Thesis (M. Phil.)--University of Hong Kong, 1988.

Thesis (M.B.A.)--University of Hong Kong, 1988.

published_or_final_version / Civil and Structural Engineering / Doctoral / Doctor of Philosophy

published_or_final_version / Electrical and Electronic Engineering / Doctoral / Doctor of Philosophy

No description available.

This thesis systematically identifies the building blocks of a knowledge based system for simulation and modelling. We present the design and implementation of Controlled Environment Simulation System (CESS), which bridges a discrete event simulation system (DEVS-SCHEME) and a continuous simulation system (TRNSYS). The rationale behind the approach is that a discrete or a continuous model can be abstracted to a level at which the uniform treatment on these two kinds of models is possible. A top-down approach to model creation (abstraction) is proposed, in contrast to the traditional bottom-up approach. CESS is implemented on an object-oriented programming environment (SCOOPS on TI-SCHEME). A knowledge representation scheme known as System Entity Structure is employed for MODEL management, recording system structural knowledge, and the utilization of techniques in Artificial Intelligence. Some prospective research topics are also brought up.

Institutions of higher learning are currently faced with the crisis of finding appropriate criteria for undergraduate admission. This concern has been sparked by the fact that matriculation grades are no longer seen as an accurate reflection of students' academic potential. As tertiary education is becoming more expensive, it is therefore becoming more and more important to select only students who have a realistic chance of being successful in their studies. The main aim of this study is to validate the Situation Specific Evaluation Expert (SPEEX) as a predictor of academic success of first year students of Mechanical Engineering at the Vaal Triangle Technikon. The design used in this study is a non-experimental correlational design. This design was selected because the investigation of this study is aimed at determining the presence or absence of the relationship between the independent and dependent variables without specific reference to causality. The sample of this study consisted of a total of 140 mechanical engineering student at the Vaal Triangle Technikon. This sample was the total number of students from the Mechanical Engineering department who enrolled for mechanical engineering courses for the year 2000. The sample consisted of 94% males and females 6% females. Subject matter experts from industry as well as those involved in the training of Mechanical Engineering at the Vaal Triangle Technikon selected competencies, which were hypothesised to be indicative of a potentially successful student. Based on the selected competencies the assessment battery was compiled with the selected indices being considered as predictor variables. A multiple regression analysis was performed on data in order to establish the predictive validity of the assessment battery. SPEEX 2502 (Language proficiency) consistently showed a positive correlation on the prediction of academic success. / Thesis (MA (Industrial Psychology))--North-West University, Potchefstroom Campus, 2004.

An analysis of the organizational field of B.C. agriculture was conducted to explore the politics of the innovation process. Agricultural innovations in organic farming, synthetic agrichemicals and biogenetic engineering were studied at the individual, organizational and interorganizational levels. Research questions regarding the innovation decision—making process, innovation championship, organizational politics, organization theory and interorganizational networks were explored. A total of 137 persons (organic and conventional farmers, BCMAFF employees, farm organization employees) were interviewed in this research study. Data was collected via semi—structured interviews, questionnaires, and analysis of publications to investigate a total of 28 research questions. Similarities and differences between organic and conventional farmers in respect to their socioeconomic characteristics, motivations, actions and environmentalist beliefs were identified. Organic farmers basis for their innovation adoption decisions was found to be largely informed by their environmentalist philosophy whereas the primary motivating factor for conventional farmers was economic rather than ideological. Case studies of 33 farm organizations (20 conventional and 13 organic) were conducted. Organizational fields were found to be defined not only in terms of products, services and geographic location but also in terms of ideology. Within the conventional agriculture organizational field there was a high degree of homogeneity in organizational structures and decision making processes as well as close collaboration with government policy makers. Within the organic agriculture organizational field there was homogeneity in production practices, but heterogeneity in organizational structures, goals and decision making processes based on the radicalness of the environmentalist philosophy of an organization’s membership. The formation and operation of interorganizational networks in each organizational field confirmed previous findings of the critical problems in overorganized and underorganized networks. A longitudinal analysis of organizational politics in the organic agriculture organizational field revealed that institutionalization processes engender political contests among competing interests. The successful championship of an innovative government regulatory system was attributed to the early use of a wide variety of collaborative and competitive political games. Opponents’ efforts to neutralize champions’ escalation of commitment during the later stages of the innovation development process proved to be ineffective.

The number of Institutional Repositories (IRs), such as DalSpace, has been growing in the past few years. However, most IRs are not widely used by the intended end users. Evaluating the user interfaces is an essential part of any process to increase users' acceptance of IRs. There are two foci of this thesis: to evaluate the usability of DalSpace's interface using Nielsen's heuristics to uncover usability problems for development purposes and to examine the differences between user-interface experts and non-experts in uncovering problems with the interface. To apply the heuristics to the interface, I formed user profiles (also known as personas) to represent potential end-users groups. These profiles helped to communicate users' needs, abilities, tasks, and problems. To produce a reliable list of usability problems by applying the heuristic evaluation approach, I examined the impact of expertise on the quality of the results. From the individual heuristic analyses (by both experts and novices), I distilled 66 usability problems classified by severity. Also, the frequency of each violated heuristic was used to assign priority to the uncovered usability problems as well as the severity level. The results of applying the heuristic evaluation show that both experts and non-experts can uncover usability problems. In fact, the ability to find difficult and easy problems was recorded for both types of evaluators. However, experts tend to reveal more serious problems, while novices uncover less severe problems. Interestingly, the best evaluator (who found 21% of the total number of problems) was a novice. Our results indicate that we cannot rely on one evaluator even if the evaluator is an expert. The administrative interface is out of the scope of the thesis; however, the usability of the interface should be examined for improvement purposes. More user profiles should be formed to represent additional user groups for more interfaces including the end user and staff's interfaces. Both results from the user profiles and the list of usability problems will be given as tools to the development team for improvement.

The objectives of this project are to rate three approaches for deriving effluent requirements; create a prioritized list of improvement steps; apply a watershed-level model to determine treatment plant requirements; and provide recommendations for deriving effluent requirements in Ontario. Results of a two-part stakeholder survey show that until advancements are made, compromise is necessary when selecting an approach for deriving effluent requirements, as no one approach meets all the evaluation criteria. However, the necessary steps toward improvement are relatively clear and require multi-disciplinary input. When the watershed-level modelling approach was applied, it was found that although there are challenges that must be addressed, overall, it appears advantageous to use a tool such as watershed-level models for the purpose of deriving effluent requirements in Ontario. / Natural Sciences and Engineering Research Council of Canada, Ontario Graduate Scholarship Program