Grapheme-based speech recognition systems are faster to develop, but typically do not reach the same level of performance as phoneme-based systems. Using Afrikaans speech recognition as a case study, we first analyse the reasons for the discrepancy in performance, before introducing a technique for improving the performance of standard grapheme-based systems. It is found that by handling a relatively small number of irregular words through phoneme-to-grapheme (P2G) transliteration – transforming the original orthography of irregular words to an ‘idealised’ orthography – grapheme-based accuracy can be improved. An analysis of speech recognition accuracy based on word categories shows that P2G transliteration succeeds in improving certain word categories in which grapheme-based systems typically perform poorly, and that the problematic categories can be identified prior to system development. An evaluation is offered of when category-based P2G transliteration is beneficial and methods to implement the technique in practice are discussed. Comparative results are obtained for a second language (Vietnamese) in order to determine whether the technique can be generalised. / MSc (Computer Science) North-West University, Vaal Triangle Campus, 2014

Grapheme-based speech recognition systems are faster to develop, but typically do not reach the same level of performance as phoneme-based systems. Using Afrikaans speech recognition as a case study, we first analyse the reasons for the discrepancy in performance, before introducing a technique for improving the performance of standard grapheme-based systems. It is found that by handling a relatively small number of irregular words through phoneme-to-grapheme (P2G) transliteration – transforming the original orthography of irregular words to an ‘idealised’ orthography – grapheme-based accuracy can be improved. An analysis of speech recognition accuracy based on word categories shows that P2G transliteration succeeds in improving certain word categories in which grapheme-based systems typically perform poorly, and that the problematic categories can be identified prior to system development. An evaluation is offered of when category-based P2G transliteration is beneficial and methods to implement the technique in practice are discussed. Comparative results are obtained for a second language (Vietnamese) in order to determine whether the technique can be generalised. / MSc (Computer Science) North-West University, Vaal Triangle Campus, 2014

Färdigheten att avkoda ord är en grundläggande förutsättning för förmågan att läsa och har därmed en avgörande roll i elevers läsutveckling. Framgångsrik undervisning krävs för att eleverna ska ges möjlighet och förutsättningar att utveckla färdigheten att avkoda, det vill säga förståelse för sambandet mellan fonem och grafem. Med detta som utgångspunkt är syftet med studien att undersöka hur sju lärare i årskurs 1 beskriver att de utformar undervisningen i avkodning för att skapa förutsättningar för att eleverna ska utveckla färdigheten att avkoda ord. Studien utgår från en kvalitativ forskningsansats samt ett sociokulturellt perspektiv på lärande och utveckling. Empiriskt material har samlats in genom semi-strukturerade intervjuer där totalt sju lärare i årskurs 1 medverkade. Lärarna var slumpvis utvalda från olika skolor i två medelstora städer i Sverige. Resultatet visar att lärare använder olika medierande verktyg i form av digitala redskap, läromedel och aktiviteter i undervisningen, där samtalet och det sociala samspelet får stort utrymme. Det framkommer även att lärare använder sig själv som verktyg i avkodningsundervisningen där läraren stöttar, utmanar och fungerar som förebild för eleverna.  Vid utformandet av undervisningen framgår dessutom att forskning och beprövad erfarenhet samt elevers olika förutsättningar är någonting lärare förhåller sig till. Resultatet synliggör att undervisning kring avkodning varierar mellan olika lärare eftersom elevers individuella behov ligger till grund vid val av verktyg och vid utformningen av undervisningen. / Decoding skills are a crucial part of students reading development, since the ability to decode words is a fundamental qualification for reading. To give the pupils the opportunity to develop knowledge of phoneme-grapheme correspondences, an effective instruction is required. In light of this, the purpose of this paper is to examine how first grade teachers describe how they design instruction to create conditions for the students to develop decoding skills. The paper is based on a qualitative research approach and a socio-cultural perspective on learning and development. Through semi-structured interviews, empirical material has been collected from a total of seven participating first grade teachers. The teachers were randomly selected from different schools in two medium-sized cities in Sweden. The results show that teachers use mediating tools in form of digital gears, teaching materials and activities, where social interaction is emphasized. Teachers also function as tools, because they support, challenge and serve as role models for the pupils. Furthermore, the results reveal that research and proven experience as well as pupils’ different pre-conditions and needs affect teachers when they create decoding instruction. The results show that pupils’ individual needs are the starting point when teachers choose tools and create decoding instruction, which is why decoding instruction varies between different teachers.