Automatic subject indexing of Chinese documents

Automatic subject indexing is a process to produce automatically a set of attributes that represent the content or topic of a document. In this paper, two approaches of automatic subject indexing based on VSM (vector space model) and subject words segmentation respectively are presented. The experimental results show that the first approach based on VSM is appropriate when the documents, which are indexed, are concentrative and the subject words available are less. The second approach based on subject words segmentation improves greatly efficiency of indexing and inter-indexer consistency.