멀티미디어 환경에서 텍스트코퍼스의 교육적 활용 방법

김명관 한국프랑스학회 2009 한국프랑스학논집 Vol.68 No.-

L'objectif de cet article est de montrer comment on peut intégrer les résultats des nouvelles analyses de corpus parallèles. Nous voulons illustrer d'autre part comment des types de corpus bilingues ouvrent de nouvelles perspectives aux apprenants et aux enseignants. Le corpus parallèle devient un outil pédagogique multifonctionnel et flexible, beaucoup mieux adapté aux besoins d'utilisateurs très divers. Nous exposons un usage pédagogique de corpus bilingues alignés dans la mise en oeuvre d'un outil de lecture de textes français sur le Web. Nous présentons d'abord les principales fonctionnalités des corpus textuels. Nous décrivons ensuite le rôle spécifique des corpus textuels dans un environnement multimédia. Nous montrons que les corpus parallèles peuvent aider les apprenants à traduire des textes littéraires. Nous mentionnons donc les principaux bénéfices de l'utilisation de corpus parallèles. On a amplement démontré, depuis plus de dix ans, l’intérêt que pouvait avoir l’utilisation d’un corpus dans le processus de traduction. Le corpus textuel est cependant souvent présenté comme un outil permettant aux enseignants et aux apprenants de trouver des réponses à leurs questionnements. Du point de vue des concepteurs de l'outil (enseignants), l'acquisition systématique de très grands corpus parallèles a été facilitée par une série d'outils d'extraction de textes bilingues sur le Web. Du point de vue des utilisateurs finaux (apprenants), les mots des leçons sont systématiquement illustrés à l'aide d'exemples issus de ces corpus parallèles. L'apprenant peut ainsi observer le mot recherché sous plusieurs formes fléchies et sa traduction dans plusieurs contextes minimaux et différenciés, qui sont spécifiques au contexte du mot candidat. Cette aide lexicale désambiguïsée et contextualisée constitue une des plus fortes valeurs ajoutées du système.

Comparisons of Soluble Klotho Concentration Between Healthy and Patient Cohorts

김명관,정동주 대한의생명과학회 2023 Biomedical Science Letters Vol.29 No.1

Since its first identification in 1995, klotho (KL) has become the most promising gene to consider for suppressing aging and aging-related diseases. KL knockout mice exhibited similar phenotypes found in human with premature aging such as short lifespan, osteoporosis, arteriosclerosis and hearing loss. Genetically modified mice overexpressing KL prolonged lifespan more than 20%. Also, clinical reports have indicated decreased concentration of the circulating KL protein in blood, which is called soluble klotho (sKL), is closely related to development of senile diseases. The best way to discover significance of sKL on the development of the diseases might be comparison of sKL concentration between controls and patients. Here we analyzed published clinical reports identified sKL concentration in the cohorts. The sKL concentrations were displayed using heatmap for better comparison. In most of the senile diseases, disease progression was inversely related with sKL concentration. Hypertension was the only disease had no relationship, while schizophrenia was the only disease had direct proportion to the disease progression. Overall, sKL concentration in blood could be a marker to determine current severity of the senile diseases and even to estimate disease progression for the patients at the onset of their senile diseases.

灰色의 세월

김명관 한국외국어대학 학도호국단 1975 외대 Vol.10 No.1

교육 시설 안전 등급 인증과 학교 안전도 심사에 대한 연구

김명관,권영국 대한인간공학회 2023 대한인간공학회 학술대회논문집 Vol.2023 No.5

유해 사진 탐지 도구 개발에 관한 연구

김명관 凡石學術奬學財團 1999 凡石學術論文集 Vol.3 No.-

병렬 코퍼스를 활용한 어휘 분석

김명관 한국프랑스학회 2018 한국프랑스학논집 Vol.101 No.-

Nous pouvons faire des recherches lexicales sur un corpus comparable ou un corpus parallèle. Dans cet article, nous proposons de nous concentrer sur les corpus parallèles. Cet article montre comment les listes de mots-clés sont affectées par les choix de concordance, corpus de référence et test statistique. Cette étude considére les mots-clés statistiquement remarquables dans le roman "Poil de carotte" en utilisant le concordancier. Il vise à trouver un certain nombre de mots-clés du roman par rapport au corpus de référence de RFI Info. Avec le fait que les significations des mots dans les textes sont différentes de leurs significations isolées, les mots-clés pourraient dans une certaine mesure être influencés par ces aspects. Bien que les résultats ne soient pas généralisables, car le nombre de mots-clés analysés n'est pas suffisant, cette brève analyse révèle un certain nombre d'aspects qui favorisent la compréhension de ce texte. Il s´agit, dans la plupart des cas, des verbes qui sont les plus fréquemment employés. Pour cela, nous avons prêté attention au repérage de la fréquence des verbes employés dans le corpus. Cette analyse s’appuie sur des algorithmes qui utilisent les fréquences et les distributions des unités textuelles prises comme points de repère pour l’identification et l’extraction des correspondances. Pour pouvoir comparer les résultats, nous avons créé des tableaux dans lesquels il est possible d´étudier la fréquence des verbes. En conclusion, nous avons testé cette méthode sur un roman de Jules Renard pour identifier les mots-clés.

웹 문서 정보추출과 자연어처리를 통한 온톨로지 자동구축에 관한 연구

김명관,이영우 한국인터넷방송통신학회 2009 한국인터넷방송통신학회 논문지 Vol.9 No.3

인터넷의 발달로 전자문서가 증가함에 따라, 정보검색기술의 중요성도 함께 증가하게 되었다. 본 연구는 비정형 텍스트 웹 문서로부터 사용자가 요구하는 핵심 의미 지식을 추출하기 위하여 LGG (Local Grammar Graph) 구축에 기반 하여 보다 효율적이고 정확한 지식구축을 가능하게 한다. 주가등락이라는 특정 분야의 패턴을 추출하여 만든 패턴 문법을 사용해서 OWL(Web Ontology Language) 기반의 온톨로지를 구축하였다. 특정 분야의 온톨로지를 구축함으로써 기존 검색에서 할 수 없었던 지식의 의미 검색이 가능하며 나아가 사용자가 원하는 질의에 대한 정보의 추론이 가능할 것 이다.

감정요소를 사용한 정보검색에 관한 연구

김명관,박영택,Kim, Myung-Gwan,Park, Young-Tack 한국정보처리학회 2003 정보처리학회논문지B Vol.10 No.6

감정요소를 사용한 정보검색시스템은 감정에 기반한 정보검색을 수행하기 위하여 감정시소러스를 구성하였으며 이를 사용한 감정요소추출기를 구현하였다. 감정요소추출기는 기본 5가지 감정 요소를 해당 문서에서 추출하여 문서를 벡터화시킨다. 벡터화시킨 문서들은 k-nearest neighbor, 단순 베이지안 및 상관계수기법을 사용한 2단계 투표방식을 통해 학습하고 분류하였다. 실험결과 분류 방식과 K-means를 이용한 클러스터링에서 감정요소에 기반한 방식이 더 우수하다는 결과와 5,000 단어 미만의 문서 검색에 감정기반 검색이 유리하다는 것을 보였다. In this paper, we propose a novel approach to employ emotional features to document retrieval systems. Fine emotional features, such as HAPPY, SAD, ANGRY, FEAR, and DISGUST, have been used to represent Korean document. Users are allowed to use these features for retrieving their documents. Next, retrieved documents are learned by classification methods like cohesion factor, naive Bayesian, and, k-nearest neighbor approaches. In order to combine various approaches, voting method has been used. In addition, k-means clustering has been used for our experimentation. The performance of our approach proved to be better in accuracy than other methods, and be better in short texts rather than large documents.

질려자에 대한 고찰

김명관,Kim, Myeong-Gwan 대한한약협회 2003 大韓漢藥 Vol.7 No.-

유 · 무기접착제가 압출법단열재 연소에 미치는 영향

김명관,허정필,최우석,이정섭 한국화재감식학회 2017 한국화재감식학회 학회지 Vol.8 No.3

In this study, we have examined the possibility of fire spread by the extruded polystyrene on the external wall of construction site. Through burn pattern tests, it was confirmed that the polystyrene could be difficult to burn continuously without the adhesives and that the mechanisms of combustion reactions how the adhesives were applied. In Limited Oxygen Index test, the oxygen index of organic adhesive appeared at 19%, and total amount of heat release of the organic adhesive was appeared as 75.29MJ/㎡, which was 16.68MJ/㎡ more than that of the inorganic adhesive, and the persistence of heat release rate of organic adhesive was 162s longer than that of the inorganic adhesive in cone-calorimeter test. As the results of these tests, the potential of fire spread by extruded polystyrene alone or with inorganic adhesive were lower than it with organic adhesive. In addition, there would be high risk of fire spread depending on the method of application of organic adhesive for the extrude polystyrene.