http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
- 中文 을 입력하시려면 zhongwen을 입력하시고 space를누르시면됩니다.
- 北京 을 입력하시려면 beijing을 입력하시고 space를 누르시면 됩니다.
RISS 인기검색어
- 검색키워드
- 저자 : 정천영 (검색결과 34 건)
- 무료
- 기관 내 무료
- 유료
-
정천영,서영훈,임희동 충북대학교 산업과학기술연구소 1999 산업과학기술연구 논문집 Vol.13 No.1
This paper describes Korean analyzer based on core words for the machine translation of spoken language. Because the grammar based on core words can describe grammar regardless of word order, it can solve the partially-free word order feature of Korean and some problems occurring when parsing spoken language containing unnecessary syllables such as meaningless words, repetitions speech, etc. Also, because unnecessary syllables are removed by the grammar based on core words, we can analyze spoken language robustly without the mechanism such as skip of syllables. The data used for the experiment are scheduling domain of ETRI, and the parsing performance is 99.0% for trained utterances, 88.1% for untrained utterances.
-
정천영,서영훈,Jung, Cheon-Young,Seo, Young-Hoon 한국정보처리학회 1998 정보처리논문지 Vol.5 No.9
본 논문에서는 한국어대화체를 분석하고 의미패턴에 기반한 대화체 한영 기계번역 시스템에 대하여 기술한다. 한영 기계번역에서 구문정보를 이용한 한국어 모호성은 의미패턴을 이용하여 해결할 수 있다. 따라서 대화체 번역을 위하여 한국어 스케쥴링 도메인으로부터 추출된 의미패턴에 기반한 시스템을 구성한다. 번역의 강건함을 위하여 한국어 문장 분석시 음절을 건너뛰어 분석할 수 있도록 하였으며, 패턴수를 줄이기 위하여 의미패턴에 옵션을 부가하였다. 실험을 위하여 사용된 데이터는 스케쥴링 도메인으로 실험결과 88%의 번역율을 보인다. This paper analyzes Korean spoken language and describes the machine translation o[ Korean to-English spoken language based on semantic patterns, In Korean-to-English machine translation. ambiguity of Korean sentence analysis using syntactic information can be resolved by semantic patterns, Therefore, for machine translation of spoken language, we estabilish the system based on semantic patterns extracted from Korean scheduling domain, This system obtains the robustness by skip ability of syllables in analysis of Korean sentence and we add options to semantic patterns in order to reduce pattern numbers, The data used [or the experiment are scheduling domain and performance of Korean-to-English translation is 88%.
-
-
-
유도결합 플라즈마와 이중관 반응기를 이용하여 제조한 보론-도핑된 결정질 실리콘 나노입자의 합성
정천영,구정분,장보윤,이진석,김준수,한문희,Jung, Chun-Young,Koo, Jeong-Boon,Jang, Bo-Yun,Lee, Jin-Seok,Kim, Joon-Soo,Han, Moon-Hee 한국전기전자재료학회 2014 전기전자재료학회논문지 Vol.27 No.10
B-doped Si nanoparticles were synthesized by using inductive coupled plasma and specially designed double tube reactor, and their microstructures were investigated. 0~10 sccm of $B_2H_6$ gas was injected during the synthesis of Si nanoparticles from $SiH_4$ gas. Highly crystalline Si nanoparticles were synthesized, and their crystallinity did not change with increase of $B_2H_6$ flow rates. From SEM measurement, their particle sizes were approximately 30 nm regardless of $B_2H_6$ flow rates. From SIMS analysis, almost saturation of B in Si nanoparticles was detected only when 1 sccm of $B_2H_6$ was injected. When $B_2H_6$ flow rate exceeded 5 sccm, higher concentration of B than solubility limit was detected even if any secondary phase was not detected in XRD or HR-TEM results. Due to their high electronic conductivity, those heavily B-doped Si nanoparticles can be a potential candidate for an active material in Li-ion battery anode.
-
전처리와 조사/어미의 결합 제약을 이용한 한국어 형태소 해석의 중의성 해소
정천영,서영훈,김남철 충북대학교 산업과학기술연구소 1997 산업과학기술연구 논문집 Vol.11 No.2
There are many morphological ambiguities for analysing Korean language, such as part-of-speech smbiguity, stem-type ambiguity, morpheme segmentation ambiguity, morpheme-length ambiguity, irregular stem ambiguity, morpheme-drop ambiguity. If the ambiguities are not resolved in the morphological analysis step, the considerable amount of ambiguities will be generated in the next analysis step, for example, syntactic analysis, semantic analysis. In Korean, some josas and eomis need specific stems. This characteristic is useful for disambiguation of analysis. In this paper, the combination restriction of josa/eomi is used for disambiguating Korean analysis. The pre-processing of Korean analysis is proposed for analysing properly the word that is mixed with Korean and non-Korean characters or is including the quotation marks, parenthesis marks, or bracket marks. In many words that have the morphological ambiguity, the ambiguities were resolved, and the average success rate was 92%.
-
정천영,장수진,진성일 충남대학교 자연과학연구소 1992 忠南科學硏究誌 Vol.19 No.1
The purpose of this paper is to suggest an automatic indexing system for Korean documents with an auxiliary word and the ending of a word in thesis. This system is structured with an auxiliary word dictionary, a negative dictionary, a declinable word list and an abbreviation dictionary. In this thesis, we suggest methods to analyze short clauses, remove negative words, restore abbreviations to original words and extract compound phrases.
-
정천영(Jung Cheon Young),성영훈(Seo Young Hoon) 한국정보처리학회 1998 정보처리학회논문지 Vol.5 No.9
본 논문에서는 한국어 대화체를 분석하고 의미패턴에 기반한 대화체 한영 기계번역 시스템에 대하여 기술한다. 한영 기계번역에서 구문정보를 이용한 한국어 분석의 모호성은 의미패턴을 이용하여 해결할 수 있다. 따라서 대화체 번역을 위하여 한국어 스케쥴링 도메인으로부터 추출된 의미패턴에 기반한 시스템을 구성한다. 번역의 강건함을 위하여 한국어 문장 분석시 음절을 건너뛰어 분석할 수 있도록 하였으며, 패턴수를 줄이기 위하여 의미패턴에 옵션을 부가하였다. 실험을 위하여 사용된 데이터는 스케쥴링 도메인으로 실험결과 88%의 번역율을 보인다. This paper analyzes Korean spoken language and describes the machine translation of Korean-to-English spoken language based on semantic patterns. In Korean-to-English machine translation, ambiguity of Korean sentence analysis using syntactic information can be resolved by semantic patterns. Therefore, for machine translation of spoken language, we establish the system based on semantic patterns extracted from Korean scheduling domain. This system obtains the robustness by skip ability of syllables in analysis of Korean sentence and we add options to semantic patterns in order to reduce pattern numbers. The data used for the experiment are scheduling domain and performance of Korean-to-English translation is 88%.
-
정천영(Cheon-Young Jung),서영훈(Young-Hoon Seo) 한국정보과학회 2001 정보과학회논문지 : 소프트웨어 및 응용 Vol.28 No.4
본 논문에서는 대화체 번역을 위한 논항 구조에 기반한 한국어 분석에 대하여 기술한다. 논항구조 기반 문법은 순서에 관계없이 기술된다. 따라서 한국어의 부분 자유 어순 특성으로 문법이 방대해지는 문제점을 해결할 수 있다. 또한, 서술어가 지배하는 논항이 문법으로부터 선택됨으로서 대화체가 갖는 특성인 간투어나 중복 발화 현상 등을 효과적으로 해결할 수 있다. 실험을 위하여 사용된 데이타는 ‘여행 안내’ 영역 중에서 1,335개의 훈련된 발화문과 420개의 훈련되지 않은 발화문이다. 실험 결과 훈련된 발화문에서는 99.7%, 훈련되지 않은 발화문에서는 93.3%의 분석 성공률을 보였다. This paper describes a Korean analysis based on argument structures for spoken language translation. A rule in our grammar denotes only immediate dominance relation. That is, the order of components in a grammar rule is meaningless. Owing to characteristic of this grammar the our grammar size can be greatly reduced. Also, a problem of ungrammatical constituents is resolved because only arguments are described in a rule, and the parser ignores words in input utterance when they do not match in the rule. Experimental domain is travel arrangement. The total number of utterances in domain corpus is 1,755. We train a system by use of 1,335 utterances, and remain 420 utterances for test. Experimental result shows that 99.7% for trained utterances and 93.3% for untrained ones are analyzed successfully.
-
연관 검색어 추천
이 검색어로 많이 본 자료
활용도 높은 자료
