混合属性网络多维多层关联数据智能挖掘算法

Intelligent mining algorithm for multi-dimensional and multi-layer association data in hybrid attribute networks

针对传统关联数据挖掘算法,强项集挖掘后产生大量候选项集,导致挖掘耗时长、挖掘精度低等问题,提出一种混合属性网络多维多层关联数据智能挖掘算法(Multidimensional Multilayer Associative Data Intelligent Mining Algorithm,MMAD-IM)。计算混合属性网络中随机数据到簇中心的距离,将目标数据分配到距离簇中心最近的簇中,使簇中心固定,完成混合属性网络数据的聚类分析。从聚类完成的数据中提取出有效的基本频繁向量,同时计算数据的候选项集,对哈希表进行扫描,利用改进Apriori算法完成强项集挖掘。以此为基础构建空间关系,获取近似区域与近似点之间的距离,形成待挖掘数据并计算数据的隶属度数值,完成智能挖掘。实验结果表明,所提算法具有较好的数据聚类效果,强项集挖掘后剩余的候选项集数量较少,整体数据挖掘耗时远低于传统算法,挖掘精度高达90%。

For traditional association data mining algorithms,strong itemset mining generates a large number of candidate itemsets,leading to long mining time,low mining accuracy,and other problems.A hybrid attribute network multi-dimensional multi-layer association data intelligent mining algorithm is proposed.The distance from the random data in the mixed attribute network to the cluster center is calculated,and the target data is assigned to the cluster closest to the cluster center,so that the cluster center is fixed and the clustering analysis of the mixed attribute network data is completed.The effective basic frequent vectors are extracted from the data completed by clustering,and the candidate itemsets of the data are also calculated,the hash table is scanned,and the strong itemset mining is completed by using the improved Apriori algorithm.Based on this,the spatial relationship is constructed,the distance between the approximate region and the approximate point is obtained,the data set to be mined is formed and the affiliation value of the data is calculated to complete the intelligent mining.The experimental results show that the proposed algorithm has better data clustering effect,the number of remaining candidate itemsets after strong itemset mining is smaller,the overall data mining time is much lower than the traditional method,and the mining accuracy is up to 90%.