• Annual Conference on Human and Language Technology
• /
• 2020.10a
• /
• pp.501-506
• /
• 2020

질의 난이도 분석 문제는 자연어 질의문을 답변할 때 어려움의 정도를 측정하는 문제이다. 질의 난이도 분석 문제는 문서 독해, 의학 시험, 비디오 질의 등과 같은 다양한 데이터셋에서 연구되어 왔다. 본 논문에서는 질의문과 질의문에 응답하기 위한 정보들 간의 관계를 파악하는 것으로 질의 난이도 분석 문제를 접근하여 이를 BERT와 Dual Multi-head Attention을 사용하여 모델링 하였다. 본 논문에서 제안하는 모델의 우수성을 증명하기 위하여 최근 자연언어이해 부분에서 높은 성능을 보여주는 기 학습 언어 모델과 이전 연구의 질의 난이도 분석 모델과의 성능을 비교하였고, 제안 모델은 대표적인 비디오 질의 응답 데이터셋인 DramaQA의 Memory Complexity에서 99.76%, Logical Complexity에서는 89.47%의 정확도로 가장 높은 질의 난이도 분석 성능을 보여주었다.

• The Journal of the Korea Contents Association
• /
• v.19 no.4
• /
• pp.207-217
• /
• 2019

The current study aimed to examine how the caregivers and speech language pathologists (SLPs) would perceive the important communication contexts for 3~6 year-old children in the script-based intervention. Analytic Hierarchy Process analyses were conducted for the obtained survey data from SLPs and caregivers. Results showed that caregivers perceived 'educational and social life' and SLPs perceived 'family life' as the most important contexts from among three high-level contexts. Within each high-level context, 'meal time' in 'family life', 'playing with friends' in 'educational and social life', and 'playground' in 'culture and leisure life' were the most important communication contexts. Also, from among 35 global priority rank, caregivers perceived 'playing with friends' while SLPs perceived 'meal time' as the most important contexts, and 'family life' was high priority among first five ranked context both in caregivers and SLPs. In caregiver groups by children's age, 3 year- and 6 year-caregiver groups perceived 'family life' and 4 year- and 5 year-caregiver groups perceived 'educational and social life' as the most important contexts. There were also slight differences between caregiver groups by age in the rank order of communication contexts across the high level contexts. These results suggest that SLPs may need to consider not only child's age and developmental level but also caregiver's demands in selecting the appropriate communication contexts when delivering the script-based speech language intervention.

• The Journal of the Korea Contents Association
• /
• v.15 no.2
• /
• pp.281-291
• /
• 2015

This study investigated the relationships between factors of aggression and satisfaction with life and the differences of irrational belief, primary secondary anger thought, and satisfaction with life among four groups classified based on the levels of hostility and verbal aggression. 268 middle school boys in Chungnam areas served as subjects. The results showed that hostility showed the most negative correlation with satisfaction with life. Stepwise regression analyses showed that hostility negatively predicted satisfaction with life, whereas verbal aggression positively predicted satisfaction with life. Factors of irrational belief, primary/secondary anger thought were negatively correlated with satisfaction with life. Analyses of variances revealed that covert hostility group showed higher scores on self-depreciation, disregard and disappointment in intimate relationship situation, hopelessness, whereas overt hostility group showed higher scores on derogation of others/revenge. In addition, covert hostility group showed lower scores on satisfaction with life than other groups. Finally, limitations of this study and suggestions for future study were discussed.

• Annual Conference on Human and Language Technology
• /
• 2022.10a
• /
• pp.284-288
• /
• 2022

최근 다양한 언어모델의 구축 및 발전으로 개체명 인식 시스템의 성능은 최고 수준에 도달했다. 하지만 이와 관련된 대부분의 연구는 데이터가 충분한 언어에 대해서만 다루기 때문에, 양질의 지도학습 데이터의 존재를 가정한다. 대부분의 언어에서는 개체 유형에 대한 언어의 잠재적 특성을 충분히 학습할 수 있는 지도학습 데이터가 부족하기 때문에, 종종 자원 부족의 어려움에 직면한다. 본 논문에서는 Masked language modeling 기반 언어 간 전이학습을 이용한 개체명 인식 방법론에 대한 분석을 수행한다. 이를 위해 전이를 수행하는 소스 언어는 고자원 언어로 가정하며, 전이를 받는 타겟 언어는 저자원 언어로 가정한다. 본 논문에서는 언어모델의 토큰 사전에 언어 독립적인 가상의 자질인 개체 유형에 대한 프롬프트 토큰을 추가하고 이를 소스 언어로 학습한 뒤, 타겟 언어로 전이하는 상황에서 제안하는 방법론에 대한 평가를 수행한다. 실험 결과, 제안하는 방법론은 일반적인 미세조정 방법론보다 높은 성능을 보였으며, 한국어에서 가장 큰 영향을 받은 타겟 언어는 네덜란드어, 한국어로 전이할 때 가장 큰 영향을 준 소스 언어는 중국어인 결과를 보였다.

• Journal of KIISE:Software and Applications
• /
• v.26 no.3
• /
• pp.431-440
• /
• 1999

기존 시각언어 컴파일러 자동화 도구는 시각 구문의 그래픽 정의부분을 명세하고 확장 및 수정하는 방법의 한계로 인해 어휘분석단계를 처리하는데 어려움이 있다. 이러한 문제점을 해결하기 위해 본 논문에서는 객체지향 시각언어 문법을 기반으로 한 새로운 방식의 문법 기반 자동 생성 시스템인 시각언어 컴파일러 자동 생성 환경(CGE-VL)을 구축한다. 이 시스템은 객체지향 패러다임을 기초로 한 시각언어 컴파일러 구축을 위한 실질적인 도구로 기존의 시각언어 컴파일러 자동화 도구와는 구별되는 객체지향 시각언어의 효과적인 명세 방법 제공, 어휘분석기를 이용한 파싱 시간의 단축, 객체지향 파서를 자동 생성하는 등의 새로운 특성을 제공한다. 이러한 특성은 특히 동일한 의미를 가지는 언어 구성요소가 그 쓰임에 따라 서로 다른 구체화된 의미 및 행동을 가지는 객체지향 시각언어에서는 언어 명세에서 최종 산물인 컴파일러에 이르는 전 단계에 동일한 객체 지향 패러다임을 사용함으로써 일관된 개발방법을 제공하여 그 효과가 두드러진다.

• Journal of Digital Convergence
• /
• v.10 no.11
• /
• pp.279-285
• /
• 2012

This study classified the categorization process of convergence products as a rule-based and a similarity-based categorization process. And we examined that how the categorization process was determined according to information types(visual vs. visual + verbal) about the components of two prototypes before convergence and thinking styles(holistic vs. analytic). The result of this study showed: (1) The rule-based categorization process appeared more in case of visual information with verbal information than only visual information. (2) Analytic thinkers chose a rule-based categorization process more than holistic thinkers. These findings provide the theoretical and practical implications to comprehend the categorization process of convergence products and the judgement for consideration set from various convergence products.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.21 no.1
• /
• pp.477-486
• /
• 2020

This study was undertaken to analyze the correlation between depression and memory, by considering the occurrence of ApoE ɛ4 and clinical dementia rating in the elderly with Alzheimer's dementia. This study included 50 participants over 65 years of age, evaluated with CDR 0.5 to 2. We performed CDR, SVLT-E, RCFT, SGDS-K, and ApoE genotyping. Spearman's correlation analysis was used for determining the correlation between depression and memory. The results indicate a significant negative correlation between depression and immediate recall verbal memory in the CDR 1 and 2 without ApoE ɛ4 carrier group (p<0.05). Furthermore, a significant negative correlation was also determined between depression and delayed recall verbal memory in the CDR 1 of the same group. Ed. Notes: The previous sentence already shows this correlation. I suggest this should be deleted from this statement. However, no significant correlation was observed between depression and visual memory. This study found a significant correlation between depression and immediate recall verbal memory. Also, the presence of ApoE ɛ4 indicates a significant correlation between depression and delayed verbal recall memory. Taken together, our results indicate that verbal memory training rather than visual memory training can be more effective in early AD. Also, the treatment of depression will provide a complementary effect.

• Proceedings of the Korean Institute Of Construction Engineering and Management
• /
• autumn
• /
• pp.519-524
• /
• 2001

Recently, our construction market recognizes the necessity of risk management, however the application of practical system is still limited on the construction site because the methodology for analyzing and quantifying construction risk and for building actual risk factors is not easy. This study suggests a risk management method by fuzzy theory, which is using subjective knowledge of an expert and linguistic value, to analyze and Quantify risk. The result of study is expected to improve the accuracy of risk analysis because three factors, such as probability, impact and frequency, for estimating membership function are introduced to quantify each risk factor.

• Journal of Digital Convergence
• /
• v.19 no.4
• /
• pp.49-55
• /
• 2021

This study analyzed the relationship between the characteristics of learners and the achievement level according to the language they learn in programming education for non-major students. The learner's characteristics were set as mastery goal, situaltional interest, cognitive engagement, and self-efficacy for performance, and path analysis was conducted to understand the relationship between the language used and the learning achievement. Situaltional interest and cognitive engagement were found to have an effect on self-efficacy for performace, and self-efficacy for performance had an effect on academic achievement. At this time, the type of programming language to be studied showed a moderating effect on learning achievement. Based on the results of this study, it is suggested that appropriate instruction is needed for each language used in programming lessons for software non-majors.

• The Journal of the Convergence on Culture Technology
• /
• v.7 no.1
• /
• pp.234-239
• /
• 2021

The purpose of this paper is to find out the nature of language variation by exploring the ways of the progress of the language variation that produces all English-lects, i.e., the World Englishes. The study of language variation in linguistics is a hybrid enterprise, so the study of World Englishes has led to the recognition of a highly diverse set of all English-lects, encompassing regional dialects, sociolects, ethnolects and (post-)colonial dialects of World Englishes. In this paper, we propose a hybrid language variation model with three interacting factors of social distancing, on/off-contact, and linguistic diversity to examine the characteristics of language variation. In the context of World Englishes, the social distance is typically low in terms of their local location (country/speech) for local purposes. The social distance also varies based on online/offline communication modes and other social factors like gender, age and ethnic groups, resulting in all English-lects. To clarify the nature of World Englishes, the core Englishes, BrE, AmE and CanE are discussed here.