News topics categorization using latent Dirichlet allocation and sparse representation classifier

Abstract

Recently, subscribing news from websites has become a new trend for many Internet users. In a news reading browser, it is essential all the news documents are properly categorized. For automatically categorizing the news topics, this paper presents a news categorization method using latent Dirichlet allocation (LDA) and sparse representation classifier (SRC). In our work, the LDA is used as the feature learning method. The multinomial distribution of the news topics is regarded as the feature of the document. These features are stacked as an over-complete dictionary, permitting us to perform SRC-based categorization. The experimental results show that the proposed method outperforms the traditional method.