<p>One of the notable developments in current natural language processing is the practical efficacy of probabilistic word representations, where words are embedded in high-dimensional continuous vector spaces that are optimized to reflect their distributional relationships. For sequences of words, such as phrases and sentences, distributional representations can be estimated by combining word embeddings using arithmetic operations like vector averaging or by estimating composition parameters from data using various objective functions. The quality of these compositional representations is typically estimated by their performance as features in extrinsic supervised classification benchmarks. Word and compositional embeddings for a single language can be induced without supervision using a large training corpus of raw text. To handle multiple languages and dialects, bilingual dictionaries and parallel corpora are often used for learning cross-lingual embeddings directly or to align pre-trained monolingual embeddings. In this work, we explore and develop various cross-lingual alignment techniques, compare the performance of the resulting cross-lingual embeddings, and study their characteristics. We pay particular attention to the bilingual data requirements of each approach since lower requirements facilitate wider language expansion. To begin with, we analyze various monolingual general-purpose sentence embedding models to better understand their qualities. By comparing their performance on extrinsic evaluation benchmarks and unsupervised clustering, we infer the characteristics of the most dominant features in their respective vector spaces. We then look into various cross-lingual alignment frameworks with different degrees of supervision. We begin with unsupervised word alignment, for which we propose an approach for inducing cross-lingual word mappings with no prior bilingual resources. We rely on assumptions about the consistency and structural similarities between the monolingual vector spaces of different languages. Using comparable monolingual news corpora, our approach resulted in highly accurate word mappings for two language pairs: French to English, and Arabic to English. With various refinement heuristics, the performance of the unsupervised alignment methods approached the performance of supervised dictionary mapping. Finally, we develop and evaluate different alignment approaches based on parallel text. We show that incorporating context in the alignment process often leads to significant improvements in performance. At the word level, we explore the alignment of contextualized word embeddings that are dynamically generated for each sentence. At the sentence level, we develop and investigate three alignment frameworks: joint modeling, representation transfer, and sentence mapping, applied to different sentence embedding models. We experiment with a matrix factorization model based on word-sentence co-occurrence statistics, and two general-purpose neural sentence embedding models. We report the performance of the various cross-lingual models with different sizes of parallel corpora to assess the minimal degree of supervision required by each alignment framework.

Inaug.-Diss.--Marburg. / Vita.

Thesis (Ph.D.)--Syracuse University, 2008. / "Publication number: AAT 3333568."

Thesis (Ph. D.)--Catholic University of America, 1943. / Reproduced from type-written copy. "Selected bibliography": p. xiii-xx.

Thesis (Ph.D.)--Victoria University (Melbourne, Vic.), 2010.

Orientador: Regina Yu Shon Chun / Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Ciências Médicas / Made available in DSpace on 2018-08-17T18:44:27Z (GMT). No. of bitstreams: 1 DiGiulio_RafaelaMarques_M.pdf: 1132136 bytes, checksum: c7ecacdc23c99838a12fbf1e78c299bc (MD5) Previous issue date: 2011 / Resumo: INTRODUÇÃO: O acidente vascular cerebral (AVC) representa a terceira causa de morte em países industrializados. Além de alta incidência, a sequela mais comum de AVC é a afasia, o que traz sérios impactos sobre a qualidade de vida dos indivíduos. Os comprometimentos lingüístico-cognitivos e psíquicos repercutem nas relações entre a pessoa com afasia e seus familiares, geram a demanda de cuidados diferenciados e implicam a participação de um cuidador, geralmente um dos membros familiares. No âmbito da Fonoaudiologia, existem poucos estudos voltados aos cuidadores e/ou familiares na afasia. OBJETIVOS: investigar o impacto da afasia no cuidar de sujeitos afásicos fluentes e não-fluentes bem como nas relações de comunicação entre eles e seus cuidadores e/ou familiares. Esse estudo também procura verificar como a atuação fonoaudiológica, na ótica dos cuidadores e/ou familiares, pode contribuir para o processo de comunicação. SUJEITOS E MÉTODO: Participaram da pesquisa 18 cuidadores e/ou familiares de pessoas afásicas participantes de um Centro de Convivência de Afásicos do Instituto de Estudos da Linguagem/UNICAMP e do Grupo Fonoaudiológico de Comunicação Suplementar e/ou Alternativa do CEPRE/FCM - UNICAMP. Para a coleta de dados foram realizados quatro grupos focais (com uma média de três a sete integrantes), conduzidos pela pesquisadora. Os encontros foram gravados em vídeo e transcritos. Além disso, foram realizadas leituras do material, utilizando-se os critérios de repetição e de relevância para estabelecimento das categorias de análise. Foram estabelecidos dois grandes eixos temáticos para análise, o impacto da afasia (i) no processo de cuidar e (ii) nas relações de comunicação entre os sujeitos afásicos e seus familiares/cuidadores. RESULTADOS: Os resultados evidenciam o impacto que a afasia causa para o próprio indivíduo e para o seu cuidador e/ou familiar, com influências nas relações entre eles e repercussão no dia-a-dia da dinâmica familiar. Embora essa situação traga conseqüências para toda a família, comumente, o cuidado principal é assumido por um único familiar, geralmente cônjuges e filhos, o que gera mudanças no dia-a-dia dos cuidadores, com sobrecargas física e emocional, capazes de alterar sua qualidade de vida. Em relação à linguagem, verifica-se que a dificuldade em se expressar faz com que o cuidador, algumas vezes, exclua a pessoa afásica de conversas. Observa-se também que foram estabelecidas entre os afásicos e seus cuidadores e/ou familiares formas próprias de comunicação, sendo a CSA um recurso que se mostrou um importante instrumento para o favorecimento da linguagem do sujeito afásico. CONCLUSÃO: Os achados evidenciam que a dificuldade de linguagem afeta as relações sociais entre o sujeito e sua família/cuidador. Ademais, os resultados reiteram a necessidade de espaços que promovam momentos de escuta e acolhimento das demandas dos familiares, bem como salientam a importância da atuação fonoaudiológica com esse grupo / Abstract: INTRODUCTION: The cerebral vascular accident (AVC) is the third cause of death in industrialized countries. Besides its high incidence, the most prevalent sequel of AVC is aphasia, an acquired communication disorder that impairs a person?s ability to process language. Linguistic-cognitive impairments and psychological impacts on the relationships between the individual with aphasia and his family (the focus of this study) generate a need for special care, which involves the participation of a caregiver, usually a family member. From the perspective of Speech, Language and Hearing Science, few studies have focused on caregivers and family members in the treatment of aphasia. OBJECTIVES: To analyze the impacts of aphasia on the care process of patients (non-fluent and fluent), and to identify the impacts of aphasia on the communication between these patients and their caregivers and/or family members. This study also seeks to verify how the Speech-Language Pathology can contribute in the communication process from the perspective of caregivers and/or family members. SUBJECTS AND METHOD: Eighteen caregivers and/or family members of aphasic patients participated in this study. They are involved with Aphasia Center (CCA) of Institute of Language Studies (IEL-UNICAMP) and the group of Speech-Language Pathology of Augmentative and Alternative Communication (AAC) - (CEPRE/ FCM - UNICAMP). To collect the data we carried out four focus groups (with an average of three to seven members) led by the researcher. The meetings were videotaped and transcribed. In addition, there were readings of the material, using the criteria of repetition and relevance to establish the categories of analysis. We established two major themes for analysis: (i) the impact of aphasia on the care process and (ii) the impact of aphasia on the communication between the aphasic patients and their caregivers/family members. RESULTS: The results show that aphasia brings serious impacts on the relations between patients and their caregivers and/or families, which influence their routine and family dynamics. We also observed that even though the impacts of aphasia brings consequences for the whole family, usually the main care is assumed by a single member (spouses and daughters/sons), which generates changes in caregivers? routine with physical and emotional consequences, affecting their quality of life. Considering some difficulties of communication and comprehension, we observe that caregivers/family members, in general, exclude aphasic patients of meetings and regular conversations. We also observed that aphasic patients have tried to establish their own forms of communication. In these cases, it is common that the patients adopt the Augmentative and Alternative Communication (AAC), which is an important instrument for facilitating the language of individuals with aphasia. CONCLUSION: These findings highlight how difficulties of language can affect social relations between individuals with aphasia and their families/caregivers. Moreover, the results confirm the need for aphasia groups and aphasia centers, which can provide moments to listen to and embrace the demands of families of aphasic patients. The results also emphasize the important role of the Speech-Language Pathology practice with these all individuals, which can help reconnect persons with aphasia to their family members, and bring positive repercussions not only for the patients but also for those who are responsible for their care process / Mestrado / Saude, Interdisciplinaridade e Reabilitação / Mestre em Saúde, Interdisciplinaridade e Reabilitação

This study draws upon the framework for analyzing the discourse of a science classroom introduced by Mortimer and Scott (2003). It has been used for many studies before on high school youths, but never on adults. Therefore, this study focuses on the interaction between teacher and students in an adult education on high school level in physics, and how students can change the classroom discourse. Observations took place in two different schools for adults, in four different teaching groups. The result was analyzed by pattern of interactions, communicative approach, teaching purpose, type of student questions and their effect on the classroom discourse. The result was that, although students initiated the most discussions, the teacher is in charge of their discussion. The students could mostly influence the physics discourse by asking questions that the teacher did not expect and therefore the teacher needed to change his way of teaching. These changes would only last for a short while, and then would the teacher return to the original way. The students also had a small influence over the discourse when they participate during the discussions that the teacher had planned beforehand, but the teacher would according to my study have the ultimate power in the classroom. / Denna studie utgår ifrån ett ramverk som utvecklats av Mortimer och Scott (2003) för att analysera diskursen i naturvetenskapliga klassrum. Detta ramverk har tidigare endast använts på gymnasieelever, men aldrig tidigare på vuxna studenter. Studien fokuserar därför på interaktionerna mellan lärare och studenter på komvux i fysikkurserna på gymnasienivå. Observationer skedde på två olika komvux, i fyra olika undervisningsgrupper. Data analyserades utifrån interaktionsmönster, kommunikativ ansats, lärarens syfte, samt studenternas frågor och dess påverkan på klassrumsdiskursen. Resultatet visar på att trots att studenterna stod för de flesta diskussionsinitieringar, så hade läraren stor kontroll över diskursen och undervisningen. Studenterna kunde påverka fysikdiskursen med frågor som läraren inte genast visste hur han skulle kunna besvara, och därmed uppmuntra läraren att förändra undervisningen. Denna förändring varade bara ett kort tag, tills frågan var besvarad och läraren gick tillbaka till sitt ursprungliga undervisningssätt. Studenter kunde även påverka diskursen genom att med ett samspel med läraren skapa de diskussioner som läraren tycker passar in i undervisningen, men den slutliga makten över diskursen finns hos läraren.

Orientador: Eliane Martins / Dissertação (mestrado) - Universidade Estadual de Campinas, Instituto de Computação / Made available in DSpace on 2018-08-24T16:35:59Z (GMT). No. of bitstreams: 1 Letizio_CarolineCastello_M.pdf: 1670478 bytes, checksum: c2f330cd004addd705aef75384effcd2 (MD5) Previous issue date: 2013 / Resumo: Teste baseado em modelo é uma técnica na qual um sistema é modelado e geram-se testes a partir do modelo. Esta técnica apresenta como uma das maiores dificuldades a modelagem do sistema. Alguns dos empecilhos encontrados ao modelar são o que considerar do sistema e como representar essas considerações no modelo, incluindo em aplicações Web. Sendo assim, este trabalho propõe um método para auxiliar na modelagem de aplicações Web. Esta escolha é motivada pelo fato que esse tipo de sistema tem sido cada vez mais utilizado nos últimos anos e, com a sua evolução, sua complexidade tem aumentado consideravelmente, devido à dinamicidade e interatividade que esse tipo de aplicação oferece. A abordagem proposta é baseada em trabalhos correlatos, para se saber o que modelar da aplicação Web e como criar o modelo, descrevendo-se o passo-a-passo para se gerar os testes a partir dos modelos desenvolvidos e aplicá-los no sistema. A abordagem deste trabalho sugere tanto formas manuais como automáticas para os testes baseados em modelos. O método foi proposto para testar uma aplicação real: o ambiente de ensino à distância TelEduc. Os resultados, obtidos a partir da aplicação da proposta deste trabalho por um analista de teste e pela própria equipe do TelEduc, são analisados, a fim de validar a proposta e auxiliar no processo de desenvolvimento e aplicação dos testes de equipes de qualidade de software, principalmente, daquelas que não possuem nenhuma forma de organização de testes / Abstract: Model based testing is a technique in which the system is modeled and, from this model, tests can be generated. The major difficulty about this technique is to generate the model, since it is problematic to define what should be considered from the system and insert these characteristics into the model. This technique is also used in Web applications which presents the same challenges. Therefore, the present work proposes a method to assist in the modeling of Web applications. This sort of system has been increasingly used in recent years, where the system¿s complexity has considerably grown due to the dynamic and interactivity that this kind of application offers. The proposed method is based on related works in order to know what to model from the applications and how to create this model. The method contains the steps to generate tests cases from the created model, and how to execute them in the system. This work also suggests manual and automatic ways to execute the method. The proposal has been applied to test a real application: the e-learning Web environment TelEduc. The results obtained from the method¿s application by a test analyst and by the TelEduc team, are here analyzed to validate the method and to assist in the process development and tests execution by quality teams, especially those teams where there are no means of tests organization. / Mestrado / Ciência da Computação / Mestra em Ciência da Computação

En ce début de XXIème siècle, toutes les sociétés multilingues et multiculturelles, soit pratiquement tous les pays du monde, sont concernées par la politique linguistique scolaire. La gestion des langues de scolarisation, en particulier, est devenue l’affaire de tous les États soucieux (ou contraints) d’assurer le vivre-ensemble et de contribuer au développement humain, social, économique tout en ménageant des identités culturelles labiles et plurielles. Le XXème siècle a été celui de l’organisation des sociétés en unités étatiques sur fond d’utopie démocratique. Il a été celui d’une métamorphose du paysage politique avec l’apparition des organismes supranationaux et le réveil de la société civile. Dans un contexte global qui semble vouloir en finir avec la malédiction de Babel, nous tentons de problématiser la dimension concrète de la fabrique des politiques linguistiques scolaires. Nous soutenons dans ce travail qu’une politique linguistique scolaire est le résultat de l’interaction complexe entre une multitude d’acteurs au poids variable. En nous intéressant au cas mexicain et plus spécifiquement à celui de l’État du Jalisco, nous nous proposons d’interroger la fabrique de la politique d’éducation bilingue interculturelle mexicaine comme un espace de négociations au carrefour des recommandations globales, des choix étatiques et des revendications des populations indigènes.Cette thèse est une réflexion théorique et pragmatique sur la fabrique des politiques linguistiques scolaires. Elle est entièrement tournée vers l’objectif de prendre en considération la complexité inhérente à la conception des politiques linguistiques scolaires dans un monde glocalisé. / At this start of the XXIst century, all multilingual and multicultural societies, that is to say almost all countries in the world, are concerned with school language policies. In particular, schooling language management is a shared concern by all states involved in (or obliged to) ensure the living together and to contribute to the human, social and economic development while maintaining cultural identities both labile and plural. During the XXth century societies organised themselves into state units with a democratic utopia in the background. It was a century where the political scene got transformed with the appearance of supranational organizations and the reawakening of the civil society.In a global context that seems “to want” to finish with the Babel’s curse, we try to analyse the practical dimensions of schooling language policies’ construction.In this work we support that school language policies are the results of a complex interaction between numerous actors with variable power of action.With our focus on the Mexican case and more specifically on the case of the State of Jalisco, we try to analyse the construction of the intercultural and bilingual education policy - considering it as a space for negotiations at the crossroad between global recommendations, State choices and the expectations of the native populations.This thesis is a theoretical and pragmatic reflection on the “making” of schooling language policies. It is completely turned to the objective to take into consideration the inherent complexity in the conception of the school language policies in a glocalized world. / En esos principios del siglo XXI, cualquier sociedad multilingüe y multicultural, es decir casi todos los países del mundo, se debe de enfocar en una política lingüística escolar. El manejo de los idiomas de escolarización, en especial, se convirtió en un asunto central para todos los Estados preocupados (o forzados) por asegurar la convivencia y contribuir al desarrollo humano, social y económico, respetando a la vez identidades culturales cambiantes y plurales. El siglo XX vio las sociedades organizarse en entidades estatales en medio de una utopía democrática. Asimismo, vio el profundo cambio del paisaje político con la aparición de organismos supranacionales y el despertar de la sociedad civil. Dentro de un contexto global que parece querer ponerle fin a la maldición de Babel, tratamos aquí de plantear la creación / fabricación de las políticas lingüísticas escolares. En contra de cierta tradición en la investigación sobre política y planificación lingüística que se esmera en analizar una política a través de un análisis estático, sostenemos en el presente trabajo la idea que una política lingüística escolar es el resultado de una compleja interacción entre una multitud de actores de importancia variable. Enfocándonos en el caso mexicano, y en especial en el Estado de Jalisco, nos proponemos cuestionar la creación de la política lingüística escolar en México - política de educación bilingüe intercultural - como un espacio de negociaciones entre las recomendaciones globales, las elecciones de los estados, y las reivindicaciones de los pueblos indigenas. La presente tesis es una reflexión teórica y a la vez pragmática sobre la creación de las políticas lingüísticas escolares. Tiene como objetivo fundamental tomar en consideración la complejidad inherente a la concepción de esas políticas en un mundo glocalizado.

Une théorie empiriste de la connaissance affirme que le langage est doté de la capacité à « représenter » quelque chose d'autre que le langage lui-même. A la fin du XVIIe siècle, en effet, Locke et ses successeurs ont analysé empiriquement ce à quoi nos idées ressemblent et que les mots représentent mentalement. Autour des années trente, Schlick et ses camarades du Cercle de Vienne se servent des acquis linguistico-logiques de Frege, Russell et Wittgenstein pour analyser logiquement les propositions scientifiques, leur correspondance avec les états de choses qu'elles représentent. Avec eux, la philosophie analytique devra dorénavant ramener les problèmes de la connaissance au niveau de l'expérience langagière. C'est là que réside le caractère radical et réductionniste d'une démarche empiriste : les empiristes classiques ont fixé dans la perception la genèse chronologique de nos idées et nos pensées ; à leur tour, les Viennois déterminent le sens d'une proposition en la traduisant en propositions élémentaires, dites « Protokollsätze », qui renvoient directement au donné empirique. A la même période, Carnap se démarque des autres empiristes logiques en posant les problèmes du langage de la science en termes de reconstruction rationnelle des concepts et, cela, en introduisant de nouveaux concepts à partir de ceux déjà connus comme concepts de base phénoméniste (dans l'Aufbau en 1928) ou physicaliste (au cours des années trente). Tel est son réductionnisme logique qu'il renforce par la syntaxe logique. L'analyse logique ne porte pas sur le donné mais sur la proposition, et en particulier sur les propriétés formelles et des relations purement logiques qu'entretiennent les propositions dans un système. Carnap tente de libéraliser la démarche empiriste en assouplissant le critère de scientificité et, conformément à son « Principe de Tolérance », en adjoignant à l'ancrage empirique des questions de conventions pour la structure des propositions scientifiques. Il modère la critique viennoise de la métaphysique en attribuant à la philosophie, devenue « la logique de la science », la tâche de méthodes et formes adéquates pour la construction du langage de la science. / An empiricist theory of knowledge argues that language has the capacity to represent something other than itself. Beginning from the end of the 17th century, Locke and his successors analysed empirically what our ideas are reflections of and what words represent in the mind. At the beginning of the 1930s, Schlick and his Vienna Circle colleagues exploited the logico-linguistic advances of Frege, Russell and Wittgenstein in the logical analysis of scientific propositions, i.e. their correspondence with the states of affairs they represent. Following their example, analytic philosophy would henceforth reduce problems of knowledge to the level of linguistic experience. This is where the radical and reductionist character of the empiricist perspective is to be found: the classical empiricists identified the chronological genesis of our ideas and thoughts in perception; the Vienna Circle philosophers in their turn determine the meaning of a proposition by translating it into elementary propositions or Protokollsätze, which reflect directly an empirical “given”. During the same period, Carnap distances himself from his fellow empiricists by formulating the formulating the problems of scientific language in terms of the rational reconstruction of concepts by introducing new concepts based on those already identified as fundamental from a phenomenalist (the Aufbau, 1928) or physicalist (during the 1930s) perspective. This is the logical reductionism that Carnap reinforces through logical syntax. For him, logical analysis is less concerned with the given than with the proposition, and in particular with the formal properties and purely logical relations between propositions within a system. Carnap attempts in this way to “liberalize” empiricism, both by rendering more flexible and open its claim to be scientific and, in conformity to his “Tolerance Principle”, in adding questions of conventions to the empirical grounding of scientific propositions. In attributing to philosophy, understood as “the logic of science”, the task of elaborating forms and methods adequate to the construction of the language of science, he thus moderates the Viennese critique of metaphysics.