作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

计算机工程 ›› 2011, Vol. 37 ›› Issue (17): 133-135. doi: 10.3969/j.issn.1000-3428.2011.17.045

• 人工智能及识别技术 • 上一篇    下一篇

基于NPE的Web文本分类方法研究

徐海瑞,张文生,吴 双   

  1. (中国科学院自动化研究所,北京 100190)
  • 收稿日期:2010-12-27 出版日期:2011-09-05 发布日期:2011-09-05
  • 作者简介:徐海瑞(1984-),男,硕士研究生,主研方向:机器学习,数据挖掘;张文生,研究员、博士;吴 双,硕士研究生
  • 基金资助:
    国家自然科学基金资助项目(90924026)

Research of Web Text Classification Method Based on Neighborhood Preserving Embedding

XU Hai-rui, ZHANG Wen-sheng, WU Shuang   

  1. (Institute of Automation, Chinese Academy of Sciences, Beijing 100190, China)
  • Received:2010-12-27 Online:2011-09-05 Published:2011-09-05

摘要: 提出一种基于流形学习的文本分类方法以解决高维文本数据分类问题。利用近邻保持嵌入流形学习算法获得高维Web文本空间中的低维流形结构,采用K近邻分类器对低维流形进行分类。实验结果表明,基于流形学习的方法能获得较好的分类效果,具有稳定的性能。

关键词: 近邻保持嵌入算法, 流形学习, 文本分类, 特征提取, K近邻

Abstract: To efficiently resolve the high dimensional Web text classification problem, a novel classification algorithm is proposed in this paper on the basis of manifold learning. The algorithm can explore and preserve the inherent structure on high dimensional Web text space, and the classification and predication in the lower dimension feature space are implemented with K-Nearest Neighbor(KNN). Experimental results show that the algorithm achieves higher classification accuracy and stability.

Key words: Neighborhood Preserving Embedding(NPE) algorithm, manifold learning, text classification, feature extraction, K-Nearest Neighbor(KNN)

中图分类号: