WO2015143985A1 - Procédé et appareil de détermination de vecteur de résultat - Google Patents

Procédé et appareil de détermination de vecteur de résultat Download PDF

Info

Publication number
WO2015143985A1
WO2015143985A1 PCT/CN2015/073794 CN2015073794W WO2015143985A1 WO 2015143985 A1 WO2015143985 A1 WO 2015143985A1 CN 2015073794 W CN2015073794 W CN 2015073794W WO 2015143985 A1 WO2015143985 A1 WO 2015143985A1
Authority
WO
WIPO (PCT)
Prior art keywords
matrix
vector
column
query
matrix block
Prior art date
Application number
PCT/CN2015/073794
Other languages
English (en)
Chinese (zh)
Inventor
颜友亮
郑荣福
Original Assignee
华为技术有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 华为技术有限公司 filed Critical 华为技术有限公司
Publication of WO2015143985A1 publication Critical patent/WO2015143985A1/fr

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/33—Querying
    • G06F16/3331—Query processing
    • G06F16/334—Query execution
    • G06F16/3347—Query execution using vector based model

Definitions

  • Directed graphs and undirected graphs can be used to represent different types of relationships, for example, vertices are used to represent accounts in a network platform, and edges in undirected graphs can be used to indicate that there is a friend relationship between accounts, directed graphs. The side in can be used to indicate that one account is interested in another.
  • the method further includes:
  • a determining module configured to obtain a target vertex corresponding to the query condition, and determine a query vector according to a position of the corresponding row or column in the adjacency matrix after the clustering process
  • the determining module is configured to:
  • the query vector is determined according to the position of the corresponding row or column in the adjacency matrix after the clustering process of the target vertex.
  • the pre-processing module is further configured to:
  • FIG. 2 is a schematic diagram of clustering processing and blocking processing according to an embodiment of the present invention.
  • Step 1 According to the established Bloom filter, determine the position of the corresponding row or column in the adjacency matrix of the target vertex after the clustering process.
  • the matrix block related to the vector segment operation is a matrix block that needs to be operated (multiplied or transposed) after the clustering process and the query vector are operated.
  • the corresponding processing may be: obtaining a non-zero matrix block in the determined matrix block row by using the second linked list array. Specifically, the data storage address of each non-zero matrix block in the matrix block row may be queried in the second linked list array, and then the data of the corresponding matrix block is read in the queried data storage address.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • Databases & Information Systems (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

La présente invention concerne le domaine technique des ordinateurs. La présente invention concerne plus précisément un procédé et un appareil de détermination de vecteur de résultat. Le procédé comprend les étapes consistant à : obtenir un sommet cible correspondant à une condition d'interrogation et déterminer un vecteur d'interrogation en fonction de la position du sommet cible dans une ligne ou une colonne correspondante dans une matrice adjacente sur laquelle un traitement de regroupement est effectué ; diviser le vecteur d'interrogation en de multiples segments de vecteur d'après une longueur latérale prédéfinie ; dans de multiples blocs de matrice, sélectionner, en fonction d'une position d'un segment de vecteur comportant une valeur non nulle dans le vecteur d'interrogation, un bloc de matrice qui comporte une valeur non nulle et est associé au calcul du segment de vecteur comportant la valeur non nulle ; puis effectuer un calcul en fonction du bloc de matrice sélectionné et du segment de vecteur comportant la valeur non nulle et déterminer un vecteur de résultat. La présente invention peut être utilisée pour accroître l'efficacité d'interrogation.
PCT/CN2015/073794 2014-03-24 2015-03-06 Procédé et appareil de détermination de vecteur de résultat WO2015143985A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201410112336.7A CN104951442B (zh) 2014-03-24 2014-03-24 一种确定结果向量的方法和装置
CN201410112336.7 2014-03-24

Publications (1)

Publication Number Publication Date
WO2015143985A1 true WO2015143985A1 (fr) 2015-10-01

Family

ID=54166105

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2015/073794 WO2015143985A1 (fr) 2014-03-24 2015-03-06 Procédé et appareil de détermination de vecteur de résultat

Country Status (2)

Country Link
CN (1) CN104951442B (fr)
WO (1) WO2015143985A1 (fr)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112560273A (zh) * 2020-12-21 2021-03-26 北京轩宇信息技术有限公司 面向数据流模型的模型组件执行顺序确定方法及装置

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108664447B (zh) * 2017-03-31 2022-05-17 华为技术有限公司 一种矩阵与矢量的乘法运算方法及装置
CN108062551A (zh) * 2017-06-28 2018-05-22 浙江大学 一种基于邻接矩阵的图特征提取系统、图分类系统和方法
CN108520275A (zh) * 2017-06-28 2018-09-11 浙江大学 一种基于邻接矩阵的连接信息规整系统、图特征提取系统、图分类系统和方法
CN111460108A (zh) * 2019-01-22 2020-07-28 北京京东尚科信息技术有限公司 智能应答的方法和装置
CN111798363B (zh) * 2020-07-06 2024-06-04 格兰菲智能科技有限公司 图形处理器

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20040230409A1 (en) * 1998-08-31 2004-11-18 Nutech Solutions, Inc. Method for performing social computation
US20110307685A1 (en) * 2010-06-11 2011-12-15 Song William S Processor for Large Graph Algorithm Computations and Matrix Operations
US20120143932A1 (en) * 2010-12-06 2012-06-07 International Business Machines Corporation Data Structure For Tiling And Packetizing A Sparse Matrix

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP2406787B1 (fr) * 2009-03-11 2014-05-14 Google, Inc. Classification audio pour la récupération d'information utilisant des éléments épars
CN101986299A (zh) * 2010-10-28 2011-03-16 浙江大学 基于超图的多任务个性化网络服务方法
WO2013124522A1 (fr) * 2012-02-22 2013-08-29 Nokia Corporation Système et procédé pour fournir une prédiction pour commander un système
CN103336758B (zh) * 2013-06-29 2016-06-01 中国科学院软件研究所 一种采用带有局部信息的压缩稀疏行的稀疏矩阵存储方法及基于该方法的SpMV实现方法

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20040230409A1 (en) * 1998-08-31 2004-11-18 Nutech Solutions, Inc. Method for performing social computation
US20110307685A1 (en) * 2010-06-11 2011-12-15 Song William S Processor for Large Graph Algorithm Computations and Matrix Operations
US20120143932A1 (en) * 2010-12-06 2012-06-07 International Business Machines Corporation Data Structure For Tiling And Packetizing A Sparse Matrix

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112560273A (zh) * 2020-12-21 2021-03-26 北京轩宇信息技术有限公司 面向数据流模型的模型组件执行顺序确定方法及装置
CN112560273B (zh) * 2020-12-21 2023-11-10 北京轩宇信息技术有限公司 面向数据流模型的模型组件执行顺序确定方法及装置

Also Published As

Publication number Publication date
CN104951442B (zh) 2018-09-07
CN104951442A (zh) 2015-09-30

Similar Documents

Publication Publication Date Title
WO2015143985A1 (fr) Procédé et appareil de détermination de vecteur de résultat
US11888702B2 (en) Intelligent analytic cloud provisioning
CN109993299B (zh) 数据训练方法及装置、存储介质、电子装置
Wang et al. Performance prediction for apache spark platform
CN106980623B (zh) 一种数据模型的确定方法及装置
WO2019056681A1 (fr) Procédé, dispositif, appareil terminal et support de stockage pour surveillance de données en temps réel
JP2019530068A5 (fr)
JP2013529814A5 (fr)
AU2016270327B2 (en) Methods and systems for selectively retrieving data to provide a limited dataset for incorporation into a pivot table
JP2018190450A (ja) 基数推定を介した結合パスの効率的な判定
WO2019085709A1 (fr) Procédé et système de regroupement appliqués à un réseau neuronal convolutif
CN105373517A (zh) 基于Spark的分布式稠密矩阵求逆并行化运算方法
CN104239470A (zh) 一种面向分布式环境的空间数据复合处理系统和方法
CN108182204A (zh) 基于房产交易多维度数据的数据查询的处理方法及装置
CN107341193B (zh) 路网中移动对象查询方法
Wang Analysis on the construction of computer data processing mode based on era of big data
US10013393B2 (en) Parallel computer system, parallel computing method, and program storage medium
US9626397B2 (en) Discounted future value operations on a massively parallel processing system and methods thereof
CN110888909A (zh) 一种评估内容的数据统计处理方法及装置
CN110032445B (zh) 大数据聚集计算方法及装置
Stojanovic et al. A hybrid MPI+ OpenMP application for processing big trajectory data
CN111399838A (zh) 一种基于SparkSQL和物化视图的数据建模方法及装置
Rattanaopas et al. A comparison of orc-compress performance with big data workload on virtualization
CN111143328A (zh) 一种敏捷商业智能数据构建方法、系统、设备、存储介质
Shah et al. A systematic review of in-memory database over multi-tenancy

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 15768420

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 15768420

Country of ref document: EP

Kind code of ref document: A1