摘要
针对当前几种常用文本检索方法的不足,文中基于统计模型和小波变换,提出了一种新的文本检索方法。与传统方法的主要区别在于:1)利用小波变换把输入信号引入到频域进行处理,消除了交叉比较运算的巨大计算量;2)在进行相关度计算时,同时考虑了检索词的出现次数和出现位置因素,有效提高了检索精确度。理论分析和实验结果表明该方法较传统方法在查准率和查询速度上均有所提高。
A novel document retrieval method based on statistical model and wavelet transform is proposed after analyzing the disadvantage of several common used document retrieval methods. First,It analyses the input signal in frequency domain where cross compare computations were avoided;Second,it considers both the term count and position when calculating the similarity,which gets a high precision. Experimental results illustrate it a efficient method considering both precision and speed when compare with common used document retrieval methods.
作者
魏彬
张军
项颖
WEI Bin, ZHANG Jun, XIANG Ying (Department of Information Engineering, Guangdong University of Technology, Guangzhou 510006, China)
出处
《电脑知识与技术》
2009年第3期1686-1687,1698,共3页
Computer Knowledge and Technology
关键词
文本检索
词向量
谱向量
小波变换
相似度
document retrieval
term vector
spectra vector
wavelet transform
similarity