Protocolo OAI-PMH e Sistemas Federados de Informação │ OAI-PMH protocol and Federated Information Systems

Resumo O movimento de Arquivos Abertos através do protocolo OAI-PMH tem facilitado a criação de federações de bibliotecas digitais que permitem ampla agregação de produção científica, criando condições para a construção de importantes bases de dados para pesquisa bibliométrica, cientométrica, análise de redes sociais e outros modos de análise. Como um meio privilegiado para avaliação e identificação de padrões de produção da informação em diferentes comunidades científicas, as bibliotecas digitais federadas podem ser utilizadas como fontes de dados para pesquisa científica. A construção dessas bases de dados envolve diversos passos iniciando pelo modo com que a informação deve ser estruturada pelas bibliotecas locais, a forma como deve ser compartilhada em um ambiente de rede e sintetizada de modo a garantir visualizações integradas da informação. Questões relevantes sobre a qualidade dos metadados também são abordadas do ponto de vista metodológico e do modo como impactam as possibilidades de análise científica. Apresentamos neste artigo os fundamentos de como essa arquitetura de informação foi aplicada na construção do portal Univerciencia.org, com foco específico na produção científica da área da Ciências da Comunicação, bem como a tecnologia utilizada, as formas de estruturação e tratamento dos metadados, além dos primeiros resultados da análise de dados coletados e sua qualidade.Palavras-chave bibliotecas digitais, sistemas federados, OAI-PMH, repositórios abertos, cientometria, bibliometriaAbstract The Open Archives Initiative movement using the OAI-PMH protocol has allowed the production of federated digital libraries that provide wide aggregation of scientific production, creating conditions for the construction of important databases that enable the development of bibliometric research, scientometric, social network analysis and other modes of analysis. As a privileged means for evaluating and identifying patterns of information production in different scientific communities, federated digital libraries can be used as a source to study these communities. The construction of these databases involves several steps, beginning with the way information should be structured by local libraries, shared in a network environment and synthesized in order to ensure an integrated view of information. Relevant questions on the quality of such data directly impact on the possibilities of scientific analysis. We present here the foundations of how that information architecture has been implemented in the construction of the portal Univerciencia.org, which is focused on the scientific production of Communication Science, as well as the technologies involved, the ways of structuring metadata and the first results of analyzes of the data collected and their metadata quality.Keywords digital libraries, federated systems, OAI-PMH open repositories, scientometrics, bibliometrics

Merging and Outsourcing Information Systems with UML

Businesses can change their business structure by merging with other companies or, on the other end of the spectrum, by smoothly outsourcing some of their business processes to other more specialized parties. In this paper we will concentrate on conceptual modelling of merging and outsourcing information systems. Merging of a collection of information systems will be defined as the construction of a global information system that contains exactly the functionality of the original collection of systems. Such global information systems are called federated information systems, when we wish to address the situation where the component systems are so-called legacy systems; i.e. systems that are given beforehand and which are to interoperate in an integrated single framework in which the legacy systems are to maintain as much as possible their respective autonomy. Two major problems in constructing federated information systems concern achieving and maintaining consistency and a uniform representation of the data on the global level of the federation. The process of creation of uniform representations of data is known as data extraction, whereas data reconciliation is concerned with resolving data inconsistencies. Outsourcing of an information system, on the other hand, will be defined as the handing over of part of the functionality of the original system to an outside party (the supplier). Such functionality typically involves one or more operations, where each operation satisfies certain input- and output requirements. These requirements will be defined in terms of the ruling service level agreements (SLAs). We will provide a formal means to ensure that the outsourcing relationship between outsourcing party and supplier, determined by a SLA, satisfies specific correctness criteria. Formal specifications as offered in this paper can prove their value in the setup and evaluation of outsourcing contracts. We shall describe a uniform semantic framework for specification of both federated and outsourced information systems based on the UML/OCL data model. In particular, we will show that we can represent so-called exact views in UML/OCL, providing the means to capture the duality relation between federating and outsourcing.

Extending Loosely Coupled Federated Information Systems Using Agent Technology

Different FIPA-compliant agent development platforms are available for developing multiagent systems. FIPA compliance ensures interoperability among agents across different platforms. Although most agent implementation platforms provide some form of white- and yellow-page functionalities to advertise and identify agent roles and descriptions, there are no clear architectural standards that define how an agent community can effortlessly adapt to operate in a federated information system (FIS) where new content sources are constantly added or changes are made to existing content sources. This chapter presents a framework based on the semantic Web vision to address extensibility in a loosely coupled FIS.

Merging and Outsourcing Information Systems with UML

Businesses can change their business structure by merging with other companies or, on the other end of the spectrum, by smoothly outsourcing some of their business processes to other more specialized parties. In this paper we will concentrate on conceptual modelling of merging and outsourcing information systems. Merging of a collection of information systems will be defined as the construction of a global information system that contains exactly the functionality of the original collection of systems. Such global information systems are called federated information systems, when we wish to address the situation where the component systems are so-called legacy systems; i.e. systems that are given beforehand and which are to interoperate in an integrated single framework in which the legacy systems are to maintain as much as possible their respective autonomy. Two major problems in constructing federated information systems concern achieving and maintaining consistency and a uniform representation of the data on the global level of the federation. The process of creation of uniform representations of data is known as data extraction, whereas data reconciliation is concerned with resolving data inconsistencies. Outsourcing of an information system, on the other hand, will be defined as the handing over of part of the functionality of the original system to an outside party (the supplier). Such functionality typically involves one or more operations, where each operation satisfies certain input- and output requirements. These requirements will be defined in terms of the ruling service level agreements (SLAs). We will provide a formal means to ensure that the outsourcing relationship between outsourcing party and supplier, determined by a SLA, satisfies specific correctness criteria. Formal specifications as offered in this paper can prove their value in the setup and evaluation of outsourcing contracts. We shall describe a uniform semantic framework for specification of both federated and outsourced information systems based on the UML/OCL data model. In particular, we will show that we can represent so-called exact views in UML/OCL, providing the means to capture the duality relation between federating and outsourcing.

Merging and Outsourcing Information Systems with UML

Businesses can change their business structure by merging with other companies or, on the other end of the spectrum, by smoothly outsourcing some of their business processes to other more specialized parties. In this paper we will concentrate on conceptual modelling of merging and outsourcing information systems. Merging of a collection of information systems will be defined as the construction of a global information system that contains exactly the functionality of the original collection of systems. Such global information systems are called federated information systems, when we wish to address the situation where the component systems are so-called legacy systems; i.e. systems that are given beforehand and which are to interoperate in an integrated single framework in which the legacy systems are to maintain as much as possible their respective autonomy. Two major problems in constructing federated information systems concern achieving and maintaining consistency and a uniform representation of the data on the global level of the federation. The process of creation of uniform representations of data is known as data extraction, whereas data reconciliation is concerned with resolving data inconsistencies. Outsourcing of an information system, on the other hand, will be defined as the handing over of part of the functionality of the original system to an outside party (the supplier). Such functionality typically involves one or more operations, where each operation satisfies certain input- and output requirements. These requirements will be defined in terms of the ruling service level agreements (SLAs). We will provide a formal means to ensure that the outsourcing relationship between outsourcing party and supplier, determined by a SLA, satisfies specific correctness criteria. Formal specifications as offered in this paper can prove their value in the setup and evaluation of outsourcing contracts. We shall describe a uniform semantic framework for specification of both federated and outsourced information systems based on the UML/OCL data model. In particular, we will show that we can represent so-called exact views in UML/OCL, providing the means to capture the duality relation between federating and outsourcing.