Dictionary-based text categorization of chemical web pages

  • Authors:
  • Chun-Yan Liang;Li Guo;Zhao-Jie Xia;Feng-Guang Nie;Xiao-Xia Li;Liang Su;Zhang-Yuan Yang

  • Affiliations:
  • Key Laboratory of Multiphase Reactions, Institute of Process Engineering, Chinese Academy of Sciences, Beijing, China;Key Laboratory of Multiphase Reactions, Institute of Process Engineering, Chinese Academy of Sciences, Beijing, China;Key Laboratory of Multiphase Reactions, Institute of Process Engineering, Chinese Academy of Sciences, Beijing, China;Key Laboratory of Multiphase Reactions, Institute of Process Engineering, Chinese Academy of Sciences, Beijing, China;Key Laboratory of Multiphase Reactions, Institute of Process Engineering, Chinese Academy of Sciences, Beijing, China;Key Laboratory of Multiphase Reactions, Institute of Process Engineering, Chinese Academy of Sciences, Beijing, China;Key Laboratory of Multiphase Reactions, Institute of Process Engineering, Chinese Academy of Sciences, Beijing, China

  • Venue:
  • Information Processing and Management: an International Journal
  • Year:
  • 2006

Quantified Score

Hi-index 0.00

Visualization

Abstract

A new dictionary-based text categorization approach is proposed to classify the chemical web pages efficiently. Using a chemistry dictionary, the approach can extract chemistry-related information more exactly from web pages. After automatic segmentation on the documents to find dictionary terms for document expansion, the approach adopts latent semantic indexing (LSI) to produce the final document vectors, and the relevant categories are finally assigned to the test document by using the k-NN text categorization algorithm. The effects of the characteristics of chemistry dictionary and test collection on the categorization efficiency are discussed in this paper, and a new voting method is also introduced to improve the categorization performance further based on the collection characteristics. The experimental results show that the proposed approach has the superior performance to the traditional categorization method and is applicable to the classification of chemical web pages.