会员体验
专利管家(专利管理)
工作空间(专利管理)
风险监控(情报监控)
数据分析(专利分析)
侵权分析(诉讼无效)
联系我们
交流群
官方交流:
QQ群: 891211   
微信请扫码    >>>
现在联系顾问~
热词
    • 44. 发明专利
    • k-NEAREST NEIGHBOR SEARCH METHOD AND k-NEAREST NEIGHBOR SEARCH PROGRAM AND k-NEAREST NEIGHBOR SEARCH DEVICE
    • k-NEAREST邻居搜索方法和k最近邻域搜索程序和k最近邻域搜索设备
    • JP2009199151A
    • 2009-09-03
    • JP2008037362
    • 2008-02-19
    • Hitachi LtdHitachi Software Eng Co Ltd日立ソフトウエアエンジニアリング株式会社株式会社日立製作所
    • HAYASHI HIDEKIITO DAISUKETANIZAKI MASAAKIKIMURA KOJIKAJIYAMA HISANORI
    • G06F17/30G09B29/10
    • G06F17/30241G06F17/30539G06K9/6276
    • PROBLEM TO BE SOLVED: To suppress memory usage in search processing while shortening the retrieval time of a space database. SOLUTION: Provided is a k-nearest neighbor search method of searching for the high order k of nearest point data to an arbitrary point in a DBMS for creating a spatial index from multidimensional point data, comprising steps for setting a query point and the number of items of query to search conditions, determining which of a lowest branch and an intermediate branch of the spatial index a nearest region to the query point is, calculating a distance between the query point and the child region of the nearest region when the nearest region is determined to be the lowest branch from the determination of the lowest branch, storing the information of a divided region which has become the calculation target of region distance calculation, obtaining the nearest region, calculating a distance between the query point and point data included in the nearest region when the nearest region is determined to be the intermediate branch from the determination of the lowest branch, storing the point data which have become the calculation target of the distance calculation, finishing search processing when the search conditions are satisfied, and obtaining the records of the stored point data as a search result from the DBMS. COPYRIGHT: (C)2009,JPO&INPIT
    • 要解决的问题:在缩短空间数据库的检索时间的同时,抑制搜索处理中的存储器使用。 解决方案:提供一种搜索来自DBMS中的任意点的最近点数据的高阶k的k-最近邻搜索方法,用于从多维点数据创建空间索引,包括用于设置查询点和 查询到搜索条件的项目的数量,确定空间索引的最小分支和中间分支中的哪个最靠近查询点的区域是:计算最近区域的查询点和子区域之间的距离,当 最小区域被确定为从最低分支的确定中的最低分支,存储已经成为区域距离计算的计算目标的分割区域的信息,获得最近的区域,计算查询点和点数据之间的距离 当从最小分支的确定确定最近区域作为中间分支时,包含在最近的区域中,存储该点 已经成为距离计算的计算对象的数据,满足搜索条件时的整理搜索处理,以及从DBMS获取作为搜索结果的存储点数据的记录。 版权所有(C)2009,JPO&INPIT
    • 46. 发明专利
    • System, detection method and program
    • 系统,检测方法和程序
    • JP2007018349A
    • 2007-01-25
    • JP2005200293
    • 2005-07-08
    • Internatl Business Mach Corp インターナショナル・ビジネス・マシーンズ・コーポレーションInternational Business Maschines Corporation
    • YOSHIDA ITSUSEI
    • G06F17/30G06F19/00G06Q10/00G06Q30/02G06Q50/00
    • G06F17/30539G06Q30/0201Y10S707/99936
    • PROBLEM TO BE SOLVED: To seek a closed pattern in a second item space that results from the addition of a new item to a first item space, according to a closed pattern determined in the first item space. SOLUTION: A system is provided to detect the closed patterns, which are maximal sets of items, according to history information recorded an item set which is transaction processing object in each transaction. The system records at least one first closed pattern determined in a first item space. An additional item excluded from the first item space is added to any first closed pattern to generate an additional pattern, and a set of transactions including the entire additional pattern is compared with a set of transactions including the entire first closed pattern. If the compared sets are the same, the sum set of the first closed pattern and additional pattern is detected as a second closed pattern in a second item space. COPYRIGHT: (C)2007,JPO&INPIT
    • 要解决的问题:根据在第一项目空间中确定的关闭模式,寻找由第二项目空间中的新项目添加到第一项目空间中的封闭模式。 解决方案:提供一种系统,用于根据记录在每个交易中作为事务处理对象的项目集的历史信息来检测封闭图案,其是最大项目集合。 系统记录在第一项目空间中确定的至少一个第一封闭模式。 将从第一项目空间排除的附加项目添加到任何第一封闭模式以生成附加模式,并将包括整个附加模式的一组事务与包括整个第一封闭模式的一组事务进行比较。 如果比较集合相同,则第一关闭模式和附加模式的总和被检测为第二项目空间中的第二封闭模式。 版权所有(C)2007,JPO&INPIT
    • 47. 发明专利
    • Data analysis method and data analysis program
    • 数据分析方法和数据分析方案
    • JP2006338265A
    • 2006-12-14
    • JP2005161395
    • 2005-06-01
    • Fujitsu Ltd富士通株式会社
    • TSUDA HIDETAKASHIRAI EIDAI
    • G06F19/00G06F17/30
    • G06F17/30539
    • PROBLEM TO BE SOLVED: To efficiently extract a correlation among variables in a record group. SOLUTION: A record group sorting means of a computer sorts an object record group 1 by designated variables x, y and z. For example, a record group dividing/extracting means for performing sorting by a magnitude relation (x3>x1>x2...) of the variable x divides the sorted object record group 2 by a designated dividing method (for example, four divisions, eight divisions, etc., ) and extracts divided record groups G1 to Gm. A correlation calculating means calculates a correlation among designated variables for each of the divided record groups G1 to Gm. COPYRIGHT: (C)2007,JPO&INPIT
    • 要解决的问题:为了有效地提取记录组中的变量之间的相关性。

      解决方案:计算机的记录组排序装置通过指定的变量x,y和z对对象记录组1进行排序。 例如,用于通过变量x的大小关系(x3> x1> x2 ...)执行排序的记录组分割/提取装置通过指定的分割方法(例如,四个分区)来划分排序的对象记录组2, 八个部门等),并提取分割记录组G1至Gm。 相关计算装置计算每个分割记录组G1至Gm的指定变量之间的相关性。 版权所有(C)2007,JPO&INPIT