Information technology plays an increasingly important part in representing, managing, and driving the field of sustainable energy. However, current paradigms for representing much of this information can be fragmented, singular, and extremely domain focused. Linkage with wider concepts, for example between energy supply and demand data, can be minimal. This dissertation investigates ways in which such data linkages can be expanded upon, applying the latest concepts of Semantic Web technology to the area. This dissertation examines the role of the Semantic Web in representing information relevant to sustainable energy, with a particular focus on energy policy, energy supply, and the demands of the built environment. An approach for representing such information is outlined in the dissertation, which introduces new ontologies for representing energy policy and building information data and methodologies for modelling such data. Existing ontologies for representing energy supply are discussed, as are common connections between these areas and a server platform for knowledge storage and presentation. Additionally, some focus is directed towards the usability and accessibility of such data and the implementation of proof of concept applications targeted at specific areas within sustainable energy are presented. Using the outlined approach, energy information can be interlinked to allow multilevel data navigation from international policy data, through energy infrastructure, to individual energy demands, and ultimately to extremely detailed building component levels of granularity. Such data can be interlinked into wider linked data initiatives, increasing usefulness and expanding the scope for increased analysis. The implications of the outlined approach are discussed and evaluated with regard to various identified use cases requiring different levels of data granularity, in addition to impact on the wider domain of information management. This dissertation demonstrates, at a proof of concept level, that Semantic Web technology can be of significant benefit across the domain of sustainable energy.

A integração e a troca de informação na Web é uma necessidade reconhecida no mundo atual nos mais diversos domínios de aplicação. A Web constitui hoje o principal meio de acesso a informações, disponibilizando várias formas de acesso a esse conteúdo. No entanto, disponibilizar essa informação de forma integrada para o usuário ainda é um desafio. O Serviço Web Semântico é uma tecnologia de Inteligência Artificial promissora a qual pode contribuir significativamente com o processo de integração tanto de dados como de processos disponíveis em portais Web. O objetivo do presente trabalho é realizar a integração das diferentes funcionalidades oferecidas por portais disponibilizados na Web, fazendo uso de ontologias de domínio através de Serviços Web Semânticos. Para alcançar o objetivo proposto, um sistema de busca e integração, denominado W-KIS, foi projetado e desenvolvido, fornecendo uma visão unificada e transparente ao usuário de toda informação disponível em um determinado número de portais Web, contextualizados a um domínio de aplicação específico. Dois domínios de aplicação foram explorados neste trabalho através de dois estudos de caso: a Biomedicina e a Geologia, cada um fornecendo sua própria ontologia de domínio. Serviços Web Semânticos foram desenvolvidos para encapsular portais Web em ambos os domínios, permitindo ao sistema realizar a integração em cada domínio, separadamente. A aplicação proposta foi validada através da avaliação de sua atuação em ambos os domínios, além de testes com usuários finais e comparações com outras abordagens para integração na Web e motores de busca tradicionais. Através da utilização de Serviços Web Semânticos, W-KIS é capaz de localizar e executar os serviços de forma automática, além de lidar com a heterogeneidade semântica entre as fontes de informação por meio das ontologias de domínio. Desta forma, o sistema fornece apenas informação contextualizada ao interesse do usuário. / Integration and exchange of information across the Internet is a universally recognized need, in a wide variety of domains. Nowadays, the Web is the main source of information, providing many ways to access this content. However, making this information available to the user in an integrated way is still a challenge. New promising application fields such as Semantic Web Services can improve the integration process of data and processes available in Web portals. The goal of the present research is to accomplish the integration of the different functionalities offered by portals available in the Web, making use of domain ontologies through Semantic Web Services. In order to achieve the proposed goal, a search and integration system, named W-KIS, was designed and developed. The system provides the user with a unified and transparent view of the whole information available in a number of Web portals, related to a specific application domain. Two application domains were explored in this work through two case studies: the Biomedicine and the Geology, each one providing its own domain ontology. Semantic Web Services were designed to encapsulate Web portals in both domains, allowing the system to accomplish the integration in each domain, separately. The application proposed in this project was validated through the evaluation of its behavior in both domains, besides other tests with final users and comparisons with other approaches for integration in the Web, and traditional search engines. Making use of Semantic Web Services, W-KIS is able to find and execute the services in an automatic way, besides dealing with the semantic heterogeneity in the information sources by means of domain ontologies. This way, the system provides the user with information contextualized to its interest.

abstract: Computational tools in the digital humanities often either work on the macro-scale, enabling researchers to analyze huge amounts of data, or on the micro-scale, supporting scholars in the interpretation and analysis of individual documents. The proposed research system that was developed in the context of this dissertation ("Quadriga System") works to bridge these two extremes by offering tools to support close reading and interpretation of texts, while at the same time providing a means for collaboration and data collection that could lead to analyses based on big datasets. In the field of history of science, researchers usually use unstructured data such as texts or images. To computationally analyze such data, it first has to be transformed into a machine-understandable format. The Quadriga System is based on the idea to represent texts as graphs of contextualized triples (or quadruples). Those graphs (or networks) can then be mathematically analyzed and visualized. This dissertation describes two projects that use the Quadriga System for the analysis and exploration of texts and the creation of social networks. Furthermore, a model for digital humanities education is proposed that brings together students from the humanities and computer science in order to develop user-oriented, innovative tools, methods, and infrastructures. / Dissertation/Thesis / Doctoral Dissertation Biology 2014

abstract: The processing of large volumes of RDF data require an efficient storage and query processing engine that can scale well with the volume of data. The initial attempts to address this issue focused on optimizing native RDF stores as well as conventional relational databases management systems. But as the volume of RDF data grew to exponential proportions, the limitations of these systems became apparent and researchers began to focus on using big data analysis tools, most notably Hadoop, to process RDF data. Various studies and benchmarks that evaluate these tools for RDF data processing have been published. In the past two and half years, however, heavy users of big data systems, like Facebook, noted limitations with the query performance of these big data systems and began to develop new distributed query engines for big data that do not rely on map-reduce. Facebook's Presto is one such example. This thesis deals with evaluating the performance of Presto in processing big RDF data against Apache Hive. A comparative analysis was also conducted against 4store, a native RDF store. To evaluate the performance Presto for big RDF data processing, a map-reduce program and a compiler, based on Flex and Bison, were implemented. The map-reduce program loads RDF data into HDFS while the compiler translates SPARQL queries into a subset of SQL that Presto (and Hive) can understand. The evaluation was done on four and eight node Linux clusters installed on Microsoft Windows Azure platform with RDF datasets of size 10, 20, and 30 million triples. The results of the experiment show that Presto has a much higher performance than Hive can be used to process big RDF data. The thesis also proposes an architecture based on Presto, Presto-RDF, that can be used to process big RDF data. / Dissertation/Thesis / Masters Thesis Computing Studies 2014

Submitted by Valquíria Barbieri (kikibarbi@hotmail.com) on 2017-11-07T20:58:42Z No. of bitstreams: 1 DISS_2012_Luzo Vinicius Pedroso Reis.pdf: 29664723 bytes, checksum: 82045e3b6a8556861ffadbeadc17eba7 (MD5) / Approved for entry into archive by Jordan (jordanbiblio@gmail.com) on 2017-12-15T14:44:30Z (GMT) No. of bitstreams: 1 DISS_2012_Luzo Vinicius Pedroso Reis.pdf: 29664723 bytes, checksum: 82045e3b6a8556861ffadbeadc17eba7 (MD5) / Made available in DSpace on 2017-12-15T14:44:30Z (GMT). No. of bitstreams: 1 DISS_2012_Luzo Vinicius Pedroso Reis.pdf: 29664723 bytes, checksum: 82045e3b6a8556861ffadbeadc17eba7 (MD5) Previous issue date: 2012-10-16 / Este trabalho analisa como os tutoriais que circulam no ciberespaço engendram novos espaços-tempo de aprendizagem e com isso contribuem para a produção do mundo contemporâneo, com suas características de velocidade e movimento. Para isso, entendemos o tutorial como uma técnica, como uma dimensão da constituição humana. Assim como a escola tradicional produz um determinado tipo de sociedade (a sociedade disciplinar), as atuais técnicas contemporâneas da educação a distância e, de modo mais radical, os tutoriais participam da produção da sociedade de controle aberto e contínuo. Por trás desses modos de ensinar, estão objetos técnicos que operam em sua relação com os sujeitos ora o confinamento e a disciplina (escola tradicional), ora o movimento controlado (tutoriais), especialmente hoje com a chamada web semântica. Nesse sentido, partimos da observação do funcionamento de máquinas e sujeitos para entender o atual agenciamento homem-técnica. Um agenciamento em que se destaca o movimento, os procedimentos de filtragem e busca por informações. / This work analyzes how the tutorials in cyberspace engender new learning spaces and times and contribute to the production of the contemporary world, with their velocity and movement. In this sense, we understand a tutorial as a technique, as a human dimension. As traditional school produces a certain kind of society (The disciplinary society), the current techniques of distance education (e-learning) and on a radical way, tutorials participate in the production of the society of open and continuous control. Underneath these tuition procedures there are technical objects that operate, in relation to subjects either confinement and discipline (traditional school), or controlled movement (tutorials from cyberspace), specially now with semantic web. We start from the observation of the operation of the machines and subjects to understand the actual human-technique agency. An agency that highlights the characteristics of movement, the filtering procedures and the search for information.

Dans cette thèse, nous étudions plusieurs approches destinées à aider les utilisateurs à trouver des informations utiles et fiables dans le Web de données, en utilisant les technologies du Web sémantique. Nous abordons pour cela deux thèmes de recherche: le liage de données dans le Linked-Data et la confiance dans les réseaux P2P sémantiques. Nous modélisons le problème de liage dans le Web de données comme un problème de raisonnement sur des données incomplètes, qu'il s'agit d'enrichir en interrogeant de façon précise et pertinente le cloud du Linked Data. Nous avons conçu et implémenté un nouvel algorithme qui, à partir d'une requête de liage (du type et d'une base de règles modélisant de manière uniforme diverses connaissances du domaine (contraintes du schéma, axiomes d'inclusion ou d'exclusion d'une ontologie, règles expertes, mappings), construit itérativement des requêtes SPARQL pour importer des sources externes pertinentes du Linked Data les données utiles pour répondre à la requête de liage. Les expérimentations que nous avons menées sur des données réelles ont démontré la faisabilité de cette approche et son utilité dans la pratique pour le liage de données et la résolution d'homonymie. En outre, nous proposons une adaptation de cette approche pour prendre en compte des données et des connaissances éventuellement incertaines, avec en résultat l'inférence de liens ‘sameAs' et ‘differentFrom' associés à des poids de probabilité. Dans cette adaptation nous modélisons l'incertitude comme des valeurs de probabilité. Nos expérimentations ont montré que notre approche passe à l'échelle pour des bases de connaissances constituées de plusieurs millions de faits RDF et produit des poids probabilistes fiables. Concernant la confiance, nous introduisons un mécanisme de confiance permettant de guider le processus de réponse aux requêtes dans des Réseaux P2P sémantiques. Les différents pairs dans les réseaux P2P sémantiques organisent leur information en utilisant des ontologies distinctes et d épendent d'alignements entre ontologies pour traduire leurs requêtes. La notion de confiance dans un tel contexte est subjective ; elle estime la probabilité qu'un pair apportera des réponses satisfaisantes pour les requêtes spécifiques dans les interactions futures. Le mécanisme proposé de calcul de valeurs de confiance combine les informations fournies par les alignements avec celles provenant des interactions passées entre pairs. Les valeurs de confiances calculées sont affinées progressivement à chaque cycle de requête/réponse en utilisant l'inférence bayésienne. Pour l'évaluation de notre mécanisme, nous avons construit un système P2P de partage de signets sémantiques (TrustMe) dans lequel il est possible de faire varier différents paramètres quantitatifs et qualitatifs. Les résultats expérimentaux montrent la convergence des valeurs de confiance ;.ils mettent également en évidence le gain en terme de qualité des réponses des pairs - mesurées selon la précision et le rappel- lorsque le processus de réponse aux requêtes est guidé par notre mécanisme de confiance. / In this thesis we investigate several approaches that help users to find useful and trustful informationin the Web of Data using the Semantic Web technologies. In this purpose, we tackle tworesearch issues: Data Linkage in Linked Data and Trust in Semantic P2P Networks. We model the problem of data linkage in Linked Data as a reasoning problem on possibly decentralized data. We describe a novel Import-by-Query algorithm that alternates steps of subquery rewriting and of tailored querying the Linked Data cloud in order to import data as specific as possible for inferring or contradicting given target same-as facts. Experiments conducted on real-world datasets have demonstrated the feasibility of this approach and its usefulness in practice for data linkage and disambiguation. Furthermore, we propose an adaptation of this approach to take into account possibly uncertain data and knowledge, with a result the inference of same-as and different-from links associated with probabilistic weights. In this adaptation we model uncertainty as probability values. Our experiments have shown that our adapted approach scales to large data sets and produces meaningful probabilistic weights. Concerning trust, we introduce a trust mechanism for guiding the query-answering process in Semantic P2P Networks. Peers in Semantic P2P Networks organize their information using separate ontologies and rely on alignments between their ontologies for translating queries. Trust is such a setting is subjective and estimates the probability that a peer will provide satisfactory answers for specific queries in future interactions. In order to compute trust, the mechanism exploits the information provided by alignments, along with the one that comes from peer's experiences. The calculated trust values are refined over time using Bayesian inference as more queries are sent and answers received. For the evaluation of our mechanism, we have built a semantic P2P bookmarking system (TrustMe) in which we can vary different quantitative and qualitative parameters. The experimental results show the convergence of trust, and highlight the gain in the quality of peers' answers —measured with precision and recall— when the process of query answering is guided by our trust mechanism.

Orientador: Rogério Eduardo Garcia / Banca: Maria Cristina Ferreira de Oliveira / Banca: Celso Olivete Júnior / Resumo: A exploração de múltiplas visões de conjuntos de dados pode auxiliar no processo de descoberta de associações ocultas entre os elementos desses conjuntos. O emprego de uma técnica de coordenação permite associar esses elementos em visões distintas. Normalmente, técnicas de coordenação que mapeiam os elementos apenas considerando os atributos dos dados limitam a descoberta de associações entre os elementos visuais. Neste trabalho é apresentada a técnica Semantic Coordination, um mecanismo de coordenação que utiliza ontologias para associar elementos de dados com base nos mapeamentos fornecidos pela semântica. Com isso, a visualização exploratória pode se beneficiar da representação semântica para criar os mapeamentos, pois são estabelecidas associações entre os elementos visuais além das existentes pela relação dos dados. Também, são apresentados e analisados os resultados de dois estudos de caso, aplicando a Semantic Coordination e técnicas tradicionais na exploração de coleções textuais. Após a análise, os resultados obtidos mostram que o mecanismo de coordenação proposto estende o processo de exploração de múltiplas visões de conjuntos de dados, em comparação a técnicas de coordenação tradicionais / Abstract: Exploring multiple views of data sets can provide assistance on discovering unforeseen associations among their elements. Employing a coordination technique creates associations among elements across related views. Usually, the coordination process uses a mapping among data elements, that are constrained by using data attributes, which is a constraint to user exploration process (only linked data can be usually associated). In this work it is presented Semantic Coordination, a mechanism that applies ontologies to link data elements based on mappings provided by semantic. Representing the underlying data into ontology, the exploratory visualization can be benefited from a semantic representation to create the mappings, what might be helpful to establish associations on multiple views. Also, it is shown the results of two case studies applying the Semantic Coordination and traditional techniques on exploring text collections. The analysis of the obtained results, have shown the proposed coordination mechanism improves the exploration using multiple views, in comparison with traditional techniques / Mestre

Orientador: Plácida Leopoldina Ventura Amorim da Costa Santos / Resumo: Com o desenvolvimento das Tecnologias de Informação e Comunicação, a Ciência da Informação precisou repensar a postura tradicional de gerenciamento informacional. O hipertexto, advindo principalmente por meio do ambiente Web, elevou ainda mais a complexidade do tratamento informacional. A popularização da Internet fez com que a Web se tornasse mais interativa e colaborativa tornando comum a utilização de sistemas wiki para o gerenciamento informacional. Idealizada por Tim Berners-Lee, surge a iniciativa da Web Semântica, na qual as máquinas são capazes de analisar os dados presentes na rede. Nesse contexto aparecem os wikis semânticos, caracterizados por fazerem uso das tecnologias da Web Semântica. Diante desse cenário, considera-se que a Ciência da Informação, que já se preocupa com o desenvolvimento da Web e da Web Semântica, deve também se preocupar com os wikis semânticos. Assim, por meio de uma pesquisa descritiva e exploratória, objetivou-se explorar, apresentar e descrever as características dos wikis semânticos com enfoque nas atividades de descrição, de recuperação e de intercâmbio de informações apoiadas em tecnologias da Web Semântica, visando a favorecer o entendimento, a discussão e o uso dessas tecnologias em ambientes informacionais digitais. Após uma apresentação sobre as raízes históricas da Web Semântica, são destacados os padrões de representação, codificação, descrição, relação e consulta de dados estruturados (URI, XML, RDF, RDFS, OWL e SPARQL) que, junto a outras tecnologias, formam a base da Web Semântica e apoiam o funcionamento dos wikis semânticos. Os wikis semânticos são apresentados e definidos como sistemas wiki que se utilizam de tecnologias da Web Semântica para incorporar conhecimento formalizado, conteúdo, estruturas e links em suas páginas. Após essa etapa, são descritas as principais atividades de descrição, de recuperação e de intercâmbio de... / Abstract: Due to the development of Information and Communication Technologies, Information Science has been forced to rethink the traditional posture of information management. Hypertext, arising mainly through the Web environment, further increased the complexity of the information handling. The popularization of the Internet has led the Web to a more interactive and a more collaborative environment, bringing wiki systems, for example, to manage information in a collaborative way. Conceived by Tim Berners-Lee, there is the Semantic Web initiative in which machines are able to analyze data on the network. In this context, semantic wikis arise: wikis characterized by the use of Semantic Web technologies. Therefore, we believe that Information Science, which cares about the development of the Web and the Semantic Web, should also care about semantic wikis. Thus, by a descriptive and an exploratory research, the objective is to explore, to present and to describe the characteristics of the semantic wikis on the activities of representation, retrieval and exchange of information supported by Semantic Web technologies in order to facilitate the understanding, the discussion, and the use of these technologies in digital information environments. After a presentation about the origins of the Semantic Web, we highlight the data representation, encoding, description, relation, and query standards (URI, XML, RDF, RDFS, OWL and SPARQL) which, with other technologies, form the basis of the Semantic Web and support the functioning of semantic wikis. Semantic wikis are presented and defined as wiki systems that use Semantic Web technologies in order to incorporate formalized knowledge, content, structure and links on their pages. After that, we describe the main activities for information description, retrieval and interchange on Semantic MediaWiki, the most popular and most used semantic wiki engine so far. As conclusion, we consider that semantic wikis can... / Mestre

Submitted by JESSICA OLIVEIRA DE SOUZA null (osz.jessica@gmail.com) on 2017-06-09T12:04:24Z No. of bitstreams: 1 Dissertação-Jessica-Melo.pdf: 5257476 bytes, checksum: 21d6468b47635a4df09d971c6c0bb581 (MD5) / Approved for entry into archive by LUIZA DE MENEZES ROMANETTO (luizamenezes@reitoria.unesp.br) on 2017-06-12T12:21:39Z (GMT) No. of bitstreams: 1 melo_josf_me_mar.pdf: 5257476 bytes, checksum: 21d6468b47635a4df09d971c6c0bb581 (MD5) / Made available in DSpace on 2017-06-12T12:21:39Z (GMT). No. of bitstreams: 1 melo_josf_me_mar.pdf: 5257476 bytes, checksum: 21d6468b47635a4df09d971c6c0bb581 (MD5) Previous issue date: 2017-05-09 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / A Web Semântica sugere a utilização de padrões e tecnologias que atribuem estrutura e semântica aos dados, de modo que agentes computacionais possam fazer um processamento inteligente, automático, para cumprir tarefas específicas. Neste contexto, foi criado o projeto Linked Open Data (LOD), que consiste em uma iniciativa para promover a publicação de dados linkados (Linked Data). Com o evidente crescimento dos dados publicados como Linked Data, a qualidade tornou-se essencial para que tais conjuntos de dados (datasets) atendam os objetivos básicos da Web Semântica. Isso porque problemas de qualidade nos datasets publicados constituem em um empecilho não somente para a sua utilização, mas também para aplicações que fazem uso de tais dados. Considerando que os dados disponibilizados como Linked Data possibilitam um ambiente favorável para aplicações inteligentes, problemas de qualidade podem também dificultar ou impedir a integração dos dados provenientes de diferentes datasets. A literatura aplica diversas dimensões de qualidade no contexto do Linked Data, porém indaga-se a aplicabilidade de tais dimensões para avaliação de qualidade de dados linkados. Deste modo, esta pesquisa tem como objetivo propor uma metodologia para avaliação de qualidade nos datasets de Linked Data, bem como estabelecer um modelo do que pode ser considerado qualidade de dados no contexto da Web Semântica e do Linked Data. Para isso adotou-se uma abordagem exploratória e descritiva a fim de estabelecer problemas, dimensões e requisitos de qualidade e métodos quantitativos na metodologia de avaliação a fim de realizar a atribuição de índices de qualidade. O trabalho resultou na definição de sete dimensões de qualidade aplicáveis ao domínio do Linked Data e 14 fórmulas diferentes para a quantificação da qualidade de datasets sobre publicações científicas. Por fim realizou-se uma prova de conceito na qual a metodologia de avaliação de qualidade proposta foi aplicada em um dataset promovido pelo LOD. Conclui-se, a partir dos resultados da prova de conceito, que a metodologia proposta consiste em um meio viável para quantificação dos problemas de qualidade em datasets de Linked Data, e que apesar dos diversos requisitos para a publicação deste tipo de dados podem existir outros datasets que não atendam determinados requisitos de qualidade, e por sua vez, não deveriam estar inclusos no diagrama do projeto LOD. / The Semantic Web suggests the use of patterns and technologies that assign structure and semantics to the data, so that computational agents can perform intelligent, automatic processing to accomplish specific tasks. In this context, the Linked Open Data (LOD) project was created, which consists of an initiative to promote the publication of Linked Data. With the evident growth of data published as Linked Data, quality has become essential for such datasets to meet the basic goals of the Semantic Web. This is because quality problems in published datasets are a hindrance not only to their use but also to applications that make use of such data. Considering that data made available as Linked Data enables a favorable environment for intelligent applications, quality problems can also hinder or prevent the integration of data coming from different datasets. The literature applies several quality dimensions in the context of Linked Data, however, the applicability of such dimensions for quality evaluation of linked data is investigated. Thus, this research aims to propose a methodology for quality evaluation in Linked Data datasets, as well as to establish a model of what can be considered data quality in the Semantic Web and Linked Data context. For this, an exploratory and descriptive approach was adopted in order to establish problems, dimensions and quality requirements and quantitative methods in the evaluation methodology in order to perform the assignment of quality indexes. This work resulted in the definition of seven quality dimensions applicable to the Linked Data domain and 14 different formulas for the quantification of the quality of datasets on scientific publications. Finally, a proof of concept was developed in which the proposed quality assessment methodology was applied in a dataset promoted by the LOD. It is concluded from the proof of concept results that the proposed methodology consists of a viable means for quantification of quality problems in Linked Data datasets and that despite the diverse requirements for the publication of this type of data there may be other datasets that do not meet certain quality requirements, and in turn, should not be included in the LOD project diagram.

Made available in DSpace on 2014-06-11T19:29:40Z (GMT). No. of bitstreams: 0 Previous issue date: 2011-02-04Bitstream added on 2014-06-13T18:59:21Z : No. of bitstreams: 1 derossi_lh_me_sjrp.pdf: 1783577 bytes, checksum: 3017df90989f2b4b05b6bc1d87eb42ce (MD5) / Nas últimas duas décadas, diversos sistemas de ensino para Web foram desenvolvidos e amplamente utilizados. Com o surgimento da Web Semântica, a qual integra os conceitos da Web e da área de Inteligência Artificial, foi possível conceber novos sistemas de ensino, os Sistemas de Ensino Baseados na Web Semântica (Semantic Web-based Educational Systems - SWBES). Os SWBES são caracterizados por representar o conteúdo e o conhecimento na forma de ontologias e contar com agentes inteligentes que permitem às máquinas entenderem o conteúdo publicado na Web. Neste trabalho são apresentados e discutidos os padrões arquiteturais utilizados para desenvolvimento dos sistemas baseados na Web Semântica, em especial os sistemas educacionais. Como resultado do estudo é proposta uma arquitetura para os SWBES. As arquiteturas dos SWBES que utilizam a abordagem multiagente, normalmente, são concebidas de forma empírica, ou seja, apenas com a experiência do seu desenvolvedor. Neste trabalho, para a concepção do modelo multiagentes adotado na arquitetura do SWBES, foi utilizada uma metodologia de desenvolvimento de sistemas multiagentes, o framework Tropos. São apresentadas as fases do framework Tropos para o projeto de um modelo multiagentes utilizado na arquitetura de um SWBES. Utilizando a arquitetura desenvolvida foi implementado um SWBES. Para estudo de caso dessa arquitetura, foi realizada uma implementação de um Sistema de Ensino de Rochas Carbonáticas. Para esse sistema foram definidas três ontologias: ontologia de Domínio de Rochas Carbonáticas, ontologia de Perfil e ontologia de Material de Aprendizagem. Os comportamentos dos agentes e as interações entre eles são detalhadas. Finalmente, é apresentado um estudo de casos da utilização do sistema no ensino do curso de Rochas Carbonáticas / In the past two decades, several Web educational systems were designed and widely used. With the appearing of the Semantic Web, which integrates Web and Artificial Intelligence area concepts, it was possible to conceive new educational concepts, the Educational Systems Based on the Semantic Web (Semantic Web-based Educational Systems - SWBES). The SWBES represent the content and the knowledge in the form of ontology and count on intelligent agents that allow machines understand the content published on the web. In this work, architectural standards used to develop new systems based on the Semantic Web will be presented and discussed, especially the educational ones. As a result of this study, it is proposed the architecture for the SWBES. The architecture for the SWBES that use multi-agent approach generally is conceived empirically, that is, only with its developer experience. In this work, for the conception of the multi-agents model adopted in the architecture of the SWBES, a methodology of multi-agents systems development was used, the Tropos framework. The phases of Tropos framework for the project of a multi-agent model used in the architecture of SWBES are presented. A SWBES was implemented by using the architecture that was developed. For the study case of this architecture, it was held an implementation of the Educational System of Carbonate Rock. For this system, three ontologies were defined: Domain Ontology of Carbonate Rocks, Profile Ontology and Learning Material Ontology. The behavior of agents and interaction among them are detailed