CAJ | 학술논문

采用动态链地址法建立Hk 表，提出DLDHP算法，克服DHP算法不能给Ck 中的每个项集单独计数的缺点，在增加Hk 表的空间时也会删除链表中的结点空间，从而Hk 表不会占用较大的内存，最终直接由Hk 表得到频繁k-项集的集合Lk ，不用再次扫描数据库。实验结果表明：DLDHP算法是有效可行的。
채용동태련지지법건립Hk 표，제출DLDHP산법，극복DHP산법불능급Ck 중적매개항집단독계수적결점，재증가Hk 표적공간시야회산제련표중적결점공간，종이Hk 표불회점용교대적내존，최종직접유Hk 표득도빈번k-항집적집합Lk ，불용재차소묘수거고。실험결과표명：DLDHP산법시유효가행적。
DLDHP algorithm solves DHP algorithm’s problem that can’t separate count for each candidate itemsets by using dynamic chain address to create Hash table. Increasing the table space and also deleting the nodes make Hash table won’t occupy large space. Finally,frequent itemsets directly by the Hash table without scanning the da-tabase once again has been got. Examples show that DLDHP algorithm is effective and feasible.