En utilisant les méthodes de la linguistique interactionnelle, notamment de la « syntaxe en ligne », et de l’analyse conversationnelle, notre étude s’est focalisée sur les co-constructions syntaxiques en allemand parlé, qui se caractérisent par la co-construction d’une structure syntaxique par plusieurs locuteurs. L’analyse est basée sur un corpus de 199 occurrences issues de conversations quotidiennes et d’interviews informels. Premièrement, une analyse en termes de projection et rétraction a révélée la structuration syntaxique et temporelle de ce phénomène. La syntaxe s’est révélée cruciale comme ressource pour la co-construction d’un énoncé complet, mais les projections au niveau prosodique, sémantique et pragmatique sont également une ressource importante vers laquelle les locuteurs s’orientent. Nous avons également pu montrer comment les différents types de rétractions pouvaient s’expliquer d’une part par la syntaxe de la co-construction, d’autre part par ses fonctions interactionnelles. Deuxièmement, nous nous sommes focalisée sur les caractéristiques actionnelles de la séquence de co-construction. Nous discutons trois manières principales dont le receipt slot est rempli dans notre corpus: les complétions différées (affiliatives et non-affiliatives), l’acquiescement de la complétion et des nouveaux tours de parole du premier locuteur. Enfin, nous avons constaté que la complétion est une stratégie multifonctionnelle du second locuteur qui accomplit souvent une action collaborative mais qui peut également être compétitive.En somme, notre étude a montré que la structure syntaxique et séquentielle d’un phénomène spécifique de l’oral tel que les co-constructions n’est à concevoir que si on l’analyse dans son émergence temporelle. Cela nous emmène à discuter la manière dont la grammaire est exploitée pour construire des actions sociales qui sont pertinentes et reconnaissables pour les participants. / Applying the methods of Conversation Analysis and Interactional Linguistics, especially the methodological framework of Auer`s on-line syntax, this study focused on syntactic co-constructions in spoken German i.e. the production of one syntactic gestalt by two speakers.The study is based on a data set of 199 co-constructed syntactic completions which are taken from audio and video recordings of German everyday conversation and informal interviews. First, we present a syntactic and prosodic analysis of the co-constructed syntactic gestalts. An analysis in terms of projections and retractions revealed the syntactic and temporal organization of the co-constructed units. Syntax turned out to be a crucial resource speakers rely on in order to complete the preliminary speaker’s syntactic gestalt. However, the prosodic, semantic and pragmatic levels are also important for the second speaker to anticipate a possible completion. Furthermore, the study also showed that retractions in the second speaker’s completion can be explained by syntactic constraints and by the interactional function of the completion. Second, the analysis focused on the sequential organization of the co-constructions. We discussed three ways in which the receipt slot, i.e. the sequential position where the first speaker can accept or reject the suggested completion, can be filled: delayed completions (affiliative and non-affiliative), acknowledgment of the completion and a new turn of the first speaker. Finally the interactional functions of the co-constructions were described. The co-construction of one syntactic unit by two speakers is a multifunctional strategy which is often used to show understanding and agreement, but which can also be found in competitive contexts and can be exploited for turn-taking. Our study shows that the syntactic and sequential structure of interactional phenomena like co-constructions can only be analyzed if we consider their temporal emergence. This allows us to make a contribution to the description of how grammar is exploited to build accountable actions.

Este trabalho analisa diacronicamente o parâmetro do sujeito nulo no português do Brasil, mediante a utilização de documentação jurídica - mais especificamente, os depoimentos prestados em processos e inquéritos policiais. Defendemos que esse tipo de documentação consiste em significativa representação da língua falada para épocas pretéritas. Após, verificamos o que os dados neles contidos poderiam informar a respeito do parâmetro do sujeito nulo no português do Brasil na época em que foram produzidos, ou seja, final do século XIX e meados do século XX. Limitamo-nos ao estudo do parâmetro em questão apenas em orações subordinadas, mais especificamente no que denominamos encadeamento de subordinadas, estrutura muito comum no tipo de documento por nós estudado até hoje. Tal estrutura consiste em um grande número de orações subordinadas ligadas a uma mesma oração principal por meio do conectivo que. Por exemplo, \"a testemunhai disse que não cvi sabe quanto a autora tinha no Banco Mauá, [...]; que o que ellai ouvio dizer pelo finado Cavalheiro [...]\" Inicialmente, comparamos tais dados com aqueles contidos em estudo diacrônico clássico sobre o assunto, da autoria de Duarte (1993). Também verificamos o que poderiam nos informar se submetidos às hipóteses formuladas por Barbosa, Duarte e Kato (2005). A análise dos dados permite concluir pela existência de um comportamento esperado para o período analisado, mas ao mesmo tempo contém nuances indicativas das transformações pelas quais viria a passar o estatuto do sujeito nulo no Brasil. / This work analyzes diachronically the pro-drop parameter in Brazilian Portuguese through statements given in a court of law or during police investigation. We believe that this kind of document shows past spoken language. After concluding that these statements can really be a source of spoken language, we extracted data to guide us through the situation of the pro -drop parameter in Brazilian Portuguese, at the time the statements were taken - end of 19th century/middle of 20th century. We only studied subordinate clauses, in a structure we call subordinate clause chain, which is very common in the type of document we chose until today. It consists in many subordinate clauses connected to the same matrix clause by que [that]. E. g., \"a testemunhai disse [the witness said] que não cvi sabe quanto a autora tinha no Banco Mauá, [...]; [that she doesn\'t know how much the complainant had in Mauá Bank,...] que o que ellai ouvio dizer pelo finado Cavalheiro [...]\" [that what she heard the dead Cavalheiro say...] We compared these data with those which supported a classical study in the subject by Duarte (1993). And we also checked what they could inform us if submitted to the hypotheses created by Barbosa, Duarte and Kato (2005). We concluded that the data extracted from the corpus behave just as it was expected for that period of time, but with slight indications of the changes that would occur ahead in time in Brazil.

Submitted by Luciana Ferreira (lucgeral@gmail.com) on 2015-05-14T13:00:07Z No. of bitstreams: 2 Tese - Elisandra Filetti-Moura - 2014.pdf: 4169845 bytes, checksum: dee0cd4deff91a05213918126ceccb87 (MD5) license_rdf: 23148 bytes, checksum: 9da0b6dfac957114c6a7714714b86306 (MD5) / Approved for entry into archive by Luciana Ferreira (lucgeral@gmail.com) on 2015-05-14T13:03:56Z (GMT) No. of bitstreams: 2 Tese - Elisandra Filetti-Moura - 2014.pdf: 4169845 bytes, checksum: dee0cd4deff91a05213918126ceccb87 (MD5) license_rdf: 23148 bytes, checksum: 9da0b6dfac957114c6a7714714b86306 (MD5) / Made available in DSpace on 2015-05-14T13:03:56Z (GMT). No. of bitstreams: 2 Tese - Elisandra Filetti-Moura - 2014.pdf: 4169845 bytes, checksum: dee0cd4deff91a05213918126ceccb87 (MD5) license_rdf: 23148 bytes, checksum: 9da0b6dfac957114c6a7714714b86306 (MD5) Previous issue date: 2014-12-05 / This paper proposes to examine the Implicit Objects in the Portuguese spoken in Goias, from a functionalist conception of language, based on the work of Hopper and Thompson (1980, 2001); Dik (1989, 1997); Hengeveld and Mackenzie (2008); García Velasco and Portero Muñoz (2002), among others. The study of Implicit Objects is supported by the studies that discusses the Transitive phenomenon, as well as research on the Implicit Object behavior in Brazilian Portuguese and in others natural languages. These phenomena come from a pragmatic intention´s speakers that trigger the production and reception of linguistic phenomena, according to the grammatical terms available in a given language. Thus, it is considered that the language system as a whole is guided by pragmatic conditions grammatically organizes the languages. The phenomenon of verbal transitivity is fundamental to the understanding of Implicit Objects, its presents a notion scale, continuous and not categorical, according to which degrees of Transitive also reflect the cognitive salience of knowledge and experiences of the speakers. The selection or not a verbal argument to the object position seems to transcend the limits of a sentence, extending throughout the discursive context. So, we tried to explain how the production of Implicit Objects, as well as the recovery of the sense of the implicit elements, by the speakers, is processed in the grammar of the Brazilian Portuguese. For this, we used the corpus "The Contemporary Portuguese spoken in Goias." The Data on Implicit Objects focus on direct transitive sentences, from which settled the subdivision of Implicit Objects in definite and indefinite. The distribution in these two great classes allowed us to analyze how the referent of these implicit elements can be resumed at the interaction. Furthmore, we developed a discussion about how to Functional Grammar (DIK, 1989, 1997) and the Functional Discourse Grammar (Hengeveld and Mackenzie, 2008) have theoretical support for the analysis of Implicit Objects, from a pragmatic and discoursive orientation of this phenomenon. The Implicit Objects in the Goiás speech are represented by 65.45% Defined Implicit Objects, whose anaphoric reference can be found in the linguistic context, and by 34, 55% of Implicit Objects white noise, to which reference can be deduced by inference or from terms of undefined nature with which the implicit element establishes thematic relationship. Understand how events are motivated discoursive, noting how the degrees of transitivity in larger discursive units that prayer is central to a broader reading of the phenomenon and how their occurrence in the language layers, promotes the interpretation of Implicit Objects as the dialogue strategies of Portuguese spoken in Goias. / Este trabalho propõe-se a analisar os Objetos Implícitos no Português Falado em Goiás, a partir de uma concepção funcionalista de linguagem, fundamentado nos trabalhos de HOPPER e THOMPSON (1980, 2001); DIK (1989, 1997); HENGEVELD e MACKENZIE (2008); GARCÍA VELASCO e PORTERO MUÑOZ (2002), entre outros. O estudo dos Objetos Implícitos apoia-se em trabalhos que discutem o fenômeno de Transitividade, assim como em pesquisas acerca do comportamento de Objetos Implícitos no Português do Brasil e em outras línguas naturais. Esses fenômenos são oriundos de uma intenção pragmática dos falantes, que desencadeiam a produção e recepção dos fenômenos linguísticos, de acordo com as condições gramaticais disponíveis numa dada língua. Assim, considera-se que o sistema linguístico, como um todo, é orientado por condições pragmáticas que organizam gramaticalmente as línguas. O fenômeno da Transitividade verbal é fundamental para o entendimento dos Objetos Implícitos, pois apresenta uma noção escalar, contínua e não categórica, segundo a qual graus de Transitividade refletem também a saliência cognitiva dos conhecimentos e das experiências vivenciadas pelos falantes. A seleção ou não de um argumento verbal para a posição objeto parece transcender os limites oracionais, estendendo-se a todo o contexto discursivo. Assim, procurou-se explicar como a produção de Objetos Implícitos, assim como a recuperação do sentido dessa categoria implícita, por parte dos falantes, é processada na gramática do Português do Brasil. Para isso, foi utilizado o corpus “O Português Contemporâneo Falado em Goiás”. Os dados relativos aos Objetos Implícitos incidem sobre orações transitivas diretas, a partir das quais se estabeleceu a subdivisão dos Objetos Implícitos em definidos e indefinidos. A distribuição nessas duas grandes classes permitiu analisar de que modo o referente desses elementos implícitos pode ser retomado na interação. Além disso, foi desenvolvida uma discussão acerca do modo como a Gramática Funcional (DIK, 1989, 1997) e a Gramática Discursivo-Funcional (HENGEVELD e MACKENZIE, 2008) apresentam suporte teórico para a análise dos Objetos Implícitos, a partir de uma orientação pragmático-discursiva desse fenômeno. Os Objetos Implícitos na fala goiana são representados por 65,45% de Objetos Implícitos Definidos, cuja referência anafórica pode ser encontrada no contexto linguístico, e por 34, 55% de Objetos Implícitos Indefinidos, cuja referência pode ser depreendida, mediante inferenciação ou a partir de termos de natureza indefinda com os quais o element implícito estabelece relação temática. Compreender o modo como ocorrências são motivadas discursivamente, observando-se como os graus de Transitividade, em unidades discursivas maiores que a oração, são fundamentais para uma leitura mais ampla do fenômeno e como sua ocorrência nas camadas linguísticas, favorece a interpretação dos Objetos Implícitos como as estratégias de interlocução dos falantes no Português Goiano Falado.

A presente dissertação tem como objetivo categorizar as estratégias de comunicação (EC) encontradas no curta-metragem em língua italiana Le Avventure di Marco Polo, realizado em 2011 por alunos do 6º ano do ensino fundamental de um colégio particular na cidade de Atibaia, participantes do Projeto Filmes em Italiano (PFI). Apresentaremos o modelo de operação global para o ensino de línguas (MOGEL), proposto por Almeida Filho, buscando demonstrar como a interação de seus níveis principais ocorre no âmbito do PFI. Procuraremos justificar o PFI como projeto comunicativo, através do postulado de que a autonomia do aprendiz, como preconizado pelo ensino comunicativo de línguas, é potencializada por meio do engajamento em projetos (project work) e em tarefas (tasks). O trabalho por projetos e as tarefas colocam o aluno como centro das atividades didáticas e estão relacionados aos novos cenários comunicativos audiovisuais-multimídia, acessíveis através de ferramentas tecnológicas que já fazem parte do dia a dia das novas gerações. Durante a apresentação do PFI 2011, introduziremos o conceito de falas guiadas, termo que cunhamos para definir as produções orais resultantes dos diversos tipos de apoio que fornecemos aos alunos durante as filmagens, buscando auxiliá-los na produção em língua italiana diante de problemas comunicativos de diversas ordens, favorecendo, assim, o uso de EC. O corpus da análise é constituído pelas produções orais coletadas no curta-metragem e é dividido em duas partes, sendo que a primeira contempla as produções surgidas na filmagem das lacunas inseridas no script, idealizadas para propiciar momentos de interação espontânea entre os alunos, e a segunda contempla as produções orais relativas às falas do script, baseadas no texto escrito previamente ensaiado pelos estudantes. No tocante à literaturasobre EC, apresentamos as teorias de três teóricos: Tarone (1977), Faerch & Kasper (1983), e Poulisse (1993), discutindo suas contribuições e justificando nossa escolha pela taxonomia lexical de Poulisse para a categorização das duas partes do corpus. Nossos resultados parecem demonstrarque os discentesutilizam-se, igualmente, de EC nas falas do script e nos momentos de interação espontânea,o que ressalta a validade da problematização como um dos critérios principais para que haja uso de EC,e não a interação nos moldes tradicionais.Além disso, o corpus coletado parece-nos apontar a importância, muitas vezes desconsiderada, da transferência linguística na constituição da interlíngua (IL) do aprendizeno uso de EC. Também nos parece que o contato com mais de uma LE, como ocorre com os alunos do PFI, pode propiciar o desenvolvimento da competência estratégica deles, aumentando as possibilidades de transferência linguística e de fontes para a criação de itens lexicais, bem como seus recursos para sanar problemas comunicativos. / This dissertation aims to categorize the communication strategies (CS) found in the short film in Italian Le Avventure di Marco Polo, conducted in 2011 by students of the 6th grade of elementary education at a private school in the city of Atibaia, who participates of the project Movies in Italian (PFI). We will present the global operating model for teaching languages (MOGEL), proposed by Almeida Filho, to demonstrate how the interaction of its main levels occurs within the PFI. We will seek to justify PFI as a communicative project through the postulate that the learner autonomy, as recommended by the communicative language teaching is enhanced by the engagement projects (project work) and tasks (tasks). The work by projects and tasks put students at the center of educational activities and are related to new-media audiovisual communicative scenarios, accessible through technological tools that are already part of everyday life of the new generations. During the PFI 2011 presentation, we introduce the concept of guided speech, a term coined to define the oral productions resulting from various types of support we provide to students during filming, seeking to assist them to produce in Italian facing a variety of communicative problems, thus favoring the use of CS. The analysis corpus consists of utterances collected in the short film and is divided into two parts, the first of which includes productions filming in emerging gaps inserted into the script, devised to provide moments of spontaneous interaction among students, and the second includes oral production statements regarding the script, based on the previous written text rehearsed by students. Regarding the CS literature, we present three theorists: Tarone (1977), Faerch & Kasper (1983), and Poulisse (1993), discussing their contributions and justifying our choice of Poulisse lexical taxonomy to categorizing the two parts of corpus. Our results seem to demonstrate that learners use also the CS in the script speeches and in the moments of spontaneous interaction, underscoring the validity of questioning as one of the main criteria for use of the CS there, not in the traditional interaction. Moreover, the corpus collected seems to point out the importance often neglected, of the transfer language to constitute the learner interlanguage and in the usage of the CS. It also seems that the contact with more than one LE, as with students of PFI, can encourage the development of their strategic competence, increasing the possibilities of language transfer and sources for creating lexical items as well as their resources to solve communication problems.

The present thesis is concerned with the phenomenon of language transfer of fluency in sponteneous speech of advanced learners of English, whose first language is Czech. Particularly, it considers the transfer of speech rate and two selected speech management strategies: filled pauses and repeats. These strategies commonly occur in the speech of native speakers. Non-native speakers, however, tend to overuse these and the difference in distribution as compared to native speech can negatively influence their fluency. The thesis examines the hypothesis that speech rate and the use of filled pauses and repeats are affected by transfer from L1 alongside increased planning pressures, leading to overuse of these strategies in the speech of advanced learners and the differences in distribution. The data for analysis were taken from English and Czech recordings of eight advanced learners.

Les systèmes de dialogue homme-machine ont pour objectif de permettre un échange oral efficace et convivial entre un utilisateur humain et un ordinateur. Leurs domaines d'applications sont variés, depuis la gestion d'échanges commerciaux jusqu'au tutorat ou l'aide à la personne. Cependant, les capacités de communication de ces systèmes sont actuellement limités par leur aptitude à comprendre la parole spontanée. Nos travaux s'intéressent au module de compréhension de la parole et présentent une proposition entièrement basée sur des approches stochastiques, permettant l'élaboration d'une hypothèse sémantique complète. Notre démarche s'appuie sur une représentation hiérarchisée du sens d'une phrase à base de frames sémantiques. La première partie du travail a consisté en l'élaboration d'une base de connaissances sémantiques adaptée au domaine du corpus d'expérimentation MEDIA (information touristique et réservation d'hôtel). Nous avons eu recours au formalisme FrameNet pour assurer une généricité maximale à notre représentation sémantique. Le développement d'un système à base de règles et d'inférences logiques nous a ensuite permis d'annoter automatiquement le corpus. La seconde partie concerne l'étude du module de composition sémantique lui-même. En nous appuyant sur une première étape d'interprétation littérale produisant des unités conceptuelles de base (non reliées), nous proposons de générer des fragments sémantiques (sous-arbres) à l'aide de réseaux bayésiens dynamiques. Les fragments sémantiques générés fournissent une représentation sémantique partielle du message de l'utilisateur. Pour parvenir à la représentation sémantique globale complète, nous proposons et évaluons un algorithme de composition d'arbres décliné selon deux variantes. La première est basée sur une heuristique visant à construire un arbre de taille et de poids minimum. La seconde s'appuie sur une méthode de classification à base de séparateurs à vaste marge pour décider des opérations de composition à réaliser. Le module de compréhension construit au cours de ce travail peut être adapté au traitement de tout type de dialogue. Il repose sur une représentation sémantique riche et les modèles utilisés permettent de fournir des listes d'hypothèses sémantiques scorées. Les résultats obtenus sur les données expérimentales confirment la robustesse de l'approche proposée aux données incertaines et son aptitude à produire une représentation sémantique consistante / Spoken dialog systems enable users to interact with computer systems via natural dialogs, as they would with human beings. These systems are deployed into a wide range of application fields from commercial services to tutorial or information services. However, the communication skills of such systems are bounded by their spoken language understanding abilities. Our work focus on the spoken language understanding module which links the automatic speech recognition module and the dialog manager. From the user’s utterance analysis, the spoken language understanding module derives a representation of its semantic content upon which the dialog manager can decide the next best action to perform. The system we propose introduces a stochastic approach based on Dynamic Bayesian Networks (DBNs) for spoken language understanding. DBN-based models allow to infer and then to compose semantic frame-based tree structures from speech transcriptions. First, we developed a semantic knowledge source covering the domain of our experimental corpus (MEDIA, a French corpus for tourism information and hotel booking). The semantic frames were designed according to the FrameNet paradigm and a hand-craft rule-based approach was used to derive the seed annotated training data.Then, to derive automatically the frame meaning representations, we propose a system based on a two decoding step process using DBNs : first basic concepts are derived from the user’s utterance transcriptions, then inferences are made on sequential semantic frame structures, considering all the available previous annotation levels. The inference process extracts all possible sub-trees according to lower level information and composes the hypothesized branches into a single utterance-span tree. The composition step investigates two different algorithms : a heuristic minimizing the size and the weight of the tree ; a context-sensitive decision process based on support vector machines for detecting the relations between the hypothesized frames. This work investigates a stochastic process for generating and composing semantic frames using DBNs. The proposed approach offers a convenient way to automatically derive semantic annotations of speech utterances based on a complete frame hierarchical structure. Experimental results, obtained on the MEDIA dialog corpus, show that the system is able to supply the dialog manager with a rich and thorough representation of the user’s request semantics

La présente thèse est une étude des fonctions et des formes du discours rapporté (DR) en allemand oral à partir de deux corpus audio : des entretiens biographiques et des spectacles comiques. Après une discussion théorique préliminaire portant sur les conceptualisations classiques du phénomène de DR et les théories modernes, inspirées de Bakhtin/ Vološinov, nous envisageons le DR comme une stratégie dont les emplois sont à appréhender au niveau du discours afin de décrire les formes qu’il prend à l’oral. En nous appuyant sur les théories de syntaxe fonctionnelle, nous partons du principe que le discours oral peut s’analyser en unités de communication identifiées comme des gestes/paragraphes oraux, constitués d’une ou plusieurs propositions réalisant un acte de prédication principal. L’analyse s’effectue à deux niveaux : au niveau global du discours, et à l’intérieur du geste/paragraphe à l’échelle inter- et intra-propositionnelle. Dans un premier temps, une analyse énonciative et rhétorique permet de dégager les valeurs discursives du DR dans les deux corpus. Cette analyse met en évidence l’importance de l’usage figuratif de la stratégie de DR en tant que mise en scène explicite de polyphonie énonciative. Dans un deuxième temps, nous analysons l’intégration discursive et syntaxique du DR dans le geste/paragraphe et au niveau de la proposition, en examinant la démarcation du propos rapporté (ou mis en scène) au sein du geste/paragraphe, son rôle dans l’acte prédicatif, et les éléments lexicaux employés pour le situer et le composer. L’analyse révèle que l’aspect figuratif de nombreux DR explique une bonne part des agencements syntaxiques et de l’usage de lexèmes composant le discours rapporté à l’oral souvent non décrits dans les ouvrages de référence de l’allemand standard. / This study based on two corpora of oral communication (interviews and stand up comedy shows) investigates the functions and forms of reported speech in spoken German. We first present the classical and modern conceptualizations of reported speech as found in grammars of the German language, linguistic typology and in the more recent theories of J. Authier-Revuz, L. Rosier, and linguists of the interactional sociolinguistics and conversation analysis orientation (D. Vincent, S. Günthner, D. Tannen), all of which can be seen as inspired in some way by the views of Bakhtin/Vološinov. Following their footsteps, we argue that reported speech can be seen as a discourse strategy that explicitly stages polyphony and the layering of voices within discourse and therefore must be investigated with respect to its rhetorical effects. Adopting the perspective of the functional theories of syntax, we see oral discourse as a process whose product can be divided in moves (M) and discourse acts, expressed by propositions as predication units within the move. The study of reported speech must therefore be analyzed on two different levels: at the discourse level we seek to evaluate the rhetorical and pragmatic effects achieved by the use of (mostly direct) reported speech in the two corpora. Then we describe the role played by reported speech in the predication at the inter- and intrapropositional level within the move. As the discourse analysis shows, reported speech in informal oral communication can be a highly figurative device. This aspect of the use of reported speech accounts in part for specific syntactic constructions that have not always been taken into account in the grammars of standard German.

Cette thèse s’intéresse aux réseaux de confusion comme représentation compacte et structurée des hypothèses multiples produites par un moteur de reconnaissance de parole et transmises à un module de post-traitement applicatif. Les réseaux de confusion (CN pour Confusion Networks) sont générés à partir des graphes de mots et structurent l’information sous la forme d’une séquence de classes contenant des hypothèses de mots en concurrence. Le cas d’usage étudié dans ces travaux est celui des hypothèses de reconnaissance transmises à un module de compréhension de la parole dans le cadre d’une application de dialogue déployée par France Telecom. Deux problématiques inhérentes à ce contexte applicatif sont soulevées. De façon générale, un système de dialogue doit non seulement reconnaître un énoncé prononcé par un utilisateur, mais aussi l’interpréter afin de déduire sons sens. Du point de vue de l’utilisateur, les performances perçues sont plus proches de celles de la chaîne complète de compréhension que de celles de la reconnaissance vocale seule. Ce sont ces performances que nous cherchons à optimiser. Le cas plus particulier d’une application déployée implique de pouvoir traiter des données réelles et donc très variées. Un énoncé peut être plus ou moins bruité, dans le domaine ou hors-domaine, couvert par le modèle sémantique de l’application ou non, etc. Étant donnée cette grande variabilité, nous posons la question de savoir si le fait d’appliquer les mêmes traitements sur l’ensemble des données, comme c’est le cas dans les approches classiques, est une solution adaptée. Avec cette double perspective, cette thèse s’attache à la fois à enrichir l’algorithme de construction des CNs dans le but d’optimiser globalement le processus de compréhension et à proposer une stratégie adéquate d’utilisation des réseaux de confusion dans le contexte d’une application réelle. Après une analyse des propriétés de deux approches de construction des CNs sur un corpus de données réelles, l’algorithme retenu est celui du "pivot". Nous en proposons une version modifiée et adaptée au contexte applicatif en introduisant notamment un traitement différencié des mots du graphe qui privilégie les mots porteurs de sens. En réponse à la grande variabilité des énoncés à traiter dans une application déployée, nous proposons une stratégie de décision à plusieurs niveaux qui vise à mieux prendre en compte les spécificités des différents types d’énoncés. Nous montrons notamment qu’il est préférable de n’exploiter la richesse des sorties multiples que sur les énoncés réellement porteurs de sens. Cette stratégie permet à la fois d’optimiser les temps de calcul et d’améliorer globalement les performances du système / The work presented in this PhD deals with the confusion networks as a compact and structured representation of multiple aligned recognition hypotheses produced by a speech recognition system and used by different applications. The confusion networks (CN) are constructed from word graphs and structure information as a sequence of classes containing several competing word hypothesis. In this work we focus on the problem of robust understanding from spontaneous speech input in a dialogue application, using CNs as structured representation of recognition hypotheses for the spoken language understanding module. We use France Telecom spoken dialogue system for customer care. Two issues inherent to this context are tackled. A dialogue system does not only have to recognize what a user says but also to understand the meaning of his request and to act upon it. From the user’s point of view, system performance is more accurately represented by the performance of the understanding process than by speech recognition performance only. Our work aims at improving the performance of the understanding process. Using a real application implies being able to process real heterogeneous data. An utterance can be more or less noisy, in the domain or out of the domain of the application, covered or not by the semantic model of the application, etc. A question raised by the variability of the data is whether applying the same processes to the entire data set, as done in classical approaches, is a suitable solution. This work follows a double perspective : to improve the CN construction algorithm with the intention of optimizing the understanding process and to propose an adequate strategy for the use of CN in a real application. Following a detailed analysis of two CN construction algorithms on a test set collected using the France Telecom customer care service, we decided to use the "pivot" algorithm for our work. We present a modified and adapted version of this algorithm. The new algorithm introduces different processing techniques for the words which are important for the understanding process. As for the variability of the real data the application has to process, we present a new multiple level decision strategy aiming at applying different processing techniques for different utterance categories. We show that it is preferable to process multiple recognition hypotheses only on utterances having a valid interpretation. This strategy optimises computation time and yields better global performance

In this research, we trace the sociolinguistic profile of the speakers from the state of Alagoas concerning the variation of the verbs "ter" and "haver" in existential constructions, in order to analyse how that variation occurs and verify if the alternance of those verbs, in the alagoana speech, reflects either a process of stable variation or a change in progress. For the development of this work, we appeal to the Theory of Linguistics Variation (LABOV, 2008[1972]) which deals the variation and linguistics change and includes the variable use of the language in its social context. In this way, after the delimitation of the dependent variable and of the independent variables selected as potentially relevant in the variation under analysis, namely, animacity of the internal argument, nature of the internal argument, verbal tense, gender, age and scholarity, we stratify our sample and we did the gathering and transcription of data, as well as, the analysis and codification of all existential constructions formed with the verbs present in the corpus. For the quantitative analysis, we use the software VARBRUL, which delimited the variables statically nonsignificant and the variables statically significant in variation under study, showing not only that there is variation “ter” and “haver” existentials in the community studied and that the uses of “ter” existential is greater than the uses of “haver” existential, as well as that such variation is conditionated by the factor groups scholarity, verbal tense, age and nature of internal argument, leading us to argue that, in the alagoana speech, such variation reflects a process of change in cours in the direction of “ter” existential. / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / Nesta pesquisa, traçamos o perfil sociolinguístico dos falantes alagoanos em relação à variação dos verbos ter e haver em construções existenciais, com o intuito de analisar como essa variação ocorre e de verificar se a alternância desses verbos, na fala alagoana, reflete um processo de variação estável ou de mudança em progresso. Para o desenvolvimento deste estudo, recorremos à Teoria da Variação Linguística (LABOV, 2008[1972]) que trata da variação e da mudança linguística e contempla os usos variáveis da linguagem em seu contexto social. Dessa forma, após a delimitação da variável dependente e das variáveis independentes selecionadas como potencialmente relevantes na variação em estudo, a saber, animacidade do argumento interno, natureza do argumento interno, tempo verbal, sexo, faixa etária e escolaridade, estratificamos nossa amostra e fizemos não só a coleta e a transcrição dos dados, mas também a análise e a codificação de todas as construções existenciais formadas com os verbos ter e haver presentes no corpus. Para a análise quantitativa dos dados, utilizamos o programa computacional VARBRUL, que delimitou as variáveis estatisticamente não significativas e as variáveis estatisticamente significativas na variação em estudo, mostrando não só que há variação ter e haver existenciais na comunidade estudada e que o uso de ter existencial é bem maior do que o uso de haver existencial, como também que tal variação é condicionada pelos grupos de fatores escolaridade, tempo verbal, faixa etária e natureza do argumento interno, levando-nos a argumentar que, na fala alagoana, tal variação reflete um processo de mudança em curso na direção de ter existencial.

Mastery of basic concepts is an academic building block for preschool children in early education programs. Research shows that understanding basic concepts (e.g. top, under, fast, now, all, behind, full and short) is important for academic success and higher order thinking. Experts in the field of concept acquisition agree on six strategies for teaching basic concepts. These strategies include: using positive examples and negative examples, highlighting critical features of concepts through continuous conversion, isolating the concept, the order in which the examples are presented, and teaching generalization. This study investigated the extent to which nine preschool teachers of children with hearing loss used four of the six strategies (using examples, non examples, continuous conversion, and isolating the concept) during a 20-minute lesson in which a new basic concept was taught. Results indicated that teachers do well with using examples to teach basic concepts, but they lack sufficient use of the other three strategies for teaching basic concepts.