• Journal of Korea Society of Industrial Information Systems
• /
• v.9 no.3
• /
• pp.64-70
• /
• 2004

Korean and Japanese have same structure of sentences because they belong to same family of languages. So, The transfer driven machine translation is most efficient to translate each other. This paper introduce a method which creates a transfer dictionary for Token Based Transfer Driven Koran-Japanese Machine Translation(TB-TDMT). If the transfer dictionaries are created well, we get rid of useless effort for traditional parsing by performing shallow parsing. The semi-parser makes the dependency tree which has minimum information needed output generating module. We constructed the transfer dictionaries by using the corpus obtained from ETRI spoken language database. Our system was tested with 900 utterances which are collected from travel planning domain. The success-ratio of our system is $92\%$ on restricted testing environment and $81\%$ on unrestricted testing environment.

• The Journal of the Korea Contents Association
• /
• v.17 no.4
• /
• pp.331-341
• /
• 2017

This paper aims to analyze English vocabulary of the North Korean textbooks diachronically using the constructed English textbook corpus. The North Korea English textbooks attained from Information Center on North Korea of the Ministry of Unification are divided into before and after Kim Jong-Il era for the year of 1996 in which the curriculum revision has been conducted. They are stored as text files to analyse vocabularies using WordSmith Tools 7.0. The vocabulary size of the revised textbooks increased after the curriculum reorganization, but the number of vocabulary types and vocabulary diversity decreased. After the curriculum revision, it was found that lots of vocabulary related to the establishment of the Kim Jong-Il system appeared as the keyword. It was also found that some vocabularies reflected the economic and social life of North Korea. In addition, through comparison of the 100 high-frequency word list and keywords, it can be concluded that the vocabulary of the English textbooks of North Korea is gradually changing into communicative contents from contents related with written language.

• KIPS Transactions on Software and Data Engineering
• /
• v.2 no.12
• /
• pp.881-888
• /
• 2013

A pre-analyzed dictionary is used to increase the speed and the accuracy of morphological analyzers and to decrease the over-generation. However, if the dictionary includes 'Insufficiently-analyzed word-phrases', which do not include all the possible analysis of the word-phrase, it may cause the decrease of the analysis accuracy. In this paper, we measure the accuracy changes according to the number of word-phrase frequency and the size changes of corpus by Sejong corpus. And performance of integrate system(SMA with pre-dictionary) is highest when sufficient analysis rate of pre-dictionary is more than 99.82%. Also pre-dictionary is constructed with word-phrase that frequency more than 32(64) when size of corpus is 1,600,000(6,300,000) word-phrase.

• Annual Conference on Human and Language Technology
• /
• 2023.10a
• /
• pp.432-436
• /
• 2023

초거대언어모델(LLM)의 도래에 따라 다양한 과업들이 도메인 관계 없이 제로샷으로 추론이 가능해짐에 따라서 LLM이 다양한 산업분야에 적용되고 있다. 대표적으로 ChatGPT와 GPT-4는 상용 API로 서비스를 제공하여 용이한 서비스 접근으로 다양한 이용층을 끌어들이고 있다. 그러나 현재 상용 API로 제공되고 있는 ChatGPT 및 GPT-4는 사용자의 대화 내역 데이터를 수집해 기업의 보안 문제를 야기할 수 있고 또한 생성된 결과물의 환각 문제로 인한 기업 문서의 신뢰성 저하를 초래할 수 있다. 특히 LLM 생성 글은 인간의 글과 유사한 수준으로 유창성을 확보한만큼 산업현장에서 LLM 작성 글이 판별되지 못할 경우 기업 활동에 큰 제약을 줄 수 있다. 그러나 현재 한국어 LLM 작성 글 탐지 서비스가 전무한 실정이다. 본 논문에서는 한국어 초거대언어모델 작성 글 판별기: KoCheckGPT 를 제안한다.KoCheckGPT는 산업현장에서 자주 사용되는 문어체, 개조식 글쓰기로 작성된 문서 도메인을 목표로 하여 글 전체와 문장 단위의 판별 정보를 결합하여 주어진 문서의 LLM 작성 여부를 효과적으로 판별한다. 다국어 LLM 작성 글 판별기 ZeroGPT와의 비교 실험 결과 KoCheckGPT는 우수한 한국어 LLM 작성 글 탐지 성능을 보였다.

• Journal of the Korea Convergence Society
• /
• v.12 no.5
• /
• pp.23-29
• /
• 2021

Recent deep learning-based natural language processing studies are conducting research to improve performance by training large amounts of data from various sources together. However, there is a possibility that the methodology of learning by combining data from various sources into one may prevent performance improvement. In the case of machine translation, data deviation occurs due to differences in translation(liberal, literal), style(colloquial, written, formal, etc.), domains, etc. Combining these corpora into one for learning can adversely affect performance. In this paper, we propose a new Corpus Weight Balance(CWB) method that considers the balance between parallel corpora in machine translation. As a result of the experiment, the model trained with balanced corpus showed better performance than the existing model. In addition, we propose an additional corpus construction process that enables coexistence with the human translation market, which can build high-quality parallel corpus even with a monolingual corpus.

• Journal of KIISE:Software and Applications
• /
• v.37 no.6
• /
• pp.455-463
• /
• 2010

In this paper, we present an approach to sentence boundary detection for web documents that builds on statistical-based methods and uses rule-based correction. The proposed system uses the classification model learned offline using a training set of human-labeled web documents. The web documents have many word-spacing errors and frequently no punctuation mark that indicates the end of sentence boundary. As sentence boundary candidates, the proposed method considers every Ending Eomis as well as punctuation marks. We optimize engine performance by selecting the best feature, the best training data, and the best classification algorithm. For evaluation, we made two test sets; Set1 consisting of articles and blog documents and Set2 of web community documents. We use F-measure to compare results on a large variety of tasks, Detecting only periods as sentence boundary, our basis engine showed 96.5% in Set1 and 56.7% in Set2. We improved our basis engine by adapting features and the boundary search algorithm. For the final evaluation, we compared our adaptation engine with our basis engine in Set2. As a result, the adaptation engine obtained improvements over the basis engine by 39.6%. We proved the effectiveness of the proposed method in sentence boundary detection.

• The Mathematical Education
• /
• v.61 no.4
• /
• pp.559-579
• /
• 2022

Research on lexicon and language provides insights into the interests, values and practices of a community where individuals use the language. The International Classroom Lexicon Project, in which ten countries participated, identified own country's mathematics teaching and learning lexicons by investigating mathematics classroom instruction from teachers' perspectives in a speaking-oriented community. This study, as an extension of the International Classroom Lexicon Project research, investigated pedagogical lexicons used in 「Mathematics and Education」 journals specialized for Korean professional mathematics teachers published by the Korean Society of Teachers of Mathematics. Using the text mining approach, we also traced how these pedegogical lexicons have changed quantitatively over the past 10 years with a diachronic perspective. As a results, several novel terms were found in the writing-oriented community, which were not identified in the speaking-oriented community. In addition, we could discover some pedagogical lexicons have increased statistically significantly and some lexicons appeared(increased) rapidly across years. This implies the teacher community's values and zeitgeist by reflecting these changes in the sociocultural, incidental and social changing (i.e., periodical change) contexts. This study has value as a first step in understanding zeitgeist for mathematics education in Korean mathematics teacher community according to changes of times over the past 10 years. Also, this study contributes to the methodological insights: the text mining technique provides a methodological contribution to researching changes in interests, values and zeitgeist according to these changes in the times.

• Journal of the East Asian Society of Dietary Life
• /
• v.6 no.3
• /
• pp.365-379
• /
• 1996

본 연구에서는 우리나라 영남 지방에 소재하는 네 가정을 표집하여 각 가정에서 행 하여지고 있는 유교적 차례와 제사에 차려지는 제물을 실증적으로 연구, 고찰하였으며, 또 이들을 문헌적으로 비교 검토하여 그 변천의 모습을 고찰하고저 하였다. 영남지방의 북부인 안동 하회마을과 영남의 중부지방이라고 볼 수 있는 양동 마을의 불천위제사에서 보이는 제 물조리에서의 공통점은 炙을 완전히 익히지 않고 진설하여 性으로서의 기원을 그대로 지키 고 있다는 사실이며 차이점은 바다가 가까운 양동마을에서는 하회마을보다 어물의 사용이 (대게와 생문어 등) 월등히 많다는 사실과 산골에 위치한 하회마을의 경우는 그 대신 닭을 생으로 한 마리 왼으로 쓰고 엽적에 또 한 마리를 사용하여 어물이 적은 데 비하여 닭을 더 많이 사용하고 있었다. 이러한 것은 의례가 매우 엄격한 제례에서도 지역적 특성이 많이 작 용하고 있다는 것을 나타낸다고 하겠다. 또한 진설에 있어서도 한 곳은 양위진설이고 다른 곳은 단설로 같은 불천위이지만 이러한 차이를 나타낸다고 하겠다. 또한 진설에 잇어서도 한 곳은 양위진설이고 다른 곳은 단설로 같은 불천위이지만 이러한 차이를 나타내고 있어, 이것에 있어서도 한 곳은 양위진설이고 다른 곳은 단설로 같은 물천위이지만 이러한 차이를 나타내고 있어 이것은 진설배열과 함께 가가체로 지켜져 내려 오는 것이라 하겠다. 양가 모 두 기제사 때보다 차려지는 제물의 종류가 다양하고 양도 많았다. 떡, 煎, 자반, 炙 및 나물 과 과일 등이 차려졌다. 여전히 제물을 조리할 때 사용되는 양념은 매우 제한되고 있으며, 또한 이용할 수 있는 채소의 종류는 가정마다 다양하게 다르게 제한되고 있다는 것이 특징 적이다. 사용되는 식품의 종류는 전통적으로 답습해오고 있음에 비추어, 과일은 매우 최근에 수입된 외래종의 것도 모든 가정에서 이용되고 있었다. 그러나 전체적으로 볼 때 유교식 제 사의례는 시대에 따라 거의 변화됨이 없이 지켜져 내려 오고 있었다.

• Korean Journal of Cognitive Science
• /
• v.2 no.2
• /
• pp.405-428
• /
• 1990

Language has a social function to communicate information. Linguists have gradually paid their attention to the function of language since the nineteen sixties, especially to the relationship of form, meaning and the function. The relationship could be more clearly grasped through disciyrse-based analysis than through sentence-based analysis. Many researches were centered on the discourse functional notion of topic. In the early 1970's the subject was defined as the grammatiocalized topic the topic as a discrete single constituent of the clause. In the late 1970's several lingusts including Givon suggerted that the topic was not an atomic, disctete entity, and that the clause could have more than one topic. The purpose of the present study is, following Givon, to study grammatical coding devices of topic and to measure the relative topic continuity/discontinuity of participant argu, ents in Korean narratives. By so doing, I would like to shed some light on effective ways of communicating information. The grammatical coding devices analyzed are the following eight structures: zero-anaphora, personal pronous, demonstrative pronouns, names, noun phrases following demonstratives, noun phrases following possessives, definite noun phrases and indefinite referentials. The narrative studied for the count was taken from the KoreanCIA chief's Testiomny:Revolution and Idol by Hyung Wook Kim. It was chosen because it was assumed that Kim's purpose in the novel was to tell a true story, which would not distort the natural use of language for literary effect. The measures taken in the analysis wre those of 'lookback', 'persistence', ambiguity'. The first of these, 'lookback', is a measure of the size of gap between the previous occurrence of a referent and its current occurence in the clause. The meausure of persistence, which is a measure of the speaker's topocal intent, reflects the topic's importance in the discourse. The third measure is a measure of ambiguity. This is necessary for assessing the disruptive effects that other topics within five previous clauses may have on topic identification. The more other topics are present within five previous clauses, the more difficult is the task of correct identification of a topic. The results of the present study show that the humanness of entities is the most powerful factior in topic continutiy in narrative discourse. The semantic roles of human arguments in narrative discourse tend to be agents or experiences. Since agents and experiences have high topicality in discourse, human entities clearly become clausal or discoursal topics. The results also show that the grammatical devices signal varying degrees of topic continuity discontinuity in continuous discourse. The more continuous a topic argument is, the less it is coded. For example, personal pronouns have the most continutiy and indefinite referentials have the least continutiy. The study strongly shows that topic continuity discontinutiy is controlled not only by grammatical devices available in the language but by socio-cultural factors and writer's intentions.

• The Korean Journal of Nuclear Medicine Technology
• /
• v.23 no.1
• /
• pp.29-34
• /
• 2019

Purpose An examination guide is a useful medium to provide the patient with an overview, pre- and post-test preparation, and precautions of nuclear testing. The design and arrangement of existing written texts and announcements were evaluated to elucidate the comprehension of patients undergoing testing. Materials and Methods Informational material describing bone scanning and $^{201}thallium$ myocardium perfusion single-photon emission computed tomography (SPECT), as a secondary examination, which accounts for the largest portion of gamma imaging at Asan Hospital (Seoul, South Korea), was selected as an improvement target in consultation with a national innovation center. Existing informational material was dispensed to patients scheduled for bone scans from November 2016 to February 2017 and the revised material was issued from March 2017 to May 2017. A survey was conducted of 200 patients who underwent $^{201}thallium$ myocardium perfusion SPECT before and after the revisions (n = 100 each time period) to assess the patients' understanding of the informational material. Results When comparing the use of the conventional vs. revised material, the number of patients who received treatment before bone scanning had decreased from 130 to 60, while the number of those who required additional imaging decreased from 53 to 14. Prior to the revision, 43% of patients underwent testing before preparation and 18% underwent additional testing. The decreased need for additional image acquisition after revision of the informational material resulted in a decrease in acquisition time of about 2 min, from 16.5 to 14.2 min. In the case of $^{201}thallium$ myocardium perfusion SPECT, patient comprehension of all five items surveyed had increased, while the number of patients who had repeatedly asked about various facets of the procedure pre- and post-testing had decreased from 36% to 16% and 31% to 14%, respectively. Conclusion Lower patient comprehension is accompanied by a decrease in image quality due to non-compliance during pre-testing and may lead to repetitive questions from the patient, which may also negatively affect the fatigue and work efficiency of the examiner. Improved readability and visibility of informational material through visualization was correlated with greater patient comprehension as well as improved image quality and acquisition time.