Mémoire numérisé par la Direction des bibliothèques de l'Université de Montréal.

This diploma thesis examines functions of the modal verb should in nominal content clauses introduced by the conjunction that. The Czech counterparts of the English sentences are considered as well. The research focuses on the so-called putative should which occurs after main clauses with directive, epistemic, attitudinal, evaluative and volitional expressions. Should expressing intrinsic (root) modality (expressing permission, obligation or ability) is left out of account. The work pays attention to the basic classification of nominal content clauses deriving from a verbal form alternating with putative should. Two kinds of putative should will be dealt with, namely should after directive and volitional expressions which could have its alternative form in the present subjunctive and should after epistemic, attitudinal and evaluative expressions which could possibly alternate with the indicative. The work examines these verbal forms in relation to the intentional modality of the sentence in an independent form (e.g. declarative, interrogative or imperative sentence). The work is divided into two parts: theoretical and practical. The theoretical part of the work describes the basic classification of all central modal verbs in English and putative should in terms of its relation towards the...

Thesis advisor: James Russell / Adults' gender stereotypes of emotion have been investigated with a variety of methods, but those methods do not provide a strong test of the stereotype: The participants were presented only with cues to the gender or to the emotion; or when both cues were available, gender was confounded with poser. This study examined the effects of attributed gender on adults' perception of emotion in facial expressions and stories when presented with clear versus ambiguous cues to both emotion and gender. College students (n = 90) were first asked to label the emotion of either a man (Timothy) or a woman (Sophia) with identical prototypical and “mixed” facial expressions and, separately, to Free Label stories about emotions. The same students were then to choose from a list of ten emotion labels the one that best described the protagonist's emotion for the same stimuli. Results showed that, for ambiguous cues to emotion, participants labeled facial expressions according to gender stereotypes. However, for the stimuli with clear cues to both emotion and gender of the poser, a reverse effect of gender stereotypes was observed for anger, fear, shame, and compassion due to an expectancy violation. / Thesis (BA) — Boston College, 2004. / Submitted to: Boston College. College of Arts and Sciences. / Discipline: Psychology. / Discipline: College Honors Program.

Thesis advisor: James A. Russell / Prior research showing that children recognize emotional expressions has used a choice-from-array style task; for example, children are asked to find the fear face in an array of several expressions. However, these choice-from-array tasks allow for the use of a process of elimination strategy in which children could select an expression they are unfamiliar with when presented a label that does not apply to other expressions in the array. Across six studies (N = 144), 80% of 2- to 4-year-olds selected a novel expression when presented a target label and performed similarly when the label was novel (such as <italic>pax</italic>) or familiar (such as <italic>fear</italic>). In addition, 46% of children went on to freely label the expression with the target label in a subsequent task. These data are the first to show that children extend the process of elimination strategy to facial expressions and also call into question the findings of prior choice-from-array studies. / Thesis (PhD) — Boston College, 2011. / Submitted to: Boston College. Graduate School of Arts and Sciences. / Discipline: Psychology.

Tem sido freqüentemente discutido na literatura científica o papel que desempenha cada hemisfério cerebral no processamento da informação emocional. O estudo realizado teve por objetivo investigar o padrão de dominância hemisférica para a percepção das expressões faciais de alegria, tristeza, raiva e medo. Em dois experimentos realizados foi utilizada a técnica de estudo campo visual dividido com a apresentação taquitoscópica de estímulos por 150 ms na tela de um monitor. Os estímulos foram compostos com fotografias de faces de quatro indivíduos (2H, 2M) retiradas da série Pictures of Facial Affect. Vinte e um observadores destros (9H, 12M) participaram do experimento 1. Em cada tentativa eram apresentadas duas fotografias de faces, uma à esquerda e outra à direita do ponto de fixação na tela do computador, em quatro diferentes condições: 1) face com emoção à esquerda e face neutra à direita, 2) face neutra à esquerda e face com emoção à direita, 3) face com emoção à direita e à esquerda, 4) face neutra à direita e à esquerda. Em cada tentativa, os observadores determinaram o lado em que havia sido apresentada a face que aparentava expressar mais emoção. Dezessete observadores destros (8H, 9M) participaram do experimento 2. Em cada apresentação de estímulo, uma foto de face era apresentada à direita ou à esquerda do ponto de fixação, localizado no centro da tela, e do lado oposto era apresentado um retângulo cinza. Foram elaboradas as seguintes condições de estímulo: 1) face com emoção à esquerda e retângulo cinza à direita, 2)retângulo cinza à esquerda e face com emoção à direita, 3) face neutra à esquerda e retângulo cinza à direita, 4) retângulo cinza à esquerda e face neutra à direita. Em cada tentativa, os observadores determinaram se a face apresentada aparentava ou não possuir emoção. Os tempos de reação e os erros de julgamento foram submetidos a ANOVAs para medidas repetidas. No primeiro experimento, a emoção foi em geral detectada mais rapidamente em faces apresentadas no campo visual esquerdo (p<0,01). As expressões de tristeza e raiva também foram percebidas mais rapidamente quando apresentadas no campo visual esquerdo (p<0,05). Em ambos os experimentos, as expressões de alegria e medo foram percebidas mais rapidamente e mais acuradamente que as expressões de tristeza e raiva (p<0,001). A expressão de tristeza foi detectada mais facilmente em faces femininas e a expressão de raiva em faces masculinas (p<0,05). De maneira geral, entretanto, a emoção foi detectada mais facilmente em faces femininas. Em ambos os experimentos houve diferenças entre as faces dos quatro indivíduos que representavam as expressões faciais. O hemisfério direito mostrou-se superior ao esquerdo na percepção das expressões faciais, especialmente na percepção das expressões de tristeza e raiva. A vantagem perceptiva do hemisfério direito é mais evidente para as expressões que são detectadas com maior dificuldade. A percepção de expressões faciais pode ser afetada pelo gênero da face e pelas singularidades da expressão facial individual. / The role that each brain hemisphere plays in the processing of emotional information has been frequently discussed in the scientific literature. The aim of this study was to investigate the pattern of hemispheric dominance for the perception of the facial expressions of happiness, sadness, anger and fear. In two experiments the divided-visualfield technique was used with the taquitoscopic presentation of stimuli on a computer screen for 150 ms. The stimuli were composed with pictures of faces of four people (2M,2F) taken from the series Pictures of Facial Affect. Twenty one right-handed observers (9M,12F) took part in the experiment I. In each trial two pictures of faces were presented on the computer screen, one of them placed on the left side and the other one on the right side of the fixation point, in four different conditions: 1) face with emotion on the left and neutral face on the right, 2) neutral face on the left and face with emotion on the right, 3)face with emotion on the right and on the left, 4) neutral face on the right and on the left. In each trial the observers determined the side on which the face seemed to show greater emotional intensity. Seventeen right-handed observers (8M, 9F) took part in the experiment II. In each stimulus presentation, a picture of a face was presented either on the right or left side of the fixation point, placed on the center of the screen and, on the opposite side, a gray rectangle was presented. The following stimuli conditions were elaborated: 1) face with emotion on the left and gray rectangle on the right, 2) gray rectangle on the left and face with emotion on the right, 3) neutral face on the left and gray rectangle on the right, 4) gray rectangle on the left and neutral face on the right. In each trial the observers determined if the face presented had emotion or not. Time reactions and judgement errors were submitted to ANOVAs for repeated measures. In the first experiment, emotion was generally detected more quickly in faces presented on the left 10 visual field (p<0,01). The expressions of sadness and anger were also perceived more quickly when presented on the left visual field (p<0,05). In both experiments, expressions of happiness and fear were perceived more quickly and more accurately than expressions of sadness and anger (p<0,001). The expression of sadness was detected more easily in feminine faces and the expression of anger in masculine faces (p<0,05). In general, however, the emotion was detected more easily in feminine faces. In both experiments there were differences in the perception between the faces of the four individuals that represented the facial expressions. The right hemisphere showed superior to the left hemisphere in the perception of facial expressions, especially for the perception of expressions of sadness and anger. The perceptive advantage of the right hemisphere is more evident for the expressions that are detected with more difficulty. The perception of facial expressions can be affected by the gender of the face and the singularities of the individual facial expression.

Sistemas de processamento de linguagem natural baseiam-se com frequência na hipótese de que a linguagem humana é composicional, ou seja, que o significado de uma entidade linguística pode ser inferido a partir do significado de suas partes. Essa expectativa falha no caso de expressões multipalavras (EMPs). Por exemplo, uma pessoa caracterizada como pão-duro não é literalmente um pão, e também não tem uma consistência molecular mais dura que a de outras pessoas. Técnicas computacionais modernas para inferir o significado das palavras com base na sua distribuição no texto vêm obtendo um considerável sucesso em múltiplas tarefas, especialmente após o surgimento de abordagens de word embeddings. No entanto, a representação de EMPs continua a ser um problema em aberto na área. Em particular, não existe um método consolidado que prediga, com base em corpora, se uma determinada EMP deveria ser tratada como unidade indivisível (por exemplo olho gordo) ou como alguma combinação do significado de suas partes (por exemplo tartaruga marinha). Esta tese propõe um modelo de predição de composicionalidade de EMPs com base em representações de semântica distribucional, que são instanciadas no contexto de uma variedade de parâmetros. Também é apresentada uma avaliação minuciosa do impacto desses parâmetros em três novos conjuntos de dados que modelam a composicionalidade de EMP, abrangendo EMPs em inglês, francês e português. Por fim, é apresentada uma avaliação extrínseca dos níveis previstos de composicionalidade de EMPs, através da tarefa de identificação de EMPs. Os resultados obtidos sugerem que a escolha adequada do modelo distribucional e de parâmetros de corpus pode produzir predições de composicionalidade que são comparáveis às observadas no estado da arte. / Natural language processing systems often rely on the idea that language is compositional, that is, the meaning of a linguistic entity can be inferred from the meaning of its parts. This expectation fails in the case of multiword expressions (MWEs). For example, a person who is a sitting duck is neither a duck nor necessarily sitting. Modern computational techniques for inferring word meaning based on the distribution of words in the text have been quite successful at multiple tasks, especially since the rise of word embedding approaches. However, the representation of MWEs still remains an open problem in the field. In particular, it is unclear how one could predict from corpora whether a given MWE should be treated as an indivisible unit (e.g. nut case) or as some combination of the meaning of its parts (e.g. engine room). This thesis proposes a framework of MWE compositionality prediction based on representations of distributional semantics, which we instantiate under a variety of parameters. We present a thorough evaluation of the impact of these parameters on three new datasets of MWE compositionality, encompassing English, French and Portuguese MWEs. Finally, we present an extrinsic evaluation of the predicted levels of MWE compositionality on the task of MWE identification. Our results suggest that the proper choice of distributional model and corpus parameters can produce compositionality predictions that are comparable to the state of the art.

Previous research demonstrates that the age of an observer, the peripheral location of a face stimulus on a display, and the intensity of the emotion expressed by the face all play a role in emotion perception. Older individuals have more difficulty identifying emotion in faces, especially at lower expressive intensities. The purpose of the current study was to understand how younger and older adults’ abilities to detect emotion in facial stimuli presented in the periphery would be affected by the intensity of the emotional expressions and the distance that the expressions are presented away from the center of the display. The current study presented facial stimuli for a short duration to bypass reactionary attentional influences. More intense fearful and angry expressions were expected to be easier to classify for both younger and older adults than lower intensity expressions, but all expressions were expected to become more difficult to classify when presented further in the periphery. Older adults and younger adults displayed similar emotion detection for typical and extreme intensity angry expressions and for high intensity fearful expressions. However, older adults struggled to detect typical intensity fear, and this deficit grew with the angle of eccentricity from which the stimuli were presented from the center of the display. Possible explanations for these age differences are discussed.

The increasing amount of electronically available information in society today is undeniable. Examples include the numbers of general web pages, scientific publications, and items in online auctions. From a user's perspective, this trend will lead to information overflow. Moreover, information publishers are compromised by this situation, as users have greater difficulty in identifying useful information. Publish-subscribe systems can be applied to cope with the reality of information overflow. In these systems, users specify their information interests as subscriptions and, subsequently, only matching information (event messages) is delivered; uninteresting information is filtered out before reaching users. In this dissertation, we consider content-based publish-subscribe systems, a sophisticated example of these systems. They perform the information-filtering task based on the content of provided information. In order to deal with high numbers of subscriptions and frequencies of event messages, publish-subscribe systems are realized as distributed systems. Advertisements---publisher specifications of potential future event messages---are optionally applied in these systems to reduce the internal distribution of subscriptions. Existing work on content-based publish-subscribe concepts mainly focuses on subscriptions and advertisements as pure conjunctive expressions. Therefore, subscriptions or advertisements using operators other than conjunction need to be canonically converted to disjunctive normal form by these systems. Each conjunctive component is then treated as individual subscription or advertisement. Unfortunately, the size of converted expressions is exponential in the worst case. In this dissertation, we show that the direct support of general Boolean subscriptions and advertisements improves the time and space efficiency of general-purpose content-based publish-subscribe systems. For this purpose, we develop suitable approaches for the filtering and routing of general Boolean expressions in these systems. Our approaches represent solutions to exactly those components of content-based publish-subscribe systems that currently restrict subscriptions and advertisements to conjunctive expressions. On the subscription side, we present an effective generic filtering algorithm, and a novel approach to optimize event routing tables, which we call subscription pruning. To support advertisements, we show how to calculate the overlap between subscriptions and advertisements, and introduce the first designated subscription routing optimization, which we refer to as advertisement pruning. We integrate these approaches into our prototype BoP (BOolean Publish-subscribe) which allows for the full support of general Boolean expressions in its filtering and routing components. In the evaluation part of this dissertation, we empirically analyze our prototypical implementation BoP and compare its algorithms to existing conjunctive solutions. We firstly show that our general-purpose Boolean filtering algorithm is more space- and time-efficient than a general-purpose conjunctive filtering algorithm. Secondly, we illustrate the effectiveness of the subscription pruning routing optimization and compare it to the existing covering optimization approach. Finally, we demonstrate the optimization effect of advertisement pruning while maintaining the existing overlapping relationships in the system.

La théorie des automates est apparue pour résoudre des problèmes aussi bien pratiques que théoriques, et ceci dès le début de l'informatique. Désormais, les automates font partie des notions fondamentales de l'informatique, et se retrouvent dans la plupart des logiciels. En 1974, Samuel Eilenberg proposa un modèle de calcul qui unifie la plupart des automates (transducteurs, automates à pile et machines de Turing) et qui a une propriété de modularité intéressante au vu d'applications reposant sur différentes couches d'automates ; comme cela peut être le cas en linguistique computationnelle. Nous proposons d'étudier les techniques permettant d'avoir des machines d'Eilenberg effectives. Cette étude commence par la modélisation de relations calculables à base de flux, puis continue avec l'étude de la simulation des machines d'Eilenberg définies avec ces relations. Le simulateur est un programme fonctionnel énumérant progressivement les solutions, en explorant un espace de recherche selon différentes stratégies. Nous introduisons, en particulier, la notion de machine d'Eilenberg finie pour laquelle nous fournissons une preuve formelle de correction de la simulation. Les relations sont une première composante des machines d'Eilenberg, la deuxième composante étant son contrôle, qui est défini par un automate fini. Dans ce contexte, on peut utiliser une expression régulière comme syntaxe pour décrire la composante de contrôle d'une machine d'Eilenberg. Récemment, un ensemble de travaux exploitant la notion de dérivées de Brzozowski, a été la source d'algorithmes efficaces de synthèse d'automates non-déterministes à partir d'expressions régulières. Nous faisons l'état de l'art de ces algorithmes, tout en donnant une implémentation efficace en OCaml permettant de les comparer les uns aux autres.

The benefits of programming in a functional style are well known. In particular, algorithms that are expressed as compositions of functions operating on series/vectors/streams of data elements are much easier to understand and modify than equivalent algorithms expressed as loops. Unfortunately, many programmers hesitate to use series expressions, because they are typically implemented very inefficiently. Common Lisp macro packages (OSS) has been implemented which supports a restricted class of series expressions, obviously synchronizable series expressions, which can be evaluated very efficiently by automatically converting them into loops. Using this macro package, programmers can obtain the advantages of expressing computations as series expressions without incurring any run-time overhead.