Based on the practical contradiction between the ever-changing employment situation and the higher employment expectation of the postgratuate, this paper divides the respondents into several groups using two different methods of the cluster analysis and the Feature Weighting Naive Bayesian Classifier. So the department of employment guidance can use different strategies to deal with different groups and to give highly targeted employment guidance and psychological counselling. For the new respondent to unknown group, two different methods can also predict his group. This respondent can quickly find the corresponding group, and the department of employment guidence carries personalized employment guidance out. Based on actuality of normal university, this paper creates the concepts of the education index and the non-education index, and improves the efficiency of the employment guidance. At last, the two different methods are compared. |