Esta tese propõe um recurso léxico de verbos com anotação de papéis semânticos, denominado VerbLexPor, baseado em recursos como VerbNet, PropBank e FrameNet. As bases teóricas da proposta são interdisciplinares e retiradas da Linguística de Corpus e do Processamento de Linguagem Natural (PLN), visando-se a contribuir para a Linguística e para a Computação. As hipóteses de pesquisa são: a) um mesmo conjunto de papéis semânticos pode ser aplicado a diferentes gêneros textuais; e b) as diferenças entre esses gêneros se destacam no ranqueamento dos papéis semânticos. O desenvolvimento do VerbLexPor se apoia em dois corpora: um especializado, com mais de 1,6 milhão de palavras, composto por artigos científicos de Cardiologia de três periódicos brasileiros; e um não especializado, com mais de 1 milhão de palavras composto por artigos do jornal popular Diário Gaúcho. Os corpora foram anotados com o parser PALAVRAS, e as informações de sentenças, verbos e argumentos foram extraídas e armazenadas em um banco de dados. O VerbLexPor tem 192 verbos e mais de 15 mil argumentos anotados distribuídos em mais de 6 mil sentenças. Observou-se que o corpus do Diário Gaúcho privilegia uma sintaxe direta e pouco uso de voz passiva e adjuntos, enquanto o corpus de Cardiologia apresenta mais voz passiva e um maior uso de INSTRUMENTOS na posição de sujeito, além de uma menor incidência de AGENTES. Foram realizados também alguns experimentos paralelos, como a anotação de papéis semânticos por vários anotadores e o agrupamento automático de verbos. Na tarefa de múltiplos anotadores, cada um anotou exatamente as mesmas 25 orações. Os anotadores receberam um manual de anotação e um treinamento básico (explicação sobre a tarefa e dois exemplos de anotação). Usou-se o cálculo de multi-π para avaliar a concordância entre os anotadores, e o resultado foi de π = 0,25. Os motivos para essa concordância baixa podem estar na falta de um treinamento mais completo. A tarefa de agrupamento de verbos mostrou que a sintaxe e a semântica são igualmente importantes para o agrupamento. Este estudo contribui para a área de Linguística, com um léxico de verbos anotados semanticamente, e também para a Computação, com dados que podem ser consultados e processados para diversas aplicações do PLN, principalmente por estarem disponíveis nos formatos XML e SQL. / This dissertation aims at developing a lexical resource of verbs annotated with semantic roles, called VerbLexPor, and based on other resources, such as VerbNet, PropBank, and FrameNet. The theoretical bases of this study lies in Corpus Linguistics and Natural Language Processing (NLP), so that it aims at contributing to both Linguistics and Computer Science. The hypotheses are: a) one set of semantic roles can be applied to different genres; and b) the differences among genres are shown by the ranking of semantic roles. The development of VerbLexPor has two corpora at the basis: a specialized one, with more than 1.6 million words, composed by scientific papers in the field of Cardiology from three Brazilian journals; and a non-specialized one, with more than 1 million words, composed by newspaper articles from Diário Gaúcho. The corpora were analyzed with the parser PALAVRAS, and sentence, verb and argument information was extracted and stored in a database. VerbLexPor has 192 verbs and more than 15 thousand arguments annotated with semantic roles, distributed among more than 6 thousand sentences. We observed that Diário Gaúcho has a more direct syntax, with less passive voice and adjuncts, while Cardiology has more passive voice and more INSTRUMENTS for subjects, and fewer AGENTS. We also conducted some parallel experiments, such as semantic role labeling with multiple annotators and automatic verbal clustering. In the multiple annotators task, each of them annotated exactly the same 25 sentences. They received an annotation manual and basic training (explanation on the task and two annotation examples). We used multi-π to evaluate agreement among annotators, and results were π = 0,25. Reasons for this low agreement may be a lack of a thoroughly developed training. The verbal clustering task showed that syntax and semantics are equally important for verbal clustering. This study contributes to Linguistics, with a verbal lexicon annotated with semantic roles, and also to Computer Science, with data that can be assessed and processed for various NLP applications, especially because the data are available in both XML and SQL formats.

abstract: <bold>your words</bold> / Dissertation/Thesis / Masters Thesis Agribusiness 2014

Programming and the possibility to express one’s intent to a machine is becoming a very important skill in our digitalizing society. Today, instructing a machine, such as a computer to perform actions is done through programming. What if this could be done with human language? This thesis examines how new technologies and methods in the form of Natural Language Processing can be used to make programming more accessible by translating intent expressed in natural language into code that a computer can execute. Related research has studied using natural language as a programming language and using natural language to instruct robots. These studies have shown promising results but are hindered by strict syntaxes, limited domains and inability to handle ambiguity. Studies have also been made using Natural Language Processing to analyse source code, turning code into natural language. This thesis has the reversed approach. By utilizing Natural Language Processing techniques, an intent can be translated into code containing concepts such as sequential execution, loops and conditional statements. In this study, a system for converting intent, expressed in English sentences, into code is developed. To analyse this approach to programming, an evaluation framework is developed, evaluating the system during the development process as well as usage of the final system. The results show that this way of programming might have potential but conclude that the Natural Language Processing models still have too low accuracy. Further research is required to increase this accuracy to further assess the potential of this way of programming.

Esta tese propõe um recurso léxico de verbos com anotação de papéis semânticos, denominado VerbLexPor, baseado em recursos como VerbNet, PropBank e FrameNet. As bases teóricas da proposta são interdisciplinares e retiradas da Linguística de Corpus e do Processamento de Linguagem Natural (PLN), visando-se a contribuir para a Linguística e para a Computação. As hipóteses de pesquisa são: a) um mesmo conjunto de papéis semânticos pode ser aplicado a diferentes gêneros textuais; e b) as diferenças entre esses gêneros se destacam no ranqueamento dos papéis semânticos. O desenvolvimento do VerbLexPor se apoia em dois corpora: um especializado, com mais de 1,6 milhão de palavras, composto por artigos científicos de Cardiologia de três periódicos brasileiros; e um não especializado, com mais de 1 milhão de palavras composto por artigos do jornal popular Diário Gaúcho. Os corpora foram anotados com o parser PALAVRAS, e as informações de sentenças, verbos e argumentos foram extraídas e armazenadas em um banco de dados. O VerbLexPor tem 192 verbos e mais de 15 mil argumentos anotados distribuídos em mais de 6 mil sentenças. Observou-se que o corpus do Diário Gaúcho privilegia uma sintaxe direta e pouco uso de voz passiva e adjuntos, enquanto o corpus de Cardiologia apresenta mais voz passiva e um maior uso de INSTRUMENTOS na posição de sujeito, além de uma menor incidência de AGENTES. Foram realizados também alguns experimentos paralelos, como a anotação de papéis semânticos por vários anotadores e o agrupamento automático de verbos. Na tarefa de múltiplos anotadores, cada um anotou exatamente as mesmas 25 orações. Os anotadores receberam um manual de anotação e um treinamento básico (explicação sobre a tarefa e dois exemplos de anotação). Usou-se o cálculo de multi-π para avaliar a concordância entre os anotadores, e o resultado foi de π = 0,25. Os motivos para essa concordância baixa podem estar na falta de um treinamento mais completo. A tarefa de agrupamento de verbos mostrou que a sintaxe e a semântica são igualmente importantes para o agrupamento. Este estudo contribui para a área de Linguística, com um léxico de verbos anotados semanticamente, e também para a Computação, com dados que podem ser consultados e processados para diversas aplicações do PLN, principalmente por estarem disponíveis nos formatos XML e SQL. / This dissertation aims at developing a lexical resource of verbs annotated with semantic roles, called VerbLexPor, and based on other resources, such as VerbNet, PropBank, and FrameNet. The theoretical bases of this study lies in Corpus Linguistics and Natural Language Processing (NLP), so that it aims at contributing to both Linguistics and Computer Science. The hypotheses are: a) one set of semantic roles can be applied to different genres; and b) the differences among genres are shown by the ranking of semantic roles. The development of VerbLexPor has two corpora at the basis: a specialized one, with more than 1.6 million words, composed by scientific papers in the field of Cardiology from three Brazilian journals; and a non-specialized one, with more than 1 million words, composed by newspaper articles from Diário Gaúcho. The corpora were analyzed with the parser PALAVRAS, and sentence, verb and argument information was extracted and stored in a database. VerbLexPor has 192 verbs and more than 15 thousand arguments annotated with semantic roles, distributed among more than 6 thousand sentences. We observed that Diário Gaúcho has a more direct syntax, with less passive voice and adjuncts, while Cardiology has more passive voice and more INSTRUMENTS for subjects, and fewer AGENTS. We also conducted some parallel experiments, such as semantic role labeling with multiple annotators and automatic verbal clustering. In the multiple annotators task, each of them annotated exactly the same 25 sentences. They received an annotation manual and basic training (explanation on the task and two annotation examples). We used multi-π to evaluate agreement among annotators, and results were π = 0,25. Reasons for this low agreement may be a lack of a thoroughly developed training. The verbal clustering task showed that syntax and semantics are equally important for verbal clustering. This study contributes to Linguistics, with a verbal lexicon annotated with semantic roles, and also to Computer Science, with data that can be assessed and processed for various NLP applications, especially because the data are available in both XML and SQL formats.

Submitted by Renata Lopes (renatasil82@gmail.com) on 2016-10-18T10:29:25Z No. of bitstreams: 1 fernandacamposjunqueira.pdf: 1155057 bytes, checksum: ee51ffe8a4cc117c3e53b3cfeb65c084 (MD5) / Approved for entry into archive by Adriana Oliveira (adriana.oliveira@ufjf.edu.br) on 2016-10-25T12:04:55Z (GMT) No. of bitstreams: 1 fernandacamposjunqueira.pdf: 1155057 bytes, checksum: ee51ffe8a4cc117c3e53b3cfeb65c084 (MD5) / Made available in DSpace on 2016-10-25T12:04:55Z (GMT). No. of bitstreams: 1 fernandacamposjunqueira.pdf: 1155057 bytes, checksum: ee51ffe8a4cc117c3e53b3cfeb65c084 (MD5) Previous issue date: 2008-08-28 / O presente trabalho tem como objetivo observar as interações sociais desenvolvidas entre os diversos atores numa instituição escolar estadual localizada no município de Juiz de Fora; analisar o processo de formação de grupos e identidades sociais a partir destas relações e averiguar a existência e as modalidades de expressões violentas e criminosas ocorridas na escola – bem como suas representações –, tendo como base as percepções e proposições dos diversos atores da comunidade educacional, como alunos, professores, corpo técnicopedagógico e funcionários. Iluminados pelo Interacionismo Simbólico procuramos evidenciar a maneira pela qual os distintos grupos significam e resignificam os contextos, os indivíduos e as interações sociais. Compreendemos que o universo escolar se apresenta como um espaço repleto de complexidades que implicam em encontros, desencontros, similaridades, diferenças, identificações e conflitos. As diversidades presentes nesse universo se expressam nas variadas relações e interações desenvolvidas, permitindo que tal contexto esteja em constante construção e transformação, como preconizado pelos pressupostos interacionistas. Buscamos compreender de que forma as identidades são formadas e modificadas a partir dos distintos discursos, situações e espaços. Observamos ainda que a escola, enquanto espaço de convivência, é também palco de expressões violentas, definidas diferentemente por atores distintos, além de ser um ambiente de desencontros significativos entre os grupos que se manifestam em confrontos, estigmatizações e rotulações. Dessa maneira, a instituição escolar foi analisada, considerando toda sua pluralidade, manifesta em questões, fenômenos e problemas emergentes do próprio contexto. Procuramos, com base nas diversas orientações do referencial teórico interacionista, compreender a miríade de questões peculiares ao ambiente educacional procedentes das variadas situações interacionais, respeitando as definições de cada um dos seus interagentes. / This paper aims at observing social interactions developed among the various actors in a state school located in Juiz de Fora; examining the process of formation of social groups and identities from these relationships and verifying the existence and the types of violent and criminal expressions occurring in school - and their representations -, based on the perceptions and propositions of various actors of the education community, such as students, teachers, technical-educational professionals and staff in general. Based on Symbolic Interactionism, our intention is to highlight the way in which the different groups interpret and reinterpret the contexts, individuals and social interactions. We understand the School environment as an area full of complexities involving agreements, disagreements, similarities, differences, identities and conflicts. The diversity present in this universe is expressed in various relationships and interactions developed, allowing such a context to be in constant construction and transformation, which was already assumed by interactionists. The aim was to understand how the identities are formed and modified from separate speeches, situations and spaces. The school was also observed as a social place, and the scene of violent expressions, defined differently by different actors as well as being an environment of significant disagreements between the groups which occur in clashes, stigmatization and labeling. Thus, the school was examined considering all its plurality, manifested in issues, phenomena and problems emerging from the context. The aim was to understand, based on the guidelines of various interactionist theoretical references, the myriad of peculiar issues to the educational environment coming from different interactionist situations, respecting the definitions of each of its interactors.

O enorme desenvolvimento que tem ocorrido na área alimentícia gera a necessidade de legislações que regulamentem o setor nos âmbitos de qualidade, segurança e rotulagem. Entretanto, o cumprimento das legislações muitas vezes não tem sido observado. Este trabalho tem o objetivo de verificar a conformidade da rotulagem de algumas categorias de produtos alimentícios comercializados na cidade de São Paulo frente à legislação brasileira em vigor, analisando dados quantitativos e qualitativos da conformidade e da não-conformidade dos rótulos dos alimentos frente à legislação e apresentando propostas para sua melhoria. Foram analisadas 11 categorias de produtos alimentícios de grande consumo pela população (tais como leites, chocolates e sorvetes) e 2 categorias de produtos para grupos populacionais específicos (alimentos infantis e alimentos para praticantes de atividade física). Dos 52 rótulos analisados, 80,8% apresentaram no mínimo um tipo de não-conformidade frente à legislação vigente, e apenas 19,2% estavam plenamente de acordo com o estabelecido na legislação brasileira. Entre as categorias analisadas, a categoria que apresentou maior número de itens não-conformes foi a categoria de alimentos para praticantes de atividade física. Com relação ao número de rótulos não-confomes, quatro categorias (bombons, leites UHT, biscoitos e alimentos infantis) apresentaram 100% de rótulos não-confomes, ou seja, para estas categorias todos os rótulos analisados não atendiam ao estabelecido na legislação. Já com relação ao tipo de irregularidade, a tabela nutricional foi o item avaliado que apresentou maior número de irregularidades. Desta forma, alguns itens da legislação são discutidos, tais como a previsão em legislação do uso de valores diários de referência considerando as necessidades nutricionais de crianças quando o produto for destinado ao público infantil, a fim de proporcionar uma melhor adequação dos rótulos às necessidades dos consumidores. / The huge development that has occurred in the food sector generates the necessity of laws regulating the industry in the areas of quality, safety and labeling. However, compliance with legislation is not always observed. This work is intended to verify conformity of labeling of some categories of food products sold in São Paulo city with Brazilian legislation in force, analyzing quantitative and qualitative data compliance and non-compliance of food product labels with legislation and to submit proposals for their improvement. Eleven categories of food products of large consumption by the population have been analyzed (such as milk, chocolate and ice cream) and two categories of products for specific population groups (infant food and food products for practitioners of physical activity). From the 52 labels analyzed, 80.8% presented at least one type of non-conformity with legislation and only 19.2% were fully in accordance with the established in the Brazilian legislation. Among categories analyzed, the one presenting the largest amount of non-compliant items was the food category for practitioners of physical activity. With respect to the number of noncompliant labels, four categories (chocolates, UHT milk, biscuits and infant food) presented 100% non-compliant labels, meaning that for these categories all labels analyzed did not meet the legislation standards. With respect to the type of nonconformity, the nutritional table was the item analyzed which presented the largest number of irregularities. This way, some items of the legislation are discussed, such as stipulating by legislation the use of daily values of reference considering children nutritional necessities when the food product is intended for the childrens\' segment, in order to provide a proper adequacy of labels to consumer\'s necessities.

Contusion spinal cord injury (SCI) results in devastating life-long debilitation in which there are currently no effective treatments. The primary injury site presents a complex environment marked by subsequent secondary pathophysiological cascades involving excessive reactive oxygen and nitrogen species (ROS/RNS) production, glutamate-induced excitotoxicity, calcium dysregulation, and delayed neuronal apoptosis. Many of these cascades involve mitochondrial dysfunction, thus a single mitochondrial-centric therapy that targets a variety of these factors could be far reaching in its potential benefits after SCI. As such, this dissertation examines whether transplantation of exogenous mitochondria after SCI can attenuate secondary injury cascades to decrease the spread and severity of the injury. Our first experiment tested the dose-dependent effects of mitochondrial transplantation on the ability to maintain acute overall bioenergetics after SCI. We compared transplantation of mitochondria originating from two different sources-cultured PC12 cells or rat soleus leg muscle. 24 hours after injury, State III oxygen consumption rates were maintained to over 80% of sham levels when 100ug of mitochondria was transplanted, regardless of the origin of the mitochondria. Complex I enzyme activity assays corroborated our findings that the 100ug dosage gave optimal benefits compared to vehicle injection. We also analyzed the rostral-caudal distribution and cell-type colocalization of transplanted transgenically-labeled tGFP mitochondria after SCI. There were greater volumes and rostral-caudal spread of tGFP mitochondria at the 24 hour time point compared to 7 days post injection. tGFP mitochondria had the greatest propensity to colocalize with macrophages and pericytes. Colocalization was evident in endothelial cells, oligodendrocytes and astrocytes, though no such colabeling was present in neurons. Further, colocalization of tGFP was always greater at the 24 hour time compared to 48 hour or 7days post injection time points. These data indicate that there is a cell-type difference in incorporation potential of exogenous mitochondria which changes over time. Finally, we tested the effects of mitochondrial transplantation on long term functional recovery. Animals were injected with either vehicle, 100ug cell-derived mitochondria, or 100ug muscle-derived mitochondria immediately after contusion SCI. Functional analyses including BBB overground locomotor scale and von Frey mechanical sensitivity tests did not show any differences between treatment groups. Likewise, there were no differences in tissue sparing when mitochondria were transplanted compared to vehicle injections, though there were higher neuronal cell counts in tGFP mitochondria injected groups caudal of the injury site. These studies present the potential of mitochondrial transplantation for therapeutic intervention after SCI. While our acute measures do not correspond into long term recovery, we show that at 24 hours transplanted mitochondria do have an effect on bioenergetics and that they are taken into host cells. We believe that further investigation into caveats and technical refinement is necessary at this time to translate the evident acute bioenergetic recovery into long term functional recovery.

CSR has become an important issue in the business practices of retailers worldwide. When it comes to the food sector, retailers are expected to adopt CSR practices and influence their consumers and supply chains to behave in a socially responsible way. Since the food consumption in answerable for 30 per cent of the negative environmentally impact the retailers are suggested to stimulate the green market development and sustainability in both production and consumption practices. This can be done with establishment of the responsible assortment. In relation to that, one of the main tools that food retailers use to create responsible assortment is called CSR labelling. The purpose of this study is to explore how food retail store managers engage to adopt CSR in the store assortment and how private eco-brands contribute to the green-market development in food retail. In order to achieve a widely understanding of food retailers practices when the creation of responsible assortment ant the stimulation of the green market development are in case, a qualitative method with 9 semi structured interviews was applied. The results show that retailers use CSR and eco-labells with the intention to motivate consumers to buy sustainable food products. The retailers are also showing increased attention of controlling the supply chain and influencing suppliers to behave in a socially responsible way. However, the findings indicates that both store managers and consumers need to be better informed about the benefit of the sustainable products.

Introducción: El etiquetado nutricional ayuda a los consumidores a realizar elecciones más saludables, sin embargo, este es poco leído, mal interpretado o no usado. Objetivos: Evaluar la frecuencia y factores asociados a la lectura, uso e interpretación de etiquetas nutricionales en usuarios de gimnasios en la ciudad de Huancayo, Perú. Materiales y métodos: Estudio de corte transversal en 385 usuarios de una cadena de gimnasios. Se midió la interpretación con un test a partir de dos etiquetas, lectura reportada (frecuentemente o no) y el uso reportado para elegir un producto (frecuentemente o no), así como variables demográficas, dietéticas y relacionadas al gimnasio. Se evaluó la asociación calculando las razones de prevalencias ajustadas en base a las variables significativas del modelo bivariado. Resultados: Los participantes tuvieron una edad media de 27,8 + 9,3 años, 44,7% fueron varones, 49,4% contaron con educación universitaria. El 27,5% reporta leer frecuentemente y dentro de estos el 55,7% reporta usarlas frecuentemente, asimismo el 17,4% sabe interpretar. La lectura se asoció con factores dietéticos y de uso de gimnasio y la interpretación con capacitación previa de lectura de la etiqueta. El uso fue principalmente para seleccionar productos bajos en grasa total (65,1%), alto en proteínas (64,2%) y en menor proporción, productos bajo en sodio (47,2%). Conclusiones: La frecuencia de lectura, uso e interpretación de etiquetas nutricionales es baja. La capacitación previa de lectura es un factor necesario para una elección saludable. Es necesario establecer estrategias educativas para enseñar a interpretar en forma adecuada las etiquetas nutricionales. / Introduction: The nutritional labeling helps users to make healthier choices, nevertheless, this is overseen, misunderstood or not used at all. Objectives: To evaluate the frequency and the factors associated with the reading, interpretation and use of nutritional labels by fitness center users in the city of Huancayo, Peru. Materials and methods: A cross-sectional survey was conducted on 385 users of a fitness center chain. Interpretation was measured by using a test with 2 labels, reported reading (frequently or not) and the reported use to choose a product (frequently or not), as well as demographic, dietetic and fitness-center-related variables. The association was evaluated by calculating the prevalence ratios adjusted based on the significant variables from bivariate model. Results: The participants had an average age of 27,8± 9,3 years, 44,7% were men and 49,4% had college degree. 27,5% report reading frequently and 55,7% of them reports use frecuently, furthermore 17,4% know how to interpret. The reading was associated with dietetic factors and gym use, and interpretation with prior training of reading labeling. The use was mainly high to select low fat products (65,1%), high in protein (64,2%) and was mainly low in low sodium products (47,2%). Conclusions: The reading, use and interpretation frequency of nutritional labels is low. The prior training of reading labeling is a required factor to make healthier choices. It is necessary to establish educative strategies for teaching to interprete the nutrition labels in an adequate way.

Cette thèse est consacrée à l'analyse de l'association de codes LDPC non-binaires (LDPC-NB) à des modulations d’ordre élevé. Cette association vise à améliorer l’efficacité spectrale pour les futurs systèmes de communication sans fil. Notre approche a consisté à tirer au maximum profit de l'association directe des symboles d’un code LDPC-NB sur un corps de Galois avec une constellation de même cardinalité. Notre première contribution concerne la diversité spatiale obtenue dans un canal de Rayleigh avec et sans effacement en faisant subir une rotation à la constellation. Nous proposons d’utiliser l'information mutuelle comme paramètre d’optimisation de l’angle de rotation, et ce pour les modulations de type « BICM » et les modulations codées. Cette étude permet de mettre en évidence les avantages de la modulation codée par rapport à la modulation BICM de l’état de l’art. Par simulation de Monte-Carlo, nous montrons que les gains de codage théoriques se retrouvent dans les systèmes pratiques. Notre deuxième contribution consiste à concevoir conjointement l'étiquetage des points de constellation et le choix des coefficients d'une équation de parité en fonction de la distance euclidienne, et non plus de la distance de Hamming. Une méthode d’optimisation est proposée. Les codes ainsi construits offrent des gains de performance de 0.2 dB et ce, sans ajout de complexité. / This thesis is devoted to the analysis of the association of non-binary LDPC codes (NB-LDPC) with high-order modulations. This association aims to improve the spectral efficiency of future wireless communication systems. Our approach tries to take maximum advantage of the straight association between NB-LDPC codes over a Galois Field with modulation constellations of the same cardinality. We first investigate the optimization of the signal space diversity technique obtained with the Rayleigh channel (with and without erasure) thanks to the rotation of the constellation. To optimize the rotation angle, the mutual information analysis is performed for both coded modulation (CM) and bit-interleaved coded modulation (BICM) schemes. The study shows the advantages of coded modulations over the state-of-the-art BCIM modulations. Using Monte Carlo simulation, we show that the theoretical gains translate into actual gains in practical systems. In the second part of the thesis, we propose to perform a joint optimization of constellation labeling and parity-check coefficient choice, based on the Euclidian distance instead of the Hamming distance. An optimization method is proposed. Using the optimized matrices, a gain of 0.2 dB in performance is obtained with no additional complexity.