Call number: LD2668 .T4 1986 Y84 / Master of Science / Computing and Information Sciences

D.Litt. et Phil. / The need for an authority list for South African names has been expressed on various occasions. The general aim with this study was to determine the reasons for this need and propose an effective solution to the problem. At the hand of a comprehensive literature study an overview was given of what authority control is, why name authority control is necessary and the problems experienced during name authority control. This literature study was done for two reasons, namely to: Use the information obtained in this way, as a basis to analyse the South African situation in regard to the standardisation of South African names. To clear up the existing, ignorance in South Africa about the nature, aim and necessity for standardisation of names. In the light of the poor quality of records imported by participating libraries into the South African cooperative databases, it was concluded that the participating libraries are not really aware of the impact the standard of their work has on resource sharing and shared cataloguing. By providing this background information an attempt is made to make libraries and information services aware of the importance of. the standardisation of names on national and international levels. With the basic information on name authority control as starting point, the South African situation was investigated with regard to standardisation of South African names. Important roleplayers were interviewed in order to determine how they go about standardising South African names and to identify the problems experienced with the standardisation of names. In order to get a better understanding of the problems the roleplayers experience, a sample of names was taken from the name authority file of the South African National Bibliography (SANB). The sample of names was identified by random sampling. The minimum size of the sample was determined by using, s statistical formula. The sample of names was analysed regarding_ variations in names, the same name for different people as well as changes in corporate names. A critical analysis of the South African situation regarding the standardisation of South African names was done, using all the information gathered. In order to find a solution to all the problems experienced, two models were proposed, namely a traditional model and a model consisting of an alphanumeric code. The two models were compared to each other in order to determine which one of the two models would be the most effective for the standardisation of South African names. In the presentation of the one model a programme was developed in order to practically demonstrate the model and to test its viability. During the comparison of the two models the ordinal and interval scales of _ measurement were used. At the hand of the results of the measurements, the most effective model for the standardisation of South African names was proposed. Thus, an answer was provided to the original problem statement namely: How can a standardised list of South African names be compiled cost-effectively in order to make names available timeously for use by libraries and information services nationally and internationally?

Organizational project knowledge is not being captured, consolidated, and organized, making it difficult to learn from past projects, expose the knowledge of the most experienced people, or share experience across geographic project locations. The lack of an ontology for representing this comprehensive project store inhibits its creation and the development of tools to operate on it. Process-orientation links organizational resources or artifacts with process phases and workflow. A process-orientation in knowledge management can be used to add contextual metadata to knowledge artifacts. Context can be used to improve information retrieval precision. Therefore, the study proposed a process-oriented ontology to improve the transfer of software engineering project knowledge. Four questions guided the research: What knowledge about projects should be captured? Are all project artifacts necessary and are they all equally valuable? How can process-orientation be applied to a software engineering project knowledge ontology? Are current knowledge representation languages appropriate for the task? Can software development project knowledge, as represented by this ontology, be captured and retrieved effectively in a KMS? Literature research and an empirical laboratory study answered all of the questions: Four areas of project knowledge are particularly valuable in terms of their impact on project success; requirements, revisions, risks, and resolutions. These areas also cover a meaningful breadth of software engineering project knowledge. A process abstraction was created that breaks a project down into eleven phases. These phases were the basis for a class definition that was added as a peer class to the knowledge artifacts. Using Protégé, the Process-Oriented Ontology for Software Engineering (POSE) was successfully implemented in OWL-DL. Project knowledge from a software organization was used to construct two knowledgebases: one using Google Desktop and the other using Protégé and POSE. Results demonstrated that software engineering project knowledge, as represented by POSE, can be effectively captured and retrieved. POSE-enhanced search was superior to keyword search. Google was comparable in broad text search. But the benefits of metadata and semantics proved to have significant advantages for ontologies. Process-orientation was also validated as a contributor to improved classification and retrieval.

The vast majority of text freely available on the Internet is not available in a form that computers can understand. There have been numerous approaches to automatically extract information from human- readable sources. The most successful attempts rely on vast training sets of data. Others have succeeded in extracting restricted subsets of the available information. These approaches have limited use and require domain knowledge to be coded into the application. The current thesis proposes a novel framework for Information Extraction. From large sets of documents, the system develops statistical models of the data the user wishes to query which generally avoid the lim- itations and complexity of most Information Extractions systems. The framework uses a semi-supervised approach to minimize human input. It also eliminates the need for external Named Entity Recognition systems by relying on freely available databases. The final result is a query-answering system which extracts information from large corpora with a high degree of accuracy.

This work is an investigation into reconstructing fragmented ASCII files based on content analysis motivated by a desire to demonstrate machine learning's applicability to Digital Forensics. Using a categorized corpus of Usenet, Bulletin Board Systems, and other assorted documents a series of experiments are conducted using machine learning techniques to train classifiers which are able to identify fragments belonging to the same original file. The primary machine learning method used is the Support Vector Machine with a variety of feature extractions to train from. Additional work is done in training committees of SVMs to boost the classification power over the individual SVMs, as well as the development of a method to tune SVM kernel parameters using a genetic algorithm. Attention is given to the applicability of Information Retrieval techniques to file fragments, as well as an analysis of textual artifacts which are not present in standard dictionaries.

Private Information Retrieval (PIR) protocols enable a client to access data from a server without revealing what data was accessed. The study of Computational Private Information Retrieval (CPIR) protocols, an area of PIR protocols focusing on computational security, has been a recently reinvigorated area of focus in the study of cryptography. However, CPIR protocols still have not been utilized in any practical applications. The aim of this thesis is to determine whether the Melchor Gaborit CPIR protocol can be successfully utilized in a practical manner in an anonymous peer-to-peer environment.

This report describes a distributed approach to social filtering based on the agent metaphor. Firstly, previous approaches are described, such as cognitive filtering and social filtering. Then a couple of previously implemented systems are presented and then a new system design is proposed. The main goal is to give the requirements and design of an agent-based system that recommends web-documents. The presented approach combines cognitive and social filtering to get the advantages from both techniques. Finally, a prototype implementation called WebCondor is described and results of testing the system are reported and discussed.

[pt] A seguinte dissertação tem como objetivo explorar a Mineração de Textos através de um estudo amplo e completo do que atualmente é considerado estado da arte. Esta nova área, considerada por muitos como uma evolução natural da Mineração de Dados, é bastante interdisciplinar e vem obtendo importantes colaborações de estudiosos e pesquisadores de diversas naturezas, como Lingüística, Computação, Estatística e Inteligência Artificial. Entretanto, muito se discute sobre como deve ser um processo completo de investigação textual, de forma a tirar máximo proveito das técnicas adotadas nas mais variadas abordagens. Desta forma, através de um encadeamento sistemático de procedimentos, pode-se chegar a uma conclusão do que seria a metodologia ideal para a Mineração de Textos, conforme já se chegou para a de Dados. O presente trabalho explora um modelo de processo, do início ao fim, que sugere as seguintes etapas: coleta de dados, pré-processamento textual, indexação, mineração e análise. Este sequenciamento é uma tendência encontrada em trabalhos recentes, sendo minuciosamente discutido nos capítulos desta dissertação. Finalmente, a fim de se obter enriquecimento prático, foi desenvolvido um sistema de Mineração de Textos que possibilitou a apresentação de resultados reais, obtidos a partir da aplicação de algoritmos em documentos de natureza geral. / [en] The following essay is intended to explore the area of Text Mining, through an extensive and comprehensive study of what is currently considered state of the art. This new area, considered by many as a natural evolution of the Data Mining, is quite interdisciplinary. Several scholars and researchers from fields like linguistics and computing, for instance, have contributed for its development. Nevertheless, much has been discussed on how complete dossier of textual investigation must be carried out, in order to take maximum advantage of the techniques adopted in various approaches. Thus, through a systematic sequence of procedures, one can come to a conclusion of what would be the ideal method for the Mining of documents, as one has come about Data. This work explores a model of process which suggests the following steps: collecting data, textual preprocessing, indexing, mining and analysis. This sequence is a tendency followed in some recent works and it is thoroughly discussed in the chapters to come. Finally, in order to obtain a practical enrichment, one developed a system of Mining of documents with which became possible the presentation of results, obtained from the application of algorithms in documents of a general nature.

O modelo de recuperação de informação mais amplamente utilizado no contexto de acervos digitais é o Vector Space Model. Algoritmos implementados para este modelo que aproveitam informações sobre relevância obtidas dos usuários (chamados feedbacks) na tentativa de melhorar os resultados da busca. Porém, estes algoritmos de feedback de relevância não possuem uma estratégia global e permanente, as informações obtidas desses feedbacks são descartadas para cada nova sessão de usuário (são perenes) ou não modificam os documentos como um todo (são alterações locais). Este trabalho apresenta um método de feedbacks de relevância denominado orientado a termos, permitindo que as modificações realizadas por influência dos feedbacks dos usuários sejam globais e permanentes. Foram realizados experimentos utilizando o dataset ClueWeb09 que dão evidências de que este método melhora a qualidade dos resultados da busca em relação ao modelo tradicional Vector Space Model. / The Vector Space Model is the most widely used information retrieval model within digital libraries\' systems. Algorithms developed to be used with this model use relevance information obtained from users (called feedbacks) to improve the search results. However, the relevance feedback algorithms developed are not global nor permanent, the feedbacks are discarded in users new sessions and do not affect every document. This paper presents a method that uses of relevance feedback named terms oriented. In this method, users\' feedbacks lead to modifications in the terms\' vectors representations. These modifications are global and permanent, influencing further searches. An experiment was conducted using the ClueWeb09 dataset, giving evidence that this method improves the quality of search results when compared with Vector Space Model.

Esta tese apresenta um modelo de relevância de opinião genérico e independente de domínio para usuários de Redes Sociais. O Social Opinion Relevance Model (SORM) é capaz de estimar a relevância de uma opinião com base em doze parâmetros distintos. Comparado com outros modelos, a principal característica que distingue o SORM é a sua capacidade para fornecer resultados personalizados de relevância de uma opinião, de acordo com o perfil da pessoa para a qual ela está sendo estimada. Devido à falta de corpus de relevância de opiniões capazes de testar corretamente o SORM, fez-se necessária a criação de um novo corpus chamado Social Opinion Relevance Corpus (SORC). Usando o SORC, foram realizados experimentos no domínio de jogos eletrônicos que ilustram a importância da personalização da relevância para alcançar melhores resultados, baseados em métricas típicas de Recuperação de Informação. Também foi realizado um teste de significância estatística que reforça e confirma as vantagens que o SORM oferece. / This thesis presents a generic and domain independent opinion relevance model for Social Network users. The Social Opinion Relevance Model (SORM) is able to estimate an opinions relevance based on twelve different parameters. Compared to other models, SORMs main distinction is its ability to provide customized results, according to whom the opinion relevance is being estimated for. Due to the lack of opinion relevance corpora that are able to properly test our model, we have created a new one called Social Opinion Relevance Corpus (SORC). Using SORC, we carried out some experiments on the Electronic Games domain that illustrate the importance of customizing opinion relevance in order to achieve better results, based on typical Information Retrieval metrics, such as NDCG, QMeasure and MAP. We also performed a statistical significance test that reinforces and corroborates the advantages that SORM offers.