Abstract: In recent years, associate mining based on constraint is more and more focused. From the algorithm of existing associate mining, it is easy to observe that the traditional thresholds are given mostly by expert or found after repealed test User's feedback and the support of objective evidence was lacked. According to this problem,the construetional method of thresholds catered to user requirement was proposed. To obtain the self-adaptation constraints thresholds, the theory of normal distribution is cited in this method. FGC algorithm is ameliorated by using succinct constraint. A speedy, effective and intuitive frequent items mining algorithm was proposed. Experiments show that the proposed methods can increase the system's availability and reduce the runtime of algorithm.

Key words: FP-Tree,Frequent item-sets,Succinct constraints,Adaptive threshold