摘要
传统的挖掘算法Apriori是依据统计学中的数据显著性挖掘关联规则,需多次扫描数据库,效率较低,且忽视了数据显著性与价值性不匹配的问题。针对"大数据"下容易产生数量繁多但无效的关联规则,通过采用基于布尔矩阵挖掘关联规则的算法,只扫描一次数据库,得出布尔矩阵及相应的利润矩阵,随后根据"二八法则"设定对客户最具吸引力的"最小价值度",最终挖掘出高价值的关联规则,从而提高规则挖掘的效率及价值。
The traditional association rules mining algorithm Apriori is based on the significant mining association rules in statistics. The algorithm is Inefficient because it needs to repeatedly scan the database. And it also neglects the problem that the significance of data does not match the value. Oppositely it is easy to produce excessive but Invalid association rules. The paper uses the algorithm based on hooleanmatrix to mining association rules. This algorithm draws the bool- eanmatrix and the corresponding profit matrix by scanning the database only once. Then, it sets the most attractive minimal degree for the client based on the Pareto rule. At last, it mines the high - value degree association rules and improves effi- ciency and value.
出处
《科技管理研究》
CSSCI
北大核心
2014年第6期188-191,共4页
Science and Technology Management Research
关键词
关联规则
布尔矩阵
规则相关项布尔矩阵
平均利润矩阵
最小价值度
association rules
booleanmatrix
booleanmatrix related with the items of the rules
average profit matrix
minimal degree
