Earticle

This paper investigates on how children's implicit knowledge of their mother tongue helps their language development from 12 to 48 months. For the literature review, this paper discusses kernel theoretical aspects of language development regarding three themes: 1) the relationships between implicit knowledge and language acquisition, 2) the role of self-correction for language development, and 3) the significance of the error analysis of children's utterances. To discuss these themes by analyzing empirical data, the author observed and collected data for a couple of twins' verbal utterances. The targets of analysis are their utterance errors and the courses of self -correction in three prominent grammatical aspects－i.e., negation ‘an,’ subject particle ‘i/ga,’ and deixis ‘this, that, it’－which frequently appear and are crucial grammatical morphemes in child language. Based on the discussion, this paper emphasizes that children employ and test various hypotheses on the rules of their mother tongue as the part of their implicit knowledge, and they finally gain grammatical knowledge from ‘unseen’ efforts of self-correction

This paper introduces an adaptive and integrated utterance verification (UV) framework using minimum verification error (MVE) training as a new set of solutions suitable for real applications. UV is traditionally considered an add-on procedure to automatic speech recognition (ASR) and thus treated separately from the ASR system model design. This traditional two-stage approach often fails to cope with a wide range of variations, such as a new speaker or a new environment which is not matched with the original speaker population or the original acoustic environment that the ASR system is trained on. In this paper, we propose an integrated solution to enhance the overall UV system performance in such real applications. The integration is accomplished by adapting and merging the target model for UV with the acoustic model for ASR based on the common MVE principle at each iteration in the recognition stage. The proposed iterative procedure for UV model adaptation also involves revision of the data segmentation and the decoded hypotheses. Under this new framework, remarkable enhancement in not only recognition performance, but also verification performance has been obtained.

Utterance verification aims at rejecting both out-of-vocabulary (OOV) utterances and low-confidence-scored in-vocabulary (IV) utterances. For utterance verification on Korean connected digit recognition task, we investigate several methods to construct filler and anti-digit models. In particular, we propose a substitution error correction method based on 2-best decoding results. In this method, when 1st candidate is rejected, 2nd candidate is selected if it is accepted by a specific hypothesis test, instead of simply rejecting the 1st one. Experimental results show that the proposed method outperforms the conventional log likelihood ratio (LLR) test method.

음성인식에서 발화검증은 비인식대상어휘(OOV)를 기각시키고, 인식대상어휘라도 오인식 가능성이 높은 결과를 기각시키는 기술을 말한다. 본 논문에서는 혼동가능성 높은 숫자쌍들이 존재하는 한국어 연결 숫자 인식에서 발화검증 결과로 숫자열 기각시 오인식 가능성이 높은 숫자열을 그냥 기각시키는 대신에 대체오류를 수정하여 인식성능을 향상시키고자 하였다. N-best decoding 결과에 따르면 $2^{nd}\;best$ 나 $3^{rd}\;best$안에 대부분의 제대로 된 인식결과들이 포함된다. 따라서, N-best decoding을 이용해, 숫자열 기각시 $2^{nd}\;best$ 숫자열로 대체된 것이라고 가정한 후, 개별숫자 log likelihood ratio(LLR)과 N-best 기반의 숫자열 LLR[3] 등을 함께 고려한 신뢰도 측정방식에 의해 그 가정이 맞다고 판단이 되면 $2^{nd}\;best$ 의 숫자열과 대체함으로써 부분적으로 오류를 수정하였다.

In this study, natural colloquialisms that young people in their 20s and 30s actually use in daily life in Korea were studied, and five machine translation platforms, such as Naver Papago and Kakaoi in Korea and ‘百度’, ‘腾讯’ and ‘有道’ in China, were used as subjects of study. This study examined the error aspects of colloquial machine translation between Korea and China. This study does not focus on the evaluation and comparison of machine translation platform performance, but focuses more on the inherent characteristics of Korean spoken language, looking at how AI machine translation handles subtle situational nuances or linguistic and cultural contexts, and what they have in common. Efforts were made to reveal the error patterns. While previous studies on Korean-Chinese machine translation mainly dealt with written language such as newspaper articles, editorials, speeches, and reports, and even if they dealt with colloquial language, they focused on Chinese conversation or refined dialogues appearing in HSK textbooks. It is differentiated in that machine translation was studied by taking live dialogues actually used in daily life by young Koreans in their 20s and 30s who appeared in season 2 of the Netflix original reality program <Single's Inferno>.