Please wait a minute...
J4  2010, Vol. 44 Issue (4): 732-737    DOI: 10.3785/j.issn.1008-973X.2010.04.019
计算机科学技术     
基于WordNet的无导词义消歧方法
王瑞琴1,2, 孔繁胜1, 潘俊1
1.浙江大学 人工智能研究所,浙江 杭州310027; 2.温州大学 计算机科学与工程学院,浙江 温州 325035
Unsupervised word sense disambiguation based on WordNet
WANG Ruiqin1,2, KONG Fansheng1, PAN Jun1
1. Institute of Artificial Intelligence, Zhejiang University, Hangzhou 310027, China;
2. School of Computer Science and Engineering, Wenzhou University, Wenzhou 325035, China
 全文: PDF  HTML
摘要:

有导词义消歧机器学习方法由于需要大量人力进行词义标注,难以适用于大规模词义消歧任务.提出一种避免人工词义标注的无导消歧方法.该方法综合利用WordNet知识库中的多种知识源(包括:词义定义描述、使用实例、结构化语义关系、领域属性等)描述歧义词的词义信息,生成词义的“代表词汇集”和“领域代表词汇集”,结合词汇的词频分布信息和所处的上下文环境进行词义判定.利用通用测试集Senseval3对6个典型的无导词义消歧方法进行开放实验,该方法取得平均正确率为49.93%的消歧结果.

Abstract:

Word sense disambiguation (WSD) based on supervised machine learning is hard to deal with largescale WSD because of its big labor cost. To solve this problem, an unsupervised WSD method was provided, which describes the word senses of an ambiguous word via synthesizing multiple knowledge sources in WordNet ontology, including definition glosses, samples, structured semantic relations, domain attributes, etc. From the description, a representative glossary and a domain representative glossary are deduced. The two structures together with the word sense frequency distribution and the context are used for WSD. The average disambiguation accuracy was 49.93% by this method in open test for six representative unsupervised WSD methods with Senseval3 English lexical sample data set.

出版日期: 2010-05-14
:  TP393  
通讯作者: 孔繁胜,男,教授.     E-mail: kfs@cs.zju.edu.cn
作者简介: 王瑞琴(1979—),女,内蒙古鄂尔多斯人,博士,从事人工智能、语义挖掘研究. E-mail: angelwrq@163.com
服务  
把本文推荐给朋友
加入引用管理器
E-mail Alert
RSS
作者相关文章  

引用本文:

王瑞琴, 孔繁胜, 潘俊. 基于WordNet的无导词义消歧方法[J]. J4, 2010, 44(4): 732-737.

WANG Rui-Qin, KONG Fan-Qing, BO Dun. Unsupervised word sense disambiguation based on WordNet. J4, 2010, 44(4): 732-737.

链接本文:

http://www.zjujournals.com/eng/CN/10.3785/j.issn.1008-973X.2010.04.019        http://www.zjujournals.com/eng/CN/Y2010/V44/I4/732

[1] IDE N, VERONIS J. Introduction to the special issue on word sense disambiguation: the state of the art [J]. Computational Linguistics, 1998, 24(1): 140.
[2] Cognitive Science Laboratory. WordNet—a lexical database for the English language [EB/OL]. [20061202]. http:∥wordnet.princeton.edu/.
[3] LESK M. Automatic sense disambiguation using machine readable dictionaries: how to tell a pine cone from an ice cream cone [C]∥ Fifth International Conference on Systems Documentation. Toronto Canada: ACM, 1986: 2426.
[4] WILKS Y, STEVENSON M. The grammar of sense: is wordsense tagging much more than partofspeech tagging [R]. Sheffield: University of Sheffield, 1996.[5] WU Z B, PALMER M. Verb semantics and lexical selection [C]∥ Proceedings of the 32nd Annual Meeting of the Association for Computational Linguistics. Las Cruces: [s.n.], 1994: 133138.
[6] LEACOCK C, CHODOROW M. Combining local context and WordNet similarity for word sense identification [M]. London: The MIT Press, 1998: 265283.
[7] RESNIK P. Using information content to evaluate semantic similarity in a Taxonomy [C]∥ Proceedings of the 14th International Joint Conference on Artificial Intelligence. Montreal: IEEE, 1995: 448453.
[8] JIANG J J, CONRATH D W. Semantic similarity based on corpus statistics and lexical taxonomy [C]∥ Proceedings of International Conference on Research in Computational Linguistics. Manchester: IEEE, 1997: 1933.
[9] LIN Dekang. An informationtheoretic definition of similarity [C]∥ Proceedings of the 15th International Conference on Machine Learning. Madison: Morgan Kaufmann, 1998: 296304.
[10] 王瑞琴,孔繁胜. 利用Wikipedia 的结构化信息计算语义相关性[J]. 浙江大学学报:工学版, 2009, 43(2): 315320.
WANG Ruiqin, KONG Fansheng. Computing semantic relatedness using structured information of Wikipedia [J]. Journal of Zhejiang University: Engineering Science, 2009, 43(2): 315320.
[11] AGIRRE E, RIGAU G. A proposal for word sense disambiguation using conceptual distance [C]∥ Proceedings of the International Conference on Recent Advances in NLP. Berlin/Heidelberg: Springer, 1995: 258264.
[12] BUSCALDI D, ROSSO P, MASULLI F. Integrating conceptual density with WordNet domains and CALD glosses for noun sense disambiguation [M]∥Advances in Natural Language Processing. Berlin/Heidelberg: Springer, 2004: 183194.
[13] NAVIGLI R, VELAROI P. Structural semantic interconnections: a knowledgebased approach to word sense disambiguation [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2005, 27(7): 10751086.
[14] MAGNINI B, STRAPPARAVA C, PEZZULO G, et al. The role of domain information in word sense disambiguation [J]. Natural Language Engineering, 2002, 8(4): 359373.
[15] Magnini and Cavaglia. WordNet domains [EB/OL]. \
[200005\]. http:∥wndomains.itc.it/.

[1] 尹可挺, 周波, 张帅, 徐斌, 陈一稀, 江丹. Web服务组合中基于QoS的自底向上服务替换[J]. J4, 2010, 44(4): 700-709.
[2] 周强, 应晶, 吴明晖. 基于特征分类的机会网络多因素预测路由[J]. J4, 2010, 44(3): 413-419.