Data mining is an important method of building intelligent decision support system, association rule is an important content of data mining. Apriori, the traditional algorithm, can only discovery the qualitative associated relation among data, but the quantitative associated relation is more helpful in decision making. Mapping attribute's value into discrete characters is a key step in mining quantitative association rules, in which the partition granularity of attribute's value is a key factor affecting the quality of the result of data mining— In this paper，by integrating the theory of rough sets．a method of mapping attribute’s value into discrete character with a fine value partition granularity is developed．and then a new algorithm of mining quantitative association rules．Apriori_2，is presented． M any rules which are helpfu1 in decision—making can be mined by Apriori_2