| 基于模糊理论的关键词识别算法 |
| |
| 引用本文： | 陈立伟,王文姝,王颖芳.基于模糊理论的关键词识别算法[J].应用科技,2010,37(9):5-8. |
| |
| 作者姓名： | 陈立伟 王文姝 王颖芳 |
| |
| 作者单位： | 哈尔滨工程大学,信息与通信工程学院,黑龙江哈尔滨150001 |
| |
| 基金项目： | 国家自然科学基金资助项目 |
| |
| 摘 要： | 针对关键词发音相似易混淆及反词模型难确定、难训练等问题,提出一种结合模糊理论的方法,利用模糊C均值聚类算法对候选关键词进行2次聚类,同时将新的聚类中心作为反词模型进行最后确认.实验结果表明,这种方法使识别率得到了显著的提高.
|
| 关 键 词： | 关键词识别 模糊C均值聚类 隐马尔可夫模型 反词模型 |
Keyword recognition algorithm based on fuzzy theory |
| |
| CHEN Li-wei,WANG Wen-shu,WANG Ying-fang.Keyword recognition algorithm based on fuzzy theory[J].Applied Science and Technology,2010,37(9):5-8. |
| |
| Authors: | CHEN Li-wei WANG Wen-shu WANG Ying-fang |
| |
| Institution: | (College of Information and Communication Engineering, Harbin Engineering University, Harbin 150001,China) |
| |
| Abstract: | Considering key words can be easily confused with the similarly pronounced words, anti-word model is difficult to be trained and difficult to be sure,a method combined with the fuzzy theory was proposed, which applied the fuzzy C-means clustering algorithm in the second cluster of the candidate keywords, and at the same time the new cluster center was used as an anti-word model in final confirmation. Experimental results show this method can improve the recognition rate significantly. |
| |
| Keywords: | keyword recognition fuzzy C-means clustering hidden Markov model anti-word model |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
