Verb classifications have attracted a great deal of interest in both linguistics and natural language processing (NLP). They have proved useful for important tasks and applications, including e.g. computational lexicography, parsing, word sense disambiguation, semantic role labelling, information extraction, question-answering, and machine translation (Swier and Stevenson, 2004; Dang, 2004; Shi and Mihalcea, 2005; Kipper et al., 2008; Zapirain et al., 2008; Rios et al., 2011). Particularly useful are classes which capture generalizations about a range of linguistic properties (e.g. lexical, (morpho-)syntactic, semantic), such as those proposed by Beth Levin (1993). However, full exploitation of such classes in real-world tasks has been limited because no comprehensive or domain-specific lexical classification is available. This thesis investigates how Levin-style lexical semantic classes could be learned automatically from corpus data. Automatic acquisition is cost-effective when it involves either no or minimal supervision and it can be applied to any domain of interest where adequate corpus data is available. We improve on earlier work on automatic verb clustering. We introduce new features and new clustering methods to improve the accuracy and coverage. We evaluate our methods and features on well-established cross-domain datasets in English, on a specific domain of English (the biomedical) and on another language (French), reporting promising results. Finally, our task-based evaluation demonstrates that the automatically acquired lexical classes enable new approaches to some NLP tasks (e.g. metaphor identification) and help to improve the accuracy of existing ones (e.g. argumentative zoning).

The aim of this thesis was to study change in the meaning of the word cool, comparing the 1890s, the 1950s and the 2010s. Examples where the word cool was used were obtained from the Corpus of Historical American English, which is created by Mark Davies, and analyzed in terms of in which senses and text types they were used. The results show that the biggest meaning change in cool between the 1890s and the 2010s is that the most prominent sense in the 1890s is ‘calm’, and ‘slightly cold in a pleasant way’ not far after, and in the 2010s the most prominent sense was ‘pleasant, enjoyable or likable’. The most frequent senses of cool in the 1950s were ‘slightly cold in a pleasant way’, ‘temperature that is too cold’, and ‘calm’ as the most commonly used. The most prominent text type where the word cool occurred was fiction in all three analyzed decades.

No description available.

The thesis provides an in-depth analysis of motion event descriptions of two Australian indigenous languages. Jaminjung is a highly endangered non Pama-Nyungan language with approximately 50 remaining speakers. Kriol, an English-lexified Creole, is spoken by about 20.000 people in different varieties across northern Australia. While the languages are typologically very different, occupancy of the same linguistic and cultural area provides an intriguing opportunity to examine the effects of culture and language contact on conceptual components and distribution patterns in discourse. This investigation also applies and tests a number of existing frameworks and typologies regarding the linguistic encoding of motion and space in general. The thesis first provides an overview of the encoding of motion event descriptions in Jaminjung and Kriol. It becomes clear that, concerning overt marking of case, ground-encodings follow a systematic semantic pattern with no or rare case-marking for deictic terms, optional marking for toponyms and mandatory marking for all other types of landmarks. Furthermore, the structure and semantics of the motion verb phrase is investigated. Particularly noteworthy here is a study of asymmetrical serial verb constructions in Kriol which revealed a number of previously undescribed types. Following this, various proposals for a typology of Frames of Reference are applied. The notion of ‘anchor’ is at the centre of the analysis. The investigation shows that contextual restrictions for the use of Jaminjung’s absolute terms can be accounted for by a restriction on egocentric anchoring and ‘Orientation’ settings only. Furthermore, absolute Frame of Reference is realised differently in Roper and Westside Kriol respectively, suggesting an ongoing influence of the traditional languages spoken by the respective communities rather than the lexifier English. Jaminjung and Kriol, additionally, prefer the use of absolute over relative Frame of Reference. The following chapter investigates how lexicalisation patterns influence the distribution of path and manner encodings in discourse. After concluding that Jaminjung might best be described as following an equipollently-framed pattern and Kriol as satellite-framed, path and manner salience is investigated in different types of discourse using a dataset of motion event encodings in a Frog Story collection and a general corpus of various discourse environments. It is concluded that while the two languages behave very differently with regards to frequency patterns of ground- and other path-encodings, they show remarkable similarities in distributing path and manner over larger chunks of discourse. These findings suggest that cultural influences may sometimes override structural typological constraints. Finally, motion event encodings in specific types of discourse are analysed. Regarding route descriptions, speakers show a clear preference for dynamic over static modes of presentation. This includes encoding ‘fictive motion’ events for which a figure- and ground-based distinction is introduced. Additionally, concerning the use of deictics in a comparative analysis of different types of corpora for both languages, it was shown that the distribution of absolute terms remains stable across discourse environments while deictic usage differs drastically. Lastly, the concept of ‘motion’ is abstracted and described as a kind of structuring device in narratives. It is shown that the ‘journey’ within the story world is used by speakers of both languages to bridge episodes sometimes even overriding a temporal in favour of a spatial order of events.

Dans cette thèse, nous abordons du point de vue linguistique la question de l'évaluation des bases lexicales extraites par analyse distributionnelle automatique (ADA). Les méthodes d'évaluation de ces ressources qui sont actuellement mises en œuvre (comparaison à des lexiques de référence, évaluation par la tâche, test du TOEFL...) relèvent en effet d'une approche quantitative des données qui ne laisse que peu de place à l'interprétation des rapprochements générés. De ce fait, les conditions qui font que certains couples de mots sont extraits alors que d'autres ne le sont pas restent mal connues. Notre travail vise une meilleure compréhension des fonctionnements en corpus qui régissent les rapprochements distributionnels. Pour cela, nous avons dans un premier temps adopté une approche quantitative qui a consisté à comparer plusieurs ressources distributionnelles calculées sur des corpus différents à des lexiques de références (le Dictionnaire électronique des synonymes du CRISCO et le réseau lexical JeuxDeMots). Cette étape nous a permis, premièrement, d'avoir une estimation globale du contenu de nos ressources, et, deuxièmement, de sélectionner des échantillons de couples de mots à étudier d'un point de vue qualitatif. Cette deuxième étape constitue le cœur de la thèse. Nous avons choisi de nous focaliser sur les relations lexico-sémantiques que sont la synonymie, l'antonymie, l'hyperonymie et la méronymie, que nous abordons en mettant en place quatre protocoles différents. En nous appuyant sur les relations contenues dans les lexiques de référence, nous avons comparé les propriétés distributionnelles des couples de synonymes/antonymes/hyperonymes/méronymes qui ont été extraits par l'ADA avec celles des couples qui ne l'ont pas été. Nous mettons ainsi au jour plusieurs phénomènes qui favorisent ou bloquent la substituabilité des couples de mots (donc leur extraction par l'ADA). Ces phénomènes sont considérés au regard de paramètres comme la nature du corpus qui a permis de générer les bases distributionnelles étudiées (corpus encyclopédique, journalistique ou littéraire) ou les limites des lexiques de référence. Ainsi, en même temps qu'il questionne les méthodes d'évaluation des bases distributionnelles actuellement employées, ce travail de thèse illustre l'intérêt qu'il y a à considérer ces ressources comme des objets d'études linguistiques à part entière. Les bases distributionnelles sont en effet le résultat d'une mise en œuvre à grande échelle du principe de substituabilité, ce qui en fait un matériau de choix pour la description des relations lexico-sémantiques. / In this thesis, we address the question of the evaluation of distributional thesauri from a linguistic point of view. The most current ways to evaluate distributional methods rely on the comparison with gold standards like WordNet or semantic tasks like the TOEFL test. However, these evaluation methods are quantitative and thus restrict the possibility of performing a linguistic analysis of the distributional neighbours. Our work aims at a better understanding of the distributional behaviors of words in texts through the study of distributional thesauri. First, we take a quantitative approach based on a comparison of several distributional thesauri with gold standards (the DES - a dictionary of synonyms - and JeuxDeMots - a crowdsourced lexical network). This step allowed us to have an overview of the nature of the semantic relations extracted in our distributional thesauri. In a second step, we relied on this comparison to select samples of distributional neighbours for a qualitative study. We focused on "classical" semantic relations, e.g. synonymy, antonymy, hypernymy and meronymy. We considered several protocols to compare the properties of the couples of distributional neighbours which were found in the gold standards and the others. Thus, taking into account parameters like the nature of the corpora from which were generated our distributional thesauri, we explain why some synonyms, hypernyms, etc. can be substituted in texts while others cannot. The purpose of this work is twofold. First, it questions the traditional evaluation methods, then it shows how distributional thesauri can be used for the study of semantic relations.

Made available in DSpace on 2014-06-11T19:22:19Z (GMT). No. of bitstreams: 0 Previous issue date: 2009-02-20Bitstream added on 2014-06-13T19:27:30Z : No. of bitstreams: 1 kasama_dy_me_sjrp.pdf: 2392559 bytes, checksum: 97410d6af14f5587b05c598ca84a15f0 (MD5) / Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) / O Processamento de Língua Natural (ou PLN) tem sido objeto de estudo de pesquisadores das mais diversas áreas do conhecimento. O léxico é, sem sombra de dúvida, elemento essencial para o tratamento automático de dados lingüísticos, sendo a sua análise semântica fator crucial para um efetivo processamento computacional que, não raro, encontra barreiras em questões ligadas a uma representação semântica eficaz e que permita ser representada em linguagem de máquina. Na Terminologia, esse tratamento semântico favorece o estabelecimento de relações existentes entre unidades lexicais especializadas, e determina a elaboração de definições terminológicas coerentes e representativas ao campo de especialidade ao qual pertencem. Nesse sentido, propomos neste trabalho traçar uma estrutura conceitual do domínio da Nanociência e Nanotecnologia, em língua portuguesa do Brasil, visando a criação do que modernamente se conhece por ontologias, cujos preceitos nortearam o desenvolvimento desta pesquisa. Aliada a essas práticas, encontra-se a importância da adoção de um modelo que permita representar formalmente as relações semânticas existentes entre os diversos termos que compõem essa área técnico-científica. A busca por essas unidades lexicais especializadas e suas relações deu-se em um córpus formado por textos de tipologia diversa, com o auxílio de ferramentas computacionais – de extração semiautomática de termos e um processador de córpus. A modelagem do domínio em questão e sua representação em uma linguagem corrente e atual (a saber, a linguagem OWL) fez-se com o auxílio da ferramenta Protégé. Defende-se neste trabalho a necessidade, cada vez mais crescente, da adoção de métodos eficazes para o delineamento de estruturas conceituais a fim de executar tarefas computacionais utilizando informação lingüística. Espera-se ainda... / Natural Language Processing (or NLP) has been an object of study by researchers from different fields of knowledge. Lexicon is undoubtedly an essential element for the automatic processing of language data, and its semantic analysis is a crucial factor for an effective computational processing that as often as not finds barriers in matters concerning a productive semantic representation in machine language. In terminology, this semantic treatment favors the establishment of relations between specialized lexical units and determines the development of consistent terminological definitions that may represent the field of expertise to which they belong. Accordingly, what we propose in this work is to provide a conceptual structure of the specialized subject field of Nanoscience and Nanotechnology, in Brazilian Portuguese language, aimed at creating the modernly so-called ontologies, whose principles guided the development of this research. In addition to such practices, it is important to adopt a model which allows a formal representation of the semantic relations between the terms in this domain. The specialized lexical units and the semantic relations were extracted semiautomatically from a corpus, compiled with different types of texts, using a term extractor and a corpus processor. The modeling of the area concerned and its representation in a current language (i.e., OWL language) was possible with the aid of Protégé tool. We support in this work the increasing need to adopt effective methods for the design of conceptual structures in order to carry computational tasks using linguistic information. We hope that this work will strengthen the dialogue between linguists, computational and information scientists.

The thesis belongs to the field of lexical semantics studies, associated with describing the Russian linguistic world-image. The research focuses on the universal situation of purchase and sale as reflected in the Russian lexical standard and sub-standard. The work deals also with subjects related to the sphere of social linguistics: the social stratification of the language, the structure of sub-standard, etc. The thesis is a contribution to the description of the Russian linguistic world-image as well as to the further elaboration of the conceptional analysis method. The results are applicable in teaching Russian as a foreign language, particularly in lexis and Russian culture and mentality studies.

Στην παρούσα μεταπτυχιακή εργασία, αρχικά, γίνεται λόγος για τα γενικά τυπικά χαρακτηριστικά που διέπουν το άσεμνο λεξιλόγιο, κυρίως της Ελληνικής. Στη συνέχεια, επιχειρείται μια μορφολογική ανάλυση των συστατικών που χρησιμοποιούνται για τον σχηματισμό των λέξεων του άσεμνου λεξιλογίου. Ωστόσο, βασικός στόχος είναι η διερεύνηση των σημασιολογικών και πραγματολογικών χαρακτηριστικών που διέπουν αυτό το ιδιαίτερα ενδιαφέρον τμήμα του λεξιλογίου μας. Παράλληλα, ασχολούμαι με περιπτώσεις υποκορισμού ή μεγέθυνσης μέσα στο άσεμνο λεξιλόγιο, με τους μηχανισμούς δανεισμού, με φαινόμενα ευφημισμού και δυσφημισμού και άλλα. Αναλυτικότερα, στο πρώτο κεφάλαιο, το οποίο λειτουργεί ως εναρκτήριο στάδιο ένταξης στο αντικείμενο μελέτης της εργασίας, παρουσιάζω τα περιθωριακά ιδιώματα της ελληνικής καθώς και τη δομή και λειτουργία των ειδικών λεξιλογίων στα οποία εντάσσεται και το προς εξέταση λεξιλόγιο. Στο τέλος του κεφαλαίου εξετάζω τις διαδικασίες και τους μηχανισμούς δανεισμού που υιοθετούνται για την εισαγωγή νέων λέξεων στην Ελληνική από άλλες γλώσσες, οι οποίες επηρέασαν και εμπλούτισαν με αυτόν τον τρόπο το άσεμνο λεξιλόγιο. Στο δεύτερο κεφάλαιο ασχολούμαι με τις διαδικασίες σχηματισμού των λημμάτων του άσεμνου λεξιλογίου. Μορφολογικές διαδικασίες σχηματισμού λέξεων όπως είναι η παραγωγή και η σύνθεση, αλλά και δομές όπως οι συμφυρμοί, λεξικοποιημένες εκφράσεις κ. ά, θα με απασχολήσουν ιδιαίτερα στο κεφάλαιο αυτό. Το τρίτο κεφάλαιο είναι αφιερωμένο στις εννοιακές σχέσεις, στις σχέσεις δηλαδή των λέξεων με άλλες λέξεις, με τις οποίες σχετίζονται. Το βασικό θεωρητικό πρότυπο πάνω στο οποίο θα βασιστώ για την ανάλυση της λεξικής σημασιολογίας είναι αυτό που προτείνεται από τον Cruse (1986, 2004). Εννοιακές σχέσεις, όπως αυτή της πολυσημίας, της συνωνυμίας, της μετωνυμίας, της μερωνυμίας και της μεταφοράς θα με απασχολήσουν ιδιαίτερα. Στη συνέχεια, στο τελευταίο κεφάλαιο με τίτλο «Άσεμνο λεξιλόγιο και Πραγματολογία» αφού αναφερθώ στα κίνητρα που ωθούν στην χρήση αυτού του λεξιλογίου, θα εξετάσω τις λειτουργίες και τη χρήση των λέξεων αυτών μέσα σε συγκεκριμένα εκφωνήματα. Το περιβάλλον μέσα στο οποίο εμφανίζονται οι λέξεις αυτές αλλά και οι διαφορετικοί πολιτισμοί και η νοοτροπία των ομιλητών, όπως θα δούμε στο κεφάλαιο αυτό παίζει πολύ σημαντικό ρόλο. Ολοκληρώνοντας, θα επικεντρωθώ σε λέξεις που χαρακτηρίζονται ως άσεμνες και μπορούν να θεωρηθούν, με βάση πραγματολογικές μελέτες (μεταξύ αυτών Fraser, 2008) ότι λειτουργούν ως δείκτες οργάνωσης λόγου και συγκεκριμένα ως δείκτες προσοχής (attention markers). / Words are extremely important to people. They constitute a means of communication. It is through them that we express our thoughts, feelings and emotions. They are an integral part of our everyday life and they are with us wherever we go. Every word, of any kind, deserves our full attention and needs to be studied, no matter what kind of vocabulary it may be part of. The present thesis is about a very vibrant and expressive part of our language. It concerns “obscene vocabulary”, a domain on which very few scientific studies have been carried out to date. In this thesis I decided to examine the morphological, semantic and pragmatic aspects of the “obscene vocabulary” of the Modern Greek language. In the first chapter, I present the “marginal” jargons of the Modern Greek language, of which “obscene vocabulary” constitutes a part, based on the existing literature so far. Moreover, I propose an appropriate classification of the Greek data, based on Jay (1997). I also look into cases of euphemisms and calumny, and the way they are used in Modern Greek. In addition, I present some cases of loans concerning not only words, but expressions and even affixes coming from other languages as well. Loans and calques seem to constitute a great part of the Greek “obscene vocabulary”, most of them coming from Italian, French and Turkish. The second chapter is about the way that words and expressions of the “obscene vocabulary” are formed. It seems that this kind of vocabulary is highly productive in compound words. What is more, some prefixes and suffixes of the “purist Greek” which are not very productive in the Modern Greek common vocabulary, combined with popular words, they are frequently used to form “obscene” words, resulting in funny word formations. It should be noted here, that the formation of such words is subject to the same constraints as the rest of the vocabulary of the Modern Greek language. Finally, I look into compound expressions, lexicalized phrases, blends and the diminutive and augmentative suffixes and prefixes which are used in order to make a word sound less offensive, more familiar and even positive. The third chapter consists of a semantic analysis of the “obscene vocabulary” of the Modern Greek language, based on Cruse (1986, 2004) and Veloudis (2005). I look into polysemous and synonymous pairs of words and also into notions such as meronymy, metonymy and metaphors, concerning mostly parts of the human body. Finally, I try to analyze the pragmatic aspects of the Greek “obscene vocabulary” This field concerns the circumstances in which this kind of vocabulary is used, human communication, the intonation and the gestures that accompany the use of “obscene vocabulary” and every possible kind of social and psychological reasons, as well as the motives that make someone use this kind of vocabulary. Different languages reflect different cultures and have a different degree and way of using “obscene vocabulary”.

La compréhension du langage naturel repose souvent sur des raisonnements de sens commun, pour lesquels la connaissance de relations sémantiques, en particulier entre prédicats verbaux, peut être nécessaire. Cette thèse porte sur la problématique de l'utilisation d'une méthode distributionnelle pour extraire automatiquement les informations sémantiques nécessaires à ces inférences de sens commun. Des associations typiques entre des paires de prédicats et un ensemble de relations sémantiques (causales, temporelles, de similarité, d'opposition, partie/tout) sont extraites de grands corpus, par l'exploitation de la présence de connecteurs du discours signalant typiquement ces relations. Afin d'apprécier ces associations, nous proposons plusieurs mesures de signifiance inspirées de la littérature ainsi qu'une mesure novatrice conçue spécifiquement pour évaluer la force du lien entre les deux prédicats et la relation. La pertinence de ces mesures est évaluée par le calcul de leur corrélation avec des jugements humains, obtenus par l'annotation d'un échantillon de paires de verbes en contexte discursif. L'application de cette méthodologie sur des corpus de langue française et anglaise permet la construction d'une ressource disponible librement, Lecsie (Linked Events Collection for Semantic Information Extraction). Celle-ci est constituée de triplets: des paires de prédicats associés à une relation; à chaque triplet correspondent des scores de signifiance obtenus par nos mesures.Cette ressource permet de dériver des représentations vectorielles de paires de prédicats qui peuvent être utilisées comme traits lexico-sémantiques pour la construction de modèles pour des applications externes. Nous évaluons le potentiel de ces représentations pour plusieurs applications. Concernant l'analyse du discours, les tâches de la prédiction d'attachement entre unités du discours, ainsi que la prédiction des relations discursives spécifiques les reliant, sont explorées. En utilisant uniquement les traits provenant de notre ressource, nous obtenons des améliorations significatives pour les deux tâches, par rapport à plusieurs bases de référence, notamment des modèles utilisant d'autres types de représentations lexico-sémantiques. Nous proposons également de définir des ensembles optimaux de connecteurs mieux adaptés à des applications sur de grands corpus, en opérant une réduction de dimension dans l'espace des connecteurs, au lieu d'utiliser des groupes de connecteurs composés manuellement et correspondant à des relations prédéfinies. Une autre application prometteuse explorée dans cette thèse concerne les relations entre cadres sémantiques (semantic frames, e.g. FrameNet): la ressource peut être utilisée pour enrichir cette structure par des relations potentielles entre frames verbaux à partir des associations entre leurs verbes. Ces applications diverses démontrent les contributions prometteuses amenées par notre approche permettant l'extraction non supervisée de relations sémantiques. / Natural language understanding often relies on common-sense reasoning, for which knowledge about semantic relations, especially between verbal predicates, may be required. This thesis addresses the challenge of using a distibutional method to automatically extract the necessary semantic information for common-sense inference. Typical associations between pairs of predicates and a targeted set of semantic relations (causal, temporal, similarity, opposition, part/whole) are extracted from large corpora, by exploiting the presence of discourse connectives which typically signal these semantic relations. In order to appraise these associations, we provide several significance measures inspired from the literature as well as a novel measure specifically designed to evaluate the strength of the link between the two predicates and the relation. The relevance of these measures is evaluated by computing their correlations with human judgments, based on a sample of verb pairs annotated in context. The application of this methodology to French and English corpora leads to the construction of a freely available resource, Lecsie (Linked Events Collection for Semantic Information Extraction), which consists of triples: pairs of event predicates associated with a relation; each triple is assigned significance scores based on our measures. From this resource, vector-based representations of pairs of predicates can be induced and used as lexical semantic features to build models for external applications. We assess the potential of these representations for several applications. Regarding discourse analysis, the tasks of predicting attachment of discourse units, as well as predicting the specific discourse relation linking them, are investigated. Using only features from our resource, we obtain significant improvements for both tasks in comparison to several baselines, including ones using other representations of the pairs of predicates. We also propose to define optimal sets of connectives better suited for large corpus applications by performing a dimension reduction in the space of the connectives, instead of using manually composed groups of connectives corresponding to predefined relations. Another promising application pursued in this thesis concerns relations between semantic frames (e.g. FrameNet): the resource can be used to enrich this sparse structure by providing candidate relations between verbal frames, based on associations between their verbs. These diverse applications aim to demonstrate the promising contributions provided by our approach, namely allowing the unsupervised extraction of typed semantic relations.

Dem 15. Internationalen Slavistentag im August 2013 ging im Mai 2010 der 5. Internationale Weißrussisten-Kongress voraus, dessen Beiträge 2012 in drei Bänden in der Minsker Reihe „Беларусіка-Albaruthenica“ erschienen sind und zusammen mit einschlägigen Publikationen zum Slavistentag wie zum Beispiel dem unlängst von Gerd Hentschel herausgegebenen Beiträgen zu „Variation und Stabilität in Kontaktvarietäten. Beobachtungen zu gemischten Formen der Rede in Weißrussland, der Ukraine und Schlesien“ (Studia Slavica Oldenburgensia 21) einen guten Überblick über die aktuelle Weißrussistik geben. Der erste Band ist der Literaturwissenschaft und Folkloristik gewidmet («Літаратуразнаўства і фалькларыстыка», І. Э. Багдановіч (рэд.)), der zweite Band enthält Beiträge zur Sprachwissenschaft («Мовазнаўства», С. М. Запрудскі, Г. А. Цыхун (рэд.)), im dritten werden historische Themen und kulturologische Fragestellungen behandelt («Гісторыя і культуралогія», А.А. Суша (рэд.)).