The purpose of the present study is to develop and explain a teaching aid for Spanish word recognition, to be presented as advance preparation for large-exposure reading.

Zero-resource speech processing is a growing research area which aims to develop methods that can discover linguistic structure and representations directly from unlabelled speech audio. Such unsupervised methods would allow speech technology to be developed in settings where transcriptions, pronunciation dictionaries, and text for language modelling are not available. Similar methods are required for cognitive models of language acquisition in human infants, and for developing robotic applications that are able to automatically learn language in a novel linguistic environment. There are two central problems in zero-resource speech processing: (i) finding frame-level feature representations which make it easier to discriminate between linguistic units (phones or words), and (ii) segmenting and clustering unlabelled speech into meaningful units. The claim of this thesis is that both top-down modelling (using knowledge of higher-level units to to learn, discover and gain insight into their lower-level constituents) as well as bottom-up modelling (piecing together lower-level features to give rise to more complex higher-level structures) are advantageous in tackling these two problems. The thesis is divided into three parts. The first part introduces a new autoencoder-like deep neural network for unsupervised frame-level representation learning. This correspondence autoencoder (cAE) uses weak top-down supervision from an unsupervised term discovery system that identifies noisy word-like terms in unlabelled speech data. In an intrinsic evaluation of frame-level representations, the cAE outperforms several state-of-the-art bottom-up and top-down approaches, achieving a relative improvement of more than 60% over the previous best system. This shows that the cAE is particularly effective in using top-down knowledge of longer-spanning patterns in the data; at the same time, we find that the cAE is only able to learn useful representations when it is initialized using bottom-up pretraining on a large set of unlabelled speech. The second part of the thesis presents a novel unsupervised segmental Bayesian model that segments unlabelled speech data and clusters the segments into hypothesized word groupings. The result is a complete unsupervised tokenization of the input speech in terms of discovered word types|the system essentially performs unsupervised speech recognition. In this approach, a potential word segment (of arbitrary length) is embedded in a fixed-dimensional vector space. The model, implemented as a Gibbs sampler, then builds a whole-word acoustic model in this embedding space while jointly performing segmentation. We first evaluate the approach in a small-vocabulary multi-speaker connected digit recognition task, where we report unsupervised word error rates (WER) by mapping the unsupervised decoded output to ground truth transcriptions. The model achieves around 20% WER, outperforming a previous HMM-based system by about 10% absolute. To achieve this performance, the acoustic word embedding function (which maps variable-duration segments to single vectors) is refined in a top-down manner by using terms discovered by the model in an outer loop of segmentation. The third and final part of the study extends the small-vocabulary system in order to handle larger vocabularies in conversational speech data. To our knowledge, this is the first full-coverage segmentation and clustering system that is applied to large-vocabulary multi-speaker data. To improve efficiency, the system incorporates a bottom-up syllable boundary detection method to eliminate unlikely word boundaries. We compare the system on English and Xitsonga datasets to several state-of-the-art baselines. We show that by imposing a consistent top-down segmentation while also using bottom-up knowledge from detected syllable boundaries, both single-speaker and multi-speaker versions of our system outperform a purely bottom-up single-speaker syllable-based approach. We also show that the discovered clusters can be made less speaker- and gender-specific by using features from the cAE (which incorporates both top-down and bottom-up learning). The system's discovered clusters are still less pure than those of two multi-speaker unsupervised term discovery systems, but provide far greater coverage. In summary, the different models and systems presented in this thesis show that both top-down and bottom-up modelling can improve representation learning, segmentation and clustering of unlabelled speech data.

This thesis describes a novel methodology, grounded in the distributional semantic paradigm, for building context sensitive models of word meaning, affording an empirical exploration of the relationship between words and concepts. Anchored in theoretical linguistic insight regarding the contextually specified nature of lexical semantics, the work presented here explores a range of techniques for the selection of subspaces of word co-occurrence dimensions based on a statistical analysis of input terms as observed within large-scale textual corpora. The relationships between word-vectors that emerge in the projected subspaces can be analysed in terms of a mapping between their geometric features and their semantic properties. The power of this modelling technique is its ability to generate ad hoc semantic relationships in response to an extemporaneous linguistic or conceptual situation. The product of this approach is a generalisable computational linguistic methodology, capable of taking input in various forms, including word groupings and sentential context, and dynamically generating output from a broad base model of word co-occurrence data. To demonstrate the versatility of the method, this thesis will present competitive empirical results on a range of established natural language tasks including word similarity and relatedness rating, metaphor and metonymy detection, and analogy completion. A range of techniques will be applied in order to explore the ways in which different aspects of projected geometries can be mapped to different semantic relationships, allowing for the discovery of a range of lexical and conceptual properties for any given input and providing a basis for an empirical exploration of distinctions between the semantic phenomena under analysis. The case made here is that the flexibility of these models and their ability to extend output to evaluations of unattested linguistic relationships constitutes the groundwork for a method for the extrapolation of dynamic conceptual relationships from large-scale textual corpora. This method is presented as a complement and a counterpoint to established distributional methods for generating lexically productive word-vectors. Where contemporary vector space models of distributional semantics have almost universally involved either the factorisation of co-occurrence matrices or the incremental learning of abstract representations using neural networks, the approach described in this thesis preserves the connection between the individual dimensions of word-vectors and statistics pertaining to observations in a textual corpus. The hypothesis tested here is that the maintenance of actual, interpretable information about underlying linguistic data allows for the contextual selection of non-normalised subspaces with more nuanced geometric features. In addition to presenting competitive results for various computational linguistic targets, the thesis will suggest that the transparency of its representations indicates scope for the application of this model to various real-world problems where an interpretable relationship between data and output is highly desirable. This, finally, demonstrates a way towards the productive application of the theory and philosophy of language to computational linguistic practice.

Internets framfart har skapat ändrade förhållande på marknaden för kommunikation och marknadsföring. Företag marknadsför sig alltmer på Internet och e-handeln har samtidigt ökat kraftigt. Debatten kring nya sätt att marknadsföra sig på Internet och hur konsumenterna reagerar på information är i full gång. I en undersökning gjord av Posten visar det sig att 7 av 10 svenskar gör sina distanshandelsköp på Internet och att de har en sofistikerad informationssökningsprocess. Påståendet görs att konsumenter litar mer på vänner, bekanta och andra konsumenter på ”nätet” . Här kommer begreppet word-of-web in, ett begrepp som har stor betydelse på Internet beträffande trovärdighet. Internet har banat väg för öppna diskussioner och konversationer mellan såväl konsument och konsument som mellan företag och konsument. På det sättet är Internet ett högst aktivt medium, till skillnad från traditionella marknadsföringskanaler.Problemet kring dessa moderna kommunikationsmodeller och marknadsföringskanaler är det ändrade beteende det medför hos konsumenterna desto vanare de blir vid att använda Internet. I och med detta nya fenomen som kallas Internetmognad, blir konsumenterna allt svårare för marknadsförarna och företagen att påverka och det ställer nya och höga krav på de kanaler och den information som konsumenterna tar in. Syftet med uppsatsen är att ta reda på och undersöka hur e-handelsföretag inom kläd- och konfektionsbranschen använder sig av den moderna sortens marknadsföring. Vi kommer att rikta in oss på två av dessa, affiliatemarknadsföring och marknadsföring genom bloggar. Vidare vill vi ta reda på hur kommunikationen förmedlas, kommuniceras och tas emot av kvinnliga konsumenter i åldersgruppen 20-35 år.Uppsatsen innehåller dels en kvalitativ undersökning och dels en mindre kvantitativ undersökning. Avsikten har varit att skapa en övergripande helhetsbild över problemområdet ur företagens och konsumenternas perspektiv. Undersökningarna består av två djupintervjuer med det rena e-handelsföretaget NLY Scandinavia och postorder/e-handelsföretaget Ellos AB samt en enkätstudie bestående av 75 respondenter. Därefter har det empiriska materialet redogjorts och analyserats.Undersökningarna har resulterat i att vi kunnat göra antagandet att konsumenter de facto lyssnar på vänner, bekanta och andra konsumenter. Det har också framkommit att företagen använder sig av affiliatemarknadsföringen som en ren marknadsföringsstrategi, samtidigt som den är ett komplement till övrig mer traditionell marknadsföring. Bloggarna i sin tur fungerar snarare som ett verktyg eller ett informationsredskap i hop med affiliatemarknadsföringen för att förmedla informationen. / Uppsatsnivå: C

Konkurrensen mellan aktörerna inom klädbranschen är idag stor och kommer förmodligen växa sig ännu större de kommande åren. Alla företag måste därför sträva efter att komma förbi det enorma kommersiella bruset för att finna köpare till sina egna produkter. Det gäller att finna ett unikt och annorlunda sätt att marknadsföra sig på och som står ut i mängden bland alla företag som vill nå ut till konsumenterna. Ett företag som lyckats hålla sig vid liv på den konkurrensutsatta marknaden är butikskedjan Weekday. Weekday har lyckats expandera, trots att företaget håller en ganska låg profil jämfört med övriga konkurrenter vad gäller marknadsföring. Företaget använder ingen traditionell marknadsföring, t.ex. genom annonsering och reklam, utan förlitar sig främst på ryktesspridning. Därför är Weekday ett väldigt intressant företag att studera ur marknadsföringssynpunkt.Syftet med studien är att utifrån de teorier vi använt oss av, undersöka och beskriva butikskedjan Weekdays kommunikationsstrategi, vilka eventuella brister som finns samt om kundernas attityder skulle förändras vid övergång till traditionell marknadsföring. Studien har främst haft ett beskrivande syfte, men även till viss del ett explorativt syfte. Vi har använt oss av en kvalitativ undersökningsmetod, då underlaget till empirin och analysen främst utgjorts av djupintervjuer med anställda inom företaget samt den enkätundersökning vi genomfört. Vår teoretiska referensram bygger främst på kommunikationsmodellen, marknadsmixen (4P) och olika varumärkesteorier. Den empiriska delen beskriver Weekdays bakgrund, kommunikationsstrategi, vilka kommunikationskanaler företaget använder samt vilket resultat vår enkätundersökning gav. I analysen diskuterar vi och jämför empirin och enkätundersökningen med de olika teoretiska modellerna vi valt. Vi försöker på ett konstruktivt sätt analysera både för- och nackdelar med Weekdays kommunikationsstrategi. Vi har i vår slutsats kommit fram till att Weekdays främsta kommunikationskanaler består av de fysiska butikerna, eventen, hemsidan samt den mer okontrollerbara metoden ryktesspridning. De främsta bristerna ligger i att deras nuvarande kommunikationskanaler inte räcker till i samband med den expansion företaget befinner sig i. För att kunna utvecklas behövs andra metoder än de som används idag och ett steg mot traditionell marknadsföring hade kunna vara ett alternativ. Kundernas attityder till varumärket kommer till viss del förändras, men vi anser att det är en konsekvens företaget får ta i samband med att verksamheten utvecklas. / Program: Butikschefsutbildningen

Thesis (Ph.D.)--Massachusetts Institute of Technology, Dept. of Linguistics and Philosophy, 1980. / MICROFICHE COPY AVAILABLE IN ARCHIVES AND HUMANITIES. / Vita. / Bibliography: leaves 218-223. / by Ann Kathleen Farmer. / Ph.D.

A tese procurou estudar as dualidades na poesia de Armando Freitas Filho, considerando três temas da sua obra - a palavra, a cidade e a morte -, e como, a partir desses núcleos, o corpo se manifesta na sua poesia. / The thesis sought to explore the dualities in the poetry of Armando Freitas Filho, considering three themes of his work: the word, the city and death, and from these cores the body is manifested in his poetry.

This research aims to investigate the role of consumer engagement in influencing loyalty and word-of-mouth through a user generated brand community on Facebook by collecting evidence from both quantitative and qualitative studies. A quantitative study is adopted to test the engagement dimension and its relationships with other constructs, such as participation, loyalty, and word-of-mouth, with a sample size of 551 collected among Facebook users in the UK. The empirical analysis from the quantitative data supports the ABC (i.e., affective, cognitive, and behavioural) dimensions of engagement as assumed in the study and finds a positive relationship between engagement, participation, loyalty, and word-of-mouth. Similarly, a qualitative study is adopted in the form of a semi-structured interview held with six user Apple brand champions from an Apple user generated online brand community on Facebook. A thematic analysis is conducted to analyse the engagement dimensions and their relationship to participation, loyalty, and word-of-mouth. In addition, the application of both the methods (i.e., the quantitative and qualitative studies) to investigate the main aim helps the research to attain complementarity. The combination of both methods provides evidence to justify the engagement dimensions and their relation to participation, loyalty, and word-of-mouth. The quantitative study supports the argued engagement dimensions and their relationship with other constructs, whereas the qualitative study explores other components of engagement and their relationship with similar constructs, as well as helping to enhance the relationships and dimensions of engagement. Moreover, this study contributes to marketing literature by empirically validating customer loyalty and word-of-mouth as outcomes of customer participation and engagement. No study so far has empirically investigated the effect of customer engagement on loyalty and word-of-mouth in a user-generated online brand community context. This enhanced understanding of vigour, personal identity, attention, absorption, sharing, and learning suggests that marketers should concentrate on the type of information presented, as well as the format in which information is presented outside the company’s networks. Community markers, such as feelings, emotions, excitement, contribution, and interaction with peers, contribute significantly to engaging and influencing loyalty and word-of-mouth with both the brand community and the brand itself.

This thesis explores and analyzes Origen of Alexandria’s conviction that Scripture is itself the enfleshed Christ, or that “in the Scriptures the Word became flesh that he might tabernacle among us” (Philoc 15.19). For Origen, Scripture as the “Word of the Lord” is identical to the Word who was “with God,” and who “was God” in the Johannine Prologue. The Word assumes flesh not only in his birth, but also through the words and phrases of the patriarchs, prophets, and apostles. As a result, many scholars have noted the interesting “parallel” or “analogy” Origen draws between Scripture and the Incarnation, but this study provides the first comprehensive and focused treatment of Scripture as incarnate Word in Origen’s work. Ultimately, it demonstrates that for Origen, biblical interpretation is nothing less than a direct noetic encounter with the person of Christ, allowing the reader to know him in any time or place, to see him transfigured in the movement from the letter to the spirit, and even to consume his flesh and blood. Following an introductory chapter, the project consists of two parts. Part One (Chapters Two and Three) addresses the nature of “scriptural flesh” in Origen’s work. Chapter Two seeks to articulate what it means for the Word to become “flesh” in the first place, as well as what is required to “lift the veil” and perceive that flesh as divine. By examining the role of the cross in Origen’s Christology, it demonstrates that it is only in light of the Passion, through the lens of the crucified Christ, that the divinity of both man (Jesus) and text (Scripture) is made manifest. Chapter Three looks to define this scriptural “flesh” in Origen’s thought, specifically by relying on the doctrine of the epinoiai (the “aspects” or biblical titles of Christ). It is the epinoiai that clothe Christ and give him shape through the text. Part Two (Chapters Four and Five) addresses the theological and spiritual implications for the reader and interpreter of Scripture. Chapter Four examines the “coming of Christ” (parousia) as an individualized noetic phenomenon, brought about by the Christological reading of Scripture in any time or place. Finally, Chapter Five addresses the consumption of Christ through the Scriptures, which turns out to be much more about hermeneutics than about sacramental theology. A short conclusion follows, raising some of the broader implications for Origen studies as well as for the study of early Christian biblical exegesis.

Sum Kwok-wing. / Thesis submitted in: December 2003. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2004. / Includes bibliographical references (leaves 73-79). / Abstracts in English and Chinese. / Abstract --- p.ii / Chinese Abstract (論文摘要) --- p.iii / Acknowledgements --- p.iv / Table of Contents --- p.v / List of Tables --- p.vi / List of Figures --- p.vii / Chapter Chapter 1 - --- "Significance of Onset, Rime and Tone in Monosyllab Words" --- p.1 / Chapter Chapter 2 - --- General Methods --- p.19 / Chapter Chapter 3 - --- Experiment 1 --- p.28 / Chapter Chapter 4 - --- Experiment 2 --- p.35 / Chapter Chapter 5 - --- Experiment 3 --- p.41 / Chapter Chapter 6 - --- Comparison and Summary of the First Three Experiments --- p.47 / Chapter Chapter 7 - --- Experiment 4 --- p.53 / Chapter Chapter 8 - --- General Discussion --- p.59 / References --- p.72 / Appendix I --- p.79 / Appendix II --- p.87 / Appendix III --- p.95 / Appendix IV --- p.103 / Appendix V --- p.111 / Appendix VI --- p.112 / Appendix VII --- p.113