Thesis (M.S.C.I.T.)--Regis University, Denver, Colo., 2006. / Title from PDF title page (viewed on Aug. 30, 2006). Includes bibliographical references.

Thesis (Ph.D.)--University of Missouri-Columbia, 2004. / Typescript. Vita. Includes bibliographical references. Also available on the Internet.

Thesis (doctoral)--Universität zu Lübeck, 2006. / Includes index and bibliographical references (p. [165]-175).

Thesis (M.S.) -- Worcester Polytechnic Institute. / Keywords: Continuous query processing; Adaptation policies; Partitioned window join operator. Includes bibliographical references (p.).

Thesis (Ph. D.)--University of Washington, 2000. / Vita. Includes bibliographical references (leaves 95-97).

Orientador: Carlos Roberto Valêncio / Banca: Rogéria Cristiane Gratão de Souza / Banca: Marilde Terezinha Prado Santos / Resumo: Os Sistemas Gerenciadores de Banco de Dados Não Convencionais são utilizados por aplicações que necessitam de e ciência no gerenciamento de objetos complexos. Um dos conceitos fundamentais nestes sistemas e o de identidade de objetos, pois em uma base de dados cada objeto possui um identificador unico que e utilizado para acessá-lo e referenciá-lo em relacionamentos com outros objetos. A implementação de identidade pode ser feita com OIDs físicos ou OIDs lógicos. A abordagem física apresenta o problema de fragmentação da base de dados, pois os OIDs são formados diretamente pelos endereções dos objetos. Já a abordagem lógica não tem este problema, e as técnicas são arvore-B, hashing e mapeamento direto. Cada uma destas abordagens apresenta um determinado problema: arvore-B pode ocupar muita mem oria e o tempo de mapeamento possui complexidade logar tmica; em hashing ocorrem colisões pois o conhecimento prévio do tamanho da tabela hash se torna inviável em base de dados, que crescem de maneira imprevisível; e por ultimo, mapeamento direto, que apesar de possuir o menor tempo de mapeamento dentre as três abordagens, não permite a relocação de todas as páginas da base de dados. Uma outra abordagem lógica é utilizada no Núcleo Gerenciador de Dados Multimídia (NUGEM), o qual vem sendo desenvolvido junto ao Grupo de Banco de Dados do IBILCE de São José do Rio Preto com o intuito de gerenciar dados não convencionais. Neste trabalho e proposta uma nova estrutura e funcionalidades para a t ecnica de gerenciamento de OIDs, cuja experimenta cão foi efetivada junto ao NUGEM, caracterizando uma nova abordagem com um menor tempo de manipulação dos OIDs, a qual pode ser nomeada de mapeamento indireto. E tamb em avaliado o esgotamento de OIDs, comprovando-se que este não representa um problema para a abordagem de mapeamento indireto. Em comparaçãocom as abordagens... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: Non-Conventional Database Management Systems are used for applications that require e cient management of complex objects. One fundamental concept of these systems is object identity, because in a database each object has a unique identi er that is used to access it and reference it in relationships with another objects. The implementation of identity can be made with physical OIDs or logical OIDs. The physical approach presents the problem of fragmenting the database, because the OIDs are formed directly by addresses of objects. The logical approach does not have this problem, and the techniques are B-tree, hashing and direct mapping. Each of these approaches presents a particular problem: B-tree can occupy a lot of memory and its mapping time is logarithmic, in hashing collisions occur because the prior knowledge of the table size becomes a problem in databases, which grow in unpredictable ways, and nally, direct mapping, which despite having the smallest mapping time of the three approaches, does not allow the relocation of all pages in the database. Another logical approach is used in the Multimedia Data Manager Kernel (NUGEM), which is being developed by the Database Group at IBILCE, S~ao Jos e do Rio Preto, S~ao Paulo, Brazil, in order to manage non-conventional data. This work proposes a new structure and features for the OID management, which was tested in NUGEM, featuring a new approach with a shorter handling time of OIDs, that can be named indirect mapping. It is also evaluated the exhaustion of OIDs, proving that this is not a problem for the indirect mapping approach. Compared with the approaches of the literature, the approach of indirect mapping has constant mapping time, uses less memory than the B-tree approach, has no problems like collisions and, unlike direct mapping, allows total exibility for relocating pages in the database / Mestre

Made available in DSpace on 2014-06-11T19:29:41Z (GMT). No. of bitstreams: 0 Previous issue date: 2010-05-21Bitstream added on 2014-06-13T18:39:30Z : No. of bitstreams: 1 ferrizzi_ac_me_sjrp.pdf: 598030 bytes, checksum: 77b6ef0379d56ab11ae68873cafa2f56 (MD5) / Os Sistemas Gerenciadores de Banco de Dados Não Convencionais são utilizados por aplicações que necessitam de e ciência no gerenciamento de objetos complexos. Um dos conceitos fundamentais nestes sistemas e o de identidade de objetos, pois em uma base de dados cada objeto possui um identificador unico que e utilizado para acessá-lo e referenciá-lo em relacionamentos com outros objetos. A implementação de identidade pode ser feita com OIDs físicos ou OIDs lógicos. A abordagem física apresenta o problema de fragmentação da base de dados, pois os OIDs são formados diretamente pelos endereções dos objetos. Já a abordagem lógica não tem este problema, e as técnicas são arvore-B, hashing e mapeamento direto. Cada uma destas abordagens apresenta um determinado problema: arvore-B pode ocupar muita mem oria e o tempo de mapeamento possui complexidade logar tmica; em hashing ocorrem colisões pois o conhecimento prévio do tamanho da tabela hash se torna inviável em base de dados, que crescem de maneira imprevisível; e por ultimo, mapeamento direto, que apesar de possuir o menor tempo de mapeamento dentre as três abordagens, não permite a relocação de todas as páginas da base de dados. Uma outra abordagem lógica é utilizada no Núcleo Gerenciador de Dados Multimídia (NUGEM), o qual vem sendo desenvolvido junto ao Grupo de Banco de Dados do IBILCE de São José do Rio Preto com o intuito de gerenciar dados não convencionais. Neste trabalho e proposta uma nova estrutura e funcionalidades para a t ecnica de gerenciamento de OIDs, cuja experimenta cão foi efetivada junto ao NUGEM, caracterizando uma nova abordagem com um menor tempo de manipulação dos OIDs, a qual pode ser nomeada de mapeamento indireto. E tamb em avaliado o esgotamento de OIDs, comprovando-se que este não representa um problema para a abordagem de mapeamento indireto. Em comparaçãocom as abordagens... / Non-Conventional Database Management Systems are used for applications that require e cient management of complex objects. One fundamental concept of these systems is object identity, because in a database each object has a unique identi er that is used to access it and reference it in relationships with another objects. The implementation of identity can be made with physical OIDs or logical OIDs. The physical approach presents the problem of fragmenting the database, because the OIDs are formed directly by addresses of objects. The logical approach does not have this problem, and the techniques are B-tree, hashing and direct mapping. Each of these approaches presents a particular problem: B-tree can occupy a lot of memory and its mapping time is logarithmic, in hashing collisions occur because the prior knowledge of the table size becomes a problem in databases, which grow in unpredictable ways, and nally, direct mapping, which despite having the smallest mapping time of the three approaches, does not allow the relocation of all pages in the database. Another logical approach is used in the Multimedia Data Manager Kernel (NUGEM), which is being developed by the Database Group at IBILCE, S~ao Jos e do Rio Preto, S~ao Paulo, Brazil, in order to manage non-conventional data. This work proposes a new structure and features for the OID management, which was tested in NUGEM, featuring a new approach with a shorter handling time of OIDs, that can be named indirect mapping. It is also evaluated the exhaustion of OIDs, proving that this is not a problem for the indirect mapping approach. Compared with the approaches of the literature, the approach of indirect mapping has constant mapping time, uses less memory than the B-tree approach, has no problems like collisions and, unlike direct mapping, allows total exibility for relocating pages in the database

Le choix d'un système de gestion de bases de données (SGBD) et de plateforme d'exécution pour le déploiement est une tâche primordiale pour la satisfaction des besoins non-fonctionnels(comme la performance temporelle et la consommation d'énergie). La difficulté de ce choix explique la multitude de tests pour évaluer la qualité des bases de données (BD) développées.Cette évaluation se base essentiellement sur l'utilisation des métriques associées aux besoins non fonctionnels. En effet, une mine de tests existe couvrant toutes les phases de cycle de vie de conception d'une BD. Les tests et leurs environnements sont généralement publiés dans des articles scientifiques ou dans des sites web dédiés comme le TPC (Transaction Processing Council).Par conséquent, cette thèse contribue à la capitalisation et l'exploitation des tests effectués afin de diminuer la complexité du processus de choix. En analysant finement les tests, nous remarquons que chaque test porte sur les jeux de données utilisés, la plateforme d'exécution, les besoins non fonctionnels, les requêtes, etc. Nous proposons une démarche de conceptualisation et de persistance de toutes .ces dimensions ainsi que les résultats de tests. Cette thèse a donné lieu aux trois contributions. (1) Une conceptualisation basée sur des modélisations descriptive,prescriptive et ontologique pour expliciter les différentes dimensions. (2) Le développement d'un entrepôt de tests multidimensionnel permettant de stocker les environnements de tests et leurs résultats. (3) Le développement d'une méthodologie de prise de décision basée sur un système de recommandation de SGBD et de plateformes. / Choosing appropriate database management systems (DBMS) and/or execution platforms for given database (DB) is complex and tends to be time- and effort-intensive since this choice has an important impact on the satisfaction of non-functional requirements (e.g., temporal performance or energy consumption). lndeed, a large number of tests have been performed for assessing the quality of developed DB. This assessment often involves metrics associated with non-functional requirement. That leads to a mine of tests covering all life-cycle phases of the DB's design. Tests and their environments are usually published in scientific articles or specific websites such as Transaction Processing Council (TPC). Therefore, this thesis bas taken a special interest to the capitalization and the reutilization of performed tests to reduce and mastery the complexity of the DBMS/platforms selection process. By analyzing the test accurately, we identify that tests concem: the data set, the execution platform, the addressed non-functional requirements, the used queries, etc. Thus, we propose an approach of conceptualization and persistence of all dimensions as well as the results of tests. Conseguently, this thesis leads to the following contributions. (1) The design model based on descriptive, prescriptive and ontological concepts to raise the different dimensions. (2) The development of a multidimensional repository to store the test environments and their results. (3) The development of a decision making methodology based on a recommender system for DBMS and platforms selection.

M. Ing. / Change management is a crucial component in the daily activities inherent within the civil engineering industry, which if used effectively, can mitigate the impact of project scope changes and be beneficial to the life cycle of the a project. Project scope changes regularly occur during the construction phase of civil engineering projects, and normally result in severe disruptions. Although scope changes during the implementation phase are avoided where possible, their existence is a practical reality. These changes can be managed by a Change Management System (CMS) that can minimize the impact of these changes. This dissertation investigates different change management systems, which are used by major companies to mitigate the impact of project scope changes. Case studies on change management systems are provided, and the advantages and disadvantages of these systems are discussed. These change management systems were used as a reference axis for a newly proposed project scope change management system, namely the Proactive Change Management Database System (PACMDS). The PACMDS was developed by the author, to minimize the impact of scope changes in civil projects. In a survey done to establish whether the PACMDS would be effective in mitigating project scope changes, it was found that the PACMDS would assist in managing project scope changes effectively.

Increased production and use of multimedia data has led to the development of a more advanced Database Management System (DBMS), like an Object Relational Database Management System (ORDBMS). These advanced databases are necessitated by the complexity in structure and the functionality required by multimedia data. Unfortunately, no suitable benchmarks exist with which to test the performance of databases when handling multimedia data. This thesis describes the design of a benchmark to measure the performance of basic functionality found in multimedia databases. The benchmark, called MORD (Multimedia Object Relational Databases), targets Oracle, a well known commercial Object Relational Database Management System (ORDBMS) that can handle multimedia data. Although MORD targets Oracle, it can easily be applied to other Multimedia Database Management System (MMDBMS) as a result of a design that stressed its portability, and simplicity. MORD consists of a database schema, test data, and code to simulate representative queries on multimedia databases. A number of experiments are described that validate MORD and ensure its correct design and that its objectives are met. A by-product of these experiments is an initial understanding of the performance of multimedia databases. The experiments show that with multimedia data the buffer cache should be at least large enough to hold the largest dataset, a bigger block size improves the performance, and turning off logging and caching for bulk loading improves the performance. MORD can be used to compare different ORDBMS or to assist in the configuration of a specific database.