The invention discloses a sensitive attribute identification and classification method for structured data. According to the method, the sensitive attributes in the structured data table are automatically identified and accurately classified and graded, so that the classification and grading results of the sensitive attributes are output subsequently, and data release is carried out after proper privacy protection processing is adopted. The clustering analysis is an unsupervised learning mode in machine learning. The classification of data attributes can be automatically achieved according tothe similarity between the attributes without predicting the classification situation of a sample set. In the final identification stage of the sensitive attributes, the incidence relation between thesensitive attributes and the suspected sensitive attributes is considered. The purpose is to solve the problem of link attacks, so that the attributes associated with the sensitive attributes are further mined. The privacy leakage degree is reduced. In the classification stage of the sensitive attributes, mutual information correlation between the sensitive attributes is considered. The incidencerelation between the attributes is analyzed. The sensitive attributes are classified more accurately by combining the mutual information correlation and the incidence relation.