• Child Health Nursing Research
• /
• v.28 no.3
• /
• pp.187-196
• /
• 2022

Purpose: This study aimed to clarify the factors related to mothers' decisions to provide disease explanations to their children, investigate the associations between those factors and social adaptive skills, and examine support relative to children's growth and development. Methods: Data were collected from anonymous, self-administered questionnaires answered by 71 mothers of outpatient school-aged children with congenital heart disease. The questionnaire items included characteristics of mothers and children, decision factors for providing disease explanations, and children's social adaptive skills (Asahide-Shiki social adaptive skills test). Factor analysis was performed on the decision factors, and multiple regression analysis was performed to examine the association between the identified factors and each social adaptive skill. Results: The decision factors for providing disease explanations were the mother's explanation ability, the receptive capability of the child, and the child's interest and literacy regarding the explanation. Multiple regression analysis showed that language and social life skills were significantly associated with the child's receptive capability, and language and daily life skills were significantly associated with the child's interest and literacy regarding the explanation. Conclusion: Improving children's language, social life, and daily life skills may enhance their receptive capability and literacy regarding explanations of their disease.

• Annual Conference on Human and Language Technology
• /
• 2021.10a
• /
• pp.9-15
• /
• 2021

기계번역 품질 예측 (Quality Estimation, QE)은 정답 문장에 대한 참조없이 소스 문장과 기계번역 결과를 통해 기계번역 결과에 대한 품질을 수준별 주석으로 나타내주는 태스크이며, 다양한 활용도가 있다는 점에서 꾸준히 연구가 수행되고 있다. 그러나 QE 모델 학습을 위한 데이터 구성 시 기계번역 결과에 대해 번역 전문가가 교정한 문장이 필요한데, 이를 제작하는 과정에서 상당한 인건비와 시간 비용이 발생하는 한계가 있다. 본 논문에서는 번역 전문가 없이 병렬 또는 단일 말뭉치와 기계번역기만을 활용하여 자동화된 방식으로 한국어-영어 합성 QE 데이터를 구축하며, 최초로 단어 수준의 한국어-영어 기계번역 결과 품질 예측 모델을 제작하였다. QE 모델 제작 시에는 Cross-lingual language model (XLM), XLM-RoBERTa (XLM-R), multilingual BART (mBART)와 같은 다언어모델들을 활용하여 비교 실험을 수행했다. 또한 기계번역 결과에 대한 품질 예측의 객관성을 검증하고자 구글, 아마존, 마이크로소프트, 시스트란의 번역기를 활용하여 모델 평가를 진행했다. 실험 결과 XLM-R을 활용하여 미세조정학습한 QE 모델이 가장 좋은 성능을 보였으며, 품질 예측의 객관성을 확보함으로써 QE의 다양한 장점들을 한국어-영어 기계번역에서도 활용할 수 있도록 했다.

• Language and Information
• /
• v.18 no.1
• /
• pp.53-75
• /
• 2014

This study discovers the dynamic nature of an interactional hierarchy as well as an institutional hierarchy in the use of Korean honorifics. Data was collected from the conversations of two Korean female interlocutors. The interlocutors met for the first time in the U.S. and often changed their use of honorifics. The paper examines the method in which the two interlocutors negotiate hierarchies during interaction and how the negotiation is reflected in their use of honorific shifts. The paper also investigates honorific shifts in terms of self-representation to suggest that there is another hierarchy at work other than the institutional hierarchy. An examination of the data shows that the shifts occurred not randomly but strategically. The findings suggest that 1) interlocutors may negotiate interactional hierarchy during their conversation, often in the same sentence, 2) interactional hierarchy often cross the boundary of the institutional hierarchy to obtain interactional goals, in this case, intimacy, and 3) the utterance contents may play a significant role in the interlocutors' honorific shifts.

• Journal of the Korean Society for Precision Engineering
• /
• v.9 no.1
• /
• pp.96-105
• /
• 1992

A study on the construction of a CAD/CAM system operated by 16 Bit PC basic language, for machining sculptured surface of die, was carried out. The system consists of 2 steps i.e., process for geometric modelling by wire frame and process for machining data generation. Geometric modelling for sculptured surface is made by the point data fitting, parallel sweeping, normal sweeping and linear connection of cross section curve. Machining data are gained by cutter off-set of geometric model data and machining carried out by DNC. This system is to be proved enough for rough cutting by actual machining experiment. But, for becoming a high level system, another method of cutter off-set has to be regarded and system must be reconstructed by another program language.

• Speech Sciences
• /
• v.12 no.1
• /
• pp.65-74
• /
• 2005

The aim of this paper is to suggest an index of coarticulation, proportional undershoot values, given the observation that absolute undershoot within a language tends to be proportional to target-locus difference. The target-locus proportionality predicts that a large difference between the consonant locus and the vowel target will result in a large amount of vowel undershoot, while a small difference a small amount of vowel undershoot. It turns out that the proportional undershoot is a potentially more appropriate way of comparing degree of undershoot across languages. However, even though the proportional undershoot measurement may provide a useful index comparing the overall coarticulation degree in a CV token for cross-linguistic data, it is concluded that it may potentially wrongly predict the cases of transfer or error as a progress in learning.

• International Journal of Human Ecology
• /
• v.6 no.1
• /
• pp.75-86
• /
• 2005

This study examines unique manifestations of intergenerational conflict related to the acculturation process of immigrant families. No scale that measured the acculturation aspect of intergenerational conflict exsited. Thus, a new scale was developed to investigate this unique aspect among Korean American adolescents. The study design was cross-sectional, and employed a convenience sampling method. The participants were Korean American adolescents of junior and senior high school age, 14 to18 years old. The study was conducted at eleven Korean churches and one hakwon (private out-of-school studies .institute) in Fairfax County, Virginia. Korean American adolescents expressed that the issues related to education, such as academic pressures and high expectations, caused intergenerational conflict most frequently. Unlike findings from previous studies, the participants indicated that language differences between parents and children rarely caused intergenerational conflict. Contrary to previous findings, none of the characteristics variables, such as age, gender, length of residency and language preference, were significantly correlated with this unique conflict. This study provides a rare opportunity to enhance our understanding on how Korean American adolescents interact with their immigrant parents.

• Korean Journal of English Language and Linguistics
• /
• v.3 no.3
• /
• pp.341-373
• /
• 2003

A cross-linguistic examination of motion constructions reveals that the nature of the special property of English motion verbs that Tenny (1995) discussed-namely, why English can freely append locational delimiters to manner-of-motion verbs, as in Bill swam/rowed/canoed to the end of the lake -resides not in the verbs but in the semantic structure of the prepositions that denote transition from motion to end location. It is further argued that the differentiation of bounded paths from non-bounded Ones provides a clear-cut basis on which to distinguish motion constructions from resultative constructions. This proposal provides an alternative to the analyses of resultative constructions by Wechsler (1997) and Rappaport Hovav and Levin (2001).

• Language and Information
• /
• v.20 no.1
• /
• pp.1-24
• /
• 2016

Korean is well-known for its topic-prominent properties. In this paper, we look into several subtypes of topic constructions whose grammatical complexities have received much attention in generative grammar. From a semantic/pragmatic view, topics in Korean can be classified into three different types: aboutness, contrastive, and scene-setting. Meanwhile, syntax can classify topic constructions into two types, depending on whether or not the comment clause following topic has a syntactic gap linked to the topic. In this paper, we review some key properties of these topic constructions in Korean, and suggest that each type is licensed by tight interactions between syntactic and semantic constraints. In particular, the paper tries to offer a Construction Grammar analysis where each grammatical component is interacting in non-modular ways and in which the multiple inheritance network of constructions plays an important role in capturing cross-cutting generalizations of the topic constructions.

• Annual Conference on Human and Language Technology
• /
• 2000.10d
• /
• pp.224-230
• /
• 2000

기존의 전문 검색 시스템은 문서를 단지 단어의 연속이라는 제한적 관점에서만 바라보았다. 또한 기존의 필드 검색 시스템은 고정된 필드를 색인 및 검색대상으로 하거나, 문서의 내용이 아닌 메타 정보에 관한 검색만이 가능하였다. 본 논문에서는 내용과 필드 구조를 통합하여 가변 필드 구조 문서를 색인 및 검색하는 모델인 교차 필드 검색 모델을 제안한다. 기존 정보검색 시스템의 기능을 기본으로 제공하면서 필드구조를 색인/검색하기 위한 기능적 요구사항을 제시하고, 내용 및 필드 구조를 색인하면서 동적인 삽입/삭제가 가능한 색인 구조를 제안한다. 아울러 검색시에 문서 가중치를 계산하여 문서를 순위조정하는 불리언 모델, 확장 불리언 모델, 벡터 공간 모델의 변형 모델을 제시한다. 아울러 구현 사례로 STEER-XDS 검색 시스템에 대해 알아본다.

• Annual Conference on Human and Language Technology
• /
• 2022.10a
• /
• pp.418-423
• /
• 2022

한국어 형태소 분석은 자연어 처리의 기초가 되는 태스크이므로 빠르게 결과를 출력해야 한다. 기존연구는 자동회귀 모델을 한국어 형태소 분석에 적용하여 좋은 성능을 기록하였다. 하지만 자동회귀 모델은 느리다는 단점이 있고, 이 문제를 극복하기 위해 비자동회귀 모델을 사용할 수 있다. 비자동회귀 모델을 한국어 형태소 분석에 적용하면 조화롭지 않은 시퀀스 문제와 토큰 반복 문제가 발생한다. 본 논문에서는 두 문제를 해결하기 위하여 다중 디코더 기반의 한국어 형태소 분석을 제안한다. 조화롭지 않은 시퀀스는 다중 디코더를 적용함으로써, 토큰 반복 문제는 두 개의 디코더에 서로 어텐션을 적용하여 문제를 완화할 수 있다. 본 논문에서 제안한 모델은 세종 형태소 분석 말뭉치를 대상으로 좋은 성능을 확보하면서 빠르게 결과를 생성할 수 있음을 실험적으로 보였다.