摘要
通过分析文本挖掘中的2个关键步骤——文本特征空间构造和相似距离度量,指出流行的文本挖掘过程中存在着大量同义和关联噪声。大量存在的同义词和关联词,造成文本特征空间无法准确表达文本语义以及高维计算复杂性问题。利用潜在语义分析和关联规则挖掘构造同义和关联词集,用于减少文本特征空间中的同义词和关联词,降低信息冗余,改进挖掘效率。文中对相应的算法进行了描述,实验结果令人满意。
There are many synonymy and association words in procedure of popular text mining based on analysis of two key step-construction of document feature space and calculation of similarity- in text mining. The synonymy and association words are so many that decrease veracity of semantic presentation and increase complexity of calculation. Paper integrates the latent semantic analysis with the mining of association rules and puts forward the method of construction on set of synonymy and association words.We can use two sets to reduce the redundance of information and improve the efficiency of mining. The corresponding algorithm is described and result of experiment is satisfactory.
出处
《微电子学与计算机》
CSCD
北大核心
2007年第4期118-121,共4页
Microelectronics & Computer
关键词
文本挖掘
潜在语义分析
关联规则挖掘
算法
text mining
latent semantic analysis
mining of association rule
algorithm