Today, the Transformer model, which allows parallelization and also has its own internal attention, has been widely used in the field of speech recognition. The great advantage of this architecture is the fast learning speed, and the lack of sequential operation, as with recurrent neural networks. In this work, Transformer models and an end-to-end model based on connectionist temporal classification were considered to build a system for automatic recognition of Kazakh speech. It is known that Kazakh is part of a number of agglutinative languages and has limited data for implementing speech recognition systems. Some studies have shown that the Transformer model improves system performance for low-resource languages. Based on our experiments, it was revealed that the joint use of Transformer and connectionist temporal classification models contributed to improving the performance of the Kazakh speech recognition system and with an integrated language model it showed the best character error rate 3.7% on a clean dataset.
Abstract. In this paper the results of a study conducted to determine the perception of counselling needs by primary school teachers in Malaysia and Lesotho are reported. To this end, teachers were asked about the special needs of students they teach, how they cope with these students, what assistance they would like to have in teaching, and the best approach for dealing with student concerns. Teachers in Lesotho are dealing with larger numbers of special needs students and thus are having more difficulties than Malaysian teachers. Malaysian teachers appeared to be more interested in improving their teaching effectiveness. There was some variation in the perceptions of how to best deal with student concerns.
The Malaysian government recently reverted to the English language as the medium of instruction in its education system for important content subjects such as Mathematics and the Sciences. This policy decision was made in the midst of widespread concern over the perceived falling standards of English language proficiency after the country had used Malay, the native language of the majority Malays, as the medium of instruction for almost two decades. This year marks the first year high school students will sit for the national standardized examinations in English. Dissenters to this policy shift in the language of instruction fear that many will suffer as a result of this change. In this study, the self reports of academically high and low performing students are compared in terms of their ability to perform language related tasks in Malay and the English language. Using the Chi-square statistic, the results of the yes-no responses given by the students indicate that while high performing students report no major difficulties in either language, responses from the low performing students indicate otherwise. Additionally open-ended responses to learning Science and Mathematics in a second language are also qualitatively analysed in order to obtain a more comprehensive view of the challenges these students may face. Several implications based on these results are also discussed.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.