CN106503092A - A kind of method using multidimensional technique construction Spatial Multi-Dimensional degree search tree - Google Patents
A kind of method using multidimensional technique construction Spatial Multi-Dimensional degree search tree Download PDFInfo
- Publication number
- CN106503092A CN106503092A CN201610892887.9A CN201610892887A CN106503092A CN 106503092 A CN106503092 A CN 106503092A CN 201610892887 A CN201610892887 A CN 201610892887A CN 106503092 A CN106503092 A CN 106503092A
- Authority
- CN
- China
- Prior art keywords
- node
- search
- tree
- search tree
- data
- Prior art date
- Legal status (The legal status 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 status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2453—Query optimisation
- G06F16/24534—Query rewriting; Transformation
- G06F16/24549—Run-time optimisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/22—Indexing; Data structures therefor; Storage structures
- G06F16/2228—Indexing structures
- G06F16/2246—Trees, e.g. B+trees
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2455—Query execution
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- Software Systems (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The present invention provides a kind of method using multidimensional technique construction Spatial Multi-Dimensional degree search tree, belongs to and is related to data-intensive operating technology field, and the present invention includes：Data space is created, is set and is accessed dimension；Search tree is created, node is created；Create index node list；Carry out traversal access, the optimum index path of inquiry；Return Search Results.The method can effectively lift the execution efficiency of search tree algorithm, so as to lift the performance and intelligence degree of related application.
Description
Technical field
The present invention relates to data-intensive operating technology field, more particularly to a kind of many using multidimensional technique construction space
The method of dimensional searches tree.
Background technology
Although the development of computer nowadays hardware technology is maked rapid progress, the development of basis of software algorithm is but stagnated all the time
Before not, it is difficult to obtain critical breakthrough.This aspect is because tradition based on mathematical algorithm and two-dimensionses design
System architecture substantially perfect, then be difficult to obtain important breakthrough, on the other hand also in that the mode of thinking of designer
Through tending to fixed pattern, particularly under the actual requirement that investment cycle day by day shortens, the research of basic algorithm has become not
Important again, and how the core design theory that biggest market return is only most products is quickly obtained using mature technology.
Just because of this, common on the market software and hardware unification is on the rise, and the product of different vendor is almost complete
Customized, either operational efficiency, resource occupation, or stability, difference functionally is carried out based on identical core algorithm storehouse
Different all reduce all the more, different be often only left UI designs and to third party's instrument customized on, thus result caused by institute
It is exactly the dog-eat-dog between product, the profit margin of software industry is successively decreased year by year.
The stagnation of Software Industry is additionally, since, causes hardware performance seriously superfluous, the development of whole computer industry
All receive and have a strong impact on, made slow progress because of the disappearance of basic theory with the new technology that intelligent, human nature turns to attraction, greatly
The application on a large scale such as data, cloud computing is difficult to give play to its real potentiality.
By taking traditional basic binary tree related algorithm as an example, although wherein most of algorithms possess simple structure, search speed
The characteristic such as fast is spent, but is building complication system, particularly in realization of the intelligent Application with modules such as machine learning, which is tied
The simple characteristic of structure has led to being significantly increased for search depth, so as to counteract its advantage in performance.But in multi-fork
On the structure of tree, traditional two-dimensionses multiway tree similarly cannot effectively solving traversal mode be single, searching algorithm is difficult to optimize
Defect, therefore its tree during neural Meta algorithm is built be difficult to play due effect, thus plant mode institute
The algorithm for constructing often has features such as search depth is big, flexibility is not enough, be difficult to safeguard, operation efficiency is low, thus pole
The big scale and processing speed that limit data.
Content of the invention
In order to solve above technical problem, the present invention propose one kind be easy to Dynamic Maintenance, be easily managed, recall precision high, and
And the Spatial Multi-Dimensional degree search tree constructing method suitable for large-scale data process, not only effectively can be lifted by the method
The search efficiency of complex data, and perfect by way of various dimensions can annotate association letter specific to intelligent computer
Breath group concept, and possess dimensional information management, filter capacity, information network associative ability etc..
Technical scheme comprises the steps：
Step 1：A data space is created and initialized, and sets several corresponding access dimensions as needed, such as：Three-dimensional space
Between access dimension, four-dimensional time and access dimension etc., higher dimension can access the information that more enriches.
Step 2：More than one search tree region is created as needed, such as：Data processing area（It is ranked up at scheduling algorithm
Reason）, logical process area, information storage area etc..
Step 3：The heart creates a node in the zone, as the ancestor node of multi-dimensional search tree, then saves in branch successively
Data are filled in point.
Step 4：Index node list is created for each region（Memory node data classification information and space coordinates scope, many
The index contents such as dimension coordinate scope）, using each region entirety as a node（Index node is under the node）It is associated as
The multi-dimensional search tree construction of one higher level, referred to as feature index tree.
Step 5：When needing to carry out traversal access, system finds the regional space for matching first in feature index tree
Orientation, then traverses the data classification of coupling again by index information in the region.
Step 6：In data classification, inquiry passes through if existing with the presence or absence of the optimum index path for meeting search condition
The path fast search necessary data, as do not existed, sends traversal access request.
Step 7：Return Search Results.
What the present invention was created is a kind of tree network structure based on space nodes, by a certain between different information
Specific attribute realizes interconnection, and which possesses several essential characteristics as follows：
Not only root node can be randomly assigned, and inquiry starting point, traversal direction etc. can be controlled
There is no the definition of proper direction in space tree, just do not exist yet fixed from top to bottom successively time
Go through rule.Therefore, after the tree-shaped Structure Creating in space, the request of any once traversal self-defined can specify a most probable
There is the traversal point of expected results, then with this point as the center of circle, successively outside traversal search.
There is invertibity feature, father node mutually can change as needed with child node
Because there is no the father and son's node definition under certain sense in space tree, therefore specified a certain node for search
After start node, the node just becomes the root node of this search naturally, thereafter with node that this is traveled through successively as the center of circle
Become for child node.
The observability of tree is controlled by way of various dimensions
When above-mentioned space tree is processed in the way of various dimensions, we can be different dimension sets its independent can
See content.During height data such as when we will retrieve certain 30 years old people 18 years old, time dimension at this time can be just used
The concept of degree.As the mode of operation of brain, instantly the most frequently used data only can be added the structure by space tree
Among, other non-frequently-used datas（Such as data many years ago）, it is possible to other dimensions are stored in, to reduce the data of current tree construction
Amount, to lift traversal efficiency, to save system resource.
Allow child node to return and be linked to root node, ultimately form closed-loop path
Same be not present fixed root node, there is no the characteristic of fixed-direction because of space tree, thus owns in which
Node may all there is a kind of interrelational form, the starting point is returned by different access paths finally.Due to the characteristic
Presence, the traversal request of space tree is often achieved in the way of multithreading, opened from a node that specifies
Begin, searching process is started simultaneously to each adjacent node, until process crosses, and till no longer there is new node.Pass through
Which, can be substantially improved the associated efficiency of the algorithm.
Intelligent anticipation is carried out by direction in space information
The Spatial Multi-Dimensional search tree difference maximum with two-dimensional search tree is to it comprises direction in space attribute, and traversal search is arrived
Up to the problem that can face a set direction after each space nodes, at this time except common traversal method successively it
Outward, also there is the optimum choice of various modes, and algorithm of these systems of selection just for intelligent quick traversal is realized providing
Necessary condition.
Repeat to select to optimize
When system records traversal search automatically, the node for being repeated to be chosen as correct result under given conditions is selected
Process, by these nodes are associated as a kind of relatively-stationary space tracking, when the searching request of condition is met afterwards, soon
Speed finds correct result, so as to provide the related algorithm of self-service correction class for intelligentized design, particularly machine learning behavior
Hold.
Set direction optimizes
By controlling traversal direction（Left and right, upper and lower etc.）To control the prioritizing selection behavior for traveling through, when searching request meets certain
When planting specified conditions, a certain traversal direction of prioritizing selection.
Counter push away process optimization
Without tree, and by most end end node, direct search to the result for needing, then pass through the result reverse search
The method of optimal path, we term it reverse thinking search method.
Association's model-based optimization
By a certain general-purpose attribute, be associated with similar node, such as the node remains with optimized search path record, then via
Similar node is quickly found out the process of Search Results, we term it association's pattern search method.
Spatial Dimension optimizes
Multi-dimensional search tree contains known all relevant search tree algorithms, such as binary tree algorithm etc..These algorithms are used as two dimension
Degree Index Algorithm, can be by compliant applications among each three-dimensional or more high-dimensional search tree algorithm.Likewise, such as bubbling
The related algorithm of the single dimensions such as sequence can also be used by compatibility.Therefore, during the realization of multidimensional search tree algorithm, and
Existing search tree optimized algorithm will not be repelled, conversely, more high-dimensional algorithm is realized meaning which can be with compatible more low dimensional
Under any one algorithm.
Specific embodiment
More detailed elaboration is carried out to present disclosure below：
Step 1：A data space is created and initialized, and sets multiple corresponding access dimensions as needed, such as：Three-dimensional space
Between access dimension, four-dimensional time and access dimension.
Step 2：One or more search tree regions are created as needed, such as：Data processing area, logical process area, information
Memory block.
Step 3：The heart creates a node in the zone, as the ancestor node of multi-dimensional search tree, then saves in branch successively
Data are filled in point.
Step 4：Index node list is created for each region, using each region entirety as node be associated as one higher
The multi-dimensional search tree construction of rank, referred to as feature index tree.
Step 5：When needing to carry out traversal access, system finds the regional space for matching first in feature index tree
Orientation, then traverses the data classification of coupling again by index information in the region.
Step 6：In data classification, inquiry passes through if existing with the presence or absence of the optimum index path for meeting search condition
The path fast search necessary data, as do not existed, sends traversal access request.
Step 7：Return Search Results.
Claims (3)
1. a kind of method using multidimensional technique construction Spatial Multi-Dimensional degree search tree, it is characterised in that
Comprise the steps：
Step 1：A data space is created and initialized, and sets several corresponding access dimensions as needed；
Step 2：More than one search tree region is created as needed；
Step 3：The heart creates a node in the zone, as the ancestor node of multi-dimensional search tree, then successively in branch node
Filling data；
Step 4：Index node list is created for each region, each region entirety is associated as a higher level as a node
Multi-dimensional search tree construction, referred to as feature index tree；
Step 5：When needing to carry out traversal access, system finds the regional space side for matching first in feature index tree
Position, then traverses the data classification of coupling again by index information in the region；
Step 6：In data classification, inquiry passes through the road if existing with the presence or absence of the optimum index path for meeting search condition
Footpath fast search necessary data, as do not existed, sends traversal access request；
Step 7：Return Search Results.
2. method according to claim 1, it is characterised in that accessing dimension includes：When three dimensions accesses dimension, the four-dimension
Between access dimension.
3. method according to claim 1, it is characterised in that search tree region includes：Data processing area, logical process
Area, information storage area.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610892887.9A CN106503092A (en) | 2016-10-13 | 2016-10-13 | A kind of method using multidimensional technique construction Spatial Multi-Dimensional degree search tree |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610892887.9A CN106503092A (en) | 2016-10-13 | 2016-10-13 | A kind of method using multidimensional technique construction Spatial Multi-Dimensional degree search tree |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106503092A true CN106503092A (en) | 2017-03-15 |
Family
ID=58294795
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610892887.9A Pending CN106503092A (en) | 2016-10-13 | 2016-10-13 | A kind of method using multidimensional technique construction Spatial Multi-Dimensional degree search tree |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106503092A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107239571A (en) * | 2017-06-28 | 2017-10-10 | 浪潮金融信息技术有限公司 | Index structuring method based on multidimensional data space technology |
| CN107678979A (en) * | 2017-10-19 | 2018-02-09 | 浪潮金融信息技术有限公司 | Date storage method and device |
| CN108038475A (en) * | 2017-12-29 | 2018-05-15 | 浪潮金融信息技术有限公司 | Facial image recognition method and device, computer-readable storage medium, terminal |
| CN108182242A (en) * | 2017-12-28 | 2018-06-19 | 湖南大学 | A kind of indexing means for the inquiry of magnanimity multi dimensional numerical data area |
| CN108229378A (en) * | 2017-12-29 | 2018-06-29 | 浪潮金融信息技术有限公司 | Face image data generation method and device, computer storage media, terminal |
| CN117290893A (en) * | 2023-09-25 | 2023-12-26 | 北京万里开源软件有限公司 | Database access method and system based on data tag |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6859455B1 (en) * | 1999-12-29 | 2005-02-22 | Nasser Yazdani | Method and apparatus for building and using multi-dimensional index trees for multi-dimensional data objects |
| US20050071349A1 (en) * | 2003-08-11 | 2005-03-31 | Jordan Philip Martin | Method and apparatus for accessing multidimensional data |
| CN1838124A (en) * | 2006-02-20 | 2006-09-27 | 南京联创科技股份有限公司 | Method for rapidly positioning grid + T tree index in mass data memory database |
| CN103339624A (en) * | 2010-12-14 | 2013-10-02 | 加利福尼亚大学董事会 | High efficiency prefix search algorithm supporting interactive, fuzzy search on geographical structured data |
| CN103377237A (en) * | 2012-04-27 | 2013-10-30 | 常州市图佳网络科技有限公司 | High dimensional data neighbor search method and fast approximate image search method |
-
2016
- 2016-10-13 CN CN201610892887.9A patent/CN106503092A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6859455B1 (en) * | 1999-12-29 | 2005-02-22 | Nasser Yazdani | Method and apparatus for building and using multi-dimensional index trees for multi-dimensional data objects |
| US20050071349A1 (en) * | 2003-08-11 | 2005-03-31 | Jordan Philip Martin | Method and apparatus for accessing multidimensional data |
| CN1838124A (en) * | 2006-02-20 | 2006-09-27 | 南京联创科技股份有限公司 | Method for rapidly positioning grid + T tree index in mass data memory database |
| CN103339624A (en) * | 2010-12-14 | 2013-10-02 | 加利福尼亚大学董事会 | High efficiency prefix search algorithm supporting interactive, fuzzy search on geographical structured data |
| CN103377237A (en) * | 2012-04-27 | 2013-10-30 | 常州市图佳网络科技有限公司 | High dimensional data neighbor search method and fast approximate image search method |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107239571A (en) * | 2017-06-28 | 2017-10-10 | 浪潮金融信息技术有限公司 | Index structuring method based on multidimensional data space technology |
| CN107239571B (en) * | 2017-06-28 | 2021-04-09 | 浪潮金融信息技术有限公司 | Index construction method based on multidimensional data space technology |
| CN107678979A (en) * | 2017-10-19 | 2018-02-09 | 浪潮金融信息技术有限公司 | Date storage method and device |
| CN108182242A (en) * | 2017-12-28 | 2018-06-19 | 湖南大学 | A kind of indexing means for the inquiry of magnanimity multi dimensional numerical data area |
| CN108038475A (en) * | 2017-12-29 | 2018-05-15 | 浪潮金融信息技术有限公司 | Facial image recognition method and device, computer-readable storage medium, terminal |
| CN108229378A (en) * | 2017-12-29 | 2018-06-29 | 浪潮金融信息技术有限公司 | Face image data generation method and device, computer storage media, terminal |
| CN117290893A (en) * | 2023-09-25 | 2023-12-26 | 北京万里开源软件有限公司 | Database access method and system based on data tag |
| CN117290893B (en) * | 2023-09-25 | 2024-06-11 | 北京万里开源软件有限公司 | Database access method and system based on data tag |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106503092A (en) | A kind of method using multidimensional technique construction Spatial Multi-Dimensional degree search tree | |
| CN102063486B (en) | Multi-dimensional data management-oriented cloud computing query processing method | |
| US7277876B2 (en) | Dynamic adaptive distributed computer system | |
| CN106933833B (en) | Method for quickly querying position information based on spatial index technology | |
| KR102005343B1 (en) | Partitioned space based spatial data object query processing apparatus and method, storage media storing the same | |
| CN104462351B (en) | A kind of data query model and method towards MapReduce patterns | |
| CN106095920B (en) | Distributed index method towards extensive High dimensional space data | |
| CN109840551B (en) | Method for optimizing random forest parameters for machine learning model training | |
| CN104731984B (en) | Automobile wheel hub surface sampling point R tree overflow node incremental clustering optimization method | |
| CN108717448B (en) | Key value pair storage-oriented range query filtering method and key value pair storage system | |
| CN103092992A (en) | Vector data preorder quadtree coding and indexing method based on Key / Value type NoSQL (Not only SQL) | |
| CN111177410A (en) | Knowledge graph storage and similarity retrieval method based on evolution R-tree | |
| Kadkhodaei et al. | A combination method for join ordering problem in relational databases using genetic algorithm and ant colony | |
| CN108389152A (en) | A kind of figure processing method and processing device of graph structure perception | |
| CN102831241A (en) | Dynamic index multi-target self-adaptive construction method for product reverse engineering data | |
| CN107480096B (en) | High-speed parallel computing method in large-scale group simulation | |
| CN118012602A (en) | Distributed cluster data equalization method based on balanced multi-way tree | |
| Nirmal et al. | Issues of K means clustering while migrating to map reduce paradigm with big data: A survey | |
| CN112148830A (en) | Semantic data storage and retrieval method and device based on maximum area grid | |
| Verma et al. | A novel framework for neural architecture search in the hill climbing domain | |
| Huang et al. | ACR-Tree: Constructing R-Trees Using Deep Reinforcement Learning | |
| Laskar et al. | A survey on VLSI floorplanning: its representation and modern approaches of optimization | |
| Ptiček et al. | MapReduce research on warehousing of big data | |
| Jiang et al. | Selection expressions for procedural modeling | |
| Tsitseklis et al. | Scalable community detection for complex data graphs via hyperbolic network embedding and graph databases |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB02 | Change of applicant information |
Address after: Wusong Industrial Park, Wuzhong Development District of Suzhou City, Jiangsu province 215100 Wusong Road No. 818 Applicant after: Tide Financial Information Technology Co Ltd Address before: 215104 Jiangsu province Changzhou Wuzhong Economic Development Zone the Creek Street Tower rhyme Road No. 178 Building 2 layer 1 Applicant before: Tide (Suzhou) Financial Technology Service Co., Ltd. |
|
| CB02 | Change of applicant information | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170315 |
|
| RJ01 | Rejection of invention patent application after publication |