African Journals Online
South African Computer Journal / Die Suid-Afrikaanse Rekenaartydskrif

Issue 29, December 2002
ABSTRACTS

A new classification scheme for software reuse: research article

Houhamdi, Z.; Ghoul, S.

Abstract: The Reuse Description Formalism (RDF) is a generalization of the faceted index approach to classification. It was initially designed as a tool to help increase reusability of software components at the code level (e.g. functions or subroutines). The goal of this work is to show that RDF can also be used effectively to represent and reuse other types of software knowledge. The emphasis here is on those proprieties of RDF that facilitates the representation of these objects, and not on how to define similarity distances for these objects. The problem of designing and using similarity distance comparators is addressed in another paper.

This paper demonstrates RDF's representation power by constructing sample classification taxonomy for software components to explaining informally the RDF's concepts. After that, it includes taxonomy for describing components of a commercial software library called EVB GRACE library.

Diamond terrain algorithm: continuous levels of detail for height fields: research article

Hakl, H.; Van Zijl, L.

Abstract: The generation of terrain meshes for real-time rendering of landscape data has a wide field of application ranging from military simulations and geographic information systems to flight testing and computer games. Duchaineau et al. presented a successful approach - real-time optimally adapting meshes (ROAM) - for terrain triangulation. However, the choice of geometric substructure in ROAM does not lend itself well to vertex optimizations such as triangle-strips. We present an alternative mesh representation for ROAM based on a triangle quadtree that naturally decomposes to triangle-strips. Additionally we present a new algorithm for the generation of continuous levels of detail (CLOD) for height fields that utilizes our proposed mesh representation. The algorithm is similar to ROAM in that it supports split and merge operations, but we ignore priority queueing in favor of four LIFO queues that support constant time insertion and deletion.

Directly constructing minimal DFAs: combining two algorithms by Brzozowski: research article

Watson, Bruce W.

Abstract: In this paper, we combine (and refine) two of Brzozowski's algorithms - yielding a single algorithm which constructs aminimal deterministic finite automaton (DFA) from a regular expression.

Harmonising vulnerability categories: research article

Venter, H.S.; Eloff, J.H.P.

Abstract: A legion of vulnerabilities are potentially compromising the security status of IT industries infrastructures today. Current state-of-the-art intrusion detection systems (IDSs) can potentially identify some of the vulnerabilities. Each IDS defines its own and unique list of vulnerabilities, making it cumbersome for organisations to assess the completeness and reliability of vulnerability scans. What This furthermore complicates the matter of determining the degree to which a specific IDS complies to with the security requirements of a specific organisation. This paper presents an approach to harmonise different sets of vulnerabilities as currently used by state-of-the-art IDS tools.

Hierarchical level of detail optimization for constant frame rate rendering of radiosity scenes: research article

Nirenstein, S.; Blake, E.; Winberg, S.; Mason, A.

Abstract: The predictive hierarchical level of detail optimization algorithm of Mason and Blake is experimentally evaluated in the form of a practical application to hierarchical radiosity. In a novel approach the recursively subdivided patch hierarchy generated by a perceptually refined hierarchical radiosity algorithm is treated as a hierarchical level of detail scene description. In this way we use the Mason-Blake algorithm to successfully maintain constant frame rates during the interactive rendering of the radiosity-generated scene. We establish that the algorithm is capable of maintaining uniform frame rendering times, but that the execution time of the optimization algorithm itself is significant and is strongly dependent on frame-to-frame coherence and the granularity of the level of detail description. To compensate we develop techniques which effectively reduce and limit the algorithm execution time: We restrict the execution times of the algorithm to guard against pathological situations and propose simplification transforms that increase the granularity of the scene description, at minimal cost to visual quality. We demonstrate that using these techniques the algorithm is capable of maintaining interactive frame rates for scenes of arbitrary complexity. Furthermore we provide guidelines for the appropriate use of predictive level of detail optimization algorithms derived from our practical experience.

Issues affecting the adoption of data mining in South Africa: research article

Hart, M.L.; Davies, K.; Barker-Goldie, E-J.; Theron, A.

Abstract: This research describes a study into the adoption of data mining (DM) techniques and tools in industry. They are seen by most companies interviewed as strategically important, with key applications varying per sector, but have not been seriously adopted by many. A framework reflecting issues pertaining to the adoption of data mining was proposed, and tested in interviews. Issues were divided into four main categories: Business Issues ; Data Mining Technology Issues ; Resource Issues and Organisational Issues. Results imply widely differing levels of importance attached to these, suggesting a revised framework of key barriers and enablers for successful adoption.

Limited flooding protocol for mobile ad hoc networks: research article

Denko, M.; Goddard, W.

Abstract: Mobile ad hoc networks are collections of mobile nodes without any fixed infrastructure or central co-ordinating mechanism for packet routing. Consequently, routing is a challenge. In this paper we propose a multipath routing protocol called Limited Flooding. The protocol is fully reactive and does not entail the computation of routing tables. It uses the basic features of flooding but restricts packet propagation by selecting a limited number of links. Discrete-event simulation is used to model ad hoc networks, and the performance of several variations of the protocol is evaluated. The simulation results show that Limited Flooding has better performance than pure flooding and is suitable for networks with unpredictable topological changes and highly mobile nodes.

PrivGuard: a model to protect private information based on its usage: research article

Lategan, F.A.; Olivier, M.S.

Abstract: No reliable method currently exists to safeguard the privacy of private information. Privacy policies are insufficient as compliance can not be enforced automatically. In this paper we propose a model to improve the control the owner of private information has over its protection. This is achieved by classifying private information based on the purpose it is acquired for, and then designing methods to protect each class of private information. Private information is then encrypted using homomorphic functions where such information is only required for validation. The validation can then be performed without divulging the actual private information. In cases where private information is required for other usages, a system based on Kerberos and trusted third parties is used in order to maintain as much control over private information as possible.

RDF: a formalism for describing and reusing software experience: research article

Houhamdi, Z.; Ghoul, S.

Abstract: Software reuse has been claimed to be one of the most promising approaches to enhance programmer productivity and software quality. One of the problems to be addresses to achieve high software reuse is organizing databases of software experience, in which information on software products and processes is stored and organized to enhance reuse. This paper presents a new approach to define and construct such databases called the Reuse Description Formalism (RDF). The formalism is a generalization of the faceted index approach to classification in the sense that it provides facilities to define facets, terms, and object descriptions. Unlike the faceted approach, objects in RDF can be described in terms of different sets of faceted and in terms of other object descriptions. This allows a software library to contain different classes of objects, to represent various types of relations among these classes. RDF incorporates a similarity-based retrieval mechanism that helps a reuser locate candidate reuse objects that best match the specifications of the target object. Similarity between two objects is quantified by a non-magnitude called similarity distance, which represents the estimated amount of effort required to construct one given the other. Because of this, similarity distances are not necessarily symmetric.

Specification and justification of an Authenticated Contributory Group Key Agreement Protocol: research article

Enaw, E.E.

Abstract: The purpose of this paper is twofold. First, we present a formal specification, using the RAISE Specification Language (RSL), of a practical Authenticated Contributory Group Key Agreement protocol for Dynamic Peer Groups (DPGs). During the formalisation process, we use finite state automata notations to develop an improved development method for RAISE, which addresses some limitations of the RAISE method (lack of appropriate techniques for identifying and formulating properties of the system to be specified). Second, we formally specify two desired properties of a practical Authenticated Contributory Group Key Agreement Protocol for Dynamic Peer Groups (DPGs) as theorems and provide informal justifications of their correctness. Obvious candidates of software specification and verification are sensitive software and hardware systems where: an error could have catastrophic consequences and system security is a critical issue. Distributed systems, where the global behaviour depends on parallel interaction of different sub-systems, is an example of such systems. In this paper, we use the RAISE Specification Language (RSL) to specify our system and employ the RAISE method technique of seperate development to decompose the description of our system into components and compose the final system from the (developed) components

The development of a user classification model for first year university students: research article

Streicher, M.; Wesson, J.L.; Calitz, A.P.

Abstract: A number of computerised testing systems have been developed recently without considering the users? level of computer proficiency. First year students at the University of Port Elizabeth have a diverse background, both in home language and population group. The level of computer expertise of these students differs greatly and this may influence the test scores they obtain in computerised tests. In this study, various factors were found to be significant indicators of performance when using computer-based testing systems. These factors include previous computer and software experience, attitude towards computers, self-perceived ability to work with computers, contact with technology, gender and home language. This paper discusses the development of a user classification model to classify students into three user groups, namely novice, intermediate and expert. Two methods were used for classification. The first followed a quantitative approach to user modelling and required users to perform simple computer-based tasks. The second method was qualitative in nature and used a questionnaire to assess the factors that were found to be significant indicators of performance in human-computer interaction. This research project demonstrates that a user classification model can be developed for the first year student population at the University of Port Elizabeth.

Management Information System (MIS) for university administration - the Nigerian initiative: communication article

Uwadia, C.O.; Nwamarah, G.; Eseyin, E.; Sawyerr, A.

Abstract: Quite a number of Nigerian universities have come to terms with the need for accurate storage and fast retrieval of information by putting in place some form of administrative computing for students registration, examination processing, staff records and payroll. However, there was no uniformity or standards among these universities in software development, or compatibility in the hardware procured thereby making it difficult for national supervisory and regulatory bodies to obtain the much needed statutory information from the universities.

This paper discusses the introduction of an effective MIS through the installation of a collaboratively developed standard software in some Nigerian universities. The project which is on-going, is funded by the Nigerian government, through the National Universities Commission (NUC), and was initially supported by the Overseas Development Administration (ODA), through the British Council. The authors are active participants in the design and implementation of the project.