厦门大学学报:哲学社会科学版
廈門大學學報:哲學社會科學版
하문대학학보:철학사회과학판
Journal of Xiamen University(A Quarterly for Studies in Arts & Social Sciences)
2010年
3期
137~144
,共null页
词义标注 义项 义项库
詞義標註 義項 義項庫
사의표주 의항 의항고
word sense tagging; word sense; word sense base
计算机词义标注一般都沿用传统词典的义项,随着研究的深入,对传统词典义项的设立提出了新的要求。有学者提出了以句法功能为主要依据来辨析义项。我们的研究注重词义语义特征的提取,共建构了七个资源库,互相之间主要围绕着词义语义特征展开,义项库在计算机词义标注中居于关键位置。机用义项库与传统词典义项相比,在词义标注目的、对语料的覆盖力、义项间语义关系、义项的可识别性、计算机处理能力的适应性及不同属性的义项义频的差距等方面,都存在着相当大的差别。从而,可以对基于传统词典义项上形成的机用义库提出改造与完善的若干原则与方法。
計算機詞義標註一般都沿用傳統詞典的義項,隨著研究的深入,對傳統詞典義項的設立提齣瞭新的要求。有學者提齣瞭以句法功能為主要依據來辨析義項。我們的研究註重詞義語義特徵的提取,共建構瞭七箇資源庫,互相之間主要圍繞著詞義語義特徵展開,義項庫在計算機詞義標註中居于關鍵位置。機用義項庫與傳統詞典義項相比,在詞義標註目的、對語料的覆蓋力、義項間語義關繫、義項的可識彆性、計算機處理能力的適應性及不同屬性的義項義頻的差距等方麵,都存在著相噹大的差彆。從而,可以對基于傳統詞典義項上形成的機用義庫提齣改造與完善的若榦原則與方法。
계산궤사의표주일반도연용전통사전적의항,수착연구적심입,대전통사전의항적설립제출료신적요구。유학자제출료이구법공능위주요의거래변석의항。아문적연구주중사의어의특정적제취,공건구료칠개자원고,호상지간주요위요착사의어의특정전개,의항고재계산궤사의표주중거우관건위치。궤용의항고여전통사전의항상비,재사의표주목적、대어료적복개력、의항간어의관계、의항적가식별성、계산궤처리능력적괄응성급불동속성적의항의빈적차거등방면,도존재착상당대적차별。종이,가이대기우전통사전의항상형성적궤용의고제출개조여완선적약간원칙여방법。
Computer word sense tagging set up a new requirement fot traditional dictionary.Some researchers have long been aware of this issue,Yu Shi-Wen、Wu Yun-Fang proposed to use syntactic function as the main basis.Based on previous studies,the article first introduces the method for word sense tagging and seven resources.Word Sense Base for Word Sense Tagging of Modern Chinese(WSB) is in a key position.Then it proposes the differences on semantic analysis between traditional dictionary and machine dictionary,from the purpose of word sense tagging,coverage,semantic relationships,operability,adaptability to computer processing and frequency gap between sense of different part of speech.At last it discusses the principles and methods of building new word sense base for a higher degree general,written,large-scale real corpus of modern Chinese.It should be suitable for computer processing.