| 基于词关联度的语义相关度算法研究 |
| |
| 引用本文： | 张增杰,李晓城,刘鑫,夏勇明,钱松荣.基于词关联度的语义相关度算法研究[J].微型电脑应用,2011,27(3):45-47,51,6. |
| |
| 作者姓名： | 张增杰 李晓城 刘鑫 夏勇明 钱松荣 |
| |
| 作者单位： | 1. 复旦大学通信科学与工程系,上海,200433
2. 复旦大学信息科学与工程学院,上海市,200433 |
| |
| 摘 要： | 如今网页排名算法很多,基本上可以分为两类:基于超链接和基于内容。比较成熟的算法主要有PageRank、HITS、LSI等。本文基于向量空间模型以及信息论,提出一个与文章内容相关的语义相关度算法模型。该模型将文章语义抽象为词频表,并通过机器学习构建词语之间的关联度表,以此词关联度为基础,计算文章之间的相关度。实验结果表明,文中提出的相关度算法可以有效的根据文章之间语义相关度大小来进行排名。
|
| 关 键 词： | 词关联度 语义 相关度 向量模型 信息量 概率模型 |
Research of Semantic Computation Algorithm Based on Word Relativity |
| |
| Zhang Zengjie,Li Xiaocheng,Liu Xin,Xia Yongming,Qian Songrong.Research of Semantic Computation Algorithm Based on Word Relativity[J].Microcomputer Applications,2011,27(3):45-47,51,6. |
| |
| Authors: | Zhang Zengjie Li Xiaocheng Liu Xin Xia Yongming Qian Songrong |
| |
| Affiliation: | Zhang Zengjie,Li Xiaocheng,Liu Xin,Xia Yongming,Qian Songrong(Communication Science and Engineering,Fudan University,Shanghai 200433,China) |
| |
| Abstract: | Now page rank algorithm had been well studied, basically can be divided into two categories: Hyperlink-based and content-based. There are more sophisticated algorithm PageRank, HITS, LSI and so on. Based on vector space model, and information theory, the article proposed a content-related semantic relevance algorithm model. This model calculates the relevance between articles based on the word correlation. Experimental results show that the proposed correlation algorithm can efficiently rank files according... |
| |
| Keywords: | Word Correlation Semantic Relevance Vector Model Information Theory Conditional Probability |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
