首页> 外国专利> METHOD FOR MULTI-DIMENSIONAL HISTOGRAMS USING DENSE REGIONS AND RECORDING MEDIUM STORING PROGRAM FOR THE SAME

METHOD FOR MULTI-DIMENSIONAL HISTOGRAMS USING DENSE REGIONS AND RECORDING MEDIUM STORING PROGRAM FOR THE SAME

机译:使用密集区域和相同存储介质记录程序的多维历史记录方法

摘要

The present invention relates to a multi-dimensional histogram method and a recording medium with a program stored in the high density region, resolution Technical Problems to be a search for a high-density region that is present in the given data distribution and hot spot, on the basis of the search by the bucket hotspot information in a hierarchical configuration, a decrease in the estimation accuracy due to the high density region can be reduced effectively to provide a multi-dimensional histogram method using the high-density region. ; a multi-dimensional histogram method using the high-density region in accordance with the invention for this purpose is the number of data distribution and total buckets with histogram generation request from the external factor Based on the input received first step and the second step and the Hotspot information, navigate to run the second stage repeatedly to search for hotspots in a given distribution of the data in the first stage as a factor in generating the hotspots tree the selectivity of the query corresponding to the third phase and, also selected from outside the fourth step of receiving a query as a parameter with the estimated demand and, query received by the selectivity factor estimation request and received in the fourth step the discloses a multi-dimensional histogram method using the high-density region comprising a fifth step of estimating, using the hotspot tree generated in the third step.
机译:多维直方图方法和记录介质技术领域本发明涉及一种多维直方图方法和一种具有存储在高密度区域中的程序的记录介质,解决技术问题是寻找在给定数据分布和热点中存在的高密度区域,基于分层结构中的桶热点信息的搜索,可以有效地减少由于高密度区域而导致的估计精度的降低,从而提供使用高密度区域的多维直方图方法。 ;为此目的,根据本发明,使用高密度区域的多维直方图方法是根据外部输入的直方图生成请求生成的数据分布数量和总存储桶数,基于输入的第一步和第二步以及热点信息,导航到第二阶段重复运行,以在第一阶段以给定的数据分布搜索热点,这是生成热点树的一个因素,该查询树的选择性与第三阶段相对应,并且还从外部选择在第四步中,接收具有估计需求的查询作为参数,第四步,由选择性因子估计请求接收并在第四步中接收到的查询公开了一种使用高密度区域的多维直方图方法,该方法包括第五步:使用第三步中生成的热点树。

著录项

  • 公开/公告号KR101004631B1

    专利类型

  • 公开/公告日2011-01-04

    原文格式PDF

  • 申请/专利权人

    申请/专利号KR20080112543

  • 申请日2008-11-13

  • 分类号G06F17/00;

  • 国家 KR

  • 入库时间 2022-08-21 17:50:44

相似文献

  • 专利
  • 外文文献
  • 中文文献
获取专利

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号