摘要
为了解决使用Eclat算法在挖掘事务数或项目数较多的数据时,存在效率低、系统内存不足等问题,从候选集优化和剪枝策略两方面降低算法的时间复杂度,同时采用可以降低算法空间消耗的位存储结构,并基于此提出改进算法Eclat’。通过设计对比实验,进一步证明了改进算法的有效性,算法性能较原算法提高了20.37%。并将改进算法用于真实的手机用户数据上实现手机软件(Application,APP)的合理推荐。
To solve the problems that Eclat algorithm has low efficiency and is out of memory,when transaction number or itemsets number increases,the time complexity of the algorithm was reduced from two aspects including optimized candidate set and pruning strategy.At the same time,bit storage structure was used to reduce the space consumption of the algorithm.Based on this,improved algorithm Eclat’was proposed.The effectiveness of the improved algorithm is further proved by comparison experiment,increasing 20.37%compared with the original algorithm.The improved algorithm can be used to reasonably recommend APP to users in the real data of mobile phone users.
作者
崔馨月
孙静宇
CUI Xin-yue;SUN Jing-yu(College of Computer Science and Technology,Taiyuan University of Technology,Taiyuan 030024,China)
出处
《计算机工程与设计》
北大核心
2018年第4期1059-1063,1147,共6页
Computer Engineering and Design
关键词
Eclat算法
候选集优化
剪枝策略
有效性
手机用户
Eclat algorithm
optimized candidate set
pruning strategy
effectiveness
mobile phone users
