Submitted by Renata Lopes (renatasil82@gmail.com) on 2016-04-05T11:05:51Z No. of bitstreams: 1 ludmilameireleslage.pdf: 2145852 bytes, checksum: adcbff5b4bd973b4674a037a91b50b95 (MD5) / Approved for entry into archive by Adriana Oliveira (adriana.oliveira@ufjf.edu.br) on 2016-04-24T03:42:10Z (GMT) No. of bitstreams: 1 ludmilameireleslage.pdf: 2145852 bytes, checksum: adcbff5b4bd973b4674a037a91b50b95 (MD5) / Made available in DSpace on 2016-04-24T03:42:10Z (GMT). No. of bitstreams: 1 ludmilameireleslage.pdf: 2145852 bytes, checksum: adcbff5b4bd973b4674a037a91b50b95 (MD5) Previous issue date: 2013-03-07 / CAPES - Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / Esta dissertação propõe apresentar as discussões teórico-metodológicas que embasam a implementação, na FrameNet Brasil (FN-Br), de um repertório de construções disponível online: o Constructicon do português brasileiro. A iniciativa constitui-se em um subprojeto da FN-Br, a qual, por sua vez, visa à criação de um recurso lexical online para o mesmo idioma, baseado na Semântica de Frames e sustentado por evidência em corpus. Para implementar o Constructicon, onze construções da família Para Infinitivo foram escolhidas. A escolha se deve ao fato de que essas construções formam uma rede formal e funcionalmente motivada no português do Brasil, além de evocarem frames relacionados – tais construções compartilham o esquema sintático [SN V SA/SN para (SN) Vinf] e o frame de Finalidade ou algum frame a ele relacionado. Ao longo do trabalho, realizamos análises dessas construções e discutimos quais poderiam ser tratadas como um padrão de valência, isto é, ser anotadas utilizando-se da metodologia centrada em unidades lexicais da própria FrameNet, e quais precisariam ser descritas no Constructicon. Propõe-se, dessa forma, um diálogo entre a proposta goldbergiana para o tratamento das construções e aquela idealizada por Fillmore e colaboradores. Discute-se, ainda, de que maneira e em que medida construções de estrutura argumental e unidades lexicais verbais interagem no que tange aos frames que evocam. Por fim, esta dissertação estabelece critérios que devem ser aplicados às potenciais construções do português brasileiro de modo a definir se elas devem ser tratadas como padrões de valência ou se devem figurar no Constructicon. / This dissertation presents the theoretical and methodological discussions supporting the implementation, in FrameNet Brasil (FN-Br), of a repertoire of constructions available online: the Brazilian Portuguese Constructicon. The initiative is a subproject of the FN-Br, which, in turn, aims at creating an online lexical resource for the same language, based on Frame Semantics and supported by corpus evidence. To implement the Constructicon, eleven constructions of the Para Infinitive Family were chosen. This choice is due to the fact that these constructions form a both formally and functionally motivated network in Brazilian Portuguese, besides evoking related frames – such constructions share the syntactic scheme [NP V AP/NP para (NP) Vinf] and the Purpose frame or some frame related to it. Along this work, we performed analyzes of these constructions and discussed whether they could be treated as a valence pattern, i.e. be annotated using the FrameNet methodology centered on lexical units; or be described in the Constructicon. It is proposed, therefore, a dialogue between the Goldberg proposal for the treatment of constructions and that designed by Fillmore et al. We also discuss how and to what extent argument structure constructions and verbal lexical units interact in terms of the frames they evoke. Finally, this dissertation establishes criteria which must be followed so as to determine whether a potential construction should be annotated as a valence pattern or accounted for in the Constructicon.

Submitted by Renata Lopes (renatasil82@gmail.com) on 2016-04-28T15:47:51Z No. of bitstreams: 1 igordeoliveiracosta.pdf: 2715764 bytes, checksum: c8160cddc86f647724c428607a8341c6 (MD5) / Approved for entry into archive by Adriana Oliveira (adriana.oliveira@ufjf.edu.br) on 2016-05-02T01:02:51Z (GMT) No. of bitstreams: 1 igordeoliveiracosta.pdf: 2715764 bytes, checksum: c8160cddc86f647724c428607a8341c6 (MD5) / Made available in DSpace on 2016-05-02T01:02:51Z (GMT). No. of bitstreams: 1 igordeoliveiracosta.pdf: 2715764 bytes, checksum: c8160cddc86f647724c428607a8341c6 (MD5) Previous issue date: 2015-11-30 / CAPES - Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / Este trabalho, de viés sociocognitivista e construcionista, tem como objeto analítico um nódulo da rede de Construções Quantificadoras Mórficas Sufixais, reconhecida como uma estratégia morfológica de evocação da noção de coletividade na Língua Portuguesa, como ilustram os exemplos seguir: (i) {XN-ADA} (“Com AXE vou pegar a mulherada.”); (ii) {XN-ARADA} (“No meio daquela brinquedarada toda tinha um brinquedo que, pelo menos o que eu acho hoje, era da família daqueles patinhos de borracha [...]”); (iii) {XN-AIADA} (“Após a cachaçaiada é sempre bom beber bastante liquido [...]”). Tal estudo de caso serve de estofo a um objeto teórico mais amplo, qual seja a proposição de um trato construcionista da morfologia derivacional (MIRANDA, 2013; RHODES, 1992). A Gramática das Construções Cognitiva (GOLDBERG, 1995, 2006; BOAS, 2013), entendida como um Modelo Baseado no Uso (BYBEE, 2010; CROFT E CRUSE, 2004), fornece o aparato teórico central a este estudo. Da premissa sociocognitivista que reivindica a centralidade da experiência na constituição da linguagem deriva outro fundamento crucial à sustentação das análises, qual seja a proposição de convergência entre o escopo teórico construcionista e a Semântica de Frames (FILLMORE, 1977, 2009[1982], 1985; PETRUCK, 1996; RUPENHOFER et al., 2010). Dado o relevo do uso no modelo teórico-analítico eleito, acolhe-se, em termos metodológicos, uma Linguística Cognitiva baseada em corpus (FILLMORE, 2008[1992]; MCENERY, XIAO E TONO, 2006), o que implica o uso de corpora eletrônicos e ferramentas computacionais na análise dos dados. As análises apontam, dentre outras coisas, para a consistência de um trato construcionista da morfologia derivacional, bem como para a riqueza linguística e cognitiva das construções estudadas. Os construtos, instituídos pelo esquema imagético COLEÇÃO (JOHNSON, 2005; CLAUSNER E CROFT, 1999), tem na evocação dos frames Quantidade e Desejabilidade a sua constituição semântica, apresentando-se como estratégias de quantificação e avaliação pertinentes a gêneros mais distensos da Língua Portuguesa. / This paper analyses, in a cognitive and constructionist perspective, a node of the vast network of Suffixal Morphic Quantifying Constructions, recognized as a morphological strategy evocation of the notion of collectivity in Portuguese, as illustrated with the following examples: (i) {XN-ADA} (“Com AXE vou pegar a mulherada.”); (ii) {XN-ARADA} (“No meio daquela brinquedarada toda tinha um brinquedo que, pelo menos o que eu acho hoje, era da família daqueles patinhos de borracha [...]”); (iii) {XN-AIADA} (“Após a cachaçaiada é sempre bom beber bastante liquido [...]”). This case study supports a broader theoretical object, the proposition for a constructionist approach to derivational morphology (MIRANDA, 2013; RHODES, 1992). Cognitive Construction Grammar (GOLDBERG, 1995, 2006; BOAS, 2013), a Used-based Model of Language (BYBEE, 2010; CROFT AND CRUSE, 2004), provides the central theoretical apparatus to this study. The Cognitive Linguistics premise of claiming the centrality of experience in the creation of language grants another crucial support to this analysis, which is the proposition of convergence between the constructionist approach and Frame Semantics (FILLMORE, 1977, 2009[1982], 1985; PETRUCK, 1996; RUPENHOFER et al., 2010). Given the amount of use in the theoretical and analytical model adopted, corpus-based Cognitive Linguistics (FILLMORE, 2008[1992]; MCENERY, XIAO AND TONO, 2006) was the methodology chosen. It implies the use of electronic corpora, computational tools and quantitative analysis of the data in terms of type frequency and token frequency (BYBEE, 2010; GOLDBERG, 1995). The analysis points out, among other things, to the consistency of a constructionist approach for derivational morphology, as well as the cognitive and linguistic richness of the constructional patterns under investigation. The constructs, established by the COLLECTION image schema (JOHNSON, 2005; CLAUSNER AND CROFT, 1999), evoke the Quantity and Desirability frames, presenting itself as quantifying and evaluation strategies relevant to distended genres of Portuguese.

Submitted by Renata Lopes (renatasil82@gmail.com) on 2016-05-04T13:12:02Z No. of bitstreams: 1 danielasimoesgomes.pdf: 2516376 bytes, checksum: 8f21ecce475c85eb8428cbe741299eda (MD5) / Approved for entry into archive by Adriana Oliveira (adriana.oliveira@ufjf.edu.br) on 2016-06-07T15:31:18Z (GMT) No. of bitstreams: 1 danielasimoesgomes.pdf: 2516376 bytes, checksum: 8f21ecce475c85eb8428cbe741299eda (MD5) / Made available in DSpace on 2016-06-07T15:31:18Z (GMT). No. of bitstreams: 1 danielasimoesgomes.pdf: 2516376 bytes, checksum: 8f21ecce475c85eb8428cbe741299eda (MD5) Previous issue date: 2014-08-18 / CAPES - Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / A presente dissertação apresenta a árvore conceptual dos frames que compreendem o cenário do TURISMO_COMO_EVENTO_ESPORTIVO, bem como as combinações sintático-semânticas de unidades lexicais que evocam os referidos frames. Tal trabalho traz uma apresentação dos estudos vinculados ao projeto Copa 2014 FrameNet Brasil, o qual se configura como um produto de inovação tecnológica e tem como principal objetivo desenvolver, em meio eletrônico, um dicionário temático trilíngue (Português - Inglês - Espanhol), destinado a usuários humanos, abrangendo os domínios do Futebol e do Turismo, baseado na Semântica de Frames. (SALOMÃO, M.M.M; TORRENT, T.T.; CAMPOS, F.C.A.; BRAGA, R.M.M. & VIEIRA, M.B., 2011). O projeto Copa 2014 FrameNet Brasil se caracteriza como um desdobramento do macro Projeto FrameNet Brasil que foi desenvolvido na UniversidadeFederaldeJuizdeFora em associação com o InternationalComputerScienceInstitute (EUA), que sedia o ProjetoFrameNet, que existe há mais de 10 anos na Universidade da Califórnia em Berkeley, e é liderado pelo professor Charles Fillmore. (FILLMORE, 1982, 1985) O Projeto FrameNet Brasil objetiva criar um recurso lexical on-line disponível para pesquisas sobre o Português do Brasil, baseado na semântica de frames e sustentado por evidência colhida em corpus (FILLMORE, 1982; 1985; GAWRON, 2008; e outros). Para a construção deste dicionário foi utilizada a metodologia top-down, a qual se caracteriza pela busca na FrameNet Americana de frames que atendem à temática proposta pelo mesmo, em conjunto com a metodologia bottom-up do projeto Kicktionary, na qual primeiramente as unidades lexicais (ULs) são identificadas nos corpora e em seguida atribuem-se funções micro temáticas aos elementos sintáticos relacionados e definem- se os frames evocados por aquelas ULs. As sentenças utilizadas na análise foram retiradas dos corpora FIFA e Cetenfolha, utilizando a ferramenta Word Sketch do Sketch Engine. O trabalho de descrição lexicográfica dos frames em questão contribuiu para o objetivo final do Projeto Copa 2014, que é o de descrever as estruturas conceptuais relacionadas ao domínio do Turismo e do Futebol nas três línguas alvo do dicionário. / The following paper intends to present the creation process of the frames TURISMO_COMO_EVENTO_ESPORTIVO as well as combinations of syntacticsemantic lexical units that evoke those frames. This paper brings a presentation of studies related to the project World Cup 2014 FrameNet Brazil, which is configured as a product of technological innovation and has as main objective to develop, electronically, a thematic trilingual dictionary (Portuguese - English - Spanish) for human users, covering the dominance of Soccer and Tourism, based on Semantic Frames. (SALOMÃO, M.M.M; TORRENT, T.T.; CAMPOS, F.C.A.; BRAGA, R.M.M. & VIEIRA, M.B., 2011). The project World Cup 2014 FrameNet Brazil is characterized as a deployment of the macro Project FrameNet Brazil that was being developed at the Federal University of Juiz de Fora in association with the International Computer Science Institute (USA), which hosts the FrameNet Project, which has existed for more than 10 years at the University of California in Berkeley, and is led by Professor Charles Fillmore. The Project Brazil FrameNet aims to create a lexical resource available online for research on the Brazilian Portuguese, based on the semantics of frames and supported by evidence collected in corpus (FILLMORE, 1982, 1985; Gawron, 2008, and others). For the construction of this dictionary the top-down methodology was used, which is characterized by the American FrameNet pursuit of the frames that answer to the theme proposed by such, along with the bottom-up methodology of the Kicktionary project, in which first the lexical units (ULs) are identified in the corpora and afterwards micro thematic functions are ascribed to related syntactic elements and the frames evoked by those ULs are set up. The sentences used in the analysis were taken from the FIFA corpora and Centenfolha, using the tool Word Sketch from Sketch Engine. The paper about description of lexical frames in matter contributed to the ultimate goal of the 2014 World Cup Project, which is to describe the conceptual structures related to the dominance of Tourism and Soccer in the three target languages of the dictionary.

Submitted by Renata Lopes (renatasil82@gmail.com) on 2016-08-02T14:54:54Z No. of bitstreams: 1 gabrieladasilvapires.pdf: 3876837 bytes, checksum: 354396fd2c6464588ee00023a54dc85b (MD5) / Approved for entry into archive by Diamantino Mayra (mayra.diamantino@ufjf.edu.br) on 2016-08-04T10:56:44Z (GMT) No. of bitstreams: 1 gabrieladasilvapires.pdf: 3876837 bytes, checksum: 354396fd2c6464588ee00023a54dc85b (MD5) / Made available in DSpace on 2016-08-04T10:56:44Z (GMT). No. of bitstreams: 1 gabrieladasilvapires.pdf: 3876837 bytes, checksum: 354396fd2c6464588ee00023a54dc85b (MD5) Previous issue date: 2016-06-20 / Enunciados como (1) “até que para um carro 1.0 ele anda bem na estrada” e (2) “PRA QUEM JÁ MORDEU UM CACHORRO POR COMIDA, ATÉ QUE EU CHEGUEI LONGE” têm nos chamado a atenção pelo fato de expressarem um tipo de concessividade (frustração de expectativas) ao passo que um elemento é colocado em evidência em comparação ao frame ao qual é vinculado. Ancorados nos pressupostos da Gramática das Construções (GOLDBERG, 1995, 2006; FILLMORE; LEE-GOLDMAN & RHOMIEUX, 2012), Teoria dos Espaços Mentais (FAUCONNIER, 1994, 1997; FAUCONNIER & TURNER, 2002) e igualmente norteados pelas abordagens sobre concessividade (KÖNIG & SIEMUND, 2000) e comparação linguística (HASEGAWA et al, 2010), buscamos legitimar esses tipos de ocorrências como instâncias do que chamamos de Construção Concessivo-comparativa Anteposta Enfática (CCCAE), em (1), esquematizada por “ATÉ QUE PARA X, Y”; e Construção Concessivo-comparativa Anteposta Simples (CCCAS), em (2), esquematizada por “PARA X, Y”. Nesses esquemas, “X” é marcado como um Sintagma Nominal preferencialmente Indefinido, que gera certas expectativas sobre frames; e “Y” contém um comentário avaliativo contrário às expectativas. Nesse sentido, contribuem para a pesquisa os estudos sobre Semântica de Frames (FILLMORE, 1982) e estudos complementares sobre Avaliatividade (MARTIN & WHITE, 2005). Seguindo uma metodologia empírica, constituímos um banco de dados, por meio de busca em três domínios da internet: “abril.com.br”, “blogspot.com.br” e “br.answers.yahoo.com”, com as seguintes expressões de busca: “(até que) para/pra alguém”; “(até que) para/pra quem”; “(até que) para/pra um”; “(até que) para/pra uma”. Ao final, obtivemos 19 ocorrências válidas em CCCAS e 385 ocorrências válidas em CCCAE. A partir das análises dos dados, podemos considerar alguns aspectos das construções. Foi verificado que, quando configuradas com correferencialidade (direta ou indireta) entre X e o elemento comentado em Y, essas construções atuam tipicamente para estabelecer uma relação concessivo-comparativa entre a expectativa gerada a partir de X e a quebra de expectativa feita no comentário em Y, sobre um elemento que atua como membro menos prototípico da categoria acionada pelo frame evocado em X. Quanto à CCCAE, os frames mais evocados nos comentários feitos em Y – Desirability, Success_or_failure e Aesthetics – são esquemas conceptuais ligados à avaliação, o que é um forte indício de que esta construção tenha caráter avaliativo. Como recurso avaliativo, a CCCAE é mais comumente usada para apreciar a qualidade de produtos e julgar comportamento humano em relação à capacidade. Tanto CCCAE como CCCAS parecem promover um processo de mesclagem conceptual, que relaciona escalas entre os espaços mentais e promove o surgimento de categoria ad hoc, em que a avaliação feita é relativizada a um grupo de expectativas; podendo variar entre o elogio e a crítica. Assim, essas construções, ao promoverem um tipo de crítica velada, mostram-se fortes recursos argumentativos e avaliativos da língua. / Utterances as (1) “até que para um carro 1.0 ele anda bem na estrada” and (2) “PRA QUEM JÁ MORDEU UM CACHORRO POR COMIDA, ATÉ QUE EU CHEGUEI LONGE” call our attention to the fact that they express a kind of concessivity (frustration of expectations) whilst an element is put in evidence as compared to the frame to which it is bound. Based on the assumptions of Construction Grammar (GOLDBERG, 1995, 2006; FILLMORE;-LEE GOLDMAN & RHOMIEUX, 2012), Mental Space theory (FAUCONNIER, 1994, 1997; FAUCONNIER & TURNER, 2002), and also approaches to concessivity (KÖNIG & SIEMUND, 2000) and linguistic comparison (HASEGAWA et al, 2010), we aim to legitimize these kinds of occurrences as instances of Emphatic Prefixed Concessive-comparative Construction (henceforth CCCAE), in (1), schematized by “ATÉ QUE PARA X, Y”; and Simple Prefixed Concessive-comparative Construction (henceforth CCCAS), in (2), schematized by “PARA X, Y”. In these schemes, “X” is marked as a Noun Phrase preferably Indefinite, which creates certain expectations of frames; and “Y” contains an evaluative comment contrary to expectations. In this sense, studies on Frame Semantics (FILLMORE, 1982) and further studies on Appraisal (MARTIN & WHITE, 2005) contribute to the research. Following an empirical methodology, we set up a database by searching in three areas of the Internet: “abril.com.br”, “blogspot.com.br” and “br.answers.yahoo.com” with the following search expressions: “(até que) para/pra alguém”; “(até que) para/pra quem”; “(até que) para/pra um”; “(até que) para/pra uma”. At the end, we obtained 19 valid occurrences of CCCAS and 385 valid occurrences of CCCAE. From the data analysis, we can consider some aspects of the constructions. When there is coreferenciality (directly or indirectly) between X and the element mentioned in Y, these constructions typically act to establish a concessivecomparative relationship between the expectation generated from X and the break of expectations in the comment in Y, which is about an element that acts as a less prototypical member of the category driven by the evoked frame in X. As for CCCAE, the three most mentioned frames in Y – Desirability, Success_or_failure and Aesthetics – are conceptual schemes linked to the evaluation, which is a strong indication that this construction has an evaluative character. As an evaluative resource, CCCAE is most commonly used to assess the quality of products and judge human behavior in relation to capacity. Both CCCAE and CCCAS seem to promote a conceptual blending process, which relates scales between mental spaces and promotes the emergence of ad hoc category, in which the evaluation is relativized to a group of expectations; ranging between praise and criticism. As these constructions promote a kind of veiled criticism, they act as powerful argumentative and evaluative linguistic resources.

Submitted by Renata Lopes (renatasil82@gmail.com) on 2017-06-06T17:20:33Z No. of bitstreams: 1 paulovictorhauckdaencarnacao.pdf: 5047447 bytes, checksum: ecb025b511614ad78ad5b157ec438634 (MD5) / Approved for entry into archive by Adriana Oliveira (adriana.oliveira@ufjf.edu.br) on 2017-06-07T11:03:00Z (GMT) No. of bitstreams: 1 paulovictorhauckdaencarnacao.pdf: 5047447 bytes, checksum: ecb025b511614ad78ad5b157ec438634 (MD5) / Made available in DSpace on 2017-06-07T11:03:00Z (GMT). No. of bitstreams: 1 paulovictorhauckdaencarnacao.pdf: 5047447 bytes, checksum: ecb025b511614ad78ad5b157ec438634 (MD5) Previous issue date: 2014-09-04 / CAPES - Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / O projeto FrameNet é um projeto desenvolvido pelo International Computer Science Institute (ICSI) em Berkeley, com o objetivo de documentar frames da língua inglesa baseando-se no conceito de frames semânticos da Inteligência Computacional. Sendo também estendido para outras línguas, como, por exemplo, o Projeto FrameNet-Br, de senvolvido na Universidade Federal de Juiz de Fora, que tem como foco a documentação dos frames linguísticos em português do Brasil. O recurso lexical construído pelo Frame Net pode ser utilizado em diversas aplicações de Processamento de Linguagem Natural, como traduções, sumarização, dentre outras. A utilização de tecnologias de web semân tica, como ontologias e dados ligados podem trazer diversos benefícios que colaboram com o compartilhamento das informações, principalmente considerando a possibilidade de uso de mecanismos de inferência para validação dos dados e obtenção de novas informações. Além disso, o uso de anotações semânticas em serviços Web também é considerado uma tecnologia promissora para facilitar a integração de recursos computacionais na web, ser vindo como mecanismo para facilitar a interação entre ferramentas de software. Estas anotações permitem que ferramentas possam compreender a estrutura dos serviços e como executá-los de maneira automática Desta forma, considerando as vantagens apresentadas por estas tecnologias, é possível associa-las de maneira a criar uma infraestrutura que permita a utilização de recursos lexicais em conjunto com recursos da web semântica para facilitar a compreensão e busca por informações em um dado domínio. Neste trabalho foi especificada uma infraestrutura baseada em serviços, que busca aliar as tecnologias da web semântica aos dados da base do projeto FrameNet, a partir da hipótese de que a aplicação de tecnologias como ontologias, dados ligados, e anotações em serviços Web podem contribuir com a construção e reuso de recursos lexicais baseados na semântica de frames, sendo estas contribuições tanto voltadas para a confiabilidade dos dados, quanto para o enriquecimento das informações mantidas pela base lexical. / The FrameNet project is developed by the International Computer Sciences Institute (ICSI) in Berkeley, with the goal of documenting frames of the English language based on the concept of semantic frames of Computational Inteligence. The FrameNet has been well translated for other languages, such as the FrameNet-BR Project, developed at the Universidade Federal de Juiz de Fora (UFJF), which focus on the documentation of lin guistic frames in Portuguese from Brazil. The lexical resource built by FrameNet can be used in many applications of natural language processing, such translations, summariza tion, among others. The use of semantic web technologies, such ontologies and linked data can bring many benefits that contribute to the sharing of information, specially con sidering the possibility of the use of inference mechanisms to validate data and retrieve new information. The use of semantic annotations for Web services is also considered a promising technology to facilitate the integration of computational resource on the web, serving as mechanism to facilitate interaction between software tools. These annotations allow tools understand the structure of services and how to execute them automatically. Thus, considering the advantages presented by these technologies, it is possible to asso ciate them in order to create an infrastructure that enables the use of lexical resources in conjunction with semantic web resources to facilitate understanding and search for information in a given domain. In this work was specified a service-based infrastructure that seeks to combine the technologies of semantic web to the database of the FrameNet project, considering the hypothesis that the application of technologies such as ontologies, linked data, and annotations in Web services can contribute to the construction and reuse of lexical resources based on semantic frames, and these contributions are both related to the reliability of data, and for the enrichment of the his information kept by lexical base.

Le présent travail se donne comme objectif de proposer une méthode de description complète des unités terminologiques du domaine du droit de l’Internet. Il s’agit d’un modèle hybride permettant de refléter à la fois la dimension linguistique et conceptuelle des termes et plus précisément, les relations lexico-sémantiques et les liens conceptuels que ces derniers entretiennent avec d’autres unités terminologiques et lexicales appartenant au vocabulaire du domaine et ceci par le biais de leur comportement dans l’univers discursif. Le projet, que nous avons baptisé DITerm, s’inscrit donc dans une démarche résolument descriptive dont l’ambition est de rendre compte des usages observés dans la langue de spécialité donnée. Le modèle en question est créé à partir d’un corpus de textes juridiques d’environ 5 000 000 mots et doit servir de base à la conception d’un dictionnaire spécialisé d’encodage destiné aux traducteurs dont la langue de travail est le français. Le terme, tel qu’il est envisagé dans le cadre de cette étude, constitue une unité à dimensions multiples qui doit être considérée sous différents angles. Sa description est donc basée sur l’analyse de ses relations lexicales, conceptuelles et contextuelles et s’inspire largement des modèles lexicographiques. Le projet DITerm a pour arrière-plan théorique la Lexicologie Explicative et Combinatoire, composante lexicologique de la théorie Sens-Texte (Mel’čuk et al. 1995) ainsi que la théorie des cadres et ses différentes applications (Minsky, Fillmore, Martin). La méthodologie adoptée dans ce projet s’inscrit dans le cadre de la linguistique de corpus (Sinclair). Le modèle DITerm propose de décrire chaque terme comme le centre d’une constellation autour de laquelle gravitent de nombreux termes coordonnés formant ce que Cornu (2002) appelle des familles opérationnelles. Pour modéliser les relations que partage le terme vedette avec les unités qui ont tendance à apparaître dans son univers contextuel, nous avons recours à des formules explicites fondées sur un métalangage compréhensible pour tous. Ce formalisme, basé sur la paraphrase, correspond en partie à la vulgarisation des fonctions lexicales. Il présente par ailleurs l’avantage de pouvoir s’adapter à la représentation des liens d’ordre conceptuel. Deplus, la description est enrichie par la mise en place d’un modèle d’annotation des contextes au moyen des cadres sémantiques à la FrameNet permettant de mettre en évidence les liens indirects entretenus par les termes / The aim of this work is to propose a holistic description of terminological units in the field of Internet law, based on a hybrid model that reflectis both the linguistic and the conceptual dimension of terms. To be more precise, this model aims to point out the lexical-semantic relationships and conceptual links that those terms maintain with other terminological and lexical units belonging to the Internet law vocabulary, by studying their behavior in the discursive universe. The project (which we call DITerm) is part of a descriptive approach whose ambition is to explain observed usage in the specialized language at hand. The model was created using acorpus made up of legal texts which totals about 5 000,000 words. The model will be used as the basis for the conception of a specialized encoding dictionary. This dictionary is intended for translators with French language as their working language. The term, as it is viewed in this study, is a multi-dimensional unity which should be considered from many angles. Its description is hence based on the analysis of its lexical, conceptual and contextual relationships which is widely inspired by lexicographical models. Thetheoretical background of the DITerm project is Explanatory and Combinatorial Lexicology, the lexicological component of Meaning-Text Theory (Mel’čuk et al. 1995). We also draw inspiration from the frame theory and its different applications (Minsky’s Frame System Theory,Fillmore’s Frame Semantics). The methodology adopted in this project fits into the framework of corpus linguistics (Sinclair).The DITerm model proposes to describe each term as the center of a constellation around which several coordinated terms gravitate, creating what Cornu (2002) calls operational families. In order to model the relationships the term shares with the units which tend to appear in its contextual universe, we rely on several explicit formulas based on a broadly comprehensible metalanguage. This formalism, which is based on paraphrase, correlates partly with the popularization of the lexical functions. On the other hand, it has the advantage of adapting to the representation of conceptual links. In addition, the description is enriched by the implementation of an annotation model of the contexts in which the terms occur, which makes it possible to account for indirect links that are maintained by terms. This annotation model is an application of the FrameNet methodology.

Ce projet se décline en deux parties. Dans un premier temps, il s’agit de développer une méthode de conversion automatique des textes annotés selon la sémantique des cadres dans FrameNet en représentations sémantiques de la Théorie Sens-Texte, afin de développer davantage de ressources informatiques pour assurer le développement de différents projets, notamment le réalisateur de textes GenDR. Dans un second temps, cette conversion sera mise à profit pour effectuer une analyse comparative entre les deux formalismes. Nous retiendrons que ces formalismes ne sont pas incompatibles, mais diffèrent par leurs niveaux de granularité et leurs objectifs propres. Nous tracerons quelques parallèles entre les fonctions lexicales et les relations entre cadres, et proposerons une mise en commun des formalismes afin de les enrichir. / This project is divided in two main parts. Firstly, a method allowing for an automatic conversion of FrameNet's Semantics-based text annotations into semantic representations, according to the Meaning-Text Theory framework, will be presented. This method will lead to an increased set of data usable to develop and improve various Meaning-Text Theory-based projects, including GenDR, a text realizer. Secondly, the conversion task will be used to do a comparative analysis of the two frameworks. We will conclude that the two frameworks are not incompatible, but differ in their granularity and goals. We will also draw parallels between the lexical functions and frame- to-frame relationships, and make some suggestions regarding changes to the frameworks in order to enrich them.

Meaning has been established pervasively as a central concept throughout disciplines that were involved in cognitive revolution. Its metaphoric usage comes to be, first and foremost, through the interpreter’s constraint: representational relationships and contents are considered to be in the “eye” or mind of the observer and shared properties among observers themselves are knowable through interlinguistic phenomena, such as translation. Despite the instability of meaning in relation to its underdetermination by reference, it can be a tertium comparationis or “third comparator” for extended human cognition if gauged through invariants that exist in transfer processes such as translation, as all languages and cultures are rooted in pan-human experience and, thus, share and express species-specific ontology. Meaning, seen as a cognitive competence, does not stop outside of the body but extends, depends, and partners with other agents and the environment. A novel approach for exploring the transfer properties of some constituent items of the original natural semantic metalanguage in English, that is, semantic primitives, is presented: FrameNet’s semantic frames, evoked by the primes SEE and FEEL, were extracted from EuroParl, a parallel corpus that allows for the automatic word alignment of items with their synonyms. Large Ontology Multilingual Extraction was used. Afterward, following the Semantic Mirrors Method, a procedure that consists back-translating into source language, a translatological examination of translated and original versions of items was performed. A fully automated pipeline was designed and tested, with the purpose of exploring associated frame shifts and, thus, beginning a research agenda on their alleged universality as linguistic features of translation, which will be complemented with and contrasted against further massive feedback through a citizen science approach, as well as cognitive and neurophysiological examinations. Additionally, an embodied account of frame semantics is proposed.

No description available.

L’annotation en rôles sémantiques est une tâche qui permet d’attribuer des étiquettes de rôles telles que Agent, Patient, Instrument, Lieu, Destination etc. aux différents participants actants ou circonstants (arguments ou adjoints) d’une lexie prédicative. Cette tâche nécessite des ressources lexicales riches ou des corpus importants contenant des phrases annotées manuellement par des linguistes sur lesquels peuvent s’appuyer certaines approches d’automatisation (statistiques ou apprentissage machine). Les travaux antérieurs dans ce domaine ont porté essentiellement sur la langue anglaise qui dispose de ressources riches, telles que PropBank, VerbNet et FrameNet, qui ont servi à alimenter les systèmes d’annotation automatisés. L’annotation dans d’autres langues, pour lesquelles on ne dispose pas d’un corpus annoté manuellement, repose souvent sur le FrameNet anglais. Une ressource telle que FrameNet de l’anglais est plus que nécessaire pour les systèmes d’annotation automatisé et l’annotation manuelle de milliers de phrases par des linguistes est une tâche fastidieuse et exigeante en temps. Nous avons proposé dans cette thèse un système automatique pour aider les linguistes dans cette tâche qui pourraient alors se limiter à la validation des annotations proposées par le système. Dans notre travail, nous ne considérons que les verbes qui sont plus susceptibles que les noms d’être accompagnés par des actants réalisés dans les phrases. Ces verbes concernent les termes de spécialité d’informatique et d’Internet (ex. accéder, configurer, naviguer, télécharger) dont la structure actancielle est enrichie manuellement par des rôles sémantiques. La structure actancielle des lexies verbales est décrite selon les principes de la Lexicologie Explicative et Combinatoire, LEC de Mel’čuk et fait appel partiellement (en ce qui concerne les rôles sémantiques) à la notion de Frame Element tel que décrit dans la théorie Frame Semantics (FS) de Fillmore. Ces deux théories ont ceci de commun qu’elles mènent toutes les deux à la construction de dictionnaires différents de ceux issus des approches traditionnelles. Les lexies verbales d’informatique et d’Internet qui ont été annotées manuellement dans plusieurs contextes constituent notre corpus spécialisé. Notre système qui attribue automatiquement des rôles sémantiques aux actants est basé sur des règles ou classificateurs entraînés sur plus de 2300 contextes. Nous sommes limités à une liste de rôles restreinte car certains rôles dans notre corpus n’ont pas assez d’exemples annotés manuellement. Dans notre système, nous n’avons traité que les rôles Patient, Agent et Destination dont le nombre d’exemple est supérieur à 300. Nous avons crée une classe que nous avons nommé Autre où nous avons rassemblé les autres rôles dont le nombre d’exemples annotés est inférieur à 100. Nous avons subdivisé la tâche d’annotation en sous-tâches : identifier les participants actants et circonstants et attribuer des rôles sémantiques uniquement aux actants qui contribuent au sens de la lexie verbale. Nous avons soumis les phrases de notre corpus à l’analyseur syntaxique Syntex afin d’extraire les informations syntaxiques qui décrivent les différents participants d’une lexie verbale dans une phrase. Ces informations ont servi de traits (features) dans notre modèle d’apprentissage. Nous avons proposé deux techniques pour l’identification des participants : une technique à base de règles où nous avons extrait une trentaine de règles et une autre technique basée sur l’apprentissage machine. Ces mêmes techniques ont été utilisées pour la tâche de distinguer les actants des circonstants. Nous avons proposé pour la tâche d’attribuer des rôles sémantiques aux actants, une méthode de partitionnement (clustering) semi supervisé des instances que nous avons comparée à la méthode de classification de rôles sémantiques. Nous avons utilisé CHAMÉLÉON, un algorithme hiérarchique ascendant. / Semantic role annotation is a process that aims to assign labels such as Agent, Patient, Instrument, Location, etc. to actants or circumstants (also called arguments or adjuncts) of predicative lexical units. This process often requires the use of rich lexical resources or corpora in which sentences are annotated manually by linguists. The automatic approaches (statistical or machine learning) are based on corpora. Previous work was performed for the most part in English which has rich resources, such as PropBank, VerbNet and FrameNet. These resources were used to serve the automated annotation systems. This type of annotation in other languages for which no corpora of annotated sentences are available often use FrameNet by projection. Although a resource such as FrameNet is necessary for the automated annotation systems and the manual annotation by linguists of a large number of sentences is a tedious and time consuming work. We have proposed an automated system to help linguists in this task so that they have only to validate annotations proposed. Our work focuses on verbs that are more likely than other predicative units (adjectives and nouns) to be accompanied by actants realized in sentences. These verbs are specialized terms of the computer science and Internet domains (ie. access, configure, browse, download) whose actantial structures have been annotated manually with semantic roles. The actantial structure is based on principles of Explanatory and Combinatory Lexicology, LEC of Mel’čuk and appeal in part (with regard to semantic roles) to the notion of Frame Element as described in the theory of frame semantics (FS) of Fillmore. What these two theories have in common is that they lead to the construction of dictionaries different from those resulting from the traditional theories. These manually annotated verbal units in several contexts constitute the specialized corpus that our work will use. Our system designed to assign automatically semantic roles to actants is based on rules and classifiers trained on more than 2300 contexts. We are limited to a restricted list of roles for certain roles in our corpus have not enough examples manually annotated. In our system, we addressed the roles Patient, Agent and destination that the number of examples is greater than 300. We have created a class that we called Autre which we bring to gether the other roles that the number of annotated examples is less than 100. We subdivided the annotation task in the identification of participant actants and circumstants and the assignment of semantic roles to actants that contribute to the sense of the verbal lexical unit. We parsed, with Syntex, the sentences of the corpus to extract syntactic informations that describe the participants of the verbal lexical unit in the sentence. These informations are used as features in our learning model. We have proposed two techniques for the task of participant detection: the technique based in rules and machine learning. These same techniques are used for the task of classification of these participants into actants and circumstants. We proposed to the task of assigning semantic roles to the actants, a partitioning method (clustering) semi supervised of instances that we have compared to the method of semantic role classification. We used CHAMELEON, an ascending hierarchical algorithm.