The purpose of Language Identification (LID) is to identify a specific language from a spoken utterance, automatically. Language-specific characteristics are always associated with different languages. Most existing LID approaches utilise a statistical modelling process with common acoustic/phonotactic features to model specific languages while avoiding any language-specific knowledge. Great successes have been achieved in this area over past decades. However, there is still a huge gap between these languageindependent methods and the actual language-specific patterns. It is extremely useful to address these specific acoustic or semantic construction patterns, without spending huge labour on annotation which requires language-specific knowledge. Inspired by this goal, this research focuses on the language-feature dependency. Several practical methods have been proposed. Various features and modelling techniques have been studied in this research. Some of them carry out additional language-specific information without manual labelling, such as a novel duration modelling method based on articulatory features, and a novel Frequency-Modulation (FM) based feature. The performance of each individual feature is studied for each of the language-pair combinations. The similarity between languages and the contribution in identifying a language by using a particular feature are defined for the first time, in a quantitative style. These distance measures and languagedependent contributions become the foundations of the later-presented frameworks ?? language-dependent weighting and hierarchical language identification. The latter particularly provides remarkable flexibility and enhancement when identifying a relatively large number of languages and accents, due to the fact that the most discriminative feature or feature-combination is used when separating each of the languages. The proposed systems are evaluated in various corpora and task contexts including NIST language recognition evaluation tasks. The performances have been improved in various degrees. The key techniques developed for this work have also been applied to solve a different problem other than LID ?? speech-based cognitive load monitoring.

Alcohol and tobacco are drugs of abuse which are legal to sell and consume in most western societies. Addiction to these two substances has major social and health implications worldwide. The brain structure known to mediate addictive behaviour is the dopaminergic mesocorticolimbic system. Dopaminegic neurons arise from the ventral tegmental area, project to the nucleus accumbens and interact with the amygdala and the prefrontal cortex. Chronic alcoholism elicits marked damage in the prefrontal cortex with significant loss of neurons and glia. The key components of addiction, tolerance and dependence, are thought to be the result of semipermanent adaptive changes in gene expression. Gene expression profiling of the mesocorticolimbic system from human alcoholics and alcohol-dependent animals has revealed highly region-specific alterations. How these molecular changes result in the development of alcohol dependence in humans is not fully understood. Complicating factors in human alcoholism include a high comorbidity with smoking, socioeconomic factors and the prevalence of underlying psychological pathologies. Gene expression profiling of the prefrontal cortex of six alcoholics and six controls resulted in the identification of functional gene groups sensitive to alcoholism. Mitochondrial function was found down regulated while mRNA levels of genes involved in stress response and cell protection were elevated. These results correlate with the pathology of the prefrontal cortex in chronic alcoholism. Some of the control cases used for gene expression profiling were later identified as chronic smokers, while all of the alcoholics were heavy smokers. To date the heavy co-morbidity of alcoholism with smoking has not been taken into account. Thus the expression of selected genes were investigated by realtime PCR in an extended case set of non-smoking alcoholics, smoking alcoholics, smoking non-alcoholics and non-smoking, non-alcoholics. This study revealed that alcoholism itself had a significant impact on the expression of midkine, the high affinity glial glutamate transporter, member 1 and the tissue inhibitor of the metalloproteinase 3. Heavy smoking itself led to a small but significant elevation of MDK mRNA levels as well as an increase in variation of excitatory amino acid transporter 1 and metalloproteinase inhibitor, member 3 expression. Apolipoprotein D however was induced by chronic smoking but not by alcohol dependence. These results highlight the need of careful case selection in future studies on gene expression in the human alcoholic brain. Peptide antibodies were produced to midkine and a polyclonal antibody against the excitatory amino acid transporter 1 was obtained from a collaborating laboratory. Western blots utilizing these antibodies revealed a marked increase in midkine and excitatory amino acid transporter 1 protein in alcoholics compared to non-smoking and non-drinking controls. In coronal sections of human prefrontal cortex of alcoholics and non-smoking non-drinking controls, immunofluorescence of midkine was obtained from nuclei throughout the layers of the cortex and from the cell bodies of a distinct set of astrocytes in cortical layer II. Double staining with glial fibrillary acidic protein revealed that a portion of midkine-positive nuclei were localised in glial cells. There was no difference in immunostaining of alcohol and control sections with midkine. In summary these results indicate that midkine protein is induced in the prefrontal cortex of the chronic alcoholic. However, this increase in protein may not be strong enough to be visualised by immunohistochemistry. Midkine induction may be reflective of reparative processes in the prefrontal cortex of the chronic alcoholic. Excitatory amino acid transporter 1 staining in non-alcoholic, non-smoking control cases were obtained as a confluent band in cortical layer II and sparsely in deeper cortical layers. Excitatory amino acid transporter 1 immunoreactivity overlapped partially with glial fibrillary acidic protein labelling. In chronic alcoholics, excitatory amino acid transporter 1 staining in the area between the cortical layer II and VI was significantly increased. At withdrawal, glutamate levels may reach toxic levels in the cortex. The increase in cells expressing excitatory amino acid transporter 1 throughout the cortical layers may indicate a protective measure of this brain region in the chronic alcoholic. Additionally, layer specific expression of midkine and excitatory amino acid transporter 1 in the prefrontal cortex of the healthy individual may implicate a specialised role of these astrocytes.

Thesis (D. Min.)--Westminster Theological Seminary, Philadelphia, 1991. / Includes bibliographical references (leaves 181-186).

Thesis (D. Min.)--Westminster Theological Seminary, Philadelphia, 1991. / Includes bibliographical references (leaves 181-186).

Thesis (M.A.)--Carleton University, 1988. / Also available in electronic format on the Internet.

Thesis (D. Min.)--Westminster Theological Seminary, Philadelphia, 1991. / Includes bibliographical references (leaves 181-186).

Thesis (Ph.D.) -- University of Texas Southwestern Medical Center at Dallas, 2006. / Not embargoed. Vita. Bibliography: 196-222.

This thesis provides a detailed analysis of the so-called `house money' or windfall endowment effect and its main determinants. Chapter 1 provides a detailed survey on the literature related to the house money effect. This effect according to Thaler and Johnson (1990) - refers to the situation where prior gains mitigate the influence of loss aversion and facilitate risk-seeking. The concept borrows its name from the expression employed in the gambling parlance of "playing with the house money", which is used when people gamble while ahead. As the literature has used a variety of concepts and ideas to describe the house money effect, this chapter presents and discusses them within the environment and the related literature that they have emerged. This is done in order to highlight the predominant answers to the main research questions raised in the various strands of the literature, namely: (i) whether people treat money differently depending on its origin; and (ii) the implications of the house money effect for the experimental methodology in economics. The literature is organised and presented according to the context in which the above two research questions have been examined. By presenting results in each particular context, we pin down the contextual differences that might be responsible for the presence (or absence) of the house money effect, and lay the initial ground work to answer a third research question: What drives the house money effect? In this regard, after we demonstrate the context-dependency of the house money effect we present the two main interpretations that it has received, namely that the house money effect is a result of different mental accounting over windfall gains (`windfall effect') or a result of fairness or deservingness concerns ('Lockean desert effect'). Chapter 2 re-examines the house money effect and explores its main driving forces. For that, we employ a novel experimental design utilising a within-subject approach, coupled with the use of three different contexts of economic decisions (a trust game, a set of lotteries and a public good game). Both the within-subject experimental design and the three contexts of economic decisions allow us to better test the two main interpretations of the house money effect. Our experimental data confirm the presence of the house money effect both in the decision to trust (but not in the decision of trustworthiness) in the trust game and in the decision to contribute in the public account of the public good game. However, our findings do not support the hypothesis that changes in risk behaviour of participants are due to different sources of money, suggesting that risk attitudes are robust and independent of the origin of money along the experiment. Therefore, our findings seem to favour interpretations of the house money effect as a result of 'just desert' or fairness preferences rather than the result of different mental accounting over windfall gains. Chapter 3 combines two branches of experimental literature, namely the house money effect and the literature on individual differences in social preferences. Both the house money effect and individual differences have been used extensively to explain cooperation in social dilemmas (and its decline over time). Here, we test the implications of house money on reciprocal behaviour, that is, whether participants in economic experiments are less likely to reciprocate when earned money rather than windfall money is at stake. Using the innovative experimental design of Fischbacher et al. (2001) with strategy method, we classify participants according to their behaviour in a linear public good game, and by adding the within-subject element in our experimental design we test the robustness of this classification across the different origin of endowments. Our results indicate that the types' classification is robust across the origin of money. Contrary to Harrison (2007), we find that participants' decision to free ride or not (contribute or not) is independent of the origin of money, but given that the decision to contribute has been made, contribution levels may vary -actually be lower- when money is earned rather than windfall endowed. We also elicit beliefs about others' contributions and test how these beliefs affected by the "house money" and in turn how they affect the decision to contribute. This discussion relates to what the literature has characterised so far as "anticipatory reciprocity".

Esta dissertação tem como objetivo refletir sobre as relações mobilizadas no cuidado de adultos cuja autonomia pode ser considerada limitada em razão de uma “deficiência intelectual”. A partir de uma etnografia realizada junto a três famílias de um bairro de classe popular da cidade de Porto Alegre, busca-se prestar atenção no que realmente está em jogo para as pessoas que lidam em seu cotidiano com um parente que, segundo elas, depende de cuidados em razão de ter “problemas de cabeça”. Tendo como foco da análise as práticas e narrativas de três cuidadoras, reflete-se sobre as questões que dizem respeito à obtenção e ao gerenciamento de recursos para garantir o bem estar daqueles por quem elas são responsáveis, em especial questões referentes ao Benefício da Prestação Continuada. Em uma aproximação com os teóricos dos estudos sobre a deficiência e principalmente com as pesquisas sobre o cuidado, argumenta-se que um olhar a partir do cuidado, pensado tanto enquanto uma prática como um valor moral, permite complexificar as noções de “deficiência”, “autonomia” e “negligência”. Com isso, espera-se chamar atenção para as práticas de cuidado enquanto permeadas por tensões e negociações onde diferentes versões do que é o “bom cuidado” são constantemente performadas. Nesse sentido, o cuidado, enquanto uma categoria de análise, não permite categorizações simples ou soluções pré-fixadas, pois o que a sua lógica exige são respostas locais e contingenciais a cada situação. / The purpose of this study is to reflect upon the relationships that are mobilized in the caring of adults whose autonomy can be considered limited due to “intellectual disability”. Through ethnography research with three families from a poor neighborhood in the city of Porto Alegre, Brazil, attention was paid to what is at stake for those who have to deal, in their everyday lives, with a relative that depends on their care due to “mental problems”. Focusing the analysis on the practices and narratives of three care-givers, the goal is to reflect upon questions related to obtaining and managing the welfare of those for whom they are responsible, especially regarding the Benefício de Prestação Continuada (BPC). Along the lines of disability studies and especially care research, the argument is that using care as a focal point enables an approach that complexifies the notions of “disability”, “autonomy” and “neglect”. With that, we hope to draw attention to care practices as ones embedded in tensions and negotiations where different versions of “good care” are constantly being enacted. Care, as an analytical category, does not allow simple categorizations or pre-determined solutions, but requires fluid, varied and contingent responses.

Background: Research on how rising human capital affects the consequences of population ageing rarely considers the fact that the human capital of the elderly population is composed in a specific way that is shaped by their earlier schooling and work experience. For an elderly population of a fixed size and age-sex composition, this entails that the higher its human capital, the greater the total amount of public pensions to be paid. Objective: The main purpose of this paper is to analyse the link between human capital and retiree benefits and its effect on population ageing from a demographic viewpoint. Methods: We construct an old age dependency ratio (OADR), in which each person, whether in the numerator or the denominator, is assigned the number of units corresponding to his/her level of human capital. Based on data for Italy, we study the dynamics of this human-capital-specific OADR with the help of multistate population projections to 2107. Results: Our results show that under specific conditions a constant or moderately growing human capital may aggravate the consequences of population ageing rather than alleviate them. Conclusions: With those findings, the authors would like to stimulate the debate on the search for demographic and/or socio-economic solutions to the challenges posed by population ageing.