Typescript (photocopy). / Digitized by Kansas Correctional Industries

This thesis has focused on digital born personal archives; how they are different from analog paper archives and possible ways to overbridge these differences. It is also studied how archivists and librarians are dealing with digital archives, what they think about new proactive solutions when collecting private papers and collections and positive and negative effects with this way to work. The thesis is delimitated to only private collections and digital created material. It do not concern digitized material or electronic documents in government or business organizations. The challenges in digital curation and administration are discussed in aspect of the Records Continuum Model. The empirical material is also put in relation to research in Personal Information Management and identity theory. Two qualitative methods have been used, a case study of an authors personal electronic archive and five deep-going interviews with archivists and librarians working with personal collections in university libraries. This study shows that many difficulties and problems displayed in the electronic archive could be solved in the future with a more proactive approach and early cooperation with the collector. By educating and informing creators of personal records about file formats, archival structure and the need of them to engage in their own collection, more personal digital records could be preserved. Librarians and archivists are imagining more proactive work in the future but are also asking for other solutions. A lot of the challenges united with electronic personal archives are associated with lacking continuity which is a reason to use the Records Continuum Model to explain and work with personal digital collections. This is a two year master’s thesis in archival science.

No description available.

No description available.

The construction industry has had difficulty adopting project management software effectively and in a widespread manner. This thesis seeks to uncover common problems preventing the adoption of new software for construction project management, and to find a method to design more successful software. Construction project managers working at universities in the state of Georgia were selected as the specific group to study. To better define the functions of project management software, as well as uncover benefits both realized and anticipated, research was conducted. A Situation Awareness based approach was selected to result in proposing a prototype application that would specifically match the mental model of the study group. To define the mental model, a Goal Directed Task Analysis method was used by way of an online survey and in-person verification. In order to collect the survey data, a web based Goal Directed Task Analysis application was created and tested as a part of this thesis. Using the survey data, a proposed design for a prototype application was proposed. The application design was specifically made to fulfill the needs of project managers like those in the study group. The GDTA method proved effective in producing a software design. Recommendations to create and test the proposed prototype application are provided as a next step.

The relational model has dominated the database field because of its reduced application development time and non-procedural data manipulation features. It has significant problems, however, including weak integrity constraints. This paper discusses the need for object oriented techniques to improve on these flaws. Three existing DBMS will be discussed: IRIS, ORION, and OZ. / Department of Computer Science

Transportation is one of the most frequent problems in the business world. The major feature of the problem is that there are many warehouses and many delivery places. The objective of solving this problem is to find the amount of goods that should be sent from each warehouse to each customer while minimizing cost.Unfortunately, understanding the process and interpreting the results are not easy tasks. The method is very complex. The result is in the form of a table. We might say that it is not a friendly user-interface.In this thesis, we will create an application which uses a window as an interface, and uses minimal storage.Borland C++ v.4.0 is chosen to handle the implementation, and Borland Object Windows (class) Library for C++ v.2.0, OWL, is used for the interface. Therefore, this application operates on Windows 3.1 or Windows 95, but not on DOS. / Department of Computer Science

The evidence of project management is known for centuries as can be observed from the construction of the pyramids of Egypt and the Great Wall of China. In the modern age, project-based approaches are increasingly being adopted in almost all areas of product development. Project management has emerged as a specialised discipline since the importance of completing projects within time, cost and quality constraints is crucial for organisations to continue to survive in the competitive world. Project Management, in general, has been developed under the assumption of economic, political, cultural and social rationality. Little is known about indigenous project management in developing countries. The concepts and principles put forward in Western/European countries have been applied to developing countries without much prior studies. Developing nations are struggling constantly to maintain their projects within the constraints of time, cost and quality. Software development companies have found it difficult to adopt methodologies/models/standards that have shown evidence of success in the developed world. This thesis comprises the study of software project management in developing countries. Mauritius, as an instance of developing countries, with problems related to social, economic, cultural and political conditions are discussed. However, these conditions differ from country to country. It is believed that adjustments are required in a software project management framework to fit the requirements of a country. An evolutionary software project management maturity model is proposed for managing software development in developing countries. This model adopts an evolutionary approach, whereby areas of interest (called key process areas) progressively attain maturity. Three levels of maturity are defined along with key process areas that are applicable over all the maturity levels (called the i continuous process improvement group of KPAs). The model is also applied to two software projects in Mauritius to test its effectiveness. Given the studies carried out and its successful application to the Mauritian context, this model for software project management is expected to contribute towards a higher software project success rate. Notwithstanding the application in the Mauritian context, it is plausible that other developing countries may also customise this model as similar problems occur across these countries. / Computing / D.Phil. (Computer Science)

D.Ed. / The investment in a computerised administrative system has become a common phenomenon at educational institutions. With the development of Coltech, a computerised administrative system for technical colleges, administration at colleges changed dramatically. This system held the promise of reduced administrative work for staff. The current application of the system still reflects the above-mentioned point of departure. Many educational leaders do not realise that the value of the system surpasses the mere administrative application as it is found today. The data contained within the system can be applied fruitfully during daily management of a technical college at the strategic, tactical and operational level. The quality of decision-making can be influenced directly by means of timeous and quality information being readily available to management. Management,however, currently tends to manage without using this valuable resource. The research question and related aim of this research is to determine how a computerised administrative system, (such as Coltech), as a basis for a management information system, can contribute to managerial decision-making at technical colleges. It is generally accepted that there is a direct link between information management on the one hand, and administration and administrative management on the other. The link is found in the information which flows directly from the activities known as administration. Areas such as personnel, students, external community, finance, education and curriculum as well as physical facilities need to be administered and managed to transform the data in the database to information. The transformation of data to information will provide management with the information they need to make quality decisions

We consider the problems commonly encountered in asset management such as optimal execution, portfolio construction, and trading strategy implementation. These problems are generally difficult in practice, in large part due to the uncertainties in financial markets. In this thesis, we develop data-driven approaches via machine learning to better address these problems and improve decision making in financial markets. Machine learning refers to a class of statistical methods that capture patterns in data. Conventional methods, such as regression, have been widely used in finance for many decades. In some cases, these methods have become important building blocks for many fundamental theories in empirical financial studies. However, newer methods such as tree-based models and neural networks remain elusive in financial literature, and their usabilities in finance are still poorly understood. The objective of this thesis is to understand the various tradeoffs these newer machine learning methods bring, and to what extent they can improve a market participant’s utility. In the first part of this thesis, we consider the decision between the use of market orders and limit orders. This is an important question in practical optimal trading problems. A key ingredient in making this decision is understanding the uncertainty of the execution of a limit order, that is, the fill probability or the probability that an order will be executed within a certain time horizon. Equivalently, one can estimate the distribution of the time-to-fill. We propose a data-driven approach based on a recurrent neural network to estimate the distribution of time-to-fill for a limit order conditional on the current market conditions. Using a historical data set, we demonstrate the superiority of this approach to several benchmark techniques. This approach also leads to significant cost reduction while implementing a trading strategy in a prototypical trading problem. In the second part of the thesis, we formulate a high-frequency optimal execution problem as an optimal stopping problem. Through reinforcement learning, we develop a data-driven approach that incorporates price predictabilities and limit order book dynamics. A deep neural network is used to represent continuation values. Our approach outperforms benchmark methods including a supervised learning method based on price prediction. With a historic NASDAQ ITCH data set, we empirically demonstrate a significant cost reduction. Various tradeoffs between Temporal Difference learning and Monte Carlo method are also discussed. Another interesting insight is the existence of a certain universality across stocks — the patterns learned from trading one stock can be generalized to another stock. In the last part of the thesis, we consider the problem of estimating the covariance matrix of high-dimensional asset return. One of the conventional methods is through the use of linear factor models and their principal component analysis estimation. In this chapter, we generalize linear factor models to a general framework of nonlinear factor models using variational autoencoders. We show that linear factor models are equivalent to a class of linear variational autoencoders. Further- more, nonlinear variational autoencoders can be viewed as an extension to linear factor models by relaxing the linearity assumption. An application of covariance estimation is to construct minimum variance portfolio. Through numerical experiments, we demonstrate that variational autoencoder improves upon linear factor models and leads to a more superior minimum variance portfolio.