Synaesthesia is a variation of normal human perception. A grapheme synaesthete, for example, can experience extra sensations, such as colours when seeing letters and/or numbers. Synaesthetic ability is commonly developed at an early age, and is linked to a genetic pre-disposition; however, there is a learnt component, as one must also learn to read and write to develop grapheme synaesthesia. To explore the extent to which synaesthesia can be learnt, a training method was employed, which was first used by Colizoli, Murre and Rouw (2012). In order to learn their own coloured letters a group of non-synaesthetic individuals read colour books, which are free eBooks reproduced to have four letters consistently appear in colour. Before and after reading, the participants completed a modified Stroop-design based on Mills (1999), which was used to measure if they had learnt the two key characteristics of synaesthesia, namely an involuntary and automatic reaction to letters. Both the colour reading (n=15) and control (n=6) groups did not have a significant involuntary reaction to letters. However, it was found that the participants had significantly more automatic reactions to letters. This included the control group, who did not read in colour, which suggests that merely completing the modified Stroop test is enough to learn the automatic characteristic of grapheme synaesthesia. / Psychology / M.A. (Psychology)

Submitted by Jorge Silva (jorgelmsilva@ufrrj.br) on 2018-09-12T18:04:15Z No. of bitstreams: 1 2017 - Helena Horvat de Farias Cesar.pdf: 3471223 bytes, checksum: 6d889a95cc8d6ac041268303f53a56f0 (MD5) / Made available in DSpace on 2018-09-12T18:04:15Z (GMT). No. of bitstreams: 1 2017 - Helena Horvat de Farias Cesar.pdf: 3471223 bytes, checksum: 6d889a95cc8d6ac041268303f53a56f0 (MD5) Previous issue date: 2017-04-27 / CAPES / This study investigates a specific case of overcorrection in the writing of students in the final years of elementary school: the improper addition of the grapheme <r> in the position of a syllable coda, based on present study as ?erregrafismo?. Records were found in different grammatical classes, such as "you", "boy", "to be", among others. Overcorrection is an excessive search for correction, as the term itself indicates. The phenomenon is produced when the individual interprets as incorrect a correct form of the language, consequently ends up exchanging in another form that considers cultured. This is due to linguistic insecurity, since the rules of the standard have not been internalized yet. Therefore, the present research has as general aim to minimize the cases of hypercorrection of the so-called ?erregrafismo? in the position of a syllabic coda in the writing of students of a class of 9th grade elementary school, through a proposal of pedagogical intervention that assists in due representation orthographic in regular morphological-grammatical relations. In line with this main objective, the following specific objectives have been identified: 1) to identify the contexts in which so-called ?erregrafismo? occurs most frequently; 2) develop adequate strategies to minimize the phenomenon of hypercorrection studied in the proposed research, in the different contexts in which it manifests itself. The hypothesis that guides this work is that questions of phonological and morphological awareness addressed in the intervention proposal formulated for this research, contribute to the proper application of the grapheme <r> in final position of the word, thus minimizing cases of hypercorrection in writing of students of elementary school. The methodology used is anchored to action research according to Thiollent's (2011) thinking. From the analysis of data, it was concluded that some students, due to the lack of syllabic tone perception and word segmentation in syllables, initially had difficulties in performing some proposed exercises. However, in the course of the intervention, it was found that the activities of comparison of words, seeking to perceive the similarities and / or differences between them, allowed the learners to consolidate correspondences between sound units (phonemes) and graphic units (graphemes) Thus, in the learning of morphological-grammatical phonemic-grapheme regularities. Another relevant finding was that activities involving pseudowords showed that although they did not know some invented words, the students used morphological knowledge to write such words, thus showing the importance of morphological awareness in spelling. / Este estudo investiga um caso espec?fico de hipercorre??o na escrita de estudantes dos anos finais do ensino fundamental: o acr?scimo indevido do grafema <r> em posi??o de coda sil?bica, cunhado no presente estudo como erregrafismo. Foram encontrados registros em diferentes classes gramaticais, como, por exemplo, ?voc?r? (voc?), ?meninor? (menino), ?estar? (est?), entre outros. A hipercorre??o ? uma busca excessiva de corre??o, como o termo por si s? indica. O fen?meno ? produzido quando o indiv?duo interpreta como incorreta uma forma correta da l?ngua, consequentemente acaba trocando por uma outra forma que considera culta. Isso ocorre por inseguran?a lingu?stica, uma vez que as regras da norma-padr?o ainda n?o foram interiorizadas. Sendo assim, a presente pesquisa tem como objetivo geral minimizar os casos de hipercorre??o do chamado erregrafismo em posi??o de coda sil?bica na escrita de alunos de uma turma de 9? ano do Ensino Fundamental, por meio de uma proposta de interven??o pedag?gica que auxilie na devida representa??o ortogr?fica em rela??es regulares morfol?gico-gramaticais. Em conson?ncia com esse objetivo principal, t?m-se os seguintes objetivos espec?ficos: 1) identificar os contextos nos quais o chamado erregrafismo ocorre com mais frequ?ncia; 2) desenvolver estrat?gias adequadas ? minimiza??o do fen?meno de hipercorre??o estudado na pesquisa proposta, nos diferentes contextos nos quais ele se manifesta. A hip?tese que norteia este trabalho ? que quest?es de consci?ncias fonol?gica e morfol?gica abordadas na proposta de interven??o formuladas para esta pesquisa contribuem para a devida aplica??o do grafema <r> em posi??o final de voc?bulo, minimizando, dessa forma, casos de hipercorre??o na escrita de estudantes do EF. A metodologia utilizada est? ancorada ? pesquisa-a??o, segundo o pensamento de Thiollent (2011). Da an?lise de dados, concluiu-se que alguns alunos, por apresentarem lacunas na percep??o da tonicidade sil?bica e na segmenta??o da palavra em s?labas, inicialmente tiveram dificuldades em realizar alguns exerc?cios propostos. Entretanto, no decorrer da interven??o, constatou-se que atividades de compara??o de palavras, buscando perceber as semelhan?as e/ou diferen?as entre elas, fizeram com que os aprendizes consolidassem correspond?ncias entre unidades sonoras (fonemas) e unidades gr?ficas (grafemas1), auxiliando, assim na aprendizagem das regularidades fon?mico-graf?micas morfol?gico-gramaticais. Outra constata??o relevante foi que atividades envolvendo pseudopalavras mostraram que apesar de desconhecerem algumas palavras inventadas, os alunos lan?aram m?o de conhecimentos morfol?gicos para escreverem tais voc?bulos, mostrando dessa forma a import?ncia da consci?ncia morfol?gica na ortografia.

Grapheme-to-phoneme conversion (G2P) is the task of converting a word, represented by a sequence of graphemes, to its pronunciation, represented by a sequence of phonemes. The G2P task plays a crucial role in speech synthesis systems, and is an important part of other applications, including spelling correction and speech-to-speech machine translation. G2P conversion is a complex task, for which a number of diverse solutions have been proposed. In general, the problem is challenging because the source string does not unambiguously specify the target representation. In addition, the training data include only example word pairs without the structural information of subword alignments. In this thesis, I introduce several novel approaches for G2P conversion. My contributions can be categorized into (1) new alignment models and (2) new output generation models. With respect to alignment models, I present techniques including many-to-many alignment, phonetic-based alignment, alignment by integer linear programing and alignment-by-aggregation. Many-to-many alignment is designed to replace the one-to-one alignment that has been used almost exclusively in the past. The new many-to-many alignments are more precise and accurate in expressing grapheme-phoneme relationships. The other proposed alignment approaches attempt to advance the training method beyond the use of Expectation-Maximization (EM). With respect to generation models, I first describe a framework for integrating many-to-many alignments and language models for grapheme classification. I then propose joint processing for G2P using online discriminative training. I integrate a generative joint n-gram model into the discriminative framework. Finally, I apply the proposed G2P systems to name transliteration generation and mining tasks. Experiments show that the proposed system achieves state-of-the-art performance in both the G2P and name transliteration tasks.

No description available.

Die vorliegende Arbeit beschäftigt sich mit einem datengetriebenen Verfahren zur Graphem-Phonem-Konvertierung für ein Sprachsynthesesystem. Die Aufgabe besteht darin, die Aussprache für beliebige Wörter zu bestimmen, auch für solche Wörter, die nicht im Lexikon des Systems enthalten sind. Die Architektur an sich ist sprachenunabhängig, von der Sprache abhängig sind lediglich die Wissensquellen, die zur Laufzeit des Systems geladen werden. Die Erstellung von Wissensquellen für weitere Sprachen soll weitgehend automatisch und ohne Einsatz von Expertenwissen möglich sein. Expertenwissen kann verwendet werden, um die Ergebnisse zu verbessern, darf aber keine Voraussetzung sein. Für die Bestimmung der Transkription werden zwei neuronale Netze verwendet. Das erste Netz generiert aus der Buchstabenfolge des Wortes die zu realisierenden Laute einschließlich der Silbengrenzen, und das zweite bestimmt im Anschluß daran die Position der Wortbetonung. Diese Trennung hat den Vorteil, daß man für die Bestimmung des Wortakzentes das Wissen über die gesamte Lautfolge einbeziehen kann. Andere Verfahren, die die Transkription in einem Schritt bestimmen, haben das Problem, bereits zu Beginn des Wortes über den Akzent entscheiden zu müssen, obwohl die Aussprache des Wortes noch gar nicht feststeht. Zudem bietet die Trennung die Möglichkeit, zwei speziell auf die Anforderung zugeschnittene Netze zu trainieren. Die Besonderheit der hier verwendeten neuronalen Netze ist die Einführung einer Skalierungsschicht zwischen der eigentlichen Eingabe und der versteckten Schicht. Eingabe und Skalierungsschicht werden über eine Diagonalmatrix verbunden, wobei auf die Gewichte dieser Verbindung ein Weight Decay (Gewichtezerfall) angewendet wird. Damit erreicht man eine Bewertung der Eingabeinformation während des Trainings. Eingabeknoten mit einem großen Informationsgehalt werden verstärkt, während weniger interessante Knoten abgeschwächt werden. Das kann sogar soweit gehen, daß einzelne Knoten vollständig abgetrennt werden. Der Zweck dieser Verbindung ist, den Einfluß des Rauschens in den Trainingsdaten zu reduzieren. Durch das Ausblenden der unwichtigen Eingabewerte ist das Netz besser in der Lage, sich auf die wichtigen Daten zu konzentrieren. Das beschleunigt das Training und verbessert die erzielten Ergebnisse. In Verbindung mit einem schrittweisen Ausdünnen der Gewichte (Pruning) werden zudem störende oder unwichtige Verbindungen innerhalb der Netzwerkarchitektur gelöscht. Damit wird die Generalisierungsfähigkeit noch einmal erhöht. Die Aufbereitung der Lexika zur Generierung der Trainingsmuster für die neuronalen Netze wird ebenfalls automatisch durchgeführt. Dafür wird mit Hilfe der dynamischen Zeitanpassung (DTW) der optimale Pfad in einer Ebene gesucht, die auf der einen Koordinate durch die Buchstaben des Wortes und auf der anderen Koordinate durch die Lautfolge aufgespannt wird. Somit erhält man eine Zuordnung der Laute zu den Buchstaben. Aus diesen Zuordnungen werden die Muster für das Training der Netze generiert. Um die Transkriptionsergebnisse weiter zu verbessern, wurde ein hybrides Verfahren unter Verwendung der Lexika und der Netze entwickelt. Unbekannte Wörter werden zuerst in Bestandteile aus dem Lexikon zerlegt und die Lautfolgen dieser Teilwörter zur Gesamttranskription zusammengesetzt. Dabei werden Lücken zwischen den Teilwörtern durch die neuronalen Netze aufgefüllt. Dies ist allerdings nicht ohne weiteres möglich, da es zu Fehlern an den Schnittstellen zwischen den Teiltranskriptionen kommen kann. Dieses Problem wird mit Hilfe des Lexikons gelöst, das für die Generierung der Trainingsmuster aufbereitet wurde. Hier ist eine eindeutige Zuordnung der Laute zu den sie generierenden Buchstaben enthalten. Somit können die Laute an den Schnittstellen neu bewertet und Transkriptionsfehler vermieden werden. Die Verlagsausgabe dieser Dissertation erschien 2005 im w.e.b.-Universitätsverlag Dresden (ISBN 3-937672-76-1). / The topic of this thesis is a system which is able to perform a grapheme-to-phoneme conversion for several languages without changes in its architecture. This is achieved by separation of the language dependent knowledge bases from the run-time system. Main focus is an automated adaptation to new languages by generation of new knowledge bases without manual effort with a minimal requirement for additional information. The only source is a lexicon containing all the words together with their appropriate phonetic transcription. Additional knowledge can be used to improve or accelerate the adaptation process, but it must not be a prerequisite. Another requirement is a fully automatic process without manual interference or post-editing. This allows for the adaptation to a new language without even having a command of that language. The only precondition is the pronunciation dictionary which should be enough for the data-driven approach to learn a new language. The automatic adaptation process is divided into two parts. In the first step the lexicon is pre-processed to determine which grapheme sequence belongs to which phoneme. This is the basis for the generation of the training patterns for the data-driven learning algorithm. In the second part mapping rules are derived automatically which are finally used to create the phonetic transcription of any word, even if it not contained in the dictionary. Task is to have a generalisation process that can handle all words in a text that has to be read out by a text-to-speech system.

Synaesthesia is a variation of normal human perception. A grapheme synaesthete, for example, can experience extra sensations, such as colours when seeing letters and/or numbers. Synaesthetic ability is commonly developed at an early age, and is linked to a genetic pre-disposition; however, there is a learnt component, as one must also learn to read and write to develop grapheme synaesthesia. To explore the extent to which synaesthesia can be learnt, a training method was employed, which was first used by Colizoli, Murre and Rouw (2012). In order to learn their own coloured letters a group of non-synaesthetic individuals read colour books, which are free eBooks reproduced to have four letters consistently appear in colour. Before and after reading, the participants completed a modified Stroop-design based on Mills (1999), which was used to measure if they had learnt the two key characteristics of synaesthesia, namely an involuntary and automatic reaction to letters. Both the colour reading (n=15) and control (n=6) groups did not have a significant involuntary reaction to letters. However, it was found that the participants had significantly more automatic reactions to letters. This included the control group, who did not read in colour, which suggests that merely completing the modified Stroop test is enough to learn the automatic characteristic of grapheme synaesthesia. / Psychology / M.A. (Psychology)

The study examined the typical development of early literacy in a group of typically developing preschool Sepedi first language children residing in Atteridgeville, by determining their performance on a protocol of early literacy tasks. The following aspects were included: written language awareness, narrative abilities, phonological awareness, letter name knowledge, grapheme-phoneme correspondence and literacy motivation. The performance of the participants on the various tasks was used to describe the early literacy development of the target population and to identify relevant risk criteria that may indicate delayed early literacy development in the target population. The performance of participants on these tasks differed from those of other participants in local and international studies, which underscores the necessity of culturally sensitive procedures for identifying delays in the early literacy development of children. The influence of factors such as the mother’s level of education, gender, participants’ level of engagement in literacy activities and participants' current academic performance on the development of early literacy skills were also investigated. Based on the results as well as other indications from the literature, possible risk factors for delayed early literacy development for this group are listed. / Dissertation (M (Communication Pathology))--University of Pretoria, 2007. / Speech-Language Pathology and Audiology / unrestricted

Carbon nanotubes (CNTs) and graphene nanoplatelets (GNPs) are attractive reinforcements for lightweight and high strength metal matrix composites due to their excellent mechanical and physical properties. The present work is an attempt towards investigating the effect of CNT and GNP reinforcements on the mechanical properties of nickel matrix composites. The CNT/Ni (dry milled) nanocomposites exhibiting a tensile yield strength of 350 MPa (about two times that of SPS processed monolithic nickel ~ 160 MPa) and an elongation to failure ~ 30%. In contrast, CNT/Ni (molecular level mixed) exhibited substantially higher tensile yield strength (~ 690 MPa) but limited ductility with an elongation to failure ~ 8%. The Ni-1vol%GNP (dry milled) nanocomposite exhibited the best balance of properties in terms of strength and ductility. The enhancement in the tensile strength (i.e. 370 MPa) and substantial ductility (~40%) of Ni-1vol%GNP nanocomposites was achieved due to the combined effects of grain refinement, homogeneous dispersion of GNPs in the nickel matrix, and well-bonded Ni-GNP interface, which effectively transfers stress across metal-GNP interface during tensile deformation. A second emphasis of this work was on the detailed 3D microstructural characterization of a new class of Ni-Ti-C based metal matrix composites, developed using the laser engineered net shaping (LENSTM) process. These composites consist of an in situ formed and homogeneously distributed titanium carbide (TiC) as well as graphite phase reinforcing the nickel matrix. 3D microstructure helps in determining true morphology and spatial distribution of TiC and graphite phase as well as the phase evolution sequence. These Ni-TiC-C composites exhibit excellent tribological properties (low COF), while maintaining a relatively high hardness.

Foreign words, such as names, locations or sometimes entire phrases, are a problem for any system that is meant to convert graphemes to phonemes (g2p; i.e.converting written text into phonetic transcription). In this thesis, we investigate both rule-based and neural methods of phonetically transcribing English words found in Icelandic text, taking into account the rules and constraints of how foreign phonemes can be mapped into Icelandic phonology. We implement a rule-based system by compiling grammars into finite-state transducers. In deciding on which rules to include, and evaluating their coverage, we use a list of the most frequently-found English words in a corpus of Icelandic text. The output of the rule-based system is then manually evaluated and corrected (when needed) and subsequently used as data to train a simple bidirectional LSTM g2p model. We train models both with and without length and stress labels included in the gold annotated data. Although the scores for neither model are close to the state-of-the-art for either Icelandic or English, both our rule-based system and LSTM model show promising initial results and improve on the baseline of simply using an Icelandic g2p model, rule-based or neural, on English words. We find that the greater flexibility of the LSTM model seems to give it an advantage over our rule-based system when it comes to modeling certain phenomena. Most notable is the LSTM’s ability to more accurately transcribe relations between graphemes and phonemes for English vowel sounds. Given there does not exist much previous work on g2p transcription specifically handling English words within the Icelandic phonological constraints and it remains an unsolved task, our findings present a foundation for the development of further research, and contribute to improving g2p systems for Icelandic as a whole.

The availability of the stenographic manuscripts of Astrid Lindgren have sparked an interest in the creation of a language model for stenography. By its very nature stenography is low-resource and the unavailability of data requires a tool for using normal data. The tool presented in this thesis is to create stenographic data from manipulating orthographic data. Stenographic data is distinct from orthographic data through three different types manipulations that can be carried out. Firstly stenography is based on a phonetic version of language, secondly it used its own alphabet that is distinct from normal orthographic data, and thirdly it used several techniques to compress the data. The first type of manipulation is done by using a grapheme-to-phoneme converter. The second type is done by using an orthographic representation of a stenographic alphabet. The third type of manipulation is done by manipulating based on subword level, word level and phrase level. With these manipulations different datasets are created with different combinations of these manipulations. Results are measured for both perplexity on a GPT-2 language model and for compression rate on the different datasets. These results show a general decrease of perplexity scores and a slight compression rate across the board. We see that the lower perplexity scores are possibly due to the growth of ambiguity.