This paper proposed a method to Web text categorization based on latent semantic analysis, it thinks that some context relations exist between terms, between terms and documents, and a semantic structure can be consist of respective relation between many documents and terms. The semantic structure is computed and deal with the structure to keep the most main relation between documents and terms and eliminates else huge, redundant, minor factor. The structure optimized is not only smarter than the original structure, but also keeps the most main relation, is easier to deal with the high dimensionality characteristic of the text document based on VSM, so it can mine the latent semantic relation. In sequent search, the latent similarity is computed between documents and improves the effective on the performance of the Web text categorization. |