• Proceedings of the IEEK Conference
• /
• summer
• /
• pp.382-386
• /
• 2004

Recently, with the machine learning trend, most of the machine translation systems on over the world use two syntax tree sets of two relevant languages to learn syntactic tree transfer rules. However, for the English-Vietnamese language pair, this approach is impossible because until now we have not had a Vietnamese syntactic tree set which is correspondent to English one. Building of a very large correspondent Vietnamese syntactic tree set (thousands of trees) requires so much work and take the investment of specialists in linguistics. To take advantage from our available English-Vietnamese Corpus (EVC) which was tagged in word alignment, we choose the SITG (Stochastic Inversion Transduction Grammar) model to construct English- Vietnamese syntactic tree sets automatically. This model is used to parse two languages at the same time and then carry out the syntactic tree transfer. This English-Vietnamese bilingual syntactic tree set is the basic training data to carry out transferring automatically from English syntactic trees to Vietnamese ones by machine learning models. We tested the syntax analysis by comparing over 10,000 sentences in the amount of 500,000 sentences of our English-Vietnamese bilingual corpus and first stage got encouraging result $(analyzed\;about\;80\%)[5].$ We have made use the TBL algorithm (Transformation Based Learning) to carry out automatic transformations from English syntactic trees to Vietnamese ones based on that parallel syntactic tree transfer set[6].

• Korean Journal of Community Nutrition
• /
• v.22 no.4
• /
• pp.323-335
• /
• 2017

Objectives: The purpose of this study is to analyze the current status of nutrition education programs for multicultural families and to provide policy suggestions for improvement. Methods: In-depth interviews of a total of 21 multicultural experts were conducted; 15 people were interviewed individually, while 6 people were interviewed in groups of three. Results: In-depth interviews revealed various problems related to the operation of nutrition education programs. The causes of problems were analyzed and categorized as four factors: systemic, practical, environmental and cultural. As for the systematic factors, insufficient linkage between related organizations and duplicate performance of several projects were identified as concerns Establishment of a control tower and strengthening the linkage among the related organizations may be needed to address this concern. With regard to practical factors, the study identified that language barriers, and lack of nutritional education media and tools translated into multicultural languages were limiting factors. These limitations the development of nutrition education materials that aretranslated into multiple languages, implementation of education programs that are different from the Korean education, and by providing interpreters. As for the environmental factors, low educational level and poor nutritional knowledge of multicultural women made it difficult for them to understand the contents of the education. Demonstration, practical training and urgent education on pregnancy and childbirth nutrition were identified as needs to address these concerns. Withregard to cultural factors, food culture conflict with Korean families, and difficulties in home practices were detected as concerns. Participants in the study suggested that getting education with family and facilitation of weekend and nighttime programs health of this community. Conclusions: Further studies are needed to adopt more effective and efficient nutrition intervention to promote the healthy eating of the married immigrant women based on the study results.

• Journal of Information Technology and Architecture
• /
• v.9 no.4
• /
• pp.413-422
• /
• 2012

Many researchers have been studied to reveal human natural language to let machine understand its meaning by text based, page rank based or more. Particularly, it has been considered that URL and HTML Tag information in web documents are attracting people' attention again to analyze huge amount of web document automatically. In this paper, we propose a STW (Semantic Term Weight) approach based on syntactic and linguistic structure of web documents in order to classify what genres are. For the evaluation, we analyzed more than 1,000 documents from 20-Genre-collection corpus for training the documents based on SVM algorithm. Afterwards, we tested KI-04 corpus to evaluate performance of our proposed method. This paper measured their accuracy by classifying them into an experiment using STW and one without u sing STW. As the results, the proposed STW based approach showed approximately 10.2% which Is higher than one without use of STW.

• Cross-Cultural Studies
• /
• v.27
• /
• pp.497-523
• /
• 2012

Characteristics of the Listening and Pronunciation of Korean Obstruents of Chinese Learners -Based on the Phonetic Experiments Using Kalvin and Praat- This study aims at investigating the characteristics of confrontation in three ways, lax/ fortis/ aspirated consonants, in Korean obstruents through experimental phonetic analysis for the Chinese Korean language learners. On one hand, as a result of comparing Korean and Chinese obstruent systems, there is no big difference regarding the articulatory location. On the other hand, in regards to the articulatory method there is a difference. In a Korean obstruent system, the confrontation presented in three ways by the strength of aspiration. On the contrary, the Chinese obstruent system showed confrontation in two ways by the existence of aspiration. To examine the difficulty of the learners caused by the above-mentioned reason objectively, this paper studied the relationship between input and output of sound through the experimental phonetic analysis such as Kalvin and Praat. To research the input of sound, the listening ability of the learners was examined by 'Choosing Consonant' among the Menu of Kalvin. As a result of that experiment, many errors were shown. They recognized the fortis as lax in the area of affricates and plosives. In the area of fricatives, they recognized affricatives as fricatives. To investigate the output of sound, the section of aspiration and the section of friction of a plosive, an affricate and a fricative in Praat, were expressed numerically. The learners' VOT of lax and affricate represented that lax was pronounced close to the fortis, and the VOT of fricatives was not shown the section of aspiration and friction clearly, and also the result showed that they pronounced a fricative like affricative-aspirated one. The result shows that the learners' pronunciation is related to the listening ability. The consequence is caused by the characteristics of the difference between Korean obstruents and Chinese ones. If the training pronunciation is conducted based on above result, it would be a better methodology in teaching Korean.

• Hwankyungkyoyuk
• /
• v.15 no.1
• /
• pp.115-124
• /
• 2002

The purpose of this research is to analyze the contents related to environmental education(EE) in the elementary school textbooks for the following areas: well-balanced EE, and development of EE curriculum/teachers' guide in elementary school level. For the purpose of this analysis, elementary school teachers, education administrators and EE specialists were involved. Eleven content areas of EE, namely: natural environment, artificial environment, population, industrialization, natural resources, pollution, environmental conservation, environment sanitation, environment ethics, environmentally sound and sustainable development(ESSD), and daily li(e as a consumer, were analyzed. The results of the analysis are as follows: 1. There are total 1,140 contents related to EE in the elementary school textbooks of 7th Korea National Curriculum. 2. The textbooks of grade 6 contain the most number of EE contents, while the least number is in the textbooks of grade 3. 3. The subject that includes EE contents equally in its textbooks is social studies, and the subjects that relate a lot to EE are Korean language, science, and social studies respectively. 4. The content areas that are included a lot in textbooks are' natural environment', 'pollution', and' environmental conservation' respectively, while the contents of 'population','industrialization', 'ESSD' are included to a lesser degree. The content area most frequently mentioned in the textbooks is 'pollution', and the number of the contents are increasing along with the grade level. 5. Generally, the content areas of 'population', 'industrialization', and 'natural resources' are reflected in the textbooks to a lesser degree than others. 'Industrialization' is not included in the textbooks of grade 2, while 'population' is not included in ones of grade 4. According to the result, more concern about balanced EE in content areas is needed at the elementary school level. Similar studies tot K and secondary school levels are needed. The developmental study of EE guide book and teacher training for teaching EE using the book are also recommended.

• Journal of the Korea Convergence Society
• /
• v.12 no.5
• /
• pp.1-7
• /
• 2021

In the latest trend of machine translation research, the model is pretrained through a large mono lingual corpus and then finetuned with a parallel corpus. Although many studies tend to increase the amount of data used in the pretraining stage, it is hard to say that the amount of data must be increased to improve machine translation performance. In this study, through an experiment based on the mBART model using parallel corpus filtering, we propose that high quality data can yield better machine translation performance, even utilizing smaller amount of data. We propose that it is important to consider the quality of data rather than the amount of data, and it can be used as a guideline for building a training corpus.

• The Korean Journal of Archival Studies
• /
• no.58
• /
• pp.95-145
• /
• 2018

As a communication language, non-text records and archives such as photographs, images, and videos are becoming more important than text records. As a result, domestic institution, organizations, and specialists related to record management have been emphasizing the necessity of the archive management system appropriate for the distinctive characteristics of image records. This paper summarizes the points to be considered for the establishment of the Korean audiovisual archives management system, based on the writer's experience of participating in the International Audiovisual Archives Management Training for professionals in the world(INA Frame) of 2018. In particular, various types of contents including cinema, broadcasting, cultural should be managed at the national level. Furthermore, the necessity of a new concept establishment for "public audiovisual heritage" is accentuated. In addition, the tasks regarding the establishment of foundation, such as the modification of the related systems and infrastructures, and the layouts of the institution or governance, should be reviewed and revised. Moreover, the preliminary tasks revised, should be lead to the establishment of stable management system for public audiovisual archives of Korea.

• Journal of Industrial Convergence
• /
• v.17 no.2
• /
• pp.45-52
• /
• 2019

A systematic review was conducted to identify components and convergent effects of simulation program using schizophrenia scenario in nursing education. Using 4 different databases, 226 articles were identified and 11 studies met the inclusion criteria. There were 5 qualitative studies, 5 quantitative studies and 1 study used mixed method design. The simulation incorporated various methods including standardized patients, role playing, simulator and virtual reality that majority studies(63.6%) used standardized patients. For the evaluation, studies examined diverse variables including knowledge, learning self competency, learning satisfaction and self directed learning. Considering complexity and difficulty of nursing for schizophrenia, future studies with well designed simulation program are required to prove its effectiveness.

• Journal of the Korea Convergence Society
• /
• v.10 no.5
• /
• pp.119-124
• /
• 2019

The purpose of this study is to introduce the NHK"NEWS WEB EASY"online site to intermediate level learners of Japanese language, and to teach effective methods of Blended Learning through the lesson planning method using "NEWS WEB EASY" Suggesting. First, this paper helped students cultivate various vocabulary learning ability through Blended Learning using "NEWS WEB EAS Y", Second, helped them learn about Japanese culture and Japan through various articles, Third, helped them naturally perform listening training through listening files, Fourth, helped them practice reading Kanji and improve vocabulary skills by distributing the files without Furigana to them for search, and Fifth, showed them how to improve speaking ability by reading practice through learning using "NEWS WEB EASY". We could learn the fact that the study helped students a lot understand Japan and improve their Japanese ability by learning news articles that they could not come across due to prejudice through learning using "NEWS WEB EASY".

• Phonetics and Speech Sciences
• /
• v.13 no.3
• /
• pp.71-78
• /
• 2021

Most neural-network-based speech synthesis models utilize neural vocoders to convert mel-scaled spectrograms into high-quality, human-like voices. However, neural vocoders combined with mel-scaled spectrogram prediction models demand considerable computer memory and time during the training phase and are subject to slow inference speeds in an environment where GPU is not used. This problem does not arise in linear spectrogram prediction models, as they do not use neural vocoders, but these models suffer from low voice quality. As a solution, this paper proposes a Tacotron 2 and Transformer-based linear spectrogram prediction model that produces high-quality speech and does not use neural vocoders. Experiments suggest that this model can serve as the foundation of a high-quality text-to-speech model with fast inference speed.