摘要
在文本中,常常出现一词多义的现象,本文提出一种基于语义关系图的词义消歧算法,算法首先利用Word Net的语义关系构建语义关系图;其次,通过多义词在语义关系图的上下文选择最佳语义关系。测试用Senseval-3中的全文内容作为实验测试集,结果表明,词义消歧算法的测试结果很理想。
In the text, the phenomenon of a word often has many meaning. In this paper,I make a wordsense disambiguation algorithm – based on semantic relation graph. firstly,using WordNet semanticrelation for semantic relation graph.Secondly,choosing the best semantic relation of polysemous words inthe context of semantic relation graph. in this paper, the set of test are used all the text of Senseval-3, theexperimental results display that the test results of the word sense disambiguation algorithm based onsemantic relation graph is very satisfactory.
出处
《科技通报》
北大核心
2015年第3期228-232,257,共6页
Bulletin of Science and Technology
关键词
WORDNET
语义关系图
词义消歧
语义选择
WordNet
semantic relation graph
word sense disambiguation
semantic choosing
