• English Language & Literature Teaching
• /
• v.15 no.2
• /
• pp.245-259
• /
• 2009

The purpose of this study was to investigate English teachers' perceptions about doing their own action research to find out the benefits of involvement in action research. I believe that teachers should engage in curriculum research and development because it relates to their own classrooms and because a primary aim for teacher education program is to give teachers ways of exploring their own classrooms. I focused on 17 graduate students who had undertaken action research during the fall semester of 2007 and administered a questionnaire about their perceptions of doing action research. The results revealed that their perceptions of doing action research fell into the following two categories, positive and negative aspects. For positive aspects, they experienced a sense of personal and professional growth and they underwent identity transformation from passive, etic-oriented, and uncritical to active, emic-oriented, and critical teachers. However, many of them expressed that major obstacles to doing action research were time constraints and lack of skills or training in conducting action research. Thus, it is suggested that both pre- and in-service teachers should consider conducting a language teaching diary study, doing collaborative action research, and acquiring all the necessary skills for conducting action research.

• Speech Sciences
• /
• v.1
• /
• pp.27-38
• /
• 1997

The differences in French Intonation between French and Korean learners can be applied to French intonation education. One native French speaker and three native Korean speakers who learned French language at High school were selected for this study. The subjects spoke test phrases based on the different syntactic structures. High-Speed speech Analysis system(RILP) was used for this experiment. The different intonation curves were showed at the end of phrase and at the beginning of phrase between French and Korean learners. At the end of phrases, French intonation appeared to have increasing and decending pitch contours in the case of wh-question, exclamation and finality. However, Korean learner's intonation showed only increasing pitch contours. At the beginning of phrase, French intonation shows decending pitch contours in the case of minor continuation and command. In contrast, Korean learner's intonation appeared to have increasing pitch contours. The new intonation training system using PC can have great effect on education of French as a second language.

• Annual Conference on Human and Language Technology
• /
• 2015.10a
• /
• pp.124-128
• /
• 2015

문맥의존 철자오류는 해당 단어만 봤을 때에는 오류가 아니지만 문맥상으로는 오류인 문제를 말한다. 이러한 문제를 해결하기 위해서는 문맥정보를 보아야 하지만, 형태소 분석 단계에서는 자세한 문맥 정보를 보기 어렵다. 본 논문에서는 형태소 분석 정보만을 이용한 철자오류 수정을 위한 문맥으로 사전훈련(pre-training)된 단어 표현(Word Embedding)를 사용하고, 기존의 기계학습 알고리즘보다 좋다고 알려진 딥 러닝(Deep Learning) 기술을 적용한 시스템을 제안한다. 실험결과, 기존의 기계학습 알고리즘인 Structural SVM보다 높은 F1-measure 91.61 ~ 98.05%의 성능을 보였다.

• International Journal of Computer Science & Network Security
• /
• v.21 no.11
• /
• pp.328-332
• /
• 2021

The article considers general tendencies in world and education, and also both principles and methods of forming professional, communicative, intercultural competences and in the process of teaching foreign language for professional purposes in the conditions of engineering, economic and other non-linguistic specialties at technical university. The article views some essential issues of this competence including awareness of pedagogical values, the construction of the pedagogical process, pedagogical communication and behavior; pedagogical technology, its essence, structural components, understanding of innovative components of professional activity, requirements for the design and engineering of pedagogical technologies.

• Electronics and Telecommunications Trends
• /
• v.38 no.5
• /
• pp.1-11
• /
• 2023

The emergence of generative hyperscale artificial intelligence (AI) has enabled new services, such as image-generating AI and conversational AI based on large language models. Such services likely lead to the influx of numerous users, who cannot be handled using conventional AI models. Furthermore, the exponential increase in training data, computations, and high user demand of AI models has led to intensive hardware resource consumption, highlighting the need to develop domain-specific semiconductors for hyperscale AI. In this technical report, we describe development trends in technologies for hyperscale AI processors pursued by domestic and foreign semiconductor companies, such as NVIDIA, Graphcore, Tesla, Google, Meta, SAPEON, FuriosaAI, and Rebellions.

• Annual Conference on Human and Language Technology
• /
• 2022.10a
• /
• pp.461-464
• /
• 2022

법원도서관이 발간하는 판례공보를 기반으로 판결문 자동요약을 위한 학습 데이터들이 구축되고 있다. 그런데 판결문 요약에서는 뉴스 요약과는 달리 추출요약과 생성요약 방식이 함께 사용되는 특수성이 있고, 이러한 특수성 때문에 현재 판결문 요약 데이터셋이 요약 프로그램의 성능 향상을 이끌지 못하고 있다고 생각된다. 따라서 법률가들이 판결문을 요약하는 방식을 반영하여, 추출요약 방식으로 작성된 판결요지와 생성요약 방식으로 작성된 판결요지를 분리해서 요약 데이터셋을 만들 필요가 있다. 추출요약과 생성요약에 관한 데이터셋을 따로 구축하기 위해서는 판례공보의 판결요지를 추출요약과 생성요약으로 분류하는 작업이 필요한데, 감성 분석에 사용되는 알고리즘이 판결요지의 분류 작업에 응용될 수 있다는 것을 실험 결과로 알 수 있었다.

• Journal of Korean Neurosurgical Society
• /
• v.66 no.3
• /
• pp.289-297
• /
• 2023

Technological advances in neonatology led to the improvement of the survival rate in preterm babies with very low birth weights. However, intraventricular hemorrhage (IVH) has been one of the major complications of prematurity. IVH is relevant to neurodevelopmental disorders, such as cerebral palsy, language and cognitive impairments, and neurosensory and psychiatric problems, especially when combined with brain parenchymal injuries. Additionally, severe IVH requiring shunt insertion is associated with a higher risk of adverse neurodevelopmental outcomes. Multidisciplinary and longitudinal rehabilitation should be provided for these children based on the patients' life cycles. During the infantile period, it is essential to detect high-risk infants based on neuromotor examinations and provide early intervention as soon as possible. As babies grow up, close monitoring of language and cognitive development is needed. Moreover, providing continuous rehabilitation with task-specific and intensive repetitive training could improve functional outcomes in children with mild-to-moderate disabilities. After school age, maintaining the level of physical activity and managing complications are also needed.

• Analyses & Alternatives
• /
• v.2 no.2
• /
• pp.1-30
• /
• 2018

The concept of soft power presents both theoretical and practical difficulties for researchers and policy makers. This essay examines the practical use of soft power and argues that soft power resources in education are important for attracting individuals to migrate between countries and assimilate to the new language and culture of the new location. Japan's attempts to utilize soft power resources in its educational system have provided mixed results dependent on the target population. Japan has successfully attracted individuals into fields related to higher education much more so than skilled labor programs. This essay discusses the importance of educational soft power resources in Japanese strategy to increase educated working population that is assimilated to Japanese language and culture. After reviewing the literature on soft power in Asia, and Japanese cultural integration policies, the essay examines three cases of Japanese educational soft power - the JET Programme, the caregiver-training program, and internationalization of university programs. In addition, the essay shows that Japan is more successful attracting higher educated individuals seeking higher paying employment rather than skilled labor through these programs.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2021.07a
• /
• pp.69-71
• /
• 2021

특허분야에서 자연어처리 태스크는 특허문헌의 언어적 특이성으로 문제 해결의 난이도가 높은 과제임에 따라 한국 특허문헌에 최적화된 언어모델의 연구가 시급한 실정이다. 본 논문에서는 대량의 한국 특허문헌 데이터를 최적으로 사전 학습(pre-trained)한 Korean Patent ELECTRA 모델과 tokenize 방식을 제안하며 기존 범용 목적의 사전학습 모델과 비교 실험을 통해 한국 특허문헌 자연어처리에 대한 발전 가능성을 확인하였다.

• Annual Conference on Human and Language Technology
• /
• 2019.10a
• /
• pp.236-238
• /
• 2019

신경망 기계 번역(Neural Machine Translation)은 주로 지도 학습(Supervised learning)을 이용한 End-to-end 방식의 연구가 이루어지고 있다. 그러나 지도 학습 방법은 데이터가 부족한 경우에는 낮은 성능을 보이기 때문에 BERT와 같은 대량의 단일 언어 데이터로 사전학습(Pre-training)을 한 후에 미세조정(Finetuning)을 하는 Transfer learning 방법이 자연어 처리 분야에서 주로 연구되고 있다. 최근에 발표된 MASS 모델은 언어 생성 작업을 위한 사전학습 방법을 통해 기계 번역과 문서 요약에서 높은 성능을 보였다. 본 논문에서는 영어-한국어 기계 번역 성능 향상을 위해 MASS 모델을 신경망 기계 번역에 적용하였다. 실험 결과 MASS 모델을 이용한 영어-한국어 기계 번역 모델의 성능이 기존 모델들보다 좋은 성능을 보였다.