Thesis (Ph. D.)--Ohio State University, 2006. / Available online via OhioLINK's ETD Center; full text release delayed at author's request until 2007 Aug 15

Thesis (M.Sc. (Computer Science)) --University of Limpopo, 2012 / Pronunciation dictionaries or lexicons play an important role in guiding the predictive powers of an Automatic Speech Recognition (ASR) system. As the use of automatic speech recognition systems increases, there is a need for the development of dictionaries that cover a large number of inflected word forms to enhance the performance of ASR systems. The main purpose of this study is to investigate the contribution of the morphological approach to creating a more comprehensive and broadly representative Northern Sotho pronunciation dictionary for Automatic Speech Recognition systems. The Northern Sotho verbs together with morphological rules are used to generate more valid inflected word forms in the Northern Sotho language for the creation of a pronunciation dictionary. The pronunciation dictionary is developed using the Dictionary Maker tool. The Hidden Markov Model Toolkit is used to develop a simple ASR system in order to evaluate the performance of the ASR system when using the created pronunciation dictionary.

Thesis (PhD)--Stellenbosch University, 2013. / ENGLISH ABSTRACT: This study proposes a concept model for an electronic bilingual dictionary of sign language and Afrikaans (or English, or any other spoken language) that can be applied in practice by the De la Bat School for the Deaf in Worcester – or any school of similar organisation – to compile a bilingual dictionary in which it is not only possible to look up a sign via a written word but also to look up a sign even if the user does not necessarily know the written equivalent. This is important as it will mean that, for the first time in South Africa, a Deaf user would not have to be literate in any written language to use a dictionary in his or her own language. At the same time a hearing user would be able to look up a sign of which the meaning is unknown, for example when a sign is noticed in a conversation or an unknown context. The sign language dictionaries currently available in South Africa are little more than alphabetically arranged (or sometimes thematically arranged) lists in which the written language offers the only access to the content. As per introduction the first four chapters provide general discussions of sign language, the Deaf in South Africa, lexicography and sign language dictionaries. The discussions which follow in the remaining chapters focus on the decisions which have to be made in terms of the proposed dictionary’s content and design. These chapters contain recommendations regarding equivalent relations between sign language and Afrikaans, several outer texts which could be included in the frame structure, the characteristics of the electronic database from which the dictionary will be generated, as well as decisions that have to be made about the microstructure. In the last chapter the lexicographic model for an electronic bilingual Foundation Phase dictionary of South African Sign Language and Afrikaans is provided with complete examples. / AFRIKAANSE OPSOMMING: Hierdie studie stel 'n konsepmodel voor vir 'n elektroniese tweetalige woordeboek van gebaretaal en Afrikaans (of Engels, of enige ander gesproke taal) wat deur die De la Bat Skool vir Dowes in Worcester – of enige ander skool of soortgelyke instansie – in die praktyk aangewend sou kon word om 'n tweetalige woordeboek saam te stel waarin dit nie slegs moontlik is om 'n gebaar via 'n geskrewe woord op te soek nie, maar ook om 'n gebaar op te soek sonder dat die gebruiker noodwendig die geskrewe ekwivalent ken. Dit is belangrik aangesien dit die eerste keer in Suid- Afrika sal beteken dat 'n Dowe gebruiker nie in enige geskrewe taal geletterd hoef te wees om 'n woordeboek in sy of haar eie taal te gebruik nie. Terselfdertyd sal 'n horende gebruiker wat gebaretaal leer in staat wees om 'n gebaar op te soek waarvan die betekenis onbekend is, soos wanneer 'n gebaar in 'n gesprek of onbekende konteks opgemerk word. Die gebaretaalwoordeboeke wat op die oomblik in Suid-Afrika bestaan, is weinig meer as alfabetiese (of soms tematiese) lyste waarin die geskrewe taal die enigste toegang tot die inhoud bied. Ter inleiding word algemene besprekings van gebaretaal, Dowes in Suid-Afrika, leksikografie en gebaretaalwoordeboeke in die eerste vier hoofstukke verskaf. Die besprekings wat in die hoofstukke daarop volg, fokus in meer besonderhede op die besluite wat t.o.v. die voorgestelde woordeboek se inhoud en ontwerp gemaak moet word. Dié hoofstukke bevat aanbevelings rakende ekwivalentverhoudings tussen gebaretaal en Afrikaans, verskeie buitetekste wat in die raamstruktuur gebruik sou kon word, die eienskappe van die elektroniese databasis waaruit die woordeboek gegenereer word asook besluite wat oor die mikrostruktuur geneem moet word. In die laaste hoofstuk word die leksikografiese model vir 'n elektroniese tweetalige grondslagfasewoordeboek van Suid-Afrikaanse Gebaretaal en Afrikaans met volledige voorbeelde uiteengesit. / Nasionale Navorsingstigting / Harry Crosleystigting

No description available.

Thesis (MScEng (Electrical and Electronic Engineering))--University of Stellenbosch, 2010. / ENGLISH ABSTRACT: This thesis investigates various data-driven techniques by which pronunciation dictionaries can be automatically augmented. First, well-established grapheme-to-phoneme (G2P) conversion techniques are evaluated for Standard South African English (SSAE), British English (RP) and American English (GenAm) by means of four appropriate dictionaries: SAEDICT, BEEP, CMUDICT and PRONLEX. Next, the decision tree algorithm is extended to allow the conversion of pronunciations between different accents by means of phoneme-to-phoneme (P2P) and grapheme-andphoneme- to-phoneme (GP2P) conversion. P2P conversion uses the phonemes of the source accent as input to the decision trees. GP2P conversion further incorporates the graphemes into the decision tree input. Both P2P and GP2P conversion are evaluated using the four dictionaries. It is found that, when the pronunciation is needed for a word not present in the target accent, it is substantially more accurate to modify an existing pronunciation from a different accent, than to derive it from the word’s spelling using G2P conversion. When converting between accents, GP2P conversion provides a significant further increase in performance above P2P. Finally, experiments are performed to determine how large a training dictionary is required in a target accent for G2P, P2P and GP2P conversion. It is found that GP2P conversion requires less training data than P2P and substantially less than G2P conversion. Furthermore, it is found that very little training data is needed for GP2P to perform at almost maximum accuracy. The bulk of the accuracy is achieved within the initial 500 words, and after 3000 words there is almost no further improvement. Some specific approaches to compiling the best training set are also considered. By means of an iterative greedy algorithm an optimal ranking of words to be included in the training set is discovered. Using this set is shown to lead to substantially better GP2P performance for the same training set size in comparison with alternative approaches such as the use of phonetically rich words or random selections. A mere 25 words of training data from this optimal set already achieve an accuracy within 1% of that of the full training dictionary. / AFRIKAANSE OPSOMMING: Hierdie tesis ondersoek verskeie data-gedrewe tegnieke waarmee uitspraakwoordeboeke outomaties aangevul kan word. Eerstens word gevestigde grafeem-na-foneem (G2P) omskakelingstegnieke ge¨evalueer vir Standaard Suid-Afrikaanse Engels (SSAE), Britse Engels (RP) en Amerikaanse Engels (GenAm) deur middel van vier geskikte woordeboeke: SAEDICT, BEEP, CMUDICT en PRONLEX. Voorts word die beslissingsboomalgoritme uitgebrei om die omskakeling van uitsprake tussen verskillende aksente moontlik te maak, deur middel van foneem-na-foneem (P2P) en grafeem-en-foneem-na-foneem (GP2P) omskakeling. P2P omskakeling gebruik die foneme van die bronaksent as inset vir die beslissingsbome. GP2P omskakeling inkorporeer verder die grafeme by die inset. Beide P2P en GP2P omskakeling word evalueer deur middel van die vier woordeboeke. Daar word bevind dat wanneer die uitspraak benodig word vir ’n woord wat nie in die teikenaksent teenwoordig is nie, dit bepaald meer akkuraat is om ’n bestaande uitspraak van ’n ander aksent aan te pas, as om dit af te lei vanuit die woord se spelling met G2P omskakeling. Wanneer daar tussen aksente omgeskakel word, gee GP2P omskakeling ’n verdere beduidende verbetering in akkuraatheid bo P2P. Laastens word eksperimente uitgevoer om die grootte te bepaal van die afrigtingswoordeboek wat benodig word in ’n teikenaksent vir G2P, P2P en GP2P omskakeling. Daar word bevind dat GP2P omskakeling minder afrigtingsdata as P2P en substansieel minder as G2P benodig. Verder word dit bevind dat baie min afrigtingsdata benodig word vir GP2P om teen bykans maksimum akkuraatheid te funksioneer. Die oorwig van die akkuraatheid word binne die eerste 500 woorde bereik, en n´a 3000 woorde is daar amper geen verdere verbetering nie. ’n Aantal spesifieke benaderings word ook oorweeg om die beste afrigtingstel saam te stel. Deur middel van ’n iteratiewe, gulsige algoritme word ’n optimale rangskikking van woorde bepaal vir insluiting by die afrigtingstel. Daar word getoon dat deur hierdie stel te gebruik, substansieel beter GP2P gedrag verkry word vir dieselfde grootte afrigtingstel in vergelyking met alternatiewe benaderings soos die gebruik van foneties-ryke woorde of lukrake seleksies. ’n Skamele 25 woorde uit hierdie optimale stel gee reeds ’n akkuraatheid binne 1% van di´e van die volle afrigtingswoordeboek.

This thesis examines the use of electronic dictionaries by ESL students. In particular, it considers how, when and why students use electronic dictionaries in their writing processes. It also explores the extent to which students use words found in an electronic dictionary appropriately in their texts and whether electronic dictionary use influences their long-term acquisition of vocabulary.

Submitted by Renata Lopes (renatasil82@gmail.com) on 2016-04-05T11:25:54Z No. of bitstreams: 1 mauchaandradegamonal.pdf: 5111281 bytes, checksum: 1a21973abbc0ee27bb4401bad94f7099 (MD5) / Approved for entry into archive by Adriana Oliveira (adriana.oliveira@ufjf.edu.br) on 2016-04-24T03:47:58Z (GMT) No. of bitstreams: 1 mauchaandradegamonal.pdf: 5111281 bytes, checksum: 1a21973abbc0ee27bb4401bad94f7099 (MD5) / Made available in DSpace on 2016-04-24T03:47:58Z (GMT). No. of bitstreams: 1 mauchaandradegamonal.pdf: 5111281 bytes, checksum: 1a21973abbc0ee27bb4401bad94f7099 (MD5) Previous issue date: 2013-03-11 / CAPES - Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / Esta dissertação é parte do subprojeto Copa 2014 FrameNet Brasil (SALOMÃO ET AL., 2011), iniciativa da FrameNet Brasil em parceria com o projeto FrameCorp (CHISHMAN ET AL., 2008) e com a Berkeley FrameNet (FILLMORE ET AL., 2003) que propõe a elaboração de dicionário eletrônico trilíngue – Português, Inglês, Espanhol – para os domínios da Copa do Mundo, do Futebol e do Turismo. A elaboração deste recurso se mostra diferente dos demais dicionários eletrônicos por ser estruturado a partir da teoria da Semântica de Frames (FILLMORE, 1982, 1985; PETRUCK, 1996) e da metodologia da FrameNet (FILLMORE ET AL., 2003, 2003a; RUPPENHOFER ET AL., 2010). A contribuição desta pesquisa é estabelecer as diretrizes para a estruturação deste dicionário a partir da modelagem de frames da experiência turística. Para tanto, algumas perguntas guiam o desenvolvimento deste trabalho: i) Em que medida os frames do domínio turístico modelados com corpora compilados da língua portuguesa do Brasil servem para representar os frames do Turismo para as demais línguas do dicionário? ii) Como a FrameNet responde aos desafios colocados na estruturação de recursos lexicais multilíngues? É possível utilizar frames como interlíngua? iii) Que avaliação se pode fazer do Kicktionary, dicionário multilíngue do futebol, como produto que utiliza a FrameNet e a WordNet (MILLER, 1993, 1995; FELLBAUM, 1998)? As respostas a essas reflexões apontaram que: i) os frames do domínio turístico são modelados da mesma forma pelas diferentes culturas; ii) a rede semântica FrameNet precisa se adaptar às especificidades impostas pela lexicografia multilíngue, mas, devido ao caráter transcultural do Turismo e também da Copa do Mundo e do Futebol, os frames podem ser utilizados como interlíngua; ii) o Kicktionary, como dicionário multilíngue de domínio especializado que utiliza frames e synsets, deveria rever a funcionalidade de cada teoria na estruturação dos bancos de dados. As relações intralinguísticas poderiam acontecer via WordNet, e as relações interlinguísticas, via FrameNet. / This work is part of the subproject 2014 World Cup FrameNet Brasil (SALOMÃO ET AL., 2011), an initiative of FrameNet Brazil in partnership with the FrameCorp project (CHISHMAN ET AL., 2008) and Berkeley FrameNet (FILLMORE ET AL. 2003), which proposes developing a trilingual electronic dictionary - English, Portuguese, Spanish - for the domains of the World Cup, Soccer, and Tourism. The development of this resource is different from the other electronic dictionaries because it is structured within the theory of Frame Semantics (FILLMORE, 1982, 1985; PETRUCK, 1996) and the methodology of FrameNet (FILLMORE ET AL., 2003, 2003a; RUPPENHOFER ET AL., 2010). The contribution of this research is to establish guidelines for structuring this dictionary through the analysis of tourist experience frames. Therefore, some questions guide the development of this work: i) How does this resource respond to the challenges involved in structuring multilingual lexical resources? ii) Is it possible to use frames as an interlingual representation? iii) What evaluation can be made of Kicktionary, a multilingual dictionary of football as a product that uses both FrameNet and WordNet (Miller, 1993, 1995; FELLBAUM, 1998)? The answers to these considerations have shown that: i) frames of the tourist domain are modeled in the same way by different cultures; ii) FrameNet needs to adapt to the specificities imposed by multilingual lexicography, but due to the transcultural nature of Tourism and also the Soccer and the World Cup, frames can be used as interlingua; ii) the Kicktionary, as specialized multilingual dictionary which uses frames and synsets, should review the functionality of each theory in structuring databases. Intralinguistic relations could happen via WordNet, and interlingual relations via FrameNet.