CN108874907A - A kind of data query method and apparatus, computer readable storage medium - Google Patents
A kind of data query method and apparatus, computer readable storage medium Download PDFInfo
- Publication number
- CN108874907A CN108874907A CN201810517829.7A CN201810517829A CN108874907A CN 108874907 A CN108874907 A CN 108874907A CN 201810517829 A CN201810517829 A CN 201810517829A CN 108874907 A CN108874907 A CN 108874907A
- Authority
- CN
- China
- Prior art keywords
- inquiry
- querying condition
- query
- entity
- logic
- 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
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
This application discloses a kind of data query method and apparatus, computer readable storage medium, including：Querying condition is received, the querying condition includes at least one of inquiry operation：Entity attribute inquiry, relational extensions inquiry, space-time trajectory inquiry and logic union operation；Whether detect in the querying condition includes relational extensions inquiry and/or logic union operation, in this way, using relational extensions inquiry and/or logic union operation as boundary, execution interval division is carried out to querying condition, in each execution section, carries out the merging of similar inquiry operation and the inquiry operation sequence based on cost estimation adjusts；Task queue is established according to the dependence between each inquiry operation, query process is executed, stores and/or show corresponding query result.Inquiry operation sequence of the application by carrying out the merging of similar inquiry operation in each execution section and based on cost estimation adjusts, and realizes the data fast and efficiently inquired in multidimensional map.
Description
Technical field
The present invention relates to big data technical field more particularly to a kind of data query method and apparatus, computer-readable deposit
Storage media.
Background technique
Knowledge mapping is used to describe various entities or concept present in real world and the association between them is closed
System.Wherein：Each entity or concept are identified with the ID of a globally unique determination, referred to as identifier；Each attribute-value to
Portray the intrinsic characteristic of entity, and relationship is used to connect two entities, portray the association between them.
But enriching constantly with each FIELD Data information, only defining entity, attribute and relationship can no longer meet tool
Body industry describes the demand of things.For example, certain industries have than more rich personnel time's sequence data or track data, this
A little data can neither be defined as entity, also do not represent the relationship between entity, be less entity/relationship attribute.Therefore derivative
Gone out a kind of various dimensions knowledge mapping representation as shown in Figure 1, wherein A point, B point and C point indicate entity in map,
A, between B two o'clock between A, C two o'clock while indicate relationship in map.In various dimensions knowledge mapping, either in fact
Relationship, the space-time trajectory of entity between body or entity, can be there are many different types, for example the type of entity can
To be " people ", " vehicle " etc., the relationship between entity can be " people-vehicle holding relationship ", " kinship " etc., and track can be
" train trip track ", " hotel ccommodation track ".Several " dimensions " of multidimensional map contain different information content, and each dimension is fixed
Justice is as follows：
Dimension one：Entity attributes；
Dimension two：Association between entity；
Dimension three：When (such as time point) empty (form such as latitude and longitude coordinates) track of entity.
For such multidimensional spectrum data, when needing to be implemented a complicated query process (for example, search " with
Three is having a call relationship and appear in Pekinese male in May, 2017 ", wherein the information of three dimensions is specified simultaneously, and
And the inquiry is the map inquiry of no specific starting point), traditional simple graph traversal queries for having clear starting point are
It is unable to satisfy such complex query demand.
Summary of the invention
In order to solve the above-mentioned technical problems, the present invention provides a kind of data query method and apparatus, computer-readable deposit
Storage media can be improved computational efficiency.
In order to solve the above-mentioned technical problem, the technical solution of the embodiment of the present invention is realized in：
The embodiment of the invention provides a kind of data query methods, including：
Querying condition is received, the querying condition includes the inquiry operation of at least one of：Entity attribute inquiry, relationship
Expanding query, space-time trajectory inquiry and logic union operation；
Whether include relational extensions inquiry and/or logic union operation, if including relationship if detecting in the querying condition
Expanding query and/or logic union operation, using relational extensions inquiry and/or logic union operation as boundary, to the inquiry item
Part carries out execution interval division, in each execution section of division, carries out the merging of similar inquiry operation and based on cost
The inquiry operation sequence of estimation adjusts；
Task queue is established according to the dependence between each inquiry operation, executes inquiry using the task queue of foundation
Process stores and/or shows corresponding query result.
Further, if in the querying condition not including the relational extensions inquiry and the logic union operation,
The method also includes：
The inquiry operation sequence that the querying condition is carried out the merging of similar inquiry operation and estimated based on cost is adjusted
It is whole.
Further, in detecting the querying condition whether include relational extensions inquiry and/or logic union operation it
Before, the method also includes：
Detect the querying condition whether be ring structure or with the presence or absence of conflict；
If the querying condition is ring structure or there is conflict, prompting the querying condition, there are mistakes.
Further, in detecting the querying condition whether include relational extensions inquiry and/or logic union operation it
Before, the method also includes：
Detect in the querying condition whether include query result in addition to the querying condition set；
If including the set of the query result in addition to the querying condition, looking into addition to the querying condition is detected
Whether the set for asking result has been stored in preset storage location；
If being not stored in preset storage location, by the set exhibition of the query result in addition to the querying condition
It opens.
Further, the inquiry operation sequence based on cost estimation adjusts, including：
Estimate the Query Cost T of each entity attribute inquiry executed in sectionentityWith looking into for space-time trajectory inquiry
Ask cost Tevent；
If Tentity>G*Tevent, then first carry out space-time trajectory inquiry and execute entity attribute inquiry again；
If Tevent>G*Tentity, then first carry out entity attribute inquiry and execute space-time trajectory inquiry again；
If (| Tevent–Tentity|/(Tevent+Tentity))<L then executes entity attribute inquiry parallel and space-time trajectory is looked into
It askes, and seeks common ground to the result executed parallel；
Wherein, * is multiplication sign, and G, L are preconfigured first proportionality coefficient and the second proportionality coefficient.
Further, the Query Cost T of each entity attribute inquiry executed in sectionentityIt is looked into space-time trajectory
The Query Cost T of inquiryeventRespectively：
The Query Cost T of the entity attribute inquiryentityThe average every record of=entity attribute querying condition hits *
Query time estimated value；
The Query Cost T of the space-time trajectory inquiryeventThe average every record of=space-time trajectory querying condition hits *
Query time estimated value；
Wherein, the entity attribute querying condition hits and the space-time trajectory querying condition hits pass through respectively searches
It indexes the inverted index held up to count to obtain, the query time estimated value of average every record is obtained by preparatory Performance Evaluation.
Further, the logic union operation includes at least one of：Logical AND, logic or logic NOT.
The embodiment of the invention also provides a kind of computer readable storage medium, the computer-readable recording medium storage
Have one or more program, one or more of programs can be executed by one or more processor, with realize such as with
Above the step of data query method.
The embodiment of the invention also provides a kind of data query devices, including processor and memory, wherein：
The processor is for executing the data query program stored in memory, to realize that data as described above are looked into
The step of inquiry method.
The embodiment of the invention also provides a kind of data query devices, including inquire input module, inquiry conversion module, look into
Execution module and result module for reading and writing are ask, wherein：
Input module is inquired, for receiving querying condition, the querying condition includes the inquiry operation of at least one of：
Entity attribute inquiry, relational extensions inquiry, space-time trajectory inquiry and logic union operation；
Conversion module is inquired, for whether detecting in the querying condition including relational extensions inquiry and/or logic merging
Operation is inquired with relational extensions if including relational extensions inquiry and/or logic union operation and/or logic union operation is
Boundary carries out execution interval division to the querying condition, in each execution section of division, carries out similar inquiry operation
Merge and the inquiry operation sequence based on cost estimation adjusts；
Query execution module uses foundation for establishing task queue according to the dependence between each inquiry operation
Task queue execute query process；
As a result module for reading and writing, for storing and/or showing corresponding query result.
Technical solution of the present invention has the advantages that：
Data query method and apparatus provided by the invention, computer readable storage medium, by each execution section
The interior merging for carrying out similar inquiry operation and the inquiry operation sequence estimated based on cost are adjusted, and it is quick, high to be able to use family
The data in multidimensional map are inquired to effect, the complex query demand that user is directed to multidimensional spectrum data is met.
Detailed description of the invention
The drawings described herein are used to provide a further understanding of the present invention, constitutes part of this application, this hair
Bright illustrative embodiments and their description are used to explain the present invention, and are not constituted improper limitations of the present invention.In the accompanying drawings：
Fig. 1 is one of the relevant technologies multidimensional spectrum data structural schematic diagram；
Fig. 2 is a kind of flow diagram of data query method of the embodiment of the present invention；
Fig. 3 is the directed acyclic graph schematic diagram for the inquiry operation composition that a kind of user of the embodiment of the present invention inputs；
Fig. 4 is the schematic diagram data of user query operation transmission in a kind of directed acyclic graph of the embodiment of the present invention；
Fig. 5 is the result set that back end corresponds to another querying condition in a kind of directed acyclic graph of the embodiment of the present invention
Schematic diagram；
Fig. 6 is by the directed acyclic graph structures schematic diagram after the result set expansion in Fig. 5；
Fig. 7 (a) is a kind of directed acyclic graph structures signal comprising combinable similar inquiry operation of the embodiment of the present invention
Figure；
Fig. 7 (b) is that the directed acyclic graph structures after merging the combinable similar inquiry operation in Fig. 7 (a) show
It is intended to；
Fig. 8 (a) is that a kind of directed acyclic graph structures comprising legal serial relational query of the embodiment of the present invention show
It is intended to；
Fig. 8 (b) is the directed acyclic graph knot for merging the serial relational query operating mistake in Fig. 8 (a)
Structure schematic diagram；
Fig. 9 is a kind of directed acyclic graph structures schematic diagram comprising three parallel task queues of the embodiment of the present invention；
Figure 10 is a kind of structural schematic diagram of data query device of the embodiment of the present invention.
Specific embodiment
To make the objectives, technical solutions, and advantages of the present invention clearer, below in conjunction with attached drawing to the present invention
Embodiment be described in detail.It should be noted that in the absence of conflict, in the embodiment and embodiment in the application
Feature can mutual any combination.
With reference to Fig. 2, a kind of data query method according to an embodiment of the present invention includes the following steps：
Step 201：Querying condition is received, the querying condition includes the inquiry operation of at least one of：Entity attribute
Inquiry, relational extensions inquiry, space-time trajectory inquiry and logic union operation；
In the present embodiment, the logic union operation includes at least one of：Logical AND, logic or logic NOT.
It should be noted that data query method of the invention provides the friendship for defining complex query process for user
Mutual interface, as shown in figure 3, user can pull the inquiry operation of different dimensions on the interactive interface, for each inquiry behaviour
Specific querying condition is specified, and multiple queries operating series are got up by unidirectional arrow, forms a directed acyclic graph
(Directed Acyclic Graph, DAG), the corresponding inquiry for being once directed to multidimensional spectrum data of a directed acyclic graph
Journey.Composition DAG inquiry operation type include：Entity attribute inquiry, relational extensions inquiry, space-time trajectory inquiry, logic merge
Operation.In addition, on interactive interface other than it can pull inquiry operation, can also pull out one " back end " as with
The input data of upper several inquiry operations, " back end " can specify as the unique of the group object that is present in map
The result that mark (identifier) or other DAG are calculated.
User forms the complex query condition of DAG form on interactive interface by pulling a plurality of types of inquiry operations.
In order to meet the characteristic of DAG, in user operation process, interactive interface, which can check between the inquiry operation being dragged on interface, is
It is no to form " ring ", if mistake can be shown and user is forbidden to carry out next step operation by foring " ring ".
Entity attribute inquiry settable condition include：Attribute value/range of condition, fuzzy keyword condition, such as 1 institute of table
Show：
Table 1
Relational extensions inquiry settable condition include：Relationship type condition, time range condition, attribute of a relation are fuzzy to close
Keyword condition, as shown in table 2：
Table 2
Space-time trajectory inquiry settable condition include：The time conditions of track event type condition, specifically certain track
With the attribute conditions for representing space, as shown in table 3：
Table 3
The settable condition of logic union operation includes：Intersection, union and difference set；
The set-up mode of " back end " includes：It fills in entity unique identification or uploads the Excel text comprising unique identification
The name result set that part or other DAG are calculated.
The constraint explanation of above several map inquiry operations is as follows：
Entity attribute inquiry：Entity attribute inquiry operation is equivalent to the filter operation to entity itself.If do not enabled
" external inverted index " relevant configuration, then entity attribute inquiry must have " back end " as input or connect in other inquiries
Behind operation, and cannot function as first inquiry operation (do not support " fuzzy query " in the case where no external index,
Otherwise the solid data of all specified types of map is needed to be traversed for).Be between specified multiple attribute conditions " and " logic.
Relational extensions inquiry：Relational extensions inquiry is in map for specified input solid data lookup and its
Associated other entities, Just because of this, relational extensions inquiry must be as " back end " or other inquiry operations
Subsequent node cannot independently use in the case where no input.After the operation of once relational extensions, depends on relationship and expand
When exhibition specify relationship type condition, obtained result may be with input solid data type it is identical (such as " people " warp
Cross " pair bond " extension after result still " people "), it is also possible to it is different from the solid data type of input (such as " people " process
The result is that " vehicle " entity after " man-vehicle interface " extension)." relationship type " is between multiple types if condition specifies multiple
" and " logic, i.e. the query result intersection that is the result of this variety of relationship types extension.Due to relational extensions inquiry it is special
Property, if specifying a variety of " relationship types ", interactive interface can check the reality of this variety of " relationship types " corresponding spreading result
Whether body type belongs to same type, if not can then prompt mistake and forbid carry out next step operation.
Space-time trajectory inquiry：Space-time trajectory inquiry is the equal of the abundant track characteristic information using entity to entity
It is filtered screening operation.It is similar to entity attribute query type, if do not configure relevant external inverted index information,
Space-time trajectory inquiry must also have " back end " to input or connect behind other inquiry operations and (do not support " fuzzy query ")；
Fuzzy inquiry, query result can be carried out to space-time track characteristic condition in the case where being configured with external inverted index information
To meet intended trajectory condition.Between specified a variety of different types of tracking conditions be also " and " logic.
Logic union operation：For realizing multiple result AND (" logical AND ", corresponding intersection), OR (" logic or ", it is corresponding
Union), the logical operation of NOT (" logic NOT ", corresponding difference set), therefore it must have " back end " or other inquiry operations
Result set is as input, and input node has to be larger than or be equal to 2.
The DAG query graph information that user defines in interactive interface is received by REST api interface, by interactive interface
The DAG query graph information transmitted is converted to internal data structure, and the DAG defined on interactive interface is restored from structure.Show
Example property, the DAG data representation format that DAG query graph information as shown in Figure 4 is converted into is as follows, and what which indicated is
One space-time trajectory inquiry connects an entity attribute inquiry below, and (connection relationship between operation passes through seq and inputSeq
Parameter indicates).
In the present embodiment, after the reception querying condition, the method also includes：
Detect the querying condition whether be ring structure or with the presence or absence of conflict；
If the querying condition is ring structure or there is conflict, prompting the querying condition, there are mistakes.
In the present embodiment, after the reception querying condition, the method also includes：Item is carried out to each inquiry operation
Part normalization and condition merge.
Specifically, condition normalization is carried out to the condition in entity attribute inquiry operation, condition normalization refers to and will input
Various forms of conditions be uniformly converted to one of IN and NOT IN both forms according to its type, " certain attribute "=" certain
Condition as value ", is uniformly converted to " certain attribute " IN [" certain value "], " certain attribute "！Condition as=" certain value ", it is unified
Be converted to " certain attribute " conditional forms as NOT IN [" certain value "]；The normalization of condition is in order to which the merging of subsequent condition is done
Prepare.
If current entity attribute query operation in be " certain attribute " and specify a variety of different conditions (such as specify "
=", "！=", " IN " is a variety of inside " NOT IN "), then the value range of the right operand of this multiple condition can be rushed
Prominent detection and condition merge, for example, for " name " IN [" bear is big "] and " name " IN [" bear two "] the two conditions, with regard to explanation
There is conflict, to return the result be also sky even if inquiry operation really performs, therefore can judge, will currently look into advance
Ask operation and its subsequent direct inquiry operation mark as sky " result set；For " age " IN [20,40] and " age " IN
[30,50] the two conditions can be merged into a condition " age " IN [30,40].
The normalization of condition as entity attribute inquiry operation also will do it to the condition in space-time trajectory inquiry operation
Merge with collision detection/condition.
In the present embodiment, after the reception querying condition, the method also includes：
Detect in the querying condition whether include query result in addition to the querying condition set；
If including the set of the query result in addition to the querying condition, looking into addition to the querying condition is detected
Whether the set for asking result has been stored in preset storage location；
If being not stored in preset storage location, by the set exhibition of the query result in addition to the querying condition
It opens.
Specifically, traverse DAG in " back end ", as shown in figure 5, if " back end " it is corresponding be another
The result set (set of the query result i.e. in addition to the querying condition) of DAG will judge the validity of the result set, i.e., should
Whether result set is existing, and (the result set storage location according to configuration is hard disk or memory database, removes corresponding memory bank
Middle acknowledgment of your inquiry), if there is the invalid situation of result set, then it is related to DAG " expansion " conversion.As shown in fig. 6, described
" expansion " refers to a part be included in the corresponding DAG of failed result set and calculate content as current DAG.
Step 202：Whether detect in the querying condition includes relational extensions inquiry and/or logic union operation；
Step 203：If including relational extensions inquiry and/or logic union operation in the querying condition, expanded with relationship
Exhibition inquiry and/or logic union operation are boundary, execution interval division are carried out to the querying condition, in each execution of division
In section, carries out the merging of similar inquiry operation and the inquiry operation sequence based on cost estimation adjusts；
Step 204：It is right if in the querying condition not including the relational extensions inquiry and the logic union operation
The querying condition carries out the merging of similar inquiry operation and the inquiry operation sequence based on cost estimation adjusts；
As shown in Fig. 7 (a), using " relational extensions " and " logic union operation " as boundary demarcation at multiple execution sections,
This is because the particularity of these operations, its execution sequence and position are unalterable.From correctness consider be cannot
Serial multiple relational extensions inquiry is merged (" mergings " be together it is incorrect, be equivalent to and 2 degree of figures traversed mistake
It is changed to 1 degree of figure to traverse).
As shown in Fig. 8 (a), during legal serial relational query, the input of second relational query is first
The output of a relational query, the output of first relational query are " entities-vehicle " rather than " entity-people ".As shown in Fig. 8 (b),
If two relational queries are mistakenly merged, the input of second relational query be no longer " entity-vehicle " and
It is " entity-people ", causes in logic inconsistent.
In the present embodiment, trial optimizes inquiry operation serial in DAG, and the mode of optimization includes similar inquiry behaviour
The merging of work and the search order adjustment estimated based on cost.
As shown in Fig. 7 (b), the method that similar inquiry operation merges is, can for the inquiry operation in each execution section
To merge " entity attribute inquiry " operation and " space-time trajectory inquiry " operation respectively, to simplify the structure of DAG, reduction is actually held
Capable physical queries number of operations.The concrete operations of merging process are consistent with previously described condition union operation.
In the present embodiment, the method for the search order adjustment based on cost estimation is, also in the execution divided above
(front merges in optimization the latter section at most only has a substance feature and affair character to section interior optimization execution efficiency
Inquiry), due to relational extensions operation the immutable characteristic in position, the optimization of execution efficiency be still for space-time trajectory inquiry and
Entity attribute inquiry.Query optimization based on cost estimation depends on the enabling to outside index to configure.In each execution section
Entity attribute inquiry Query Cost TentityCalculation is：
Tentity=(entity attribute querying condition hits) * (the query time estimated value of average every record)
The Query Cost T of space-time trajectory inquiry in each execution sectioneventCalculation is：
Tevent=(space-time trajectory querying condition hits) * (the query time estimated value of average every record), wherein *
For multiplication sign.
Wherein querying condition hits are the express statistic behaviour that querying condition input is passed through the inverted index of search engine
It obtains, the query time estimated value of average every record is calculated by the preparatory the performance test results under specific system deployment environment
It obtains.The logic of sequence adjustment is as follows：
If Tentity>G*Tevent, i.e. entity attribute query time is far longer than the event query time, then first carries out space-time
Track inquiry executes entity attribute inquiry again；
If Tevent>G*Tentity, then first carry out entity attribute inquiry and execute space-time trajectory inquiry again；
If (| Tevent–Tentity|/(Tevent+Tentity))<L, this means that the two execution time difference is smaller, then simultaneously
Row executes entity attribute inquiry and space-time trajectory inquiry, and seeks common ground to the result executed parallel.
Wherein, G, L are preconfigured first proportionality coefficient and the second proportionality coefficient.
Step 205：Task queue is established according to the dependence between each inquiry operation, uses the task queue of foundation
Query process is executed, stores and/or show corresponding query result.
In the present embodiment, is divided to DAG, the method for executing divided stages is according to the dependence inquired in DAG the execution stage
Relationship judges which operation can execute parallel.As shown in figure 9, three of logic union operation input (3,6,8) source respectively
In three sub- DAG, then these three sub- DAG can be executed in a parallel fashion, these three sub- DAG merge with logic between be serial
Relationship, and be another serial operation respectively inside three sub- DAG.It can be submitted and be looked into a manner of asynchronous task queue when execution
Operation is ask, the logic union operation as final result result is put into queue first and waits running, then analysis finds that it has three
A input is then respectively put into three inputs in three parallel task queues, and the inquiry operation in these three task queues is again
The forward direction for continuing to analyze them relies on operation, so analogizes and the back end inquiry most started is finally put into queue.Start to hold
After row, the inquiry operation completed in each queue, inquiring its backward node, (for example the backward node of substance feature 1 is
Logic merges), see after this to node it is all before whether all completed to relying on, its is triggered if being completed and backward is relied on
Execution.
Each number designation is the mark (do not represent and execute sequence) of each calculate node, the sequence actually executed in Fig. 9
It is：It is to execute parallel between (1,2,3), (7,8) and (4,5,6) three group polling, the execution inside (1,2,3) this group polling is suitable
Sequence is 1->2->3；Execution sequence inside (7,8) this group polling is 7->8；Execution inside (4,5,6) this group polling is suitable
Sequence is 4->5->6, it waits after three group pollings have executed above, 9 are finally performed.
Specific inquiry operation can access bottom storage assembly acquisition data result, substance feature during executing inquiry
Chart database (if external index can also be accessed by being configured with external index, to accelerate to inquire) can be accessed；Track characteristic can access
Time series database (if external index can also be accessed by being configured with external index, to accelerate to inquire)；Relational extensions can access figure number
According to library；Logic union operation can be operated real if be configured with using memory database using the high-performing sets of memory database
Now quick joint account.
After completing the inquiry operation in current DAG queue, result set data and corresponding DAG structural information are stored in firmly
Data and corresponding DAG structural information are then returned to boundary when result set is checked in interactive interface request by disk or memory database
Face is shown.
The embodiment of the invention also provides a kind of computer readable storage medium, the computer-readable recording medium storage
Have one or more program, one or more of programs can be executed by one or more processor, with realize such as with
The step of upper described in any item data query methods.
The embodiment of the invention also provides a kind of data query devices, including processor and memory, wherein：
The processor is for executing the data query program stored in memory, to realize as described in any of the above item
The step of data query method.
As shown in Figure 10, the embodiment of the invention also provides a kind of data query device, including inquiry input module 1001,
Conversion module 1002, query execution module 1003 and result module for reading and writing 1004 are inquired, wherein：
Input module 1001 is inquired, for receiving querying condition, the querying condition includes the inquiry of at least one of
Operation：Entity attribute inquiry, relational extensions inquiry, space-time trajectory inquiry and logic union operation；
Conversion module 1002 is inquired, for whether detecting in the querying condition including relational extensions inquiry and/or logic
Union operation merges behaviour if including relational extensions inquiry and/or logic union operation with relational extensions inquiry and/or logic
As boundary, execution interval division is carried out to the querying condition；In each execution section of division, similar inquiry behaviour is carried out
The merging of work and the inquiry operation sequence estimated based on cost are adjusted；
Query execution module 1003 is used for establishing task queue according to the dependence between each inquiry operation
The task queue of foundation executes query process；
As a result module for reading and writing 1004, for storing and/or showing corresponding query result.
In the present embodiment, the logic union operation includes at least one of：Logical AND, logic or logic NOT.
It should be noted that data query device of the invention provides the friendship for defining complex query condition for user
Mutual interface (interactive interface inquires input module 1001), as shown in figure 3, user can pull not on the interactive interface
With the inquiry operation of dimension, specific querying condition is specified for each inquiry operation, and by unidirectional arrow multiple queries
Operating series get up, and form the corresponding query process for being once directed to multidimensional spectrum data of directed acyclic graph DAG, a DAG.
Composition DAG inquiry operation type include：Entity attribute inquiry, relational extensions inquiry, space-time trajectory inquiry, logic merge behaviour
Make.In addition, one " back end " conduct or more can also be pulled out on interactive interface other than it can pull inquiry operation
The input data of several inquiry operations, " back end " can specify unique mark of the group object to be present in map
Know the result that (identifier) or other DAG are calculated.
Conversion module 1002 is inquired, the conversion of query process operation is responsible for, which receives the complexity that interactive interface transmits
Query structure condition, is converted to the query process condition of internal tree structure, and the execution analyzed between each query node is suitable
Sequence relies on, and attempts to merge the similar inquiry sub-operation in a branch；For out-of-order rely on inquiry sub-operation according to parallel
Mode execute；For the querying condition of multiple dimensions, the meter of different execution sequences can be calculated separately by way of estimation
Calculate cost, select it is optimal execute sequence, to realize the optimization of execution efficiency.
Query execution module 1003, the elder generation being responsible between the inquiry operation exported according to previous step inquiry conversion module 1002
Dependence executes specific inquiry operation afterwards, obtains the query result of DAG.The storage of the docking bottom of query execution module 1003
Component (chart database storage entity, relation data, time series database storage track data).Come for interactive interface submission
Inquiry, query execution module 1003 take the mode of asynchronous task queue to execute, and the inquiry operation being carrying out is in inner marker
State is RUNNING " in operation ", and inquiry operation inner marker to be performed is during WAITING is waited in the queue etc.,
The operation for having executed end is respectively labeled as " succeeding " according to state and " failure ", interactive interface are current by REST API inquiry
When the execution state of DAG, the state that query execution module 1003 can return to each operation in DAG is accordingly opened up by interactive interface
Show.
As a result module for reading and writing 1004, the calculated result that query execution module 1003 obtains will form result set
(ResultSet) it is stored in result module for reading and writing 1004.As a result module for reading and writing 1004 provide interface for interactive interface inquiry and
Show the data (and the corresponding DAG conditional information of result set) of result set.Map inquiry operation the result is that in map (one
Seed type) solid data, and a DAG only has an output result.The result set of one DAG can be further used as separately
The input data of outer DAG, that is, multiple DAG can be stitched together to form a more complicated DAG in logic.
Each result set has a unique ID mark.
In the present embodiment, the data query device further includes query configuration module, be responsible for configuration querying conversion module and
The configuration information that query execution module needs.Whether the configuration information includes the Thread Count executed parallel, using the external row of falling
Indexing component, external index name and spectrum data type mapping relations, whether using memory database accelerate inquiry, result
Collection is stored in hard disk or memory database etc..
User forms the complex query process of DAG form on interactive interface by pulling a plurality of types of inquiry operations.
In order to meet the characteristic of DAG, in user operation process, interactive interface can check between the operation being dragged on interface whether shape
At " ring ", if mistake can be shown and user is forbidden to carry out next step operation by foring " ring ".
Entity attribute inquiry settable condition include：Attribute value/range of condition, fuzzy keyword condition, such as 1 institute of table
Show：
Table 1
Relational extensions inquiry settable condition include：Relationship type condition, time range condition, attribute of a relation are fuzzy to close
Keyword condition, as shown in table 2：
Table 2
Space-time trajectory inquiry settable condition include：The time conditions of track event type condition, specifically certain track
With the attribute conditions for representing space, as shown in table 3：
Table 3
The settable condition of logic union operation includes：Intersection, union and difference set；
The set-up mode of " back end " includes：It fills in entity unique identification or uploads the Excel text comprising unique identification
The name result set that part or other DAG are calculated.
The constraint explanation of above several map inquiry operations is as follows：
Entity attribute inquiry：Entity attribute inquiry operation is equivalent to the filter operation to entity itself.If do not enabled
" external inverted index " relevant configuration, then entity attribute inquiry must have " back end " as input or connect in other inquiries
Behind operation, and cannot function as first inquiry operation (do not support " fuzzy query " in the case where no external index,
Otherwise the solid data of all specified types of map is needed to be traversed for).Be between specified multiple attribute conditions " and " logic.
Relational extensions inquiry：Relational extensions inquiry is in map for specified input solid data lookup and its
Associated other entities, Just because of this, relational extensions inquiry must be as " back end " or other inquiry operations
Subsequent node cannot independently use in the case where no input.After the operation of once relational extensions, depends on relationship and expand
When exhibition specify relationship type condition, obtained result may be with input solid data type it is identical (such as " people " warp
Cross " pair bond " extension after result still " people "), it is also possible to it is different from the solid data type of input (such as " people " process
The result is that " vehicle " entity after " man-vehicle interface " extension)." relationship type " is between multiple types if condition specifies multiple
" and " logic, i.e. the query result intersection that is the result of this variety of relationship types extension.Due to relational extensions inquiry it is special
Property, if specifying a variety of " relationship types ", interactive interface can check the reality of this variety of " relationship types " corresponding spreading result
Whether body type belongs to same type, if not can then prompt mistake and forbid carry out next step operation.
Space-time trajectory inquiry：Space-time trajectory inquiry is the equal of the abundant track characteristic information using entity to entity
It is filtered screening operation.It is similar to entity attribute query type, if do not configure relevant external inverted index information,
Space-time trajectory inquiry must also have " back end " to input or connect behind other inquiry operations and (do not support " fuzzy query ")；
Fuzzy inquiry, query result can be carried out to space-time track characteristic condition in the case where being configured with external inverted index information
To meet intended trajectory condition.Between specified a variety of different types of tracking conditions be also " and " logic.
Logic union operation：For realizing multiple result AND (" and ", corresponding intersection), OR ("or", corresponding union), NOT
The logical operation of (corresponding difference set), therefore it must have the result set of " back end " or other inquiry operations as input, and
Input node has to be larger than or is equal to 2.
The information for the DAG query graph that user defines is transferred to inquiry modulus of conversion by REST api interface by interactive interface
Block 1002, illustratively, the DAG data representation format of DAG query graph transmission as shown in Figure 4 is as follows, which indicates
Be the inquiry of space-time trajectory connect below an entity attribute inquiry (connection relationship between operation by seq and
InputSeq parameter indicates).
The DAG Content Transformation that inquiry conversion module 1002 first transmits interactive interface is its internal data structure, from
The DAG defined on interactive interface is restored in structure.
In the present embodiment, after the reception querying condition, the inquiry conversion module 1002 is also used to：
Detect the querying condition whether be ring structure or with the presence or absence of conflict；
If the querying condition is ring structure or there is conflict, prompting the querying condition, there are mistakes.
In the present embodiment, the inquiry conversion module 1002 carries out condition normalizing for each of DAG inquiry operation
Change and collision detection/condition merging treatment, condition normalization refer to various forms of conditions of input are unified according to its type
One of IN and NOT IN both forms are converted to, condition as " certain attribute "=" certain value ", are uniformly converted to " certain category
Property " IN [" certain value "], " certain attribute "！Condition as=" certain value ", is uniformly converted to " certain attribute " NOT IN [" certain value "]
Such conditional forms；The normalization of condition is in order to which the merging of subsequent condition is prepared.If current entity attribute query operates
In be " certain attribute " and specify a variety of different conditions (such as specify "=", "！=", " IN " is more inside " NOT IN "
Kind), then collision detection can be carried out to the value range of the right operand of this multiple condition and condition merges, such as " name "
IN [" bear is big "] and " name " IN [" bear two "] the two conditions, just illustrate to have occurred conflict, even if inquiry operation really executes
Returning the result is also sky, therefore can be judged in advance, and current queries operation and its subsequent direct inquiry operation is equal
Labeled as " sky " result set；For " age " IN [20,40] and " age " IN [30,50] the two conditions, one can be merged into
Condition " age " IN [30,40].
The inquiry conversion module 1002 also will do it the condition in space-time trajectory inquiry operation and inquire with entity attribute
The same condition normalization of operation and collision detection/condition merge.
In the present embodiment, after the reception querying condition, the inquiry conversion module 1002 is also used to：
Detect in the querying condition whether include query result in addition to the querying condition set；
If including the set of the query result in addition to the querying condition, looking into addition to the querying condition is detected
Whether the set for asking result has been stored in preset storage location；
If being not stored in preset storage location, by the set exhibition of the query result in addition to the querying condition
It opens.
The inquiry conversion module 1002 continues with DAG, traverses " back end " in DAG, as shown in figure 5, if
It is the result set (set of the query result i.e. in addition to the querying condition) of another DAG that " back end " corresponding,
Judge the validity of the result set, i.e., the result set it is whether existing (the result set storage location according to configuration be hard disk or
Memory database removes acknowledgment of your inquiry in corresponding memory bank), if there is the invalid situation of result set, then it is related to DAG
" expansion " conversion.It is counted as shown in fig. 6, " expansion " refers to for the corresponding DAG of failed result set to be included in as current DAG
Calculate a part of content.
As shown in Fig. 7 (a), the inquiry conversion module 1002 is using " relational extensions " and " logic union operation " as boundary
Querying condition is divided into multiple execution sections, this is because the particularity of " relational extensions " and " logic union operation ", they
Execution sequence and position be unalterable, for example, from correctness consider be that serial multiple relational extensions cannot be looked into
Inquiry merges (" merging " be together incorrect, be equivalent to and 2 degree of figures traversal mistakes are changed to 1 degree of figure traverse).Such as Fig. 8
(a) shown in, during legal serial relational query, the input of second relational query is the defeated of first relational query
Out, the output of first relational query is " entity-vehicle " rather than " entity-people ".As shown in Fig. 8 (b), if two relationships
It merges to inquiry error, then the input of second relational query is no longer " entity-vehicle " but " entity-people ", makes
It is inconsistent in logic.
The inquiry conversion module 1002 can be attempted to optimize inquiry operation serial in DAG.The mode packet of optimization
Include the merging of similar inquiry operation and the search order adjustment based on cost estimation.
In the present embodiment, if not including that the relational extensions inquiry and the logic merge behaviour in the querying condition
Make, the inquiry conversion module 1002 is also used to：
The inquiry operation sequence that the querying condition is carried out the merging of similar inquiry operation and estimated based on cost is adjusted
It is whole.
As shown in Fig. 7 (b), the method that similar inquiry operation merges is, can for the inquiry operation in each execution section
To merge " entity attribute inquiry " operation and " space-time trajectory inquiry " operation respectively, to simplify the structure of DAG, reduction is actually held
Capable physical queries number of operations.
In the present embodiment, the method for the search order adjustment based on cost estimation is, also in the execution divided above
(front merges in optimization the latter section at most only has a substance feature and affair character to section interior optimization execution efficiency
Inquiry), due to relational extensions operation the immutable characteristic in position, the optimization of execution efficiency be still for space-time trajectory inquiry and
Entity attribute inquiry.Query optimization based on cost estimation depends on the enabling to outside index to configure.In each execution section
Entity attribute inquiry Query Cost TentityCalculation is
Tentity=(entity attribute querying condition hits) * (the query time estimated value of average every record)
The Query Cost T of space-time trajectory inquiry in each execution sectioneventCalculation is
Tevent=(space-time trajectory querying condition hits) * (the query time estimated value of average every record).
Wherein querying condition hits are the express statistic behaviour that querying condition input is passed through the inverted index of search engine
It obtains, the query time estimated value of average every record is calculated by the preparatory the performance test results under specific system deployment environment
It obtains.The logic of sequence adjustment is as follows：
If Tentity>G*Tevent, i.e. entity attribute query time is far longer than the event query time, then first carries out space-time
Track inquiry executes entity attribute inquiry again；
If Tevent>G*Tentity, then first carry out entity attribute inquiry and execute space-time trajectory inquiry again；
If (| Tevent–Tentity|/(Tevent+Tentity))<L, this means that the two execution time difference is smaller, then simultaneously
Row executes entity attribute inquiry and space-time trajectory inquiry, and seeks common ground to the result executed parallel.
Wherein, G, L are preconfigured first proportionality coefficient and the second proportionality coefficient.
In the present embodiment, the Query Cost T of each entity attribute inquiry executed in sectionentityAnd space-time trajectory
The Query Cost T of inquiryeventRespectively：
The Query Cost T of the entity attribute inquiryentityThe average every record of=entity attribute querying condition hits *
Query time estimated value；
The Query Cost T of the space-time trajectory inquiryeventThe average every record of=space-time trajectory querying condition hits *
Query time estimated value；
The entity attribute querying condition hits and the space-time trajectory querying condition hits pass through search respectively and draw
The inverted index held up counts to obtain, and the query time estimated value of average every record is obtained by preparatory Performance Evaluation.
The query execution module 1003 divides the execution stage to DAG, and the method for executing divided stages is looked into according in DAG
The dependence of inquiry judges which operation can execute parallel.As shown in figure 9, three of logic union operation input (3,6,8)
Three sub- DAG are respectively derived from, then these three sub- DAG can be executed in a parallel fashion, between these three sub- DAG merge with logic
It is serial relationship, and is another serial operation respectively inside three sub- DAG.It can be with the side of asynchronous task queue when execution
Formula submits inquiry operation, and the logic union operation as final result result is put into queue first and waits running, then analysis hair
Now it is then respectively put into three inputs in three parallel task queues, looking into these three task queues there are three input
It askes operation to be further continued for analyzing their forward direction dependence operation, so analogizes and the back end inquiry most started is finally put into team
Column.After starting execution, the inquiry operation completed in each queue can inquire its backward node (such as after substance feature 1
That logic merges to node), see after this to node it is all before whether all completed to relying on, trigger it if being completed
The execution relied on backward.
Each number designation is the mark (do not represent and execute sequence) of each calculate node, the sequence actually executed in Fig. 9
It is：It is to execute parallel between (1,2,3), (7,8) and (4,5,6) three group polling, the execution inside (1,2,3) this group polling is suitable
Sequence is 1->2->3；Execution sequence inside (7,8) this group polling is 7->8；Execution inside (4,5,6) this group polling is suitable
Sequence is 4->5->6, it waits after three group pollings have executed above, 9 are finally performed.
Specific inquiry operation can access bottom storage assembly acquisition data result, substance feature during executing inquiry
Chart database (if external index can also be accessed by being configured with external index, to accelerate to inquire) can be accessed；Track characteristic can access
Time series database (if external index can also be accessed by being configured with external index, to accelerate to inquire)；Relational extensions can access figure number
According to library；Logic union operation can be operated real if be configured with using memory database using the high-performing sets of memory database
Now quick joint account.
After the query execution module 1003 completes the inquiry operation in current DAG queue, result set data and correspondence
DAG structural information hard disk or memory database are stored in by result module for reading and writing 1004, when knot is checked in interactive interface request
Data and corresponding DAG structural information are then returned to interface display when collecting by fruit.
Those of ordinary skill in the art will appreciate that all or part of the steps in the above method can be instructed by program
Related hardware is completed, and described program can store in computer readable storage medium, such as read-only memory, disk or CD
Deng.Optionally, one or more integrated circuits can be used also to realize in all or part of the steps of above-described embodiment.Accordingly
Ground, each module/unit in above-described embodiment can take the form of hardware realization, can also use the shape of software function module
Formula is realized.The present invention is not limited to the combinations of the hardware and software of any particular form.
The above is only a preferred embodiment of the present invention, and certainly, the invention may also have other embodiments, without departing substantially from this
In the case where spirit and its essence, those skilled in the art make various corresponding changes in accordance with the present invention
And deformation, but these corresponding changes and modifications all should fall within the scope of protection of the appended claims of the present invention.
Claims (10)
1. a kind of data query method, which is characterized in that including：
Querying condition is received, the querying condition includes the inquiry operation of at least one of：Entity attribute inquiry, relational extensions
Inquiry, space-time trajectory inquiry and logic union operation；
Whether include relational extensions inquiry and/or logic union operation, if including relational extensions if detecting in the querying condition
Inquiry and/or logic union operation, using relational extensions inquiry and/or logic union operation as boundary, to the querying condition into
Row executes interval division, in each execution section of division, carries out the merging of similar inquiry operation and is estimated based on cost
Inquiry operation sequence adjust；
Task queue is established according to the dependence between each inquiry operation, was inquired using the task queue execution of foundation
Journey stores and/or shows corresponding query result.
2. data query method according to claim 1, which is characterized in that if not including described in the querying condition
Relational extensions inquiry and the logic union operation, the method also includes：
To the querying condition carry out the similar inquiry operation merging and it is described based on cost estimation inquiry operation it is suitable
Sequence adjustment.
3. data query method according to claim 1, which is characterized in that in the detection querying condition whether
Before relational extensions inquiry and/or logic union operation, the method also includes：
Detect the querying condition whether be ring structure or with the presence or absence of conflict；
If the querying condition is ring structure or there is conflict, prompting the querying condition, there are mistakes.
4. data query method according to claim 1, which is characterized in that the inquiry operation based on cost estimation is suitable
Sequence adjustment, including：
Estimate the Query Cost T of each entity attribute inquiry executed in sectionentityWith the inquiry generation of space-time trajectory inquiry
Valence Tevent；
If Tentity>G*Tevent, then first carry out space-time trajectory inquiry and execute entity attribute inquiry again；
If Tevent>G*Tentity, then first carry out entity attribute inquiry and execute space-time trajectory inquiry again；
If (| Tevent–Tentity|/(Tevent+Tentity))<L then executes entity attribute inquiry and space-time trajectory inquiry parallel, and
It seeks common ground to the result executed parallel；
Wherein, * is multiplication sign, and G, L are preconfigured first proportionality coefficient and the second proportionality coefficient.
5. data query method according to claim 4, which is characterized in that each entity attribute executed in section
The Query Cost T of inquiryentityWith the Query Cost T of space-time trajectory inquiryeventRespectively：
The Query Cost T of the entity attribute inquiryentityThe average every record of=entity attribute querying condition hits * is looked into
Ask time Estimate value；
The Query Cost T of the space-time trajectory inquiryeventThe inquiry of the average every record of=space-time trajectory querying condition hits *
Time Estimate value；
Wherein, the entity attribute querying condition hits and space-time trajectory querying condition hits pass through search engine respectively
Inverted index counts to obtain, and the query time estimated value of average every record is obtained by preparatory Performance Evaluation.
6. data query method according to claim 1, which is characterized in that the logic union operation include it is following at least
One of：Logical AND, logic or logic NOT.
7. a kind of computer readable storage medium, which is characterized in that the computer-readable recording medium storage have one or
Multiple programs, one or more of programs can be executed by one or more processor, to realize such as claim 1 to 6
Any one of described in data query method the step of.
8. a kind of data query device, which is characterized in that including processor and memory, wherein：
The processor is for executing the data query program stored in memory, to realize such as any one of claims 1 to 6
The step of described data query method.
9. a kind of data query device, which is characterized in that including inquiry input module, inquiry conversion module, query execution module
With result module for reading and writing, wherein：
Input module is inquired, for receiving querying condition, the querying condition includes the inquiry operation of at least one of：Entity
Attribute query, relational extensions inquiry, space-time trajectory inquiry and logic union operation；
Conversion module is inquired, whether includes relational extensions inquiry and/or logic union operation for detecting in the querying condition,
If including relational extensions inquiry and/or logic union operation, using relational extensions inquiry and/or logic union operation as boundary,
Execution interval division is carried out to the querying condition, in each execution section of division, carries out the merging of similar inquiry operation
And the inquiry operation sequence based on cost estimation adjusts；
Query execution module uses appointing for foundation for establishing task queue according to the dependence between each inquiry operation
Business queue executes query process；
As a result module for reading and writing, for storing and/or showing corresponding query result.
10. data query device according to claim 9, which is characterized in that be in the detection querying condition
It is no include relational extensions inquiry and/or logic union operation before, the inquiry conversion module is also used to：
Detect the querying condition whether be ring structure or with the presence or absence of conflict；
If the querying condition is ring structure or there is conflict, prompting the querying condition, there are mistakes.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810517829.7A CN108874907A (en) | 2018-05-25 | 2018-05-25 | A kind of data query method and apparatus, computer readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810517829.7A CN108874907A (en) | 2018-05-25 | 2018-05-25 | A kind of data query method and apparatus, computer readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108874907A true CN108874907A (en) | 2018-11-23 |
Family
ID=64333757
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810517829.7A Pending CN108874907A (en) | 2018-05-25 | 2018-05-25 | A kind of data query method and apparatus, computer readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108874907A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111159512A (en) * | 2019-12-30 | 2020-05-15 | 北京明略软件系统有限公司 | Processing method and server of track information |
| WO2020192534A1 (en) * | 2019-03-26 | 2020-10-01 | 华为技术有限公司 | Search method, terminal, and medium |
| CN112035581A (en) * | 2020-08-21 | 2020-12-04 | 北京字节跳动网络技术有限公司 | Model-based task processing method, device, equipment and medium |
| CN112241443A (en) * | 2019-07-16 | 2021-01-19 | 中国移动通信集团浙江有限公司 | Data quality monitoring method and device, computing equipment and computer storage medium |
| CN112732821A (en) * | 2021-03-31 | 2021-04-30 | 成都新希望金融信息有限公司 | Data storage method, device and equipment and storage medium |
| CN112783922A (en) * | 2021-02-01 | 2021-05-11 | 广州海量数据库技术有限公司 | Query method and device based on relational database |
| CN113760394A (en) * | 2020-06-03 | 2021-12-07 | 阿里巴巴集团控股有限公司 | Data processing method and device, electronic equipment and storage medium |
| CN114610774A (en) * | 2022-02-10 | 2022-06-10 | 中远海运散货运输有限公司 | Method, device, electronic equipment and medium for analyzing ship passing through selected area |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104063486A (en) * | 2014-07-03 | 2014-09-24 | 四川中亚联邦科技有限公司 | Big data distributed storage method and system |
| CN105550241A (en) * | 2015-12-07 | 2016-05-04 | 珠海多玩信息技术有限公司 | Multidimensional database query method and apparatus |
| CN106610999A (en) * | 2015-10-26 | 2017-05-03 | 北大方正集团有限公司 | Query processing method and device |
| CN106649878A (en) * | 2017-01-07 | 2017-05-10 | 陈翔宇 | Artificial intelligence-based internet-of-things entity search method and system |
| CN107315744A (en) * | 2016-04-26 | 2017-11-03 | 杭州海康威视数字技术股份有限公司 | A kind of data query method and device |
-
2018
- 2018-05-25 CN CN201810517829.7A patent/CN108874907A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104063486A (en) * | 2014-07-03 | 2014-09-24 | 四川中亚联邦科技有限公司 | Big data distributed storage method and system |
| CN106610999A (en) * | 2015-10-26 | 2017-05-03 | 北大方正集团有限公司 | Query processing method and device |
| CN105550241A (en) * | 2015-12-07 | 2016-05-04 | 珠海多玩信息技术有限公司 | Multidimensional database query method and apparatus |
| CN107315744A (en) * | 2016-04-26 | 2017-11-03 | 杭州海康威视数字技术股份有限公司 | A kind of data query method and device |
| CN106649878A (en) * | 2017-01-07 | 2017-05-10 | 陈翔宇 | Artificial intelligence-based internet-of-things entity search method and system |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020192534A1 (en) * | 2019-03-26 | 2020-10-01 | 华为技术有限公司 | Search method, terminal, and medium |
| CN112241443A (en) * | 2019-07-16 | 2021-01-19 | 中国移动通信集团浙江有限公司 | Data quality monitoring method and device, computing equipment and computer storage medium |
| CN112241443B (en) * | 2019-07-16 | 2023-11-21 | 中国移动通信集团浙江有限公司 | Data quality monitoring method, device, computing equipment and computer storage medium |
| CN111159512A (en) * | 2019-12-30 | 2020-05-15 | 北京明略软件系统有限公司 | Processing method and server of track information |
| CN111159512B (en) * | 2019-12-30 | 2023-02-28 | 北京明略软件系统有限公司 | Processing method and server of track information |
| CN113760394A (en) * | 2020-06-03 | 2021-12-07 | 阿里巴巴集团控股有限公司 | Data processing method and device, electronic equipment and storage medium |
| CN112035581A (en) * | 2020-08-21 | 2020-12-04 | 北京字节跳动网络技术有限公司 | Model-based task processing method, device, equipment and medium |
| CN112783922A (en) * | 2021-02-01 | 2021-05-11 | 广州海量数据库技术有限公司 | Query method and device based on relational database |
| CN112783922B (en) * | 2021-02-01 | 2022-02-25 | 广州海量数据库技术有限公司 | Query method and device based on relational database |
| CN112732821A (en) * | 2021-03-31 | 2021-04-30 | 成都新希望金融信息有限公司 | Data storage method, device and equipment and storage medium |
| CN112732821B (en) * | 2021-03-31 | 2021-07-06 | 成都新希望金融信息有限公司 | Data storage method, device and equipment and storage medium |
| CN114610774A (en) * | 2022-02-10 | 2022-06-10 | 中远海运散货运输有限公司 | Method, device, electronic equipment and medium for analyzing ship passing through selected area |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108874907A (en) | A kind of data query method and apparatus, computer readable storage medium | |
| US9524317B2 (en) | Optimization of aggregate queries in database management systems using an early out join when processing min and max functions | |
| US9229982B2 (en) | Processing queries using oriented query paths | |
| US7895187B2 (en) | Hybrid evaluation of expressions in DBMS | |
| US6029163A (en) | Methods for collecting query workload based statistics on column groups identified by RDBMS optimizer | |
| US11436225B2 (en) | Database hierarchy-independent data drilling | |
| US10140325B2 (en) | Data source identification mapping in blended data operations | |
| US20020099688A1 (en) | Method and system for handling foreign key update in an object-oriented database environment | |
| US9317552B2 (en) | Reusing existing query plans in a database system | |
| EP2657860A1 (en) | Data flow graph optimization using rules | |
| US8903806B2 (en) | Matching queries to data operations using query templates | |
| US8090700B2 (en) | Method for updating databases | |
| CN103678589B (en) | Database kernel query optimization method based on equivalence class | |
| US20170193058A1 (en) | System and Method for Performing Blended Data Operations | |
| CN115292508B (en) | Knowledge graph construction method and system based on table data | |
| US20150007202A1 (en) | Programming Framework for Applications | |
| EP2019362A2 (en) | Spatial data validation systems | |
| KR101288208B1 (en) | System of entity-relationship model reformulation of sparql query results on rdf data and the method | |
| EP3679479A1 (en) | System for detecting data relationships based on sample data | |
| CN110807042B (en) | SQL statement generation system and method | |
| US8423523B2 (en) | Apparatus and method for utilizing context to resolve ambiguous queries | |
| CN113918606B (en) | Log query method and device | |
| US20110029480A1 (en) | Method of Compiling Multiple Data Sources into One Dataset | |
| US11636113B2 (en) | Method for performing multi-caching on data sources of same type and different types by using cluster-based processing system and device using the same | |
| US11593366B2 (en) | Techniques for pushing joins into union all views |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20181123 |
|
| WD01 | Invention patent application deemed withdrawn after publication |