Abstract:

We designed a new model via merging probability latent semantic analysis （PLSA） model and tree model, and applied it to the semantic analysis of text medical records. We solved the problem of “one meaning of multiword combination”  in the semantic analysis process of medical records so as to
reduce the semantic dimension and simplify the structure of the window semantic tree. The experiment of semantic decomposition and semantic retrieval effectively shows the advantages of PLSAtree model in the semantic analysis of text medical records.

Key words: PLSAtree model, text medical records, semantic analysis, new method