[pt] O direito ao esquecimento não é algo que surge com o advento da internet. No entanto, o crescente número de usuários que acessam a rede, além do desenvolvimento de novas tecnologias, potencializaram a disseminação do seu conceito. Alguns fatos chamaram a atenção para este tema, especialmente a nova regulamentação de proteção de dados pessoais da União Europeia, e o caso do Google Spain e Inc. vs. AEPD e Mario Costeja González, julgado pelo Tribunal de Justiça da União Europeia (TJUE). Diante das diversas controvérsias que surgem com o tema, principalmente a partir da decisão do TJUE, o presente trabalho enfrentará os desafios do direito ao esquecimento, desde a origem do seu conceito até a sua aplicação no contexto da internet, problematizando o que de fato é este direito, os fundamentos que os justificam e as obrigações que dele decorrem. A seguir, vão ser analisados casos sobre o direito ao esquecimento que foram julgados na Argentina, Colômbia e Peru, tendo como intuito identificar como que a questão está sendo apreciada no âmbito da América Latina. Serão investigadas as jurisprudências do STJ e de seis Tribunais de Justiça Estaduais (TJRJ, TJSP, TJDFT, TJPR, TJPB, TJAM) sobre o assunto, além dos projetos de lei nacionais que tentam regulamentar o direito ao esquecimento, para que se possa colocar em perspectiva o atual cenário brasileiro sobre o tema. Ao final, será explorada a seguinte hipótese: será que é necessário invocar a existência de um direito ao esquecimento no Brasil, ou é possível tutelar as pretensões que se utilizam de tal rótulo com direitos já previstos pelo ordenamento jurídico? / [en] The right to be forgotten is not a right that arises with the advent of the Internet. Nonetheless, the increasing number of users that accesses the net and the development of new technologies leveraged the dissemination of its concept. A number of events increased the awareness of this topic, such as the recent reform of the General Data Protection Regulation of the European Union, and the Google Spain and Inc. vs. AEPD and Mario Costeja González case, appreciated by the Court of Justice of the European Union (CJEU). Regarding the controversies that arises with this theme, especially with the CJEU decision, the present work has the objective of discussing the challenges of the right to be forgotten. It will explore the origin of its concept until its application to the Internet context, questioning what in fact is such right, the legal basis that such right is justified and the obligations arising therefrom. Furthermore, cases of the right to be forgotten that were judged in Argentina, Colombia and Peru will be analyzed in order to identify how such right is being appreciated within Latin America. Following this, the jurisprudence of the present subject of the STJ and six State Justice Courts (TJRJ, TJSP, TJDFT, TJPR, TJPB, TJAM) are to be discussed. Moreover, in order to better understand the Brazilian perspective, the national legislation drafts that aim to regulate the right to be forgotten will also be reviewed. Finally, we are going to stress the following hypothesis: is it necessary to invoke the existence of the right to be forgotten in Brazil or is it possible to guarantee protection of claims that use such designation with rights that are already foreseen in the local legal system?

Made available in DSpace on 2016-04-26T20:22:59Z (GMT). No. of bitstreams: 1 Pamela Gabrielle Meneguetti.pdf: 1096299 bytes, checksum: 33ba52216b79378ae9a3ac1f274fb671 (MD5) Previous issue date: 2014-08-11 / This paper aims at providing an analysis on new conflicts involving the right to information and the rights of personality, particularly after the advent and operation of search engines on the Web, with the purpose of, providing rational criteria to solve them through an appropriate technique able to settle collisions between fundamental rights. Accordingly, this paper examines in depth both the right to information and the rights of personality, particularly specifying their prima facie scope of protection and describing their key elements and historical aspects of their legal protection. The paper also addresses the search engines on the Web, by bounding their concept, the history of their development, their types and the stages involved in their operation factors whose analysis is relevant and must be considered for the solution of conflicts. Conclusively, it was observed that the restriction of links from the results of researches conducted by search engines on the Web as a way of safeguarding the rights of personality is a measure that should be balanced according to the circumstances involved in each particular case. Moreover, even when such a restriction is allowed under the circumstances of the case, the extension of the measure should be limited to specific links, under penalty of a disproportional restriction to the right of information / O presente estudo tem por objetivo examinar as novas colisões que se têm verificado entre o direito à informação e os direitos da personalidade, em especial após o advento e atuação dos mecanismos de busca na Web, com a finalidade de, ao final, fornecer critérios racionais para a sua solução, por meio de técnica adequada para a solução de colisões entre direitos fundamentais. Com base nesse objetivo e de acordo com a teoria dos princípios concebida por ROBERT ALEXY, este estudo examina em profundidade o âmbito de proteção prima facie do direito à informação e dos direitos da personalidade, traçando seus principais elementos e o histórico de sua tutela. O estudo também aborda os mecanismos de busca na Web, delimitando o seu conceito, o histórico de seu desenvolvimento, suas modalidades e as etapas envolvidas em seu funcionamento fatores cuja análise será relevante e considerada para a solução das colisões. Ao final, pôde-se verificar que a restrição de links do resultado de pesquisas com mecanismos de busca na Web como forma de tutela a direitos da personalidade é uma medida que deve ser sopesada de acordo com as circunstâncias envolvidas no caso concreto. Ademais, mesmo nas situações em que admitida tal restrição diante das circunstâncias do caso concreto, a extensão da medida deve ser limitada a links específicos, sob pena de afetar de forma desproporcional o direito à informação

Resumo: Quando fala-se de evolução da Web, poderia realmente ser mais apropriado falar de design inteligente. Com a Web se tornando a principal opção para quem produz e dissemina conteúdo digital, cada vez mais, as pessoas tomam a atenção para esse valioso repositório de conhecimento. Neste ambiente, os mecanismos de busca configuram-se em aplicativos populares, tornando-se intermediários entre os usuários e a miríade de informações, serviços e recursos disponíveis na grande rede. Neste sentido, o Webdesigner pode atuar de forma decisiva, proporcionando uma melhor resposta na classificação dos mecanismos de busca. A correta representação do conhecimento é a chave para a recuperação e para a disseminação efetiva de dados, de informação e de conhecimentos. Este trabalho apresenta um estudo que pode trazer um progresso relevante aos usuários desta grande rede, buscando apresentar uma ferramenta de domínio público que apoie a aplicação de técnicas de descrição semântica de informação na Web. No decorrer da pesquisa investigamos técnicas e metodologias capazes de otimizar a indexação dos Websites pelos mecanismos de busca, enfatizando a descrição do conteúdo nele presente, melhorando sua classificação e consequentemente colaborando com a qualidade na recuperação de informações realizadas por meio de mecanismos de buscas. Tais técnicas foram testadas em alguns Websites, obtendo resultado satisfatório, a partir de então a ferramenta foi implementada e submetida a usuários para sua validação, o resultado desta validação é apresentado demonstrando a viabilidade da ferramenta e enumeração de novas funcionalidades para trabalhos futuros / Abstract: When we speak of evolution of the Web, it might actually be more appropriate to speak of intelligent design. With the Web becoming the primary choice for those who produce and disseminate digital content , more people take attention to this valuable repository of knowledge. In this environment , search engines are configured in popular, becoming an intermediary between users and the myriad of information, service and resources available on the World Wide Web. In this sense, the Web designer can act decisively, providing a better response in the ranking of search engines. The correct representation of knowledge is the key to recovery and effective dissemination of data, information and knowledge. This paper presents a study that significant progress can bring a large network of users, seeking to present a public domain tool that supports the application of techniques for semantic description of Web information in the course of the research investigated techniques and methodologies that can optimize Website indexing by search engines, emphasizing the description of the content in it, improving your ranking and thus contributing to quality in information retrieval conducted through search engines. These techniques were tested on some websites, obtaining satisfactory results, since then the tool was implemented and submitted to users validation, the result of the validation is present demonstrating the feasibility of the tool and list of new features for future work / Orientador: João Fernando Marar / Coorientador: Ivan Rizzo Guilherme / Banca: Edson Costa de Barros Carvalho Filho / Banca: Antonio Carlos Sementille / Mestre

Submitted by Kamila Costa (kamilavasconceloscosta@gmail.com) on 2015-06-11T19:18:34Z No. of bitstreams: 1 Dissertação-Lídia L S de Carvalho.pdf: 837456 bytes, checksum: 620d89f05fc84dc2af7b89b6b6e587a0 (MD5) / Approved for entry into archive by Divisão de Documentação/BC Biblioteca Central (ddbc@ufam.edu.br) on 2015-06-15T17:53:19Z (GMT) No. of bitstreams: 1 Dissertação-Lídia L S de Carvalho.pdf: 837456 bytes, checksum: 620d89f05fc84dc2af7b89b6b6e587a0 (MD5) / Approved for entry into archive by Divisão de Documentação/BC Biblioteca Central (ddbc@ufam.edu.br) on 2015-06-15T17:57:19Z (GMT) No. of bitstreams: 1 Dissertação-Lídia L S de Carvalho.pdf: 837456 bytes, checksum: 620d89f05fc84dc2af7b89b6b6e587a0 (MD5) / Made available in DSpace on 2015-06-15T17:57:19Z (GMT). No. of bitstreams: 1 Dissertação-Lídia L S de Carvalho.pdf: 837456 bytes, checksum: 620d89f05fc84dc2af7b89b6b6e587a0 (MD5) Previous issue date: 2015-03-11 / CAPES - Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / Several research efforts have been conducted in the literature to develop methods to reduce the cost of query processing in search engines. This research aims to propose modifications to improve the performance of the block-Max WAND (BMW) algorithm, one of the most efficient algorithms proposed previously. The BMW algorithm uses heuristics to discard the documents entries at query processing, which makes it extremely fast. In this dissertation, we propose and evaluate additional heuristics to improve the perfomance of BMW and your variant BMW-CS in an attempt to both further reduces query processing times and the amount of memory required for processing queries. / Nos últimos anos, pesquisas relacionadas ao processamento de consultas em máquinas de busca têm sido realizadas com o objetivo de desenvolver métodos que reduzam o seu custo. Este trabalho visa propor modificações para melhorar o desempenho do algoritmo Block-Max WAND (BMW), um dos algoritmos mais eficientes propostos na literatura. O algoritmo BMW utiliza heurísticas para descartar documentos da resposta durante o processamento de consultas, o que torna sua execução extremamente veloz. Nesta dissertação, serão propostas e experimentadas modificações nas heurísticas de descarte de documentos e redução na quantidade de memória utilizada para processar consultas pelo algoritmo BMW e suas variantes, buscando-se assim ganhos de desempenho.

University of Macau / Faculty of Social Sciences and Humanities / Department of Communication

Στη διπλωματική μας παραθέτουμε ένα μεγάλο κομμάτι βιβλιογραφίας σχετικής με Τεχνικές Εξατομικευμένης Αναζήτησης στον Παγκόσμιο Ιστό. Επίσης περιγράφονται αναλυτικά οι εφαρμογές που κατασκευάστηκαν στην διπλωματική μας και οι τεχνολογίες που χρησιμοποιήθηκαν για την υλοποίησή τους. Συγκεκριμένα κατασκευάστηκε ένα σύστημα εξατομίκευσης, το οποίο καταγράφει την συμπεριφορά ενός χρήστη και στην συνέχεια του δίνει την δυνατότητα να εξατομικεύσει τα αποτελέσματα μιας μηχανής αναζήτησης(Google) με βάση την προηγούμενη συμπεριφορά του. Υλοποιήθηκαν τρεις διαφορετικοί αλγόριθμοι(ένας αλγόριθμος που βασίζεται στις ακμές του γράφου, ένας αλγόριθμος πρόσθεσης εικονικών κόμβων και ένας εκτεταμένος αλγόριθμος πρόσθεσης εικονικών κόμβων), οι οποίοι με βάση την προηγούμενη συμπεριφορά του χρήστη κατασκευάζουν ένα γράφημα συμπεριφοράς για αυτόν, το οποίο χρησιμοποιούν στη συνέχεια για να εξατομικεύσουν τα αποτελέσματα μιας ερώτησης. Οι αλγόριθμοι αυτοί καλούνται από το χρήστη μέσω του συστήματος εξατομίκευσης για μια ερώτηση και ταξινομούν τα αποτελέσματα της ερώτησης σύμφωνα με την νέα σειρά που υποδεικνύει το γράφημα συμπεριφοράς που έχει κατασκευαστεί για τον συγκεκριμένο χρήστη. Στη συνέχεια δίνονται αναλυτικά παραδείγματα λειτουργίας για κάθε αλγόριθμο, που αποδεικνύουν τους θεωρητικούς υπολογισμούς. Τέλος παρατίθενται βασικά κομμάτια του κώδικα που δημιουργήθηκε για την υλοποίηση του συστήματος εξατομίκευσης. / In our diplomatic we mention a big piece of bibliography relative with Techniques of Personalized Search in the World Wide Web. Also we describe analytically the applications that were manufactured in our diplomatic and the technologies that were used for their implementation. In particular, we construct a personalization system, which records the behaviour of user and then gives him the possibility of personalizing the results of a search engine (Google) with base his previous behaviour. They constructed three different algorithms(an algorithm that based on the edges of the graph, an algorithm that add virtual nodes and an extensive algorithm that add virtual nodes), which with base the previous behaviour of user construct a graph, which they use then in order to personalize the results of the query. These algorithms are called by the user via the personalization system for a query and they categorize the results of query according to the new order that indicates the graph of behaviour that has been constructed for the particular user. Then they are given analytic examples for each algorithm, which proves the theoretical calculations. Finally, they mentioned basic pieces of code that were created for the construction of the personalization system.

近年來，動漫(動畫、漫畫與線上遊戲)越來越受歡迎。全球資訊網也陸續出現收集大量包括故事情節、動漫角色、作者等動漫相關資訊的動漫網站。多數動漫網站都提供使用者文字檢索的功能，以搜尋動漫網站文字內容。但是動站網站若能提供根據文字與圖形風格來搜尋圖文內容，對於動漫使用者而言，將更為方便。圖文風格可能是漫畫人物的繪畫風格、動畫故事的敘事風格等等。 為了方便使用者以圖文風格進行搜尋相關資訊，本論文根據動漫關連探勘技術，研究並開發一個動漫網站的搜尋引擎：動漫魔鏡，以提供使用者根據圖文關連來搜尋動漫網站中風格相似的動漫資訊。本論文的搜索方法關聯了圖像特徵和文字特徵。首先，針對圖像特徵，由於動漫角色是動漫的靈魂，因此經過動漫臉部偵測後，我們以電腦視覺中的局部二值樣式(Local Binary Pattern, LBP)與灰階值分佈來抽取並表示動漫角色的臉部特徵。針對文字特徵，我們利用一般全文檢索技術來擷取文字特徵。接著，運用階層式分群技術將文字特徵與圖像特徵值轉換為文字與圖像關鍵詞。最後，以語意主題模型中的隱含狄利克雷分布(Latent Dirichlet Allocation, LDA)分析圖文關鍵詞的潛在語意，並據此計算動漫網頁之間的相似性。實驗結果顯示本論文所研發的風格搜尋，其效果優於其他三種基本作法。 / Animations, comics, and games (ACG) have become more and more popular in recent years. There exist many ACG web sites which contain lots of textual and visual information on stories, characters and authors of animations, comics and games. Most ACG web sites provide users text retrieval capability to search for textual contents. However, there is a need for users to search for textual and visual contents by styles. Examples of styles are drawing styles of comic characters, narrative styles of animation stories and so on. In order to help users to search for textual and visual contents by similar styles, this thesis investigates and develops a search engine, Comirror, for ACG web sites based on latent correlation between textual and visual contents. First, while facial styles of characters play important roles in ACG, after comic face detection, Local Binary Pattern (LBP) along with gray-value histogram is utilized to extract and represent the visual features. For the textual contents, traditional full-text indexing technique is employed to extract textual features. Then, hierarchical clustering is performed to quantize and transform the textual and visual features into textual and visual words. Finally, Latent Dirichlet Allocation (LDA) is utilized to discover the latent semantic correlation between visual and textual words. Experiments show that the developed approach performs better than the other baseline approaches.

The purpose of this mixed methods study was to investigate the Web search strategies used to retrieve information from the Web by postgraduate students at the University of KwaZulu-Natal, Pietermaritzburg campus in order to address the weaknesses of undergraduate students with regard to their Web searching strategies. The study attempted to determine the Web search tactics used by postgraduate students, the Web search strategies (i.e. combinations of tactics) they used, how they determined whether their searches were successful and the search tool they preferred. In addition, the study attempted to contribute toward building a set of best practices when searching the Web. The sample population consisted of 331 postgraduate students, yielding a response rate of 95%. The study involved a two-phased approach adopting a survey in Phase 1 and interviews in the Phase 2. Proportionate stratified random sampling was used and the population was divided into five mutually exclusive groups (i.e., postgraduate diploma, postgraduate certificate, Honours, Master’s and PhD). A pre-test was conducted with ten postgraduate students from the Pietermaritzburg campus. The study revealed that the majority of postgraduate students have been searching the Web for six years or longer and that most postgraduate students searched the Web for information from five to less than ten hours a week. Most respondents gained their knowledge on Web searching through experience and only a quarter of the respondents have been given formal training on Web searching. The Web searching strategies explored contribute to the best practices with regard to Web search strategies, as interviewees were selected based on the highest number of search tactics used and they have several years of searching experience. The study was also able to identify the most preferred Web search tool. It is envisaged that undergraduate students can potentially follow these search strategies to improve their information retrieval. This finding could also be beneficial to librarians in developing training modules that assist undergraduate students to use these Web search tools more efficiently. The final outcome of the study was an adaptation Bates’ (1979) model of Information Search Tactics to suit information searching on the Web. / Thesis (M.Com.)-University of KwaZulu-Natal, Pietermaritzburg, 2012.

Made available in DSpace on 2014-06-11T19:29:41Z (GMT). No. of bitstreams: 0 Previous issue date: 2011-08-03Bitstream added on 2014-06-13T20:59:57Z : No. of bitstreams: 1 garcia_lmls_me_sjrp.pdf: 6057674 bytes, checksum: a26fce0d239fd5ca19b1f04d3236faa6 (MD5) / Quando fala-se de evolução da Web, poderia realmente ser mais apropriado falar de design inteligente. Com a Web se tornando a principal opção para quem produz e dissemina conteúdo digital, cada vez mais, as pessoas tomam a atenção para esse valioso repositório de conhecimento. Neste ambiente, os mecanismos de busca configuram-se em aplicativos populares, tornando-se intermediários entre os usuários e a miríade de informações, serviços e recursos disponíveis na grande rede. Neste sentido, o Webdesigner pode atuar de forma decisiva, proporcionando uma melhor resposta na classificação dos mecanismos de busca. A correta representação do conhecimento é a chave para a recuperação e para a disseminação efetiva de dados, de informação e de conhecimentos. Este trabalho apresenta um estudo que pode trazer um progresso relevante aos usuários desta grande rede, buscando apresentar uma ferramenta de domínio público que apoie a aplicação de técnicas de descrição semântica de informação na Web. No decorrer da pesquisa investigamos técnicas e metodologias capazes de otimizar a indexação dos Websites pelos mecanismos de busca, enfatizando a descrição do conteúdo nele presente, melhorando sua classificação e consequentemente colaborando com a qualidade na recuperação de informações realizadas por meio de mecanismos de buscas. Tais técnicas foram testadas em alguns Websites, obtendo resultado satisfatório, a partir de então a ferramenta foi implementada e submetida a usuários para sua validação, o resultado desta validação é apresentado demonstrando a viabilidade da ferramenta e enumeração de novas funcionalidades para trabalhos futuros / When we speak of evolution of the Web, it might actually be more appropriate to speak of intelligent design. With the Web becoming the primary choice for those who produce and disseminate digital content , more people take attention to this valuable repository of knowledge. In this environment , search engines are configured in popular, becoming an intermediary between users and the myriad of information, service and resources available on the World Wide Web. In this sense, the Web designer can act decisively, providing a better response in the ranking of search engines. The correct representation of knowledge is the key to recovery and effective dissemination of data, information and knowledge. This paper presents a study that significant progress can bring a large network of users, seeking to present a public domain tool that supports the application of techniques for semantic description of Web information in the course of the research investigated techniques and methodologies that can optimize Website indexing by search engines, emphasizing the description of the content in it, improving your ranking and thus contributing to quality in information retrieval conducted through search engines. These techniques were tested on some websites, obtaining satisfactory results, since then the tool was implemented and submitted to users validation, the result of the validation is present demonstrating the feasibility of the tool and list of new features for future work

Nous cherchons dans ce travail à détricoter la relation à la fois compétitive et coopérative, indifféremment technique, économique, juridique, sociale, politique et résolument communicationnelle de Google et des éditeurs de presse. Pour cela, après avoir historicisé la rencontre de deux univers singuliers, nous décrivons ce que les éditeurs peuvent faire pour franchir le prisme du moteur de recherche et y optimiser la visibilité de leur production. Nous tâchons ensuite de décrypter ce que la firme Google peut faire faire aux éditeurs en analysant leurs relations de pouvoir, leurs incitations, leurs projets et leurs environnements informationnels respectifs. Enfin, nous rendons compte de ce que les éditeurs français issus de la presse imprimée font effectivement : ce qu’ils communiquent à Google, par quels moyens et à quel prix, pour quels résultats espérés, à l’issue de quelles concessions, quels détours, quelles contestations. Nous expliquons comment les conditions et les modalités de captation du trafic sont susceptibles d’influencer la valorisation du contenu, l’organisation e sa production, la structure du site, les pratiques journalistiques et les lignes éditoriales. Nous montrons qu’un aller-retour performatif se crée entre énoncés et conditions d’énonciation, agissant par et sur les textes, les architextes et les hypertextes. En somme, c’est à la compréhension de ce que deviennent l’actualité et eux qui la mettent en récit, dès lors qu’ils cherchent à ce que l’information qu’ils produisent soit trouvée par les utilisateurs de Google, que notre thèse est consacrée. / In this thesis, we aim to disentangle the cooperative but also competitive relationship between Google and news publishers, which is at the same time technical, economic, legal, social, political and certainly communicational. In order to do so, we trace the historical development of two singular universes, describing what publishers can do to overcome the search engine and optimize their ranking. We then analyse how Google can influence publishers’ conduct, by studying power relations, respective incentives, aims, and informational and socio-economic backgrounds. Finally, we report on actual practices of French traditional news publishers: what they communicate to Google, by which means and at what price, for which expected results, after which concessions, detours and controversies. Thus, we explain how search engine optimization is likely to affect the way content is valued, its production organisation, the website’s structure, journalists’ prac tice an editorial policy. We show a back and forth movement between performative utterances and performed circumstances, having an effect on and by texts, architexts and hypertexts. To sum up, this thesis is dedicated to understanding what happens to news and publishers once they strive for their information to be found by Google's users.