Multilabel text categorization based on a new linear classifier learning method and a category-sensitive refinement method

摘要

In this paper, we present a new approach for dealing with multilabel text categorization based on a new linear classifier learning method and a category-sensitive refinement method. We use a new weighted indexing technique to construct a multilabel linear classifier. We use the degrees of similarity between categories to adjust the relevance scores of categories with respect to a testing document. The testing document can be properly classified into multiple categories by using a predefined threshold value. We also compare the performance of the proposed method with the text categorization methods based on the Reuters-21578 ModeApte Split Text Collection. The experimental results show that the performance of the proposed method is better than the existing methods.