【作者】 廖加彬； 林世平；

【Author】 Liao Jiabin~1 Lin Shiping~2 College of mathematic and computer science,Fuzhou University 350180

【机构】 福州大学数学与计算机科学学院；

【摘要】 近年来,许多国内外自然语言理解领域的学者关注于全自动的语义模式的识别,包括单词语义、词组语义、短语语义以及句子语义等。这些研究的共同目标在于希望能够在对语义模式识别的基础上去理解语义,真正理解一个单词的意义,一个动作的结果,一个句子的含义等。本文提出了一种基于聚类的中文浅层语义模式识别的方法,希望能够用聚类这种传统的分析方法来对一些简单的语义模式进行识别。本文所提到的浅层语义模式包括语义单元共现所体现出的语义相似性以及语义单元组合形成完整意义单元所体现的模式。更多还原

【Abstract】 Recently,many researchers,all over the world,pay great attention on the automatic semantic pattern recognition,in- cluding semantic between words,phrase and sentence,and so on.The common purpose of these researches is to get the semantic basing on the recognition of semantic pattern,to get an word’s mean,to get the outcome of an action,to get the meanings of a sentence,and so on.We propose a method to recognize shallow semantic pattern in Chinese,basing on cluster.We wish to use the cluster,a classic algorithm,to analyze and recognize some simple semantic pattern.The shallow semantic patter,which we focus on in this paper,is the semantic similarity implied in the concurrence of semantic units,and the semantic similarity implied in the relation between the combination of semantic units.更多还原