CN102880629B - Accelerating query method of probabilistic database - Google Patents
Accelerating query method of probabilistic database Download PDFInfo
- Publication number
- CN102880629B CN102880629B CN201210209270.4A CN201210209270A CN102880629B CN 102880629 B CN102880629 B CN 102880629B CN 201210209270 A CN201210209270 A CN 201210209270A CN 102880629 B CN102880629 B CN 102880629B
- Authority
- CN
- China
- Prior art keywords
- calculation results
- query
- intermediate calculation
- key word
- sub
- 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.)
- Active
Links
Abstract
The invention provides an accelerating query method for a probabilistic database, which comprises the following steps: obtaining a probabilistic query keyword, wherein the probabilistic query keyword comprises a joint value of a plurality of variables; according to the probabilistic query keyword, querying in a cache; storing an intermediate operation result in the previous probabilistic query process into the cache; and if the intermediate operation result matched with the joint value of the variables is in the presence in the cache, taking the intermediate operation result as the query result of the probabilistic query. In a mode of storing the intermediate operation result in the cache, whether the matched intermediate operation result is in the presence in the cache is firstly queried when a new probabilistic query is made for the probabilistic database, and if so, the intermediate operation result is directly used as the query result of the probabilistic query so as to optimize the query speed of the probabilistic database.
Description
Technical field
The present invention relates to database technology, particularly relate to a kind of probability database speedup query method.
Background technology
Probability database, for storing uncertain data, is called probabilistic query to the inquiry that probability database carries out.The value of each variable of probability database is not what determine, and also there is the restriction of different conditions to variable, and each condition is also uncertain.Such as be arranged on the temperature in multiple wireless senser Real-time Obtaining external worlds in same place, humidity, air pressure and intensity of illumination, because physical equipment improves and limited electric power supply and network delay not, the temperature in the same place of synchronization that each wireless senser obtains, humidity, air pressure and intensity of illumination may be different.According to the information that each wireless senser is collected, a probability database can be built, the variable of this probability database is temperature, humidity, air pressure and intensity of illumination, preserves multiple value and and the probability of each value of each variable in the same place of synchronization in probability database.Owing to association may be there is between each variable, therefore in probability database, also preserve the joint distribution of associated variable, the probability of the multiple variable associating values namely associated.For example, association is there is in temperature with humidity, association is there is in temperature with intensity of illumination, association is there is in humidity with air pressure, when probabilistic query relate to intensity of illumination and air pressure associate value time, then need the joint distribution of the joint distribution according to temperature and humidity, temperature and intensity of illumination, humidity and the joint distribution of air pressure to calculate the probability associating value of intensity of illumination and air pressure.
Can find out, probabilistic query needs can obtain Query Result through very complicated matrix operation usually, causes the speed of probabilistic query slowly.
Summary of the invention
The embodiment of the present invention provides a kind of probability database speedup query method, to improve the inquiry velocity to probability database.
The embodiment of the present invention provides a kind of probability database speedup query method, comprising:
Acquisition probability key word of the inquiry, described probabilistic query key word comprises the associating value of multiple variable;
Inquire about in the buffer according to described probabilistic query key word, the intermediate calculation results of probabilistic query process before preserving in described buffer memory, if exist in described buffer memory with described multiple variable combine the intermediate calculation results that value mates, then using the Query Result of described intermediate calculation results as described probabilistic query.
The method that the embodiment of the present invention provides, by storing the mode of intermediate calculation results in the buffer, when doing new probabilistic query to probability database, first whether there is the intermediate calculation results of coupling in query caching, if exist, direct using the Query Result of this intermediate calculation results as this probabilistic query, thus optimize the inquiry velocity to probability database.
Accompanying drawing explanation
Fig. 1 is probability database speedup query method first embodiment process flow diagram of the present invention;
Fig. 2 is probability database speedup query method second embodiment process flow diagram of the present invention;
Fig. 3 is the probability database structural drawing utilizing Bayesian network to represent;
Fig. 4 is the associating tree structure diagram built probability database in Fig. 3;
Fig. 5 carries out the partition structure figure after recurrence division to the tree of combining in Fig. 4;
Fig. 6 is the tree index structural drawing of combining the probability database that the partition structure figure of tree sets up in Figure 5;
Fig. 7 is the processing flow chart of intermediate calculation results in buffer memory.
Embodiment
Fig. 1 is probability database speedup query method first embodiment process flow diagram of the present invention, as shown in Figure 1, the probability database speedup query method that the embodiment of the present invention provides can be performed by Database Modeling instrument, and this method can adopt the form of software to realize, and the method comprises:
Step S100, acquisition probability key word of the inquiry, described probabilistic query key word comprises the associating value of multiple variable;
Probabilistic query is done to probability database, in probability database, stores the marginal distribution of random occurrence, and the joint distribution of multiple association random occurrence.Wherein, random occurrence refers to that variable gets the probability that a certain fixing value and this variable equal this value.Marginal distribution refers to the probability matrix that the value of this variable and associated condition form, and joint distribution refers to the probability matrix that multiple different values that there is the variable of incidence relation each other form.Acquisition probability key word of the inquiry, wherein probabilistic query key word comprises the associating value of multiple variable.Preferably, probabilistic query key word can be query formulation, such as, stores the marginal distribution of A, B marginal distribution, C marginal distribution in probability database respectively, and the joint distribution of A, B, C, probabilistic query key word is Q(A, B), namely inquiry obtains the associating value of A and B.
Step S102, inquire about in the buffer according to described probabilistic query key word, the intermediate calculation results of probabilistic query process before preserving in described buffer memory, if exist in described buffer memory with described multiple variable combine the intermediate calculation results that value mates, then using the Query Result of described intermediate calculation results as described probabilistic query.
When doing probabilistic query to probability database, first whether exist in query caching with multiple variable in probabilistic query key word combine the intermediate calculation results that value matches, if exist with multiple variable combine the intermediate calculation results that value matches, then using the Query Result of this intermediate calculation results as this probabilistic query.
The technical scheme that the present embodiment provides is by storing the mode of intermediate calculation results in the buffer, when doing new probabilistic query to probability database, first whether there is the intermediate calculation results of coupling in query caching, if exist, direct using the Query Result of this intermediate calculation results as this probabilistic query, thus optimize the inquiry velocity to probability database.
Fig. 2 is probability database speedup query method second embodiment process flow diagram of the present invention, and as shown in Figure 2, the method comprises:
Step S201, sets up the tree index corresponding with described probability database;
Step S202, acquisition probability key word of the inquiry, described probabilistic query key word comprises the associating value of multiple variable;
Step S203, inquires about in the buffer according to described probabilistic query key word, judge whether to exist in buffer memory with multiple variable combine the intermediate calculation results that value mates, if then perform step S204, then perform step S205 if not;
Step S204, using the Query Result of described intermediate calculation results as described probabilistic query, flow process terminates.
Step S205, decomposes described probabilistic query key word according to described tree index, obtains multiple sub-key word;
Step S206, inquires about in the buffer respectively according to described multiple sub-key word, judges whether there is the intermediate calculation results with described each sub-keyword match in described buffer memory, if then perform step S207, then performs step S208 if not;
Step S207, using the intermediate calculation results of described and described each sub-keyword match as the Query Result of each sub-key word, performs step S209;
Step S208, according to described at least one sub-key word with described tree index in inquire about;
If there is not the intermediate calculation results with at least one sub-keyword match in described buffer memory, then according at least one sub-key word with described tree index in inquire about;
Step S209, according to the Query Result of described each sub-key word, calculates the Query Result of described probabilistic query.
In a preferred embodiment, set up index by the mode of combining tree to probability database, concrete grammar is as follows:
Fig. 3 is the probability database structural drawing utilizing Bayesian network to represent, as shown in Figure 3, this probability database comprises random occurrence X1 ~ X12, wherein X1 ~ X2 represents that the value of variables A is x1 or x2, X3 ~ X5 represents that the value of variable B is x3 or x4 or x5, X6 ~ X7 represents that the value of variable C is x6 or x7, X8 ~ X9 represents that the value of variables D is x8 or x9, X10 ~ X11 represents that the value of variable E is x10 or x11, X12 represents that the value of variable F is x12, R1 ~ R5 and G1 ~ G6 represents the qualifications to random occurrence respectively, this qualifications makes to there is certain incidence relation between different random event, such as, under G1 condition, X value is x1 or value is between x2, the incidence relation that there is mutual exclusion under G1 condition between X value x1 and x2, namely X can not value be x1 and x2 simultaneously.
Fig. 4 is the associating tree structure diagram built probability database in Fig. 3, and as shown in Figure 4, wherein border circular areas represents the association node of this index, and rectangular area represents the partition node of this index.By combining tree (junction tree), index is set up to probability database.Wherein association node is the marginal distribution of random occurrence, i.e. the probability matrix that forms of condition and random occurrence, and partition node is the probability matrix of the condition of association random occurrence and random occurrence.
Fig. 5 carries out the partition structure figure after recurrence division to the tree of combining in Fig. 4, Fig. 6 is the tree index structural drawing of combining the probability database that the partition structure figure of tree sets up in Figure 5, as shown in Figure 5 and Figure 6, according to the associating tree construction of probability database, tree index is set up to probability database.First be divided into different unit to combining tree according to the size of association node and partition node, each unit stores the joint distribution of all partition node in this unit around all association nodes.Carry out secondary division according to the size of unit further, set up the index structure of probability database, this index structure comprises root node, father node and child node.Preferably, to combine tree divide time, each size of node is no more than 4K.Wherein root node comprises the mark of associating tree.Father node under root node comprises the mark of random occurrence in the joint distribution of all partition node around all association nodes and its child node.Child node under father node comprises the marginal distribution of random occurrence, and associates the partition node of random occurrence corresponding to this child node.
Particularly, user is when operating probability database, and can record the query note to probability database in the inquiry log of system, query note comprises customer identification number, query time, query contents, the information such as Query Result.According to the inquiry log of probability database, the intermediate calculation results that user produces in query script can be obtained, this intermediate calculation results is stored in buffer memory.Acquisition probability key word of the inquiry, this probabilistic query key word comprises the associating value of multiple variable, according to described probabilistic query keyword search buffer memory, if exist with multiple variable combine the intermediate calculation results that value mates, then direct using the Query Result of this intermediate calculation results as this probabilistic query.
If there is not the intermediate calculation results matched, then travel through the tree index of this probability database, according to tree index, probabilistic query key word is decomposed, obtain multiple sub-key word.One by one multiple sub-key word is inquired about in the buffer respectively, judges in buffer memory, whether to there is the intermediate calculation results with each sub-keyword match, if then using with the intermediate calculation results of each sub-keyword match Query Result as each sub-key word.If certain sub-key word does not exist the intermediate calculation results matched in the buffer, then this sub-key word is inquired about in tree index, obtain the Query Result of this sub-key word, and this Query Result is saved in buffer memory as intermediate calculation results.Finally, according to the Query Result of each sub-key word, the Query Result of probabilistic query is calculated.
Due in the middle of probability database, be not separate between each variable, but there is different incidence relations.Therefore, by setting up index at probability database, and intermediate calculation results is stored in the buffer, when user again query caching time, when this probabilistic query key word or sub-key word identical with certain intermediate calculation results in buffer memory time, could directly this intermediate calculation results of acquisition as the Query Result of this probabilistic query.
When probabilistic query key word or sub-key word and intermediate calculation results incomplete same time, then directly can not obtain the Query Result of this probabilistic query.Such as, when storing the intermediate calculation results of random occurrence X2 and random occurrence X3 in buffer memory, wherein the probability of random occurrence X2 Query Result is the probability of P (X2), random occurrence X3 Query Result is P (X3).User makes new probabilistic query to probability database, need obtain Q(X2, X3) value result, owing to there is incidence relation between X2 and X3, therefore the Query Result of (X2, X3) can not directly by P (X2), P (X3) directly obtains P (X2, X3).Even if therefore store P (X2) and P (X3) in intermediate database all at buffer memory, can not directly calculate P (X2, X3).
A probabilistic query key word split, according to the incidence relation in probability database between variable, by introducing the mode of bridging amount, is multiple sub-key words, then according to sub-keyword search buffer memory by the index of traversal probability database.
Sub-key word refers to, is multiple sub-key word that is mutually related, can obtains the Query Result of former probabilistic query key word according to the Query Result of all sub-key words by probabilistic query key word split.Bridging amount refers to the condition that sub-for difference key word can be connected for probabilistic query key word.
Such as former probabilistic query key word is Q(X2, X3) there is not the intermediate calculation results matched with it in the buffer, by introducing the bridging amount R1 all relevant to random occurrence X2 and random occurrence X3, be multiple sub-key words by former probabilistic query key word split.Bridging amount refers to the condition that can connect random occurrence X2 and random occurrence X3 in queries.R1 is the bridging amount between random occurrence X2 and random occurrence X3, then by former probabilistic query key word Q(X2, X3) split into Q(X2, R1) and Q(R1, X3).
By former probabilistic query key word Q(X2, X3) Q(X2, R1 is split into) and Q(R1, X3) after, according to Q(X2, R1) and Q(R1, X3) query caching, if there is the intermediate calculation results matched with it in buffer memory, then using the Query Result of this intermediate calculation results as sub-key word, if do not exist, continue to search tree index, after obtaining the Query Result of all sub-key words, according to formula:
(X, Y are random occurrence, and S is bridging amount) obtains the Query Result of probabilistic query.
Again such as, probabilistic query key word is Q (X1, X12), and known X1 and X12 of traversal index is distributed in child node I1 and child node I3, and child node I1 is connected by child node I2 with child node I3.Therefore be multiple sub-key words by this query decomposition: Q (X1, G2), Q (G2, G4), Q (G4, X12), wherein G2 and G4 is bridging amount.Suppose, not by intermediate database, directly to inquire about the index of probability database, because Q (G2, G4) directly can be obtained by index node I2, so do not need to be decomposed into sub-key word again.
And Q (X1, G2) and Q (G4, X12) also needs to continue to decompose, Q (X1, G2) is decomposed into sub-key word Q (X1, R1) and Q (R1, G2); Q (G4, X12) is decomposed into sub-key word Q (G4, G5) and Q (G5, X12); Q (X1, R1) needs to continue to be decomposed into sub-key word Q (X1, G1) and Q (G1, R1); Q (G5, X12) needs to continue to be decomposed into sub-key word Q (G5, R6) and Q (R6, X12).
Therefore, when calculating the probability of occurrence P (X1, X12) of Q (X1, X12), the inquiry doing 11 second son key words is needed altogether.
Suppose in repeatedly inquiry before, by P (X1 in intermediate database, G2), P (G4, X12) suffered stored in buffer memory, so we just only need original Q (X1, X12) sub-key word Q (X1 is decomposed into, G2), Q (G2, G4), Q (G4, X12), and Q (G2, G4) can directly obtain from the child node I2 of index, therefore Q (X1 is directly obtained, G2), Q (G4, X12) Query Result, only need the inquiry of three second son key words namely can obtain Q (X1, X12) probability of occurrence P (X1, X12).
After obtaining the Query Result of all sub-keyword messages, according to the Query Result of the probability distribution calculating probability key word of the inquiry of the Query Result of all sub-key words, wherein Query Result is the joint distribution of multiple variable in probabilistic query key word, and specific formula for calculation is:
wherein P(X) be the joint distribution of variable multiple in probabilistic query key word, P(C) be the joint distribution of variable multiple in sub-key word, P(S) be the marginal distribution of bridging amount.
In the technical scheme that the present embodiment provides, by setting up the mode of tree index to probability database, when making probabilistic query, according to probabilistic query keyword query buffer memory, if there is the intermediate calculation results matched in buffer memory, then directly obtain the Query Result of this intermediate calculation results as probabilistic query key word; If do not exist, according to the index of probability database, by introducing the mode of bridging amount, be multiple sub-key words by probabilistic query key word split, and again search in buffer memory the intermediate calculation results whether being present in sub-key word and matching, if exist, directly obtain the Query Result of this intermediate calculation results as sub-key word.By storing the mode of intermediate calculation results in buffer memory, decrease the computing of the probabilistic query of repetition, accelerate the speed of probabilistic query, simultaneously by setting up the mode of tree index, introducing bridging amount is multiple sub-key words by probabilistic query key word split, also greatly accelerate the inquiry to probability database, accelerate the speed of probabilistic query further.
Further, Fig. 7 is the processing flow chart of intermediate calculation results in buffer memory, and as shown in Figure 7, the method comprises:
Step S300, obtains the Query Result of sub-key word;
Step S302, judges whether the Query Result of sub-key word has existed identical intermediate calculation results in the buffer, if then perform step S304, then performs step S306 if not;
Step S304, upgrades the holding time of the described intermediate calculation results recorded in buffer memory, and increases the hit-count of described intermediate calculation results;
Step S306, is saved in described intermediate calculation results in described buffer memory, record holding time, and the hit-count of intermediate calculation results described in initialization;
Step S308, according to holding time and the hit-count of intermediate calculation results each in described buffer memory, timing or periodically burin-in process is carried out to each intermediate calculation results.
Particularly, in the process of probabilistic query, be multiple sub-key words to probabilistic query key word split, obtain the Query Result of multiple sub-key word.Judge whether the Query Result of sub-key word has existed identical intermediate calculation results in the buffer.Intermediate calculation results, the holding time of intermediate calculation results last time is saved in buffer memory, and hit-count.If there is identical intermediate calculation results, is upgraded the acquisition time of sub-keyword search results for this reason the holding time of this intermediate calculation results, and one is added to the hit-count of this intermediate calculation results.
Due to a large amount of intermediate calculation results can be produced in probabilistic query process, therefore need timing to upgrade the intermediate calculation results in buffer memory, delete the intermediate calculation results that hit-count is low or be not for a long time hit.Therefore according to holding time and the hit-count of intermediate calculation results each in buffer memory, timing or periodically burin-in process is carried out to each intermediate calculation results.
One of ordinary skill in the art will appreciate that: all or part of step realizing said method embodiment can have been come by the hardware that programmed instruction is relevant, aforesaid program can be stored in a computer read/write memory medium, this program, when performing, performs the step comprising said method embodiment; And aforesaid storage medium comprises: ROM, RAM, magnetic disc or CD etc. various can be program code stored medium.
Last it is noted that above each embodiment is only in order to illustrate technical scheme of the present invention, be not intended to limit; Although with reference to foregoing embodiments to invention has been detailed description, those of ordinary skill in the art is to be understood that: it still can be modified to the technical scheme described in foregoing embodiments, or carries out equivalent replacement to wherein some or all of technical characteristic; And these amendments or replacement, do not make the essence of appropriate technical solution depart from the scope of various embodiments of the present invention technical scheme.
Claims (7)
1. a probability database speedup query method, is characterized in that, comprising:
Set up the tree index corresponding with probability database;
Acquisition probability key word of the inquiry, described probabilistic query key word comprises the associating value of multiple variable;
Inquire about in the buffer according to described probabilistic query key word, the intermediate calculation results of probabilistic query process before preserving in described buffer memory, if exist in described buffer memory with described multiple variable combine the intermediate calculation results that value mates, then using the Query Result of described intermediate calculation results as described probabilistic query; If do not exist in described buffer memory with described multiple variable combine the intermediate calculation results that value mates, according to described tree index, described probabilistic query key word is decomposed, obtains multiple sub-key word; Inquire about respectively in the buffer according to described multiple sub-key word, if there is the intermediate calculation results with each described sub-keyword match in described buffer memory, then using the intermediate calculation results of described and described each sub-keyword match as the Query Result of each sub-key word, according to the Query Result of described each sub-key word, calculate the Query Result of described probabilistic query.
2. method according to claim 1, is characterized in that,
Describedly to inquire about respectively in the buffer according to described multiple sub-key word, also comprise:
If there is not the intermediate calculation results with each described sub-keyword match in described buffer memory, then inquire about in the tree index corresponding with described probability database according to described probabilistic query key word.
3. method according to claim 1, is characterized in that, the tree index that described foundation is corresponding with probability database, comprising:
Set up and corresponding with described probability database combine tree;
The tree foundation tree index corresponding with described probability database is combined according to described.
4. method according to claim 1, is characterized in that, described inquire about respectively in the buffer according to described multiple sub-key word after, also comprise:
If there is not the intermediate calculation results with at least one sub-keyword match in described buffer memory, then according to described at least one sub-key word with described tree index in inquire about.
5. the method according to any one of claim 1-4, is characterized in that, also comprises:
The Query Result of described sub-key word is saved in described buffer memory as intermediate calculation results.
6. method according to claim 5, is characterized in that, also comprises:
When intermediate calculation results is saved in described buffer memory, judge whether preserved described intermediate calculation results in described buffer memory, if then upgrade the holding time of the described intermediate calculation results recorded in buffer memory, and increase the hit-count of described intermediate calculation results, then described intermediate calculation results is saved in described buffer memory if not, record holding time, and the hit-count of intermediate calculation results described in initialization.
7. method according to claim 6, is characterized in that, also comprises:
According to holding time and the hit-count of intermediate calculation results each in described buffer memory, timing or periodically burin-in process is carried out to each intermediate calculation results.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201210209270.4A CN102880629B (en) | 2012-06-20 | 2012-06-20 | Accelerating query method of probabilistic database |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201210209270.4A CN102880629B (en) | 2012-06-20 | 2012-06-20 | Accelerating query method of probabilistic database |
Publications (2)
Publication Number | Publication Date |
---|---|
CN102880629A CN102880629A (en) | 2013-01-16 |
CN102880629B true CN102880629B (en) | 2015-06-03 |
Family
ID=47481955
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201210209270.4A Active CN102880629B (en) | 2012-06-20 | 2012-06-20 | Accelerating query method of probabilistic database |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN102880629B (en) |
Families Citing this family (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106202254A (en) * | 2016-06-30 | 2016-12-07 | 上海斐讯数据通信技术有限公司 | A kind of querying method and data query system |
CN107426301A (en) * | 2017-06-21 | 2017-12-01 | 郑州云海信息技术有限公司 | Distributed type assemblies node information management method, system and distributed cluster system |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1947118A (en) * | 2004-04-22 | 2007-04-11 | 甲骨文国际公司 | Partial query caching |
CN101110074A (en) * | 2007-01-30 | 2008-01-23 | 浪潮乐金信息系统有限公司 | Data speedup query method based on file system caching |
CN101390080A (en) * | 2006-02-28 | 2009-03-18 | 微软公司 | Serving cached query results based on a query portion |
CN102073718A (en) * | 2011-01-10 | 2011-05-25 | 清华大学 | System and method for explaining, erasing and modifying search result in probabilistic database |
-
2012
- 2012-06-20 CN CN201210209270.4A patent/CN102880629B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1947118A (en) * | 2004-04-22 | 2007-04-11 | 甲骨文国际公司 | Partial query caching |
CN101390080A (en) * | 2006-02-28 | 2009-03-18 | 微软公司 | Serving cached query results based on a query portion |
CN101110074A (en) * | 2007-01-30 | 2008-01-23 | 浪潮乐金信息系统有限公司 | Data speedup query method based on file system caching |
CN102073718A (en) * | 2011-01-10 | 2011-05-25 | 清华大学 | System and method for explaining, erasing and modifying search result in probabilistic database |
Also Published As
Publication number | Publication date |
---|---|
CN102880629A (en) | 2013-01-16 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103488704B (en) | A kind of date storage method and device | |
JP4856627B2 (en) | Partial query caching | |
CN103794006A (en) | Method and device for processing time series data of multiple sensors | |
CN103514229A (en) | Method and device used for processing database data in distributed database system | |
JP2004518226A (en) | Database system and query optimizer | |
CN102289466A (en) | K-nearest neighbor searching method based on regional coverage | |
CN102402602A (en) | B+ tree indexing method and device of real-time database | |
CN111008521B (en) | Method, device and computer storage medium for generating wide table | |
CN108009265B (en) | Spatial data indexing method in cloud computing environment | |
CN102629269A (en) | Searching and storing method for embedded database | |
CN103678550A (en) | Mass data real-time query method based on dynamic index structure | |
CN105550368A (en) | Approximate nearest neighbor searching method and system of high dimensional data | |
CN105556474A (en) | Managing memory and storage space for a data operation | |
JP2019091257A (en) | Information processing device, information processing method, and program | |
CN113010526A (en) | Storage method and device based on object storage service | |
CN102880629B (en) | Accelerating query method of probabilistic database | |
CN101635001B (en) | Method and apparatus for extracting information from a database | |
CN103365923A (en) | Method and device for assessing partition schemes of database | |
JP2014502756A (en) | Apparatus and method for mass data storage based on tree structure | |
CN114372058A (en) | Spatial data management method and device, storage medium and block chain system | |
CN111666302A (en) | User ranking query method, device, equipment and storage medium | |
CN108647243B (en) | Industrial big data storage method based on time series | |
CN111414527A (en) | Similar item query method and device and storage medium | |
CN116561382A (en) | Data processing method, device, equipment and storage medium of graph database | |
CN116050696A (en) | Global power grid maintenance method, device, computer equipment and storage medium |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant |