Phrase-based statistical language generation using graphical models and active learning

Most previous work on trainable language generation has focused on two paradigms: (a) using a statistical model to rank a set of generated utterances, or (b) using statistics to inform the generation decision process. Both approaches rely on the existence of a handcrafted generator, which limits their scalability to new domains. This paper presents BAGEL, a statistical language generator which uses dynamic Bayesian networks to learn from semantically-aligned data produced by 42 untrained annotators. A human evaluation shows that BAGEL can generate natural and informative utterances from unseen inputs in the information presentation domain. Additionally, generation performance on sparse datasets is improved significantly by using certainty-based active learning, yielding ratings close to the human gold standard with a fraction of the data. © 2010 Association for Computational Linguistics.

Improving statistical machine translation by classifying and generalizing inflected verb forms

This paper introduces a rule-based classification of single-word and compound verbs into a statistical machine translation approach. By substituting verb forms by the lemma of their head verb, the data sparseness problem caused by highly-inflected languages can be successfully addressed. On the other hand, the information of seen verb forms can be used to generate new translations for unseen verb forms. Translation results for an English to Spanish task are reported, producing a significant performance improvement.

维持性血液透析患者认知功能特点研究

Objective: To study the episodic memory, semantic memory, cognitive planning ability and inhibition ability in MHD patients. Method: Neuropsychological research methods such as Action memory of verb-object phrase, Trail Making Test (A and B), Verbal Fluency Test, Go-No/Go test and Stroop Color Naming Task were used to investigate Episodic Memory 、Semantic Memory、Executive Function of 40 MHD and 40 NC. Self-Rating Anxiety Scale (SAS), Self-Rating Depression Scale(SDS), Social Support Scale, Life Satisfaction Scale, and biochemical examination were applied and their relationships with cognitive function were analized. The mean age and education level of MHD group and NC group have no significant difference. Result: 1.Action memory of verb-object phrase differed significantly between MHD group and NC group. 2.Two tests of Verbal Fluency differed significantly between MHD group and NC group. 3.Trail Making Test A, Trail Making Test B, the baseline condition of Go-No/Go Test and Stroop Color Naming Test differed significantly between MHD group and NC group. 4.There is no significant difference between MHD group and NC group on the correct rate of No/Go Test and the baseline condition. Both groups showed Stroop Effect in Go-No/Go test, but MHD group performed significantly worse. 5.In Stroop Color Naming Task Test, NC group showed Stroop Effect, significant Repeated Distraction Promotion Effect and significant Negative Priming Effect,while MHD group showed only Stroop Effect and no Repeated Distraction Promotion Effect and no Negative Priming Effect. There is significant difference in Stroop Effect between MHD group and NC group. Conclusion: 1.Comparing with NC group, episodic memory, semantic memory, cognitive planning ability, and inhibition ability of MHD group were impaired significantly. 2.The pathological aging of Executive Function in MHD group showed: executive Function should be a unitary system. 3.Cognitive impairment is negatively correlated with serum creatinine, blood pressure and anxiety score in MHD patients; and is related with hemoglobin, hematocrit, social support and life satisfaction. Keyword: maintenance hemodialysis, episodic memory, semantic memory, cognitive planning, inhibition ability.

Computational Structure of GPSG Models: Revised Generalized Phrase Structure Grammar

The primary goal of this report is to demonstrate how considerations from computational complexity theory can inform grammatical theorizing. To this end, generalized phrase structure grammar (GPSG) linguistic theory is revised so that its power more closely matches the limited ability of an ideal speaker--hearer: GPSG Recognition is EXP-POLY time hard, while Revised GPSG Recognition is NP-complete. A second goal is to provide a theoretical framework within which to better understand the wide range of existing GPSG models, embodied in formal definitions as well as in implemented computer programs. A grammar for English and an informal explanation of the GPSG/RGPSG syntactic features are included in appendices.

Aspects of Passives in Capeverdean (Santiago’s variety)

The present dissertation intends to study passives in Capeverdean. I argue that Capeverdean have eventive passives with specific morphology, postverbal morphemes, -du and -da, which affix to the verb to form passives and interact with the TMA morphemes available in the language in the same way as in the active voice. I also show that Capeverdean only allows short passives. However, this study demonstrates that the by-phrase, although not expressed phonologically, is implicit and can be tracked through instrumentals and Agent-oriented adverbs. In order to account for this specific property of Capeverdean passives, I assume the existence of a Voice head which introduces the external argument in all finite sentences in Capeverdean, except in unaccusatives, following proposals from Marantz (1984), Kratzer (1996), Sailor & Ahn (2010), Pratas (2014). I also assume that this Voice head is subject to a Doubly Filled Comp Filter, similar to what is proposed in Koopman (1997), which determines that either heads or specifiers can be overt, never both. In the case of passives, I propose that external argument is in Spec,Voice and the passive morphology is lexicalized in Voice0 and that while Spec,Voice is silent, Voice0 is not. This configuration can be explained if it is assumed, following Costa & Martins (2004), that in Capeverdean passives Voice0 is a strong functional head, thus requiring visibility at PF. This restriction, combined with the Doubly Filled Comp filter, imposes that Spec,Voice is silent.

Berlin, colonel Reinhard au centre, auteur de la phrase Jewish flag for the new German colours [il a comparé le nouveau drapeau allemand à un torchon juif] : [photographie de presse] / [Agence Rol]

Référence bibliographique : Rol, 58631

Interactions didactiques en classe de français : enseignement/apprentissage de l’accord du verbe en première secondaire

Notre recherche vise à décrire les connaissances grammaticales élaborées par des élèves de première secondaire au cours de l’enseignement/apprentissage de l’accord du verbe. Cette description se fonde sur l’observation des interactions didactiques entre les élèves, et leur enseignant autour de l’objet de savoir « accord du verbe » : elle concerne plus particulièrement l’interaction entre les pôles « élève » et « savoir ». Notre recherche s’inscrit dans le courant de la grammaire pédagogique moderne. La théorie de la transposition didactique de Chevallard (1985/1991) constitue également la pierre angulaire de nos travaux : les concepts de transposition didactique externe (le passage du savoir savant au savoir à enseigner) et interne (le passage du savoir à enseigner au savoir effectivement enseigné) agissent à titre d’analyseurs des interactions didactiques. L’observation, la description et la théorisation des interactions didactiques imposent une démarche écologique pour la collecte des données. Pour notre recherche, les données ont été recueillies grâce à la captation vidéo de séquences didactiques portant sur l’accord du verbe : elles consistent en des interactions verbales entre élèves ou entre les élèves et leur enseignant. L’analyse des données s’est effectuée selon une perspective macro et micro : (1) L’analyse macro indique que les connaissances antérieures des élèves résistent à l’institutionnalisation des savoirs puisque le savoir enseigné n’est pas celui qui est exclusivement mobilisé. Les élèves recourent à un vaste éventail de connaissances de types procédural et déclaratif pour l’identification du verbe et du sujet, dont la réussite n’est par ailleurs pas assurée. De plus, les connaissances qu’ils ont élaborées autour de la règle d’accord et du transfert des traits morphologiques sont également nombreuses et variées et ne les conduisent pas à accorder le verbe avec constance. (2) L’analyse micro suggère que l’élaboration des connaissances relatives à l’accord du verbe dépend de la manière dont les outils de la grammaire (manipulations syntaxiques et phrase de base) sont utilisés par les élèves. Plus précisément, le savoir piétine ou recule lorsque les manipulations syntaxiques ne sont pas appliquées dans la phrase ou qu’elles ne sont pas adaptées dans certains contextes syntaxiques; le savoir fait des bonds en avant dans les classes où les élèves sont en mesure de recourir à la phrase de base pour soutenir leur analyse grammaticale. Les descriptions proposées dans le cadre de notre thèse conduisent à discuter de leurs implications pour la transposition didactique externe et, plus généralement, pour la didactique du français et de la grammaire.

Criteria for the validation of specialized verb equivalents : application in bilingual terminography

Multilingual terminological resources do not always include valid equivalents of legal terms for two main reasons. Firstly, legal systems can differ from one language community to another and even from one country to another because each has its own history and traditions. As a result, the non-isomorphism between legal and linguistic systems may render the identification of equivalents a particularly challenging task. Secondly, by focusing primarily on the definition of equivalence, a notion widely discussed in translation but not in terminology, the literature does not offer solid and systematic methodologies for assigning terminological equivalents. As a result, there is a lack of criteria to guide both terminologists and translators in the search and validation of equivalent terms. This problem is even more evident in the case of predicative units, such as verbs. Although some terminologists (L‘Homme 1998; Lerat 2002; Lorente 2007) have worked on specialized verbs, terminological equivalence between units that belong to this part of speech would benefit from a thorough study. By proposing a novel methodology to assign the equivalents of specialized verbs, this research aims at defining validation criteria for this kind of predicative units, so as to contribute to a better understanding of the phenomenon of terminological equivalence as well as to the development of multilingual terminography in general, and to the development of legal terminography, in particular. The study uses a Portuguese-English comparable corpus that consists of a single genre of texts, i.e. Supreme Court judgments, from which 100 Portuguese and 100 English specialized verbs were selected. The description of the verbs is based on the theory of Frame Semantics (Fillmore 1976, 1977, 1982, 1985; Fillmore and Atkins 1992), on the FrameNet methodology (Ruppenhofer et al. 2010), as well as on the methodology for compiling specialized lexical resources, such as DiCoInfo (L‘Homme 2008), developed in the Observatoire de linguistique Sens-Texte at the Université de Montréal. The research reviews contributions that have adopted the same theoretical and methodological framework to the compilation of lexical resources and proposes adaptations to the specific objectives of the project. In contrast to the top-down approach adopted by FrameNet lexicographers, the approach described here is bottom-up, i.e. verbs are first analyzed and then grouped into frames for each language separately. Specialized verbs are said to evoke a semantic frame, a sort of conceptual scenario in which a number of mandatory elements (core Frame Elements) play specific roles (e.g. ARGUER, JUDGE, LAW), but specialized verbs are often accompanied by other optional information (non-core Frame Elements), such as the criteria and reasons used by the judge to reach a decision (statutes, codes, previous decisions). The information concerning the semantic frame that each verb evokes was encoded in an xml editor and about twenty contexts illustrating the specific way each specialized verb evokes a given frame were semantically and syntactically annotated. The labels attributed to each semantic frame (e.g. [Compliance], [Verdict]) were used to group together certain synonyms, antonyms as well as equivalent terms. The research identified 165 pairs of candidate equivalents among the 200 Portuguese and English terms that were grouped together into 76 frames. 71% of the pairs of equivalents were considered full equivalents because not only do the verbs evoke the same conceptual scenario but their actantial structures, the linguistic realizations of the actants and their syntactic patterns were similar. 29% of the pairs of equivalents did not entirely meet these criteria and were considered partial equivalents. Reasons for partial equivalence are provided along with illustrative examples. Finally, the study describes the semasiological and onomasiological entry points that JuriDiCo, the bilingual lexical resource compiled during the project, offers to future users.

Étude électroencéphalographique des premiers processus syntaxiques impliqués lors de la compréhension de la phrase en français québécois

Mémoire numérisé par la Division de la gestion de documents et des archives de l'Université de Montréal

L'évolution des conceptions et des pratiques relatives à la notion de phrase en FLE d’enseignants vietnamiens du secondaire engagés dans une formation continue en syntaxe

Cette recherche vise à décrire l’évolution des conceptions et des pratiques relatives à la notion de phrase de cinq enseignantes de français vietnamiennes engagées dans une formation sur la syntaxe en grammaire nouvelle et la littérature de jeunesse. Afin d’examiner l’évolution des conceptions sur la phrase des enseignantes, nous avons effectué deux entrevues semi-dirigées, l’une avant et l’autre après la formation. Pour étudier l’évolution des pratiques d’enseignement de la phrase, nous avons observé une seule enseignante à deux reprises après la formation. De plus, celle-ci a décrit ses pratiques sur la phrase lors d’une entrevue téléphonique qui se déroulait à la fin de chaque mois (novembre 2011- mars 2012). Nos résultats montrent que les enseignantes ont changé leurs conceptions sur la phrase à la suite de la formation. En effet, dans la deuxième entrevue, nous avons noté une meilleure réussite dans l’identification des phrases, des critères d’identification plus convaincants, plus de précision dans les commentaires relatifs aux définitions et dans les explications des erreurs chez les cinq enseignantes. Quant aux pratiques, l’enseignante suivie a réalisé des activités autour de la phrase après la formation, ce qui constitue une évolution dans ses pratiques, puisqu’avant la formation, cette notion était négligée. Ses pratiques observées et déclarées indiquent qu’elle a intégré dans sa classe les notions clés de la grammaire nouvelle autour de la phrase et la littérature de jeunesse au service de l’enseignement de cette notion.