• Annual Conference of KIPS
• /
• 2008.11a
• /
• pp.161-163
• /
• 2008

한국인식기술에서는 명함정보를 원격에서 관리하는 모바일 하이네임(Hi-Name)을 출시하였다. 하지만 모바일 기기에서는 명함 정보를 이용하기 위해 작은 화면에 검색어를 펜으로 입력해야 하는 불편함과 공간상의 제약성으로 음성인식엔진의 탑재가 어려우므로 분산인식환경이 필요하다. 따라서 본 연구에서는 모바일 환경에서 음성인식기술을 이용하여 명함정보를 이용할 수 있는 하이네임의 웹 서비스 환경을 구축하였다.

• Annual Conference of KIPS
• /
• 2023.11a
• /
• pp.575-576
• /
• 2023

한국 수어는 소리로 말을 배울 수 없어서 사용하는 '보이는 언어'이고 한국수화언어를 줄인 말이다. 한국어나 영어와 같이 독립된 언어로 한국어와는 문법 체계가 다른 대한민국 농인의 고유한 언어이다. 하지만, 한국 사회에서는 수어를 일상어로 사용하는 농인이 수어만으로 다른 사람과 대화하거나 서비스 등을 이용하기에는 쉽지 않은 구조이다. 이에 본 논문에서는 택시라는 상황을 가정해 택시 안에서 학습된 모델이 농인의 수어를 인식하고 택시 기사에게 해당 의미를 전달하는 시스템을 제안한다. 제안 시스템을 통해 택시 기사는 농인(수어사용자)에게 응답할 수 있다. 본 논문에서는 한국수어 번역기 웹서비스를 설계 및 구현하여 실제 환경에서의 활용 가능성을 검증한다.

• Korean Journal of Environment and Ecology
• /
• v.22 no.6
• /
• pp.616-624
• /
• 2008

The ichthyofauna and ecological community analysis in the Dong River, Korea, were investigated from May, 2006 to May, 2007. During the surveyed period, 31 species belonging 9 families were collected. Total of 31 species(primary freshwater: 26 species, peripheral freshwater: 5 species) were found with the primary freshwater fishes being highest. Dominant species was Zacco koreanus(55.73%), and subdominant species was Coreoleuciscus splendidus(7.15%). Also, Microphysogobio longidorsalis (6.12%), Hemibarbus mylodon(4.87%), Iksookimia koreensis(4.49%) were numerous. There were 19 Korean endemic species(61.29%), including Acheilognathus signifer, A. yamatsutae, Pseudopungtungia tenuicorpus, C. splendidus, Squalidus gracilis majimae, H. mylodon. Among them, 1 species was identified as qualifying for possible designation as a natural monument catergory, and 5 species were identified for designation required to protect under a wildlife category. As result of expected number of species according to rarefaction curve, the expected species of the Dong-river ranged from 19(51. 4, 5) to 22(51. 3) and result of similarity analysis ranged from 62.28%(51. 1 and 2) to 84.13%(51. 1 and 5). In conclusions, we could define that the Dong-river has a very good water conditions as habitat of fish, and habitat conditions also are very similar in every survey sites.

• Journal of fish pathology
• /
• v.4 no.2
• /
• pp.79-85
• /
• 1991

In Feb., 1990 an epizootic disease to masu salmon fry, Oncorhynchus masou cultured at the hatchery of rainbow trout in Samchok, Kwangwon-do have broken out and induced high mortality over 70%. Externally, the diseased fish showed dark discoloration, abdominal distension with ascites, slight exophthalmus and fecal casts. Internally, the gill and the liver of diseased fish were edematous and pale, and the stomach of moribund fish contained the milkish fluid. Microscopically there was extensive necrosis of the hematopoietic tissue in kidney and spleen, and scattered necrosis of pancreas, liver and lateral muscle. Especially, the necrosis of lamina propria, muscle layer and tela submucosa in the digestive tract known as the typical signs associated with infectious hematopoietic necrosis virus was seen clearly.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2000.05a
• /
• pp.410-418
• /
• 2000

In general, the experienced diver go into the water to find out Pan Shell which is spontaneously generated and brought up at the depth of 25-30 meters in the slime along the bank of an inlet of western and southern sea. However, this searching system has caused some problems in terms of economical efficiency and riskiness of diver. To overcome such problems and enhance the competence capability of fishermen, a new Pan Shell searching system is required. If an onboard searching system as a substitute of the way of diving into the deep water to identify the existence of Pan Shell is developed, it would be greatly beneficial to fishermen. The purpose of this research is to develop a new searching system of Pan Shell by making use of 50-200 Khz Ultra Sonic Signal, A/D Converter and computer process program. Based on the fact that Sonic Signal between the soft part of slime and the hard part of sand, pebble and Pan Shell is different, the possibility of this system have been approved in some degree by this time of research. However, to utilize this new system, further research on the establishment of data base and sample data is required. With the theoretical knowledge, the systematic research on the searching capability of Ultra Sonic Signal will be continued to identify the influence against the sea water subject. In this research, signal will be analyzed according to the influence range, power and sensitiveness of Ultra Sonic Generator. In addition, the radius of Ultra Sonic Signal will be included. The experimental field work will be executed at Nockdong, Pulkyo and other places well known as a habitat of Pan Shell.

• KIPS Transactions on Software and Data Engineering
• /
• v.5 no.11
• /
• pp.555-562
• /
• 2016

Semantic Role Labeling (SRL) is to determine the semantic relation of a predicate and its argu-ments in a sentence. But Korean semantic role labeling has faced on difficulty due to its different language structure compared to English, which makes it very hard to use appropriate approaches developed so far. That means that methods proposed so far could not show a satisfied perfor-mance, compared to English and Chinese. To complement these problems, we focus on suffix information analysis, such as josa (case suffix) and eomi (verbal ending) analysis. Korean lan-guage is one of the agglutinative languages, such as Japanese, which have well defined suffix structure in their words. The agglutinative languages could have free word order due to its de-veloped suffix structure. Also arguments with a single morpheme are then labeled with statistics. In addition, machine learning algorithms such as Support Vector Machine (SVM) and Condi-tional Random Fields (CRF) are used to model SRL problem on arguments that are not labeled at the suffix analysis phase. The proposed method is intended to reduce the range of argument instances to which machine learning approaches should be applied, resulting in uncertain and inaccurate role labeling. In experiments, we use 15,224 arguments and we are able to obtain approximately 83.24% f1-score, increased about 4.85% points compared to the state-of-the-art Korean SRL research.

• Journal of the Korean Institute of Traditional Landscape Architecture
• /
• v.28 no.1
• /
• pp.81-92
• /
• 2010

The purpose of this study is to find one of the standard scales of outside space. Fortunately, Korea has interesting old geographical name representing the environment where people feel comfortable by being enclosed by hills or mountains like in a room. The name of Chinese letter "谷" meaning valley is enunciated as "Gog", and Korean native equivalent word is "Sil." We thought we can find the standards of scale of the villages with name that ends with "Gog" or "Sil" by examining the scale of the village of the certain geographical name. We could collect 33 village's geographical names in Jin An Gun area, examined the scale, and considered the results from various aspects. The main conclusions are following: 1) Almost of the 33 villages with geographical names of "Gog", or "Sil" were surrounded by the hills of mountains, and the scales of the spaces surrounded were roughly 10~25ha. 2) Our ancestors felt comfortable be being enclosed with hills or mountains like in a room, and they have recognized the certain scale of the enclosed spaces. 3) After converting the 33 areas into squares for convenience, the scales of the spaces with the geographical name of "Gog" or "Sil" were roughly $300m{\times}300m{\sim}500m{\times}500m$.

• Journal of KIISE:Software and Applications
• /
• v.34 no.2
• /
• pp.95-103
• /
• 2007

This paper describes an English-to-Korean automatic translation system for patent documents which is constructed by a method customizing from a general domain to a specific domain. The customizing method consists of following steps: 1) linguistically studying about characteristics of patent documents, 2) extracting unknown words from large patent documents and terminologically constructing, 3) customizing the target language words of existing terms, 4) extracting and constructing patent translation patterns peculiar to patent documents, 5) customizing existing translation engine modules according to linguistic study about characteristics of patent documents, 6) evaluation of automatic translation results. The English-to-Korean patent machine translation system implemented by these customization steps shows a translation accuracy of 81.03% and is improving.

• Journal of the Korean Society for Library and Information Science
• /
• v.45 no.4
• /
• pp.139-155
• /
• 2011

The failure of distinguishing homographs in describing relations of individual authors and relations of authorship(i.e. distinction of persons with the same name or persons of the same literary author's name) will cause difficulties of retrieving exact information. It is because relations of automorphism cannot be formed between the two sets mentioned above. Therefore, it is ultimately necessary to set up the criteria or tools to distinguish homographs in order to retrieve more exact information. In the past, some efforts were made to develop authority data in order to solve the homograph problems by individual libraries, documents and portal sites in Korea. It is well understood that developing authority data by an individual institution was very difficult with no criteria or no rules to clarify the homograph problems at the national level. This study is to develop ways of recognizing individual names including subject words and proper nouns. The results of the study will present methods of distinguishing and describing homographs between individual author sets, and authorship sets particularly focused on the areas of arts and popular culture.

• Annual Conference on Human and Language Technology
• /
• 2006.10e
• /
• pp.57-64
• /
• 2006

본 논문에서는 웹과 같은 일반적인 도메인의 영한 자동 번역기를 특허용 영한 자동번역기로 특화하는 방법에 대해 기술한다. 특허용 영한 파동번역기로의 특화는 다음과 같은 절차에 의해 이루어진다: 1) 대용량 특허 문서에 대한 언어학적 특성 분석, 2) 대용량 특허문서 대상 전문용어 추출 및 대역어 구축, 3) 기존 번역사전 대역어의 특화, 4) 특허문서 고유의 번역 패턴 추출 및 구축, 5) 언어학적 특성 분석에 따른 번역 엔진 모듈의 특화 및 개선, 6) 특화된 번역 지식 및 번역 엔진 모듈에 따른 번역률 평가. 이와 같은 절차에 의해 만들어진 특허 영한 자동 번역기는 특허 전문번역가의 평가에 의해 전분야 평균 81.03%의 번역률을 내었으며, 분야별로는 기계분야(80.54%), 전기전자분야(81.58%), 화학일반분야(79.92%), 의료위생분야(80.79%), 컴퓨터분야(82.29%)의 성능을 보였으며 계속 개선 중에 있다. 현재 본 논문에서 기술된 영한 특허 자동번역 시스템은 산업자원부의 특허지원센터에서 변리사 및 특허 심사관이 영어 전기전자분야 특허 문서를 검색할 때 한국어 번역서비스를 제공받도록 이용되고 있으며($\underline{http://www.ipac.or.kr}$), 2007년에는 전분야 특허문서에 대한 영한 자동번역 서비스를 제공할 예정이다.