 2) the use of a more natural form of representation of information. In order to provide effective support for the software maintenance process, we believe that both factors described above should be incorporated in a software programming environment. In this paper, we present an environment which incorporate these factors and provides mechanisms for seeing and manipulating software in a much more powerful way than the current systems.

Database reasoning: extracting knowledge from databases (abstract)

As interest in developing expert systems continues to grow, it become more important for us to increase our understanding of how to successfully build these systems. The major problem that currently constricts the development of expert systems is the formidable process of eliciting knowledge from human experts. A possible solution to alleviate this knowledge-acquisition bottlenecks is to equip historical database with some deductive capabilities that will allow us to emulate expert-system reasoning entirely in the databases terms. In this paper we describe a system that tries to extend an existing database containing component test-repair history with several features to support inferencing at the reasoning level of human trouble shooters. That is, the system features automatic knowledge-base derivation. This presentation demonistrate the structure of a system that, when queried with a test step, will provide trouble shooting advice describing the most likely cause for failure.

Semantics in object-oriented data models (abstract)

Although object-oriented data models as a design tool for software development products is promising, nevertheless the semantics provided by the current contemporary object-oriented data models do not adequately support the design, evolution, and reuse of complex information systems. The semantic expressiveness of the abstraction techniques such as encapsulation, inheritance, and messaging for modeling application environment employee an overly simple object structures. The use of these simple object structures limit the capabilities of these techniques to describe the meaning of these systems which inevitably cause the loss of the information about the system. In this paper we present an object-oriented model which is based on structural constructs that are highly user-oriented and expressive of the application environment. Such as object-oriented model provide a rich set of mechanisms to allow the direct modeling of system semantics. Therefore, our goal is to establish a necessary set of semantic concepts for object-oriented data models that will enable the designer to naturally and directly incorporate more of the semantics of a system into its model. Such a semantics-based object-oriented data model will support: (1) Natural and efficient description of complex object structure; (2) Explicitly express relationships between objects; (3) Explicitly express extensive integrity constraints and exceptions on relationships between objects; (4) Express different levels of abstractions which allow the user to see objects from an operational or a structural view. Finally, we discuss reusability and extensibility of such a semantic object-oriented data model.

Extended rigerous dependency diagrams for designing fully normalized relational databases

No database management system can be implemented without the design of its database data structure. Data structures for databases intended for use by relational database management systems should be fully normalized to avoid deletion, update and insertion anomalies in such systems. Because normal form theories and guidelines are not easy to use or apply (using nonloss decomposition) and since only the first normal form is needed to get the software for relational database management systems to work, many users simply organize all data fields into a few tables and proceed with their processing without decomposing the relations into fully normalized forms. In August, 1985 a methodology was introduced by Henry C. Smith to offer a better method for database design. This new simplified methodology promised to overcome the difficulties associated with nonloss decomposition. In this methodology, from a dependency list a dependency diagram composed of interlinked bubbles and double bubbles is constructed. Fully normalized tables are driven from the dependency diagram. Although the methods has a number of positive points yet it has some drawbacks such as representation of many to many on the dependeny diagram and transformation of one to many relationships to relations and incomplete representation of the relationship between entities. In this presentation extension of the model is explored to overcome these drawbacks.

Physical design of a menu relational database system package (abstract only)

This paper outlines the design, implementation, and capabilities of a menu drive transportable relational database package. After identifying the goals of the system, a top down strategy, via data flow diagrams was used to outline the physical design of the package. During the physical design phase, the detail functions and design procedures of the DOL, DML, System support, and control routines were identified. The package is written in ANSI FORTRAN 66 and implemented on a DECSYSTEM-10. Since ANSI FORTRAN 66 compilers are available for the majority of mini and micro computer systems, this system can be implemented on different systems with a minimum of conversion efforts.

Control charts as a memory performance measurement tool (abstract only)

The needed data to analyze the utilization of the memory can be obtained from special hardware monitor devices or from software monitor programs. The software monitor programs such as IBMs RMF. Candle Corporations OMEGAMON and Boole and Babbages CPE are not as accurate as the hardware monitor devices (specially in the area of timing.) but solved all of the problems related to hardware monitors. Therefore the problem to analyze the utilization of the main memory is no longer getting the needed data. However, the presentation of the data in a format which is usable and comprehendible by both the system department (for tuning) and the data center management (for decision making) is a problem. This paper shows that the control chart technique developed by Shewart can be used as a measurement tool to analyze the utilization of the main memory to overcome data presentation problems.