• Journal of Information Processing Systems
• /
• 제18권2호
• /
• pp.282-291
• /
• 2022

This research applies a pre-trained bidirectional encoder representations from transformers (BERT) handwriting recognition model to predict foreign Korean-language learners' writing scores. A corpus of 586 answers to midterm and final exams written by foreign learners at the Intermediate 1 level was acquired and used for pre-training, resulting in consistent performance, even with small datasets. The test data were pre-processed and fine-tuned, and the results were calculated in the form of a score prediction. The difference between the prediction and actual score was then calculated. An accuracy of 95.8% was demonstrated, indicating that the prediction results were strong overall; hence, the tool is suitable for the automatic scoring of Korean written test answers, including grammatical errors, written by foreigners. These results are particularly meaningful in that the data included written language text produced by foreign learners, not native speakers.

• 한국언어정보학회:학술대회논문집
• /
• 한국언어정보학회 2007년도 정기학술대회
• /
• pp.489-497
• /
• 2007

Named entity translation plays an important role in many applications, such as information retrieval and machine translation. In this paper, we focus on translating person names, the most common type of name entity in Korean-Chinese cross language information retrieval (KCIR). Unlike other languages, Chinese uses characters (ideographs), which makes person name translation difficult because one syllable may map to several Chinese characters. We propose an effective hybrid person name translation method to improve the performance of KCIR. First, we use Wikipedia as a translation tool based on the inter-language links between the Korean edition and the Chinese or English editions. Second, we adopt the Naver people search engine to find the query name's Chinese or English translation. Third, we extract Korean-English transliteration pairs from Google snippets, and then search for the English-Chinese transliteration in the database of Taiwan's Central News Agency or in Google. The performance of KCIR using our method is over five times better than that of a dictionary-based system. The mean average precision is 0.3490 and the average recall is 0.7534. The method can deal with Chinese, Japanese, Korean, as well as non-CJK person name translation from Korean to Chinese. Hence, it substantially improves the performance of KCIR.

• 한국융합학회논문지
• /
• 제13권1호
• /
• pp.13-21
• /
• 2022

최근 딥러닝 기반의 자연어처리 연구는 사전 훈련된 언어 모델을 통해 대부분의 자연어처리 분야에서 우수한 성능을 보인다. 특히 오토인코더 (auto-encoder) 기반의 언어 모델은 다양한 한국어 이해 분야에서 뛰어난 성능과 쓰임을 증명하고 있다. 그러나 여전히 디코더 (decoder) 기반의 한국어 생성 모델은 간단한 문장 생성 과제에도 어려움을 겪고 있으며, 생성 모델이 가장 일반적으로 쓰이는 대화 분야에서의 세부 연구와 학습 가능한 데이터가 부족한 상황이다. 따라서 본 논문은 한국어 생성 모델을 위한 멀티턴 대화 데이터를 구축하고 전이 학습을 통해 생성 모델의 대화 능력을 개선하여 성능을 비교 분석한다. 또한, 검색 모델을 통해 외부 지식 정보에서 추천 응답 후보군을 추출하여 모델의 부족한 대화 생성 능력을 보완하는 방법을 제안한다.

• Journal of Audiology & Otology
• /
• 제24권2호
• /
• pp.61-70
• /
• 2020

Background and Objectives: To assess the academic performance, communication skills, and psychosocial development of prelingual deaf children with cochlear implants (CIs) attending mainstream schools, and to evaluate the impact of auditory speech perception on their classroom performance. Subjects and Methods: As participant, 67 children with CI attending mainstream schools were included. A survey was conducted using a structured questionnaire on academic performance in the native language, second language, mathematics, social studies, science, art, communication skills, self-esteem, and social relations. Additionally, auditory and speech performances on the last follow-up were reviewed retrospectively. Results: Most implanted children attending mainstream school appeared to have positive self-esteem and confidence, and had little difficulty in conversing in a quiet classroom. Also, half of the implanted children (38/67) scored above average in general academic achievement. However, academic achievement in the second language (English), social studies, and science were usually poorer than general academic achievement. Furthermore, half of the implanted children had difficulty in understanding the class content (30/67) or conversing with peers in a noisy classroom (32/67). These difficulties were significantly associated with poor speech perception. Conclusions: Improving the listening environment for implanted children attending mainstream schools is necessary.

• 대한청각학회지
• /
• 제24권2호
• /
• pp.61-70
• /
• 2020

Background and Objectives: To assess the academic performance, communication skills, and psychosocial development of prelingual deaf children with cochlear implants (CIs) attending mainstream schools, and to evaluate the impact of auditory speech perception on their classroom performance. Subjects and Methods: As participant, 67 children with CI attending mainstream schools were included. A survey was conducted using a structured questionnaire on academic performance in the native language, second language, mathematics, social studies, science, art, communication skills, self-esteem, and social relations. Additionally, auditory and speech performances on the last follow-up were reviewed retrospectively. Results: Most implanted children attending mainstream school appeared to have positive self-esteem and confidence, and had little difficulty in conversing in a quiet classroom. Also, half of the implanted children (38/67) scored above average in general academic achievement. However, academic achievement in the second language (English), social studies, and science were usually poorer than general academic achievement. Furthermore, half of the implanted children had difficulty in understanding the class content (30/67) or conversing with peers in a noisy classroom (32/67). These difficulties were significantly associated with poor speech perception. Conclusions: Improving the listening environment for implanted children attending mainstream schools is necessary.

• 말소리와 음성과학
• /
• 제12권1호
• /
• pp.51-63
• /
• 2020

음운인식은 음운론 영역의 상위언어인식 능력으로, 읽기 및 어휘력 등의 언어능력을 예측하는 것으로 알려져 있다. 본 연구는 음운인식 능력과 기타 음운처리 능력, 그리고 언어능력 간의 관계를 일반아동 집단과 언어발달지연 아동 집단 간 비교를 통해 살펴보고자 하였다. 4~6세의 언어발달지연 아동(n=15)과 일반아동(n=18)을 대상으로 음운인식 능력을 평가하기 위하여 음절수세기, 음절탈락, 음절변별 과제를 실시하였다. 또한 비단어 따라말하기, 숫자 거꾸로 회상하기의 두 가지 음운처리 과제와 수용 및 표현어휘력, 문법성판단 과제와의 상관관계를 분석하고, 언어능력을 예측하는 음운인식 하위과제가 무엇인지 검토하였다. 음운인식 하위과제 중 음절수세기를 제외한 음절탈락, 음절변별 과제 수행력의 집단 간 차이가 유의하였다. 또한 일반아동 집단은 음절탈락과 숫자 거꾸로 회상하기, 음절변별과 수용어휘력 과제 수행력 간 상관관계가 유의하였으며, 언어발달지연 아동 집단은 음절수세기 과제와 숫자 거꾸로 회상하기, 수용어휘력, 표현어휘력, 문법성판단 과제의 수행력 간 상관관계가 유의하였다. 그리고 단계적 중다회귀분석 결과 일반아동 집단은 음절변별 과제가 수용어휘력 및 문법성판단 과제 수행력을 유의하게 예측하는 것으로 나타났으며, 언어발달지연 아동 집단은 음절수세기 과제가 수용어휘력, 표현어휘력, 문법성판단 과제 수행력을 유의하게 예측하는 것으로 나타났다. 언어발달지연 아동 집단은 일반아동 집단에 비해 음절수세기를 제외한 나머지 음절 수준의 음운인식 과제의 수행력이 저조하였으며, 이러한 특징이 상관관계 분석 및 회귀분석 결과에도 반영되었다. 또한 각 집단에서 음운인식 과제 수행력이 언어능력을 유의하게 예측하는 것으로 나타난 결과는 음운론 영역의 상위언어인식 능력의 중요성을 시사한다.

• 융합신호처리학회논문지
• /
• 제22권4호
• /
• pp.206-212
• /
• 2021

사회가 점점 발전하면서 의사소통 방법이 다양한 형태로 발전하고 있다. 그러나 발전한 의사소통은 비장애인을 위한 방법이며, 청각장애인에게는 아무런 영향을 미치지 않는다. 따라서 본 논문에서는 청각장애인의 의사소통을 돕기 위한 CNN 기반의 수화번역 프로그램을 설계 및 구현한다. 수화번역 프로그램은 웹캠을 통해 입력된 수화 영상 데이터를 기반으로 의미에 맞게 번역한다. 수화번역 프로그램은 직접 제작한 24,000개의 한글 자모음 데이터를 사용하였으며, 효과적인 분류모델의 학습을 위해 U-Net을 통한 Segmentation을 진행한다. 전처리가 적용된 데이터는 19,200개의 Training Data와 4,800개의 Test Data를 통하여 AlexNet을 기반으로 학습을 진행한다. 구현한 수화번역 프로그램은 'ㅋ'이 97%의 정확도와 99%의 F1-Score로 모든 수화데이터 중에서 가장 우수한 성능을 나타내었으며, 모음 데이터에서는 'ㅣ'가 94%의 정확도와 95.5%의 F1-Score로 모음 데이터 중에서 가장 높은 성능을 보였다.

• International Journal of Contents
• /
• 제15권2호
• /
• pp.44-52
• /
• 2019

This paper uses Delphi Method and statistics provided by the State Administration of Market Regulation of China, attempts to develop the scale index of language service industry in China. Coupled with practical investigation and theoretical framework of SCP paradigm, a deep analysis on market behavior, market structure, and market performance of the industry have been explored. First, the study indicates that the scale index of Chinese language services industry has experienced an upward trend from 10.481million RMB in 2008 to 351.403 million RMB in 2017. Second, majority of language services enterprises are situated in Chinese coastal provinces and there are variations in demand for language services. Third, the standardization of language services are minimal while most of the talents involve a single discipline background. Fourth, most enterprises utilize language tools while there is lack of technological innovation for utilization of language resources and enhancement of service quality. The author mainly suggests four strategies which include: cultivating multilingual services, strengthening the development of industrial informatization and technical innovation, and optimizing the industrial talents structure while also adjusting for industrial distribution and regional coordination, which propose the meaningful implications for the development of language service industry in China.

• ETRI Journal
• /
• 제43권2호
• /
• pp.299-312
• /
• 2021

It is necessary to achieve high performance in the task of zero anaphora resolution (ZAR) for completely understanding the texts in Korean, Japanese, Chinese, and various other languages. Deep-learning-based models are being employed for building ZAR systems, owing to the success of deep learning in the recent years. However, the objective of building a high-quality ZAR system is far from being achieved even using these models. To enhance the current ZAR techniques, we fine-tuned a pretrained bidirectional encoder representations from transformers (BERT). Notably, BERT is a general language representation model that enables systems to utilize deep bidirectional contextual information in a natural language text. It extensively exploits the attention mechanism based upon the sequence-transduction model Transformer. In our model, classification is simultaneously performed for all the words in the input word sequence to decide whether each word can be an antecedent. We seek end-to-end learning by disallowing any use of hand-crafted or dependency-parsing features. Experimental results show that compared with other models, our approach can significantly improve the performance of ZAR.

• 스마트미디어저널
• /
• 제13권6호
• /
• pp.35-43
• /
• 2024

During the post-pandemic period, the interest in foreign language learning surged, leading to increased usage of language-learning apps. With the rising demand for these apps, analyzing app reviews becomes essential, as they provide valuable insights into user experiences and suggestions for improvement. This research focuses on extracting insights into users' opinions, sentiments, and overall satisfaction from reviews of HelloTalk, one of the most renowned language-learning apps. We employed topic modeling and emotion analysis approaches to analyze reviews collected from the Google Play Store. Several experiments were conducted to evaluate the performance of sentiment classification models with different settings. In addition, we identified dominant emotions and topics within the app reviews using feature importance analysis. The experimental results show that the Random Forest model with topics and emotions outperforms other approaches in accuracy, recall, and F1 score. The findings reveal that topics emphasizing language learning and community interactions, as well as the use of language learning tools and the learning experience, are prominent. Moreover, the emotions of 'admiration' and 'annoyance' emerge as significant factors across all models. This research highlights that incorporating emotion scores into the model and utilizing a broader range of emotion labels enhances model performance.