• Title/Summary/Keyword: Language convergence

Search Result 789, Processing Time 0.023 seconds

Proposed Methodology for Building Korean Machine Translation Data sets Considering Phonetic Features (단어의 음성학적 특징을 이용한 한국어 기계 번역 데이터 세트 구축 방안)

  • Zhang Qinghao;Yang Hongjian;Serin Kim;Hyuk-Chul Kwon
    • Annual Conference on Human and Language Technology
    • /
    • 2022.10a
    • /
    • pp.592-595
    • /
    • 2022
  • 한국어에서 한자어와 외래어가 차지하는 비중은 매우 높다. 일상어의 경우 한자어와 외래어의 비중이 약 53%, 전문어의 경우 약 92%에 달한다. 한자어나 외래어는 중국이나 다른 나라로부터 영향을 받아 한국에서 쓰이는 단어들이다. 한국어에서 사용되는 한자어와 외래어의 한글 표기과 원어 표기를 발음해보면, 발음이 상당히 유사하다는 것을 알 수 있다. 한자어인 도서관(图书馆)을 중국어로 발음해보면 thu.ʂu.kwan'로 해당 단어에 대한 한국 사람의 발음과 상당히 유사하다. 본 논문에서는 Source Length, Source IPA Length, Target Length, Target IPA Length, IPA Distance 등 총 5가지의 음성학적 특징을 고려한 한국어-중국어 한국어-영어 단어 기계번역 데이터 세트를 구축하고자 한다.

  • PDF

Korean Machine Reading Comprehension using Continual Learning (Continual Learning을 이용한 한국어 기계독해)

  • Shin, JoongMin;Cho, Sanghyun;Choi, Jaehoon;Kwon, Hyuk-Chul
    • Annual Conference on Human and Language Technology
    • /
    • 2021.10a
    • /
    • pp.609-611
    • /
    • 2021
  • 기계 독해는 주어진 지문 내에서 질문에 대한 답을 기계가 찾아 답하는 문제이다. 딥러닝에서는 여러 데이터셋을 학습시킬 때에 이전에 학습했던 데이터의 weight값이 점차 사라지고 사라진 데이터에 대해 테스트 하였을때 성능이 떨어진 결과를 보인다. 이를 과거에 학습시킨 데이터의 정보를 계속 가진 채로 새로운 데이터를 학습할 수 있는 Continual learning을 통해 해결할 수 있고, 본 논문에서는 이 방법을 MRC에 적용시켜 학습시킨 후 한국어 자연어처리 Task인 Korquad 1.0의 MRC dev set을 통해 성능을 측정하였다. 세 개의 데이터셋중에서 랜덤하게 5만개를 추출하여 10stage를 학습시킨 50K 모델에서 추가로 Continual Learning의 Learning without Forgetting를 사용하여 학습시킨 50K-LWF 모델이 F1 92.57, EM 80.14의 성능을 보였고, BERT 베이스라인 모델의 성능 F1 91.68, EM 79.92에 비교하였을 때 F1, EM 각 0.89, 0.22의 향상이 있었다.

  • PDF

Mutilingualism and Language Education Policy (다언어주의와 언어교육정책)

  • Kim, Yangsoon
    • The Journal of the Convergence on Culture Technology
    • /
    • v.6 no.1
    • /
    • pp.321-326
    • /
    • 2020
  • This paper is to analyze the language education policy in the context of multilingualism. As the majority of the population are multilingual, language policy should be centered on the multilingual speakers as the norm, and multilingual language policy is the best route which we can follow as a language policy in education. The motivation and legitimacy of the multilingual policies are suggested in terms of 6 different perspectives: identity, sustainability, equity, World Englishes, machine translation, and Universal Grammar (UG). As a model of language policy, the English-Plus (i.e., English+n) policy and similarly the Korean-Plus (i.e., Korean+n) policy are suggested to be the most appropriate language policies in the field of education in America and Korea respectively. These plus policies aim at bilingual fluency in both the native language and other foreign languages that are constitutive of the multilingualism of the country in which the bilingualism is treated as a variant of multilingualism. In a period of convergence and diversity in the 4th Industrial Revolution, language diversity and multilingual policy should be considered as a right to be protected or as a resource to be conserved rather than as a problem to be solved.

L2 Learners' Communication Strategies: A Data Convergence Analysis (제 2 언어 학습자의 의사소통 전략: 자료융합 분석)

  • Kang, Su-Ja
    • Journal of the Korea Convergence Society
    • /
    • v.6 no.6
    • /
    • pp.263-270
    • /
    • 2015
  • L2 learners deal with communication problems encountered during the course of their oral communication by employing what are known as communication strategies(CSs). As CSs play a significant role in communication and language learning, identifying CSs that facilitate communication has become important in L2 research. However, there have been some controversies about how to identify CSs, especially about data sources: conversation data, retrospective interview data, or combining two types of data. Drawing on the advantages of converged data, this study aimed to identify CSs that L2 learners use in their English communication through a convergence analysis of data. In this qualitative case study, the data were collected from Korean L2 learners' conversations and retrospective verbal reports in stimulated recall interviews. Using a data convergence approach, various CSs were identified: definition replacement CSs, message change CSs, time-gaining CSs, and comprehension CSs. This study suggests the effectiveness of CSs in handling L2 communication problems and the usefulness of data convergence in L2 CS research.

The evaluation of methodological quality of meta-analysis studies in speech language pathology using AMSTAR (AMSTAR에 기반한 국내 언어치료 분야 메타분석 논문의 방법론적 질평가)

  • Han, Minju;Byeon, Haewon
    • Journal of the Korea Convergence Society
    • /
    • v.11 no.2
    • /
    • pp.161-165
    • /
    • 2020
  • Although research using meta-analysis is increasing in the field of rehabilitation science, not all meta-analytical papers are of the same quality. In particular, although meta-analysis is a research method with the highest level of evidence, it may be possible to derive distorted conclusions or alternatives by simply integrating representative values without considering heterogeneity among individual studies. This study analyzed the current status of meta-analysis papers on the subject of language arbitration published in Korea from January 2010 to June 2019, and used A Measurement Tool to Assess the Methodological Quality of Systematic Review (AMSTAR). As a result of evaluating the methodological quality of the final five papers, the average of 7.4 points out of 11 points was found above average. In order to raise the qualitative level of Meta-analysis in speech-language pathology in the future, it is necessary to include verification of publication bias and specification of conflicts of interest.

The effect of animation software interface on design thinking process - Protocol analysis of Alice and KidsPlay - (애니메이션 소프트웨어 인터페이스가 디자인 사고 과정에 미치는 영향 - Alice와 KidsPlay 프로그램 프로토콜 분석을 중심으로 -)

  • Jin, Yan;Lee, Hyun Kyung;Lee, Sang Won
    • Design Convergence Study
    • /
    • v.15 no.1
    • /
    • pp.37-48
    • /
    • 2016
  • Nowadays, CAD has became an essential tool for designers due to its easy edition and manipulation as well as the capability of communication between designers. Differences between design tool interfaces can cause gaps when visualize designer's ideas. This study is about the differences between various levels of metaphor interface language which is based on the theory of the Hutchins that high-level language interface can reduce the steps of process to visualize the idea of the designers than low-level language interface. This research is based on the assumption that high-level language interface will less interrupt the flow of the design thinking and make more various visual outcomes than the low-level language interface. We verified the hypothesis by analyzing differences of design thinking flow between user groups who used two different animation software. Hopefully, the verification of the hypothesis in this study can be able to guide the use pattern of the CAD tools.

A Research on Gamification Methodology for Korean Language Education (한국어 교육을 위한 게임화 방법론 연구)

  • Kwon, Chong-San;Woo, Tack
    • Journal of Korea Game Society
    • /
    • v.13 no.1
    • /
    • pp.61-74
    • /
    • 2013
  • In this paper we proposed a gamification methodology for Korean language education. We suggested gamification elements that increased the fun factor and induced immersion by analyzing the successful gamification cases which can be classified into three types; management, Implementation, and self-expression. Then, we analyzed Korean learning theory and methods to extract four key aspects, listening, speaking, vocabulary and culture from eight Korean learning function. Finally, by using Korean Wave contents, we proposed a Korean learning gamification methodology based on the gamification elements and the extracted four key aspects.

What Kind of Fun Food Marketing Do Customers Want?

  • CHA, Seong-Soo;LEE, Min-Ho
    • The Korean Journal of Food & Health Convergence
    • /
    • v.7 no.3
    • /
    • pp.1-11
    • /
    • 2021
  • Purpose of the research: This study aims to explain the state of marketing using fun among recent popular marketing strategies. Although companies are using various differentiated marketing strategies to gain a competitive edge, among them, fun marketing has constituted the most effective area of interest recently. Research design and methodology: To extract the customer selection attributes of fun marketing, after reviewing the literature, six optional attributes were selected from the factors of fun marketing towards consumers such as funny design, language play, celebrity use, funny taste, how to eat, and newtro (new + retro). Out of 300 questionnaires, 276 were used for analysis, excluding unscrupulous or incomplete questionnaires. The results were reviewed for validity and reliability using SPSS andAMOS, and the hypothesis was verified using structural equation modelling (SEM). Principal results: The results showed that funny design, language play, and newtro statistically significantly affected customer satisfaction, but celebrity use, funny taste, and eating methods had no significant effect. It was also confirmed that satisfaction had a statistically significant effect on repurchase intention. Major conclusions: This study can serve as basic data to enhance the marketing strategy of the food service industry, and it provides theoretical and practical implications.

A Study on the Cognitiveness of Pre-Language Therapists for Language Therapy Using Smart Media (스마트미디어 활용 언어치료에 대한 언어치료전공 대학생의 인식조사)

  • Lee, Eun Kyoung;Ahn, Soo Young;Park, Sun Hee
    • Journal of Digital Convergence
    • /
    • v.16 no.3
    • /
    • pp.469-476
    • /
    • 2018
  • This study investigates the common perception of the use of preliminary language rehabilitation as a basic study to demonstrate the usefulness of media utilisation in language therapy. Based on the previous research, data were collected for 150 preliminary language rehabilitation workers and 142 questionnaires were used for data analysis. The study found that the perception of media usage in pre-language rehabilitation was necessary for language development and was also positive about the treatment. The use of media has been recognised as helpful in promoting attention, cognitive and creativity development. Also, the use of media as a treatment tool was agreed to, but it was recognized that the rules for utilizing the media were necessary and needed to know the proper site or program. Therefore, it was suggested that the development of sites and programs that can be used for language therapy is necessary.

Study on Various Factors Associated with the Effects of Cyber Home Study in Korean Language Education based on Structural Equation Model (구조방정식을 이용한 국어 사이버 가정학습의 효과 관련 요인에 관한 연구)

  • Lim, Mi-Ja;Baek, Hyeon-Gi
    • Journal of Digital Convergence
    • /
    • v.6 no.1
    • /
    • pp.83-91
    • /
    • 2008
  • The objective of this research is to assess various factors affecting E-learning in Korean language education. In this research, we hypothesize that several factors such as absorption, motivation and tutors increase the educational effects of E-learning and ultimately affect learning attitude and satisfaction of students in E-learning. To discuss the hypothesis, we analyzed survey data of 300 students who were taking E-learning class of Korean language for three weeks in October 2007 based on Structural Equation Model. The result of our analysis shows that the factors such as absorption, motivation, tutors have positive effects on E-learning in Korean language education and positive influence on learning attitude and satisfaction on students as well.

  • PDF