The invention belongs to the field of vector modeling, and particularly relates to a Chinese word vector modeling method. According to an existing Chinese word vector modeling method, only radical strokes and other information are simply introduced, and meanwhile unregistered words cannot be reasonably expressed. In order to solve the problems of inconvenience in use due to incapability of automatically updating statements, incapability of following the trend of the times and low intelligent degree in the prior art, the invention provides the following scheme that the method comprises the following steps: S1, obtaining corpus bases of form and sound characteristics of Chinese words, classifying the corpus bases, and carrying out A-type numbering on the types of the corpus bases; S2, sequentially carrying out B-type numbering on the corpus foundation in each type; and S3, mapping the number to a vector space. The unregistered words can be reasonably expressed, statements can be automatically updated, the trend of the times can be followed, the intelligent degree is high, and use is convenient.