公开(公告)号：KR1020110066705A

公开(公告)日：2011-06-17

申请号：KR1020090123461

申请日：2009-12-11

Applicant: 한국전자통신연구원 ,  전북대학교산학협력단

Inventor： 최현화 ,  김영창 ,  이미영 ,  엄정호 ,  장재우

IPC: G06F17/40 ,  G06F17/30

CPC classification number: G06F17/30784 ,  G06F17/3033 ,  G06F17/30592

Abstract: PURPOSE: An index and a searching method of high dimensional feature vector data of a Dual Min-LSH algorithm is provided to guarantee QoS to a user by keeping the accuracy over a determined level and reducing the search time. CONSTITUTION: A hash table comprises a Low hash table(226) and a High hash table(228). The Low hash table calculates and stores a MinHash value of higher ordered data on the basis of permutations of arbitrary numbers. The High hash table calculates and stores the MinHash value of higher ordered data on the basis of more permutations including the permutations of the Low hash table for data within a bucket exceeding the number of entries of a threshold in the Low hash table.

Abstract translation: 目的：提供双Min-LSH算法的高维特征向量数据的索引和搜索方法，以通过将精度保持在确定的水平并减少搜索时间来保证用户的QoS。 构成：散列表包括低散列表（226）和高散列表（228）。 低散列表基于任意数字的排列来计算并存储较高数据的MinHash值。 高散列表基于更多排列来计算和存储较高排序数据的MinHash值，其中包括对于桶内的数据的低散列表的排列超过低散列表中阈值的条目数。

公开(公告)号：KR100446639B1

公开(公告)日：2004-09-04

申请号：KR1020010042482

申请日：2001-07-13

Applicant: 한국전자통신연구원

Inventor： 박수준 ,  장재우 ,  김현진 ,  박성희 ,  장명길 ,  박상규 ,  한성근

IPC: G06F17/30

Abstract: PURPOSE: A cell based high dimensional data indexing system and method is provided to index high dimensional data based on a cell for preventing a lowering of a search efficiency in searching for high dimensional data. CONSTITUTION: The method comprises several steps. First, an N dimensional feature vector is extracted from a multimedia object via a feature vector extractor(801). A distance signature is generated via a signature generation module by using a distance between a signature on the feature vector and a cell center(802). One signature is generated by concatenating the feature vector signature and the distance signature(803), and then is stored at a signature database(804). At the same time, the feature vector is stored at a feature vector database(805). A user can perform a search operation on the stored feature vectors by using various queries like a point query, a range query or k-nearest query(806).

Abstract translation: 目的：提供一种基于单元格的高维数据索引系统和方法，用于基于单元索引高维数据，以防止在搜索高维数据时降低搜索效率。 构成：该方法包括几个步骤。 首先，经由特征向量提取器（801）从多媒体对象提取N维特征向量。 经由签名生成模块通过使用特征向量上的签名与小区中心之间的距离来生成距离签名（802）。 通过连接特征向量签名和距离签名生成一个签名（803），然后存储在签名数据库（804）。 同时，特征矢量被存储在特征矢量数据库（805）。 用户可以通过使用诸如点查询，范围查询或k-最近查询（806）的各种查询来对所存储的特征向量执行搜索操作。

公开(公告)号：KR101153966B1

公开(公告)日：2012-06-08

申请号：KR1020080131384

申请日：2008-12-22

Applicant: 한국전자통신연구원 ,  전북대학교산학협력단

Inventor： 이훈순 ,  이미영 ,  엄정호 ,  장재우

IPC: G06F17/30 ,  G06F17/00

Abstract: 본발명은고차원데이터의색인/검색하는시스템및 그방법에관한것이다. 고차원데이터의검색시스템및 그방법에서는고차원데이터를구간으로나누어클러스터링한후, 각구간에시그니처를부여하고, 고차원데이터의삽입시데이터를인접셀에도중복삽입함으로써, 사용자에게높은정확도를보장하는검색성능이제공된다.

公开(公告)号：KR1020100072855A

公开(公告)日：2010-07-01

申请号：KR1020080131384

申请日：2008-12-22

Applicant: 한국전자통신연구원 ,  전북대학교산학협력단

Inventor： 이훈순 ,  이미영 ,  엄정호 ,  장재우

IPC: G06F17/30 ,  G06F17/00

CPC classification number: G06F17/3002 ,  G06F17/10 ,  G06F17/3033

Abstract: PURPOSE: An index/search system of higher-order data and a method thereof are provided to divide higher-order data into section unit and express the data with a signature, thereby obtaining clustering effect. CONSTITUTION: A hashing operation module(220) obtains a cell. One feature vector extracted from highly dimensional data belongs to the cell. The hashing operation module generates signature for display of the cell. The hashing operation module drives hashing structure. In the hashing structure, the signature is divided in dimensional unit into different. The divided signature is stored in a plurality of indexes. A storage unit(240) stores an algorithm for inserting and searching higher-order data.

Abstract translation: 目的：提供高阶数据的索引/搜索系统及其方法，将高阶数据划分为单元，并用签名表示数据，从而获得聚类效果。 构成：散列运算模块（220）获取单元。 从高维数据中提取的一个特征向量属于该单元。 哈希操作模块生成签名以显示单元。 散列运算模块驱动散列结构。 在哈希结构中，签名被划分成不同的维度单位。 划分的签名存储在多个索引中。 存储单元（240）存储用于插入和搜索高阶数据的算法。

公开(公告)号：KR1020030006638A

公开(公告)日：2003-01-23

申请号：KR1020010042482

申请日：2001-07-13

Applicant: 한국전자통신연구원

Inventor： 박수준 ,  장재우 ,  김현진 ,  박성희 ,  장명길 ,  박상규 ,  한성근

IPC: G06F17/30

Abstract: PURPOSE: A cell based high dimensional data indexing system and method is provided to index high dimensional data based on a cell for preventing a lowering of a search efficiency in searching for high dimensional data. CONSTITUTION: The method comprises several steps. First, an N dimensional feature vector is extracted from a multimedia object via a feature vector extractor(801). A distance signature is generated via a signature generation module by using a distance between a signature on the feature vector and a cell center(802). One signature is generated by concatenating the feature vector signature and the distance signature(803), and then is stored at a signature database(804). At the same time, the feature vector is stored at a feature vector database(805). A user can perform a search operation on the stored feature vectors by using various queries like a point query, a range query or k-nearest query(806).

Abstract translation: 目的：提供一种基于单元的高维数据索引系统和方法，用于根据单元索引高维度数据，以防止在搜索高维数据时降低搜索效率。 构成：该方法包括几个步骤。 首先，经由特征向量提取器（801）从多媒体对象提取N维特征向量。 通过使用特征向量的签名与小区中心之间的距离，通过签名生成模块生成距离签名（802）。 通过连接特征向量签名和距离签名（803）生成一个签名，然后存储在签名数据库（804）。 同时，特征向量存储在特征向量数据库（805）。 用户可以通过使用诸如点查询，范围查询或k-最近查询（806）的各种查询对所存储的特征向量执行搜索操作。