Mammary cancer is a breast cells carcinogenesis disease which reduces from cells abnormal proliferation because of their normal characteristics losing in the effects of variety of both inside and outside carcinogenic factors, thus the study is based on the important factors that act as diagnostic basis of mammary image which has significant guidance for mammary cancer diagnosis.In this paper, univariate analytic method has been used to fix the important factors and to delete the unimportant factors in the process of data preprocess in order to complete the items containing missing values, thus reduce the bad influence of missing items. The items that still contain missing values will be deleted. This method can also solve the problems of support threshold setting which can help to eliminating the unimportant items. Experiment demonstrates the feasibility and correct of this method. After datum preprocessing, the associate rules to be mine could be started and rules-correlation could be evaluated. It is proved that the method that this paper adopted can effectively explore the correlative and associate rules.The main jobs of this paper are:1,Full analyzing the advantages and disadvantages of existing frequent itemsets and association rules mining algorithm.2,On this basis mentioned above, due to the deficiencies of frequent itemsets mining associate rules algorithm, Pattern Count-Correlation Frequent item Sets Algorithm have been proposed. In the end, SPSS statistics analyzing software has been used to prove the validating of this algorithm.3,Make use of this new correlative and associate rules mining algorithm to analysis mammography data and further validate the theory. |