Of key importance for metamodelling are appropriate modelling formalisms. Most metamodelling languages permit the development of metamodels that specify tree-structured models enriched with semantics like constraints, references and operations, which extend the models to graphs. However, often the semantics of these semantic constructs is not part of the metamodel, i.e., it is unspeci ed. Therefore, we propose to reuse well-known compiler construction techniques to specify metamodel semantics. To be more precise, we present the application of reference attribute grammars (RAGs) for metamodel semantics and analyse commonalities and differences. Our focus is to pave the way for such a combination, by exemplifying why and how the metamodelling and attribute grammar (AG) world can be combined and by investigating a concrete example - the combination of the Eclipse Modelling Framework (EMF) and JastAdd, an AG evaluator generator.

This thesis is focused on the issues of information behaviour of editors from publishing with an emphasis on information needs and the sources of information. The first part deals with the position in the information publishing companies, the trend for electronic publishing and its influence on the book market. The following is an overview of selected terminological terms from publishers and their analysis. The next chapter is a summary of information behaviour theories, related concepts, models and current research in the Czech republic. The results of research that was realized for the purpose of the thesis between Czech editors are summarized and analyzed in the conclusion. Based on these results, proposals for the increasing of information literacy of editors are included in the thesis.

Submitted by Priscilla Araujo (priscilla@ibict.br) on 2017-08-28T17:50:17Z No. of bitstreams: 2 license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Damasio_Tese_IBICT_2017.pdf: 4347272 bytes, checksum: aac9089b91eb4fd2d792540f78444f88 (MD5) / Made available in DSpace on 2017-08-28T17:50:17Z (GMT). No. of bitstreams: 2 license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Damasio_Tese_IBICT_2017.pdf: 4347272 bytes, checksum: aac9089b91eb4fd2d792540f78444f88 (MD5) Previous issue date: 2017-03-14 / Mudanças no forma de comunicação entre os cientistas ocorreram ao longo dos últimos séculos, mas a publicação em periódicos especializados, já no século XX, tornou-se o formato padrão e central da ciência. Em paralelo ao aumento exponencial da produção em periódicos, observou-se também um aumento no número de casos de má conduta, o que levou a uma série de iniciativas mundiais para auxiliar a elaboração de politicas e a formação de especialistas no tema, também tratado de Integridade em Pesquisa. A despeito da literatura crescente neste tema, ainda há poucos estudos sobre a responsabilização dos diferentes atores, a saber, autores, revisores e editores, em casos de má conduta. Considerando a relevância do tema e a escassez de literatura específica, este estudo parte da seguinte questão: como os editores de revistas científicas, brasileiras e de outros países da América Latina, percebem as diferentes práticas de má conduta no processo editorial? Outras questões mais específicas também direcionaram o estudo: Em que momento do fluxo editorial são identificadas as más condutas? Quais os procedimentos das revistas? As revistas têm políticas editoriais específicas para lidar com a prevenção e detecção de más condutas? Como definem plágio e outras más condutas? Qual a familiaridade dos editores com práticas de má conduta? Qual a responsabilização de autores, revisores e editores? Para responder estas questões, o estudo, de natureza quanti-qualitativa, focou na população de editores de revistas científicas das maiores coleções da Plataforma SciELO, ou seja, Brasil, Argentina, Chile, Colômbia, Cuba e México. Um questionário online foi enviado para os 858 editores-chefes, dos quais 209 retornaram. A análise das respostas utilizou tratamento estatístico para descrever as variáveis extraídas das questões fechadas e também análise de conteúdo, segundo Bardin, das questões abertas. Um primeiro conjunto de resultados mostra que 80% dos editores brasileiros e latino americanos indicam que as más condutas raramente ou nunca ocorreram em suas revistas; o momento mais frequente de identificação das más condutas é na análise dos pareceristas e os procedimentos mais frequentes é de rejeitar o artigo. Também identificou-se que a maior parte das revistas participantes do estudo têm políticas específicas de prevenção, mas 36,6% das revistas brasileiras e 24,7% das latino americanas indicaram que não as possuem. Sobre a familiaridade com algumas práticas, o editores brasileiros e latino americanos são mais familiarizados com envio simultâneo de trabalhos, conflitos de interesse e plágio. Sobre o conhecimento sobre plágio, autoplágio e redundância, identificou-se que a maior parte dos editores apresenta a definição uma clássica para estas práticas, ou seja, demonstram conhecimento sobre um conceito do senso comum. Por fim, sobre as responsabilidades nas ocorrências de fabricação, falsificação e plágio, a maior parte dos editores aponta que os autores são totalmente responsáveis, enquanto outros, especialmente os editores latino-americanos, delegam aos revisores esta responsabilidade. O trabalho, que não é exaustivo, teve o objetivo de primeiramente adentrar em temática ainda não explorada e buscar uma melhor compreensão sobre a relação entre questões éticas da comunicação cientifica, o fluxo editorial e os editores, cujas informações, espera-se, possam servir de referência para pesquisas e estudos futuros. / The way scientists communicate has changed over the last few centuries; however publication in specialized journals has become the standard and central format of science since the twentieth century. Parallel to the exponential increase of journals production, there has also been an increase in the number of misconduct cases, which has led to a series of global initiatives to assist with both, policy development and the training of experts on this theme, which is also referred to as Integrity in Research. In spite of the growing literature on this subject, there are still few studies on the accountability of the different actors, that is, authors, reviewers and editors, in misconduct cases. Considering the relevance of such a theme and the scarcity of specific literature, this study starts with the following question: how do the editors of scientific journals from Brazil and other Latin American countries perceive the different misconduct practices in the editorial process? Other more specific questions also directed the present study: at what point in the editorial flow are the misconducts identified? What are the procedures of the magazines? Do magazines have specific editorial policies to deal with the prevention and detection of misconduct? How do they define plagiarism and other misconducts? What is the publishers' familiarity with misconduct practices? What is the responsibility of authors, reviewers and editors? In order to answer such questions, this quantitative qualitative study focused on the population of scientific journals editors of the largest collections of the SciELO Platform, that is, Brazil, Argentina, Chile, Colombia, Cuba and Mexico. An online questionnaire was sent to the 858 chief editors, of which 209 returned. A statistical analysis was carried out to describe the variables extracted from the closed questions, in addition to the content of the open questions, according to Bardin. A first set of the results shows that 80% of Brazilian and Latin American publishers indicate that misconduct has rarely or never occurred in their journal. The analysis of the reviewers is the most frequent moment for identifying the misconducts; rejecting the article is the most frequent procedure. It was also seen that most of the journals participating in the study have specific prevention policies; however 36.6% of the Brazilian journals and 24.7% of the Latin American ones said that they do not. Concerning familiarity with some practices, Brazilian and Latin American publishers are more familiar with the simultaneous submission of papers, conflicts of interest, and plagiarism as well. Considering knowledge on plagiarism, self-plagiarism and redundancy, it was seen that most of the editors have the classic definition for these practices, that is, they show knowledge about a common sense concept. Finally, regarding responsibility for Fabrication, Falsification and Plagiarism, most editors point out that the authors are fully responsible, whereas others, especially Latin American publishers, delegate this responsibility to the reviewers. This non-exhaustive study had as its main purpose to enter into a not yet explored topic, in addition to seek a better understanding on the relationship among the ethical issues of scientific communication, the editorial flow and the editors, whose information is expected to be used for further research and studies.

Cette thèse présente un retour d’expérience sur le dispositif de veille et d’intelligence économique mise en place au sein de Marseille Innovation pour permettre aux éditeurs open source accompagnées dans sa pépinière d’entreprise de surveiller efficacement leur environnement pour y détecter des opportunités de développement. Ce projet s’inscrit dans le cadre d’une démarche globale, conduite par Marseille Innovation et ses partenaires pour structurer un réseau d’acteurs open source en PACA.Nous avons tout d’abord réalisé un état de l’art de la littérature open source en soulignant le paradoxe que pose le modèle open source pour les éditeurs qui choisissent d’utiliser les licences libres : l’adoption d’un modèle open source par certains éditeurs pour leurs logiciels permet une création de valeur pouvant être supérieure à celle générée dans le cadre de modèles propriétaires traditionnels, mais que cet avantage est contrebalancé par une dangereuse incertitude, quant à la captation de cette valeur, susceptible d’être récupérée par des concurrents n’ayant pourtant consentis aucun effort de développement et n’en ayant pas subis les coûts. Par la suite, nous avons présenté la naissance du logiciel libre comme le résultat des mutations de l’industrie informatique qui ont marqué le début des années 70, mutations qui ont fait passer le logiciel libre de la sphère publique à la sphère commerciale. En réaction, les défenseurs du logiciel libre ont inventé un système original de « licences publiques » pour protéger ces logiciels contre tout verrouillage technique ou légal de leur utilisation, de leur diffusion et de leur modification. Cependant, si ces licences publiques permettent de favoriser la création de valeur, à travers la coopération d’une communauté de programmeurs et la diffusion du logiciel, le caractère aléatoire de la rémunération dans le modèle open source a conduit les éditeurs à chercher et inventer de nouveaux « Bunisess Models » plus ou moins aléatoires, garantissant mieux leurs revenus.Partant des connaissances accumulées sur le modèle open source, nous avons par la suite exposé la démarche initiée par Marseille Innovation, en collaboration avec CCI de Marseille Provence et l’association Libertis, pour structurer un réseau d’acteurs open source en PACA. Cette démarche a abouti au lancement de plusieurs chantiers d’intérêt commun dont un sur la mise en place d’un dispositif de surveillance sur la filière open source.La stratégie adoptée pour mettre en place ce dispositif de surveillance consiste dans un premier temps à dresser un état de l’art des différentes approches théoriques et pratiques recensées sur le sujet dans la littérature SIC. Ensuite, une approche pragmatique de la veille a été choisie et les moyens organisationnels, processuels et techniques pour mettre en place ce dispositif de surveillance ont été étudiés. Des actions de sensibilisation à la veille et l’intelligence économique ont été également conduites auprès des membres du réseau et des startups open source accompagnées par Marseille Innovation, grâce à lesquelles des besoins en veille sont identifiés et des produits d’information sont élaborés pour permettre aux membres du réseau open source de suivre efficacement leur environnement. Des études de veille sont également réalisées pour répondre aux attentes des entreprises sur le mode de création et de captation de valeurs dans l’open source, les segments de marché porteurs, les licences informatiques et les business models qui en découlent. / This thesis presents a feedback on the competitive intelligence device set up in Marseille Innovation to allow open source software editors housed in its business incubator, to effectively monitor their environment in order to detect new opportunities for development. This project is part of a comprehensive approach, led by Marseille Innovation and its partners to structure a network of open source companies in the PACA area (south of France).First, we achieved a state of the art of open source literature emphasizing the paradox posed by the open source model for software editors who choose to use free software licenses : the adoption of an open source model by some editors for their software allows a value creation that could be greater than that generated through traditional proprietary models, but this advantage is offset by a dangerous uncertainty as to the value capture, could be recovered by competitors who have yet made no programming effort to and not having incurred costs. Subsequently, we presented the birth of free software as the result of changes in the computer industry that marked the early '70s, changes which caused the passage of free software from non commercial/public sphere to commercial/private sphere. In response, defenders of free software have invented an original system of "public licenses" to protect these software’s against any technical or legal interlocking of their use, their distribution and their modification. However, if these public licenses allow to promote a value creation, through the cooperation of a community of programmers and software distribution, the randomness of remuneration in the open source model has led software editors to seek and invent new "Business Models" more or less random, to better ensure their income.Based on knowledge accumulated on the open source model, we have subsequently explained the approach initiated by Marseille Innovation, in collaboration with Marseille-Provence Chamber of Commerce and Industry (CCI) and Libertis, to structure a network of open source companies in the PACA area. This approach led to the launch of several projects of mutual interest, including one on the setting-up of a monitoring device to monitor open source software industry.The approach adopted to setting-up this monitoring device consists in a first step to establish a state of the art of different theoretical approaches and practices identified in the information and communication sciences’ literature. Then, a pragmatic approach of monitoring has been chosen. Organizational means, processual and technical have been studied to set up this monitoring device. Raising awareness actions to competitive intelligence were also conducted with members of the network and open source startups coached by Marseille Innovation, through which information needs are identified and information products are developed to enable members of open source network to effectively monitor their environment. Bibliometric studies are also made to meet the expectations of open source companies on how to create and capture value in the open source software model, about promising market segments, software licenses and business models that result.

This master's thesis deals with the user interface for machine quality assessment of the technical groups of photos. This work is written about photos, graphics editors and their evaluation. The practical part is focused on the design and creation of application with web user interface, which evaluates and compares the quality of photos. Finally is this created application with its user interface tested by users, its benefits and opportunities for further expansion are evaluated.

In this paper I pay attention to video editation issue. Firstly the basic terminology is explained, then the history of analog editation is described and further more even digital editation as well. At this spot, very interesting history of development of this field isn't forgotten and also the architecture of video editing software is described as we know them today. The most used digital forrmats are summarized as well. In next chapters the design of simple video editing application is explained. The applied technologies are briefly described and the decision of the choice is clarified. The result of whole work is a library of tools for video processing applications development demonstrated in a simple application.