Abstract:
A quantitative method for mining positive and negative association rules is proposed. The whole structure of minin system is introduced, including support degree and confidence degree. By increasing the degree of contrast influence, this structure is optimized, which promotes the mining efficiency of interesting knowledge. Simulation experimental results show this method can eliminate some invalid association rules, and make up for the demerits of traditional ones. It has the value of application.
Key words:
data mining,
negative association rules,
contrast influence
摘要: 提出一种能够有效挖掘正、负关联规则的量化方法,介绍挖掘系统的整体架构,包括支持度和置信度,通过增加对比影响度,对其进行优化,从而提高有趣知识的挖掘效率。仿真实验结果表明,该方法可以剔除一些无效关联规则,弥补传统方法的不足,具有一定应用价值。
关键词:
数据挖掘,
负关联规则,
对比影响
CLC Number:
ZHENG Shang-zhi; LIANG Bao-hua¬; ZHAO Xiao-long; CAI Min. Quantitative Method of Positive and Negative Association Rules[J]. Computer Engineering, 2009, 35(15): 74-75,7.
郑尚志¬;梁宝华;赵小龙;蔡 敏. 正负关联规则量化方法[J]. 计算机工程, 2009, 35(15): 74-75,7.