CN109033063A - The machine inference of knowledge based map - Google Patents
The machine inference of knowledge based map Download PDFInfo
- Publication number
- CN109033063A CN109033063A CN201710433308.9A CN201710433308A CN109033063A CN 109033063 A CN109033063 A CN 109033063A CN 201710433308 A CN201710433308 A CN 201710433308A CN 109033063 A CN109033063 A CN 109033063A
- Authority
- CN
- China
- Prior art keywords
- node
- subgraph
- path
- sentence
- natural language
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2455—Query execution
- G06F16/24564—Applying rules; Deductive queries
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
- G06F40/295—Named entity recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/242—Query formulation
- G06F16/243—Natural language query formulation
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Mathematical Physics (AREA)
- Audiology, Speech & Language Pathology (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Machine Translation (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
Embodiment of the disclosure is related to the machine inference of knowledge based map.In some embodiments, a method of computer implementation is provided.This method includes in response to receiving natural language sentence, the determining predefined sentence template with the natural language statement matching.Multiple items are extracted from the natural language sentence based on the predefined sentence template.Knowledge mapping is inquired using the multiple item to obtain the reasoning for the natural language sentence, the semantic correctness of the problem or the verifying sentence related with the sentence is answered in the reasoning, and the knowledge mapping includes the node of presentation-entity or concept and the side for indicating the logical relation between node.
Description
Background technique
With the development of network, knowledge obtained by the mankind is more and more abundant, and type is also very numerous and jumbled.Data can be passed through
The form in library is correspondingly handled these knowledge to manage these knowledge, to promote human-computer interaction.For example, people can be with
To machine query, whether some logic is true in the form of natural language, or is putd question to machine to seek the solution to problem
Answer etc..Machine is basis and the important component of artificial intelligence to the understanding and processing of Human Natural Language.Although
Propose several machine interaction techniques based on natural language, but still reasoning from logic energy of the machine to Human Natural Language at present
Power is still obvious insufficient.
Summary of the invention
According to some realizations, a kind of electronic equipment is provided.The equipment includes: processing unit；And memory, it is coupled to
Processing unit and it is stored with instruction.Instruction executes following movement when being executed by processing unit: in response to receiving nature language
Speech sentence, the determining predefined sentence template with natural language statement matching；Based on predefined sentence template from natural language language
Sentence extracts multiple items；The reasoning for natural language sentence is obtained using multiple inquiry knowledge mappings, the reasoning is answered
The semantic correctness of the problem or the verifying sentence related with the sentence, knowledge mapping includes presentation-entity or concept
Node and the side for indicating the logical relation between node.
There is provided Summary is their specific realities below in order to which simplified form introduces the selection to concept
Applying in mode will be further described.Summary is not intended to identify the key feature of claimed theme or main special
Sign, is also not intended to limit the range of claimed theme.
Detailed description of the invention
Fig. 1 is shown in which that one or more exemplary computing system/server frames realized of the disclosure can be implemented
Figure；
Fig. 2 shows the block diagrams for one or more exemplary architectures realized that the disclosure wherein can be implemented；
Fig. 3 shows a part of one or more example knowledge mappings realized according to the disclosure；
Fig. 4 shows a part of one or more example knowledge mappings realized according to the disclosure；
Fig. 5 a and Figure 5b shows that a part according to one or more example knowledge mappings realized of the disclosure；And
Fig. 6 shows the flow chart of one or more logic reasonings realized according to the disclosure.
In these attached drawings, same or similar reference symbol is for indicating same or similar element.
Specific embodiment
The disclosure is discussed now with reference to several example implementations.It realizes it should be appreciated that discussing these merely to making
Obtaining those of ordinary skill in the art better understood when and therefore realize the disclosure, rather than imply to the range of this theme
Any restrictions.
As used herein, term " includes " and its variant will be read as meaning opening " including but not limited to "
Put formula term.Term "or" will be read as "and/or", unless the clear in addition instruction of context.Term "based" will be solved
It reads to be " being based at least partially on ".Term " realization " and " a kind of realization " will be read as " at least one realization ".Term
" another realization " will be read as " at least one other realization ".Term " first ", " second " etc. may refer to different
Or identical object.Hereafter it is also possible that other specific and implicit definition.Unless additionally explicitly pointing out, term is determined
Justice is consistent through specification.
Fig. 1 is shown in which that one or more exemplary computing system/servers 100 realized of the disclosure can be implemented
Block diagram.Computing system/server 100 shown in fig. 1 is only example, should not constitute the use to realization described herein
Function and range limitation.
As shown in Figure 1, computing system/server 100 is the form of universal computing device.Computing system/server 100
Component can include but is not limited to one or more processors or processing unit 100, memory 120, one or more are inputted and set
Standby 130, one or more output equipments 140, storage device 150 and one or more communication units 160.Processing unit 100 can
To be reality or virtual processor and can persistently execute various processing according to what is stored in memory 120.In multiprocessing
In system, multiplied unit executes computer executable instructions, to increase processing capacity.
Computing system/server 100 generally includes multiple computer medias.Such medium can be computing system/clothes
It is engaged in the addressable any medium that can be obtained of device 100, including but not limited to volatile and non-volatile media, detachably and not
Detachable media.Memory 120 can be volatile memory (such as register, cache, random access storage device
(RAM)), nonvolatile memory (for example, read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM),
Flash memory) or they certain combination.Storage device 150 can be detachable or non-dismountable, and may include machine readable
Medium, such as flash drive, disk or any other medium can be used in storing information and can be calculating
It is accessed in system/server 100.
Computing system/server 100 may further include other detachable/non-dismountable, volatile, nonvolatile
Computer system storage medium.Although not shown in FIG. 1, can provide for being read from detachable, non-volatile magnetic disk
The disk drive for taking or being written and the disc drives for being read out or being written from detachable, anonvolatile optical disk.At these
In situation, each driving can be connected to bus by one or more data media interfaces.Memory 120 may include at least
One program product, has (for example, at least one) program module collection, these program modules are configured as executing and be retouched herein
The function for the various realizations stated.
(for example, at least one) program module 122 can be stored in such as memory 120.Such program module
122 can include but is not limited to operating system, one or more application program, other program modules and operation data.These show
Each example or specific combination in example may include the realization of networked environment.Program module 122 can be used to execute the disclosure
Realization function and/or method.
Input equipment 130 can be one or more various input equipments.For example, input unit 130 may include user
Equipment, mouse, keyboard, trackball etc..The realization of communication unit 160 is led to other computational entity on the communication media
Letter.Additionally, the function of the component of computing system/server 100 can be come real with single computing cluster or multiple computing machines
Existing, these computing machines can fetch communication by communication link.Therefore, computing system/server 100 can be used with one or
The logical connection of other multiple servers, NetPC Network PC (PC) or another general networking node is come in networked environment
In operated.Such as, but not limited to, communication media includes wired or Wireless networking technologies.
Computing system/server 100 can also be communicated with one or more external equipment (not shown) as needed,
External equipment such as stores equipment, display equipment etc., with one or more user and computing system/server 100 is handed over
Mutual equipment is communicated, or with make appointing for computing system/server 100 and other one or more computing device communications
What equipment (for example, network interface card, modem etc.) is communicated.Such communication can be via input/output (I/O) interface
(not shown) executes.
As shown in Figure 1, program module 122 can receive natural language sentence input by user, example from input equipment 150
Such as, problem " why Einstein can think deeply? ".In addition, program module 122 can also be received from storage device 150 and knowledge
The related data of map, and obtain the reasoning results for being directed to sentence.Then program module 122 is pushed away via the output of output equipment 160
Reason is as a result, for example " Einstein is people；People has brain；Big brain-capacity thinking."
Fig. 2 shows the schematic diagrames for the exemplary architectures 200 realized according to the one or more of the disclosure.In exemplary architecture
In 200, knowledge mapping 210 is stored in memory in the form of chart database, such as memory shown in FIG. 1 120 or storage
Device 150.Knowledge mapping 210 can also be realized by distributed storage environment, to accommodate the chart database of larger capacity.With
Family can input natural language sentence by input equipment 130.In some implementations, the natural language sentence of user's input can be with
Natural language problem, for example, " why Ai Er Bert Einstein can think deeply? " figure engine 220 can be used to answer
This problem obtains the reasoning for being directed to this problem.Natural language sentence is also possible to other kinds of sentence, for example, " ratio
Your Gates has set up Microsoft." figure engine 220 can be used to judge whether the sentence true, that is, obtain be directed to the sentence
Reasoning.
Figure engine 220 is in response to receiving natural language sentence, the determining predefined sentence with natural language statement matching
Template.Figure engine 220 can processing unit 110 as shown in Figure 1 realize, can also be realized by distributed computing environment, with
Increase the requirement of calculated performance.Predefined sentence template can have multiple types, for example, " why A can B? ", " how is A
B? ", " A is B? " etc..In some implementations, it can be determined by the degree of correlation between natural language sentence and template certainly
Right language statement matches with which template.For example, " why Ai Er Bert Einstein can think natural language problem
Examine? " comprising keyword " can with ", therefore, the natural language problem and predefined sentence template " why A can B? " the degree of correlation compared with
Height can determine the natural language problem and this predefined sentence template matching.
Based on predefined sentence template, figure engine 220 extracts multiple items from natural language sentence.These may include
The item specified in predefined sentence template.For example, template " why A can B? " specify an A and B, template " A how B? " it is specified
Item A and B.Therefore, from natural language problem " why Ai Er Bert Einstein can think deeply? " in, it can extract
" Ai Er Bert Einstein " and " thinking " two.
Figure engine 220 is by obtaining the reasoning for natural language sentence using multiple inquiry knowledge mappings.Knowledge
Map includes datagram, the side of the logical relation between node and expression node with presentation-entity or concept.Data
It can have entity (also referred to as true) and concept (also referred to as common sense) in figure.In some implementations, knowledge mapping can also include patrolling
Volume rule schema, have the node for indicating abstract identifier and between logical expression side, such as super side.Super side is to connect
Connect the side of multiple nodes.
In order to the principle of the disclosure is described in further detail, in particular how utilizes and extracted from natural language sentence
Knowledge mapping is inquired, the expression of knowledge mapping is introduced below in conjunction with Fig. 3, it illustrates according to one or more of the disclosure
A part of the example knowledge mapping 300 of a realization.In the example depicted in fig. 3, knowledge mapping 300 includes three layers, entity
Layer 350, conceptual level 360 and logic rules layer 370, can also be known respectively as sterogram, concept map and logic rules figure.From reality
Body portion 350 arrives logic rules layer 370, gradually becomes more to be abstracted, correspondingly, data volume also gradually decreases.Physical layer 350
It can be linked together with conceptual level 360 by the relationship between node, to form an entirety, referred to as foregoing data
Figure.It will be appreciated, however, that the structure of Fig. 3 is merely exemplary, the expression of knowledge mapping can also include more or less
Layer or node.For example, can only include conceptual level 360 and logic rules layer 370 in the structure in figure 3.In some implementations,
Differentiation can also be not added in physical layer 350 and conceptual level 360.
Physical layer 350 is also referred to as sterogram, the side of the logical relation between node and presentation-entity including presentation-entity.
Entity can indicate specific things present in real world, for example, Ai Er Bert Einstein, Bill Gates etc..
Each entity node records each attribute relevant to the node or information.In the example of fig. 3, the work of " Paar " 351 is shown
For the entity node in physical layer 350.
Conceptual level 360 is also referred to as concept map, including indicating the node of concept and indicating the side of the logical relation between concept.Generally
Reading, which can be, is abstracted one kind of entity, for example, people, animal, dog etc..For example, Fig. 3 show " barking " 361, " animal " 362,
" dog " 363, " people " 364 and " performer " 365 are as the concept node in conceptual level 360.
Connect as described above, can be established by the logical relation between each node between physical layer 350 and conceptual level 360
It connects, so that physical layer 350 and conceptual level 360 are merged into an organic whole.This can be by opening existing and/or future
Any method of hair realizes that the disclosure is unrestricted herein.
Logic rules layer 370 includes various logic rule, and each logic rules may include abstract node and indicate section
The super side of relationship between point.Each abstract node also can specify corresponding attribute, can not also specify any attribute.Such as
Shown in Fig. 3, logic rules layer 370 includes multiple nodes, respectively by abstract identifier "? A " 310, "? B " 330, "? C " 340,
"? X " 322, "? Y " 324, "? Z " 326 etc. indicate.Relationship between node indicates by logical expression, such as: A does not have X
(A NotHasA X)；The premise of C is X (C HasPrerequisite X).In logic rules layer 370, it can predefine more
A logic rules, each logic rules can be indicated by a subgraph.For example, as shown in figure 3, for " why B can be with
C? ", which includes node 320, node 330 and node 340, define node 320 can C, node 320 is a part of B.
In some implementations, knowledge mapping 300 can be recurrence vehicle mixing (RHHG), for expressing in a uniform manner
Various types of knowledge in the real world, these knowledge can have various appearance forms.
" recurrence " expression carrys out organization knowledge in a hierarchical fashion, and each node of figure can be another figure.For example, data
At least part node in figure can be one or more datagrams, for example, one or more concept maps, one or more real
Body figure, and/or the combination of one or more concept maps and sterogram.For example, " city " node may include " Beijing " subgraph and
" Shanghai " subgraph." Beijing " subgraph may include the side of the logical relation between the nodes such as " Haidian ", " southern exposure " and expression.
" Haidian " node can also include further subgraph.Recurrence plot can imitate the organizational form of knowledge in real world.
Knowledge in " mixing " expression figure can be isomery.For example, the relationship between the node indicated by side can
Be it is deterministic, be also possible to probabilistic.For example, the side between " dog " 363 and " barking " 360 can be represented as " can,
0.8 ", that is, " dog " 363 can bark with 80% probability.Correspondingly, during traversing graph, may have 80% it is general
Rate can traverse " barking " 360 from " dog " 363.Side between " dog " 363 and " animal " 362 can be represented as " be, 0.98 ",
That is, " dog " 363 is animal, etc. with 98% probability.It, can be with during traversing graph by using probabilistic side
Relatively unessential information is filtered, to save calculation amount.
In addition, " mixing " also may indicate that the side of figure can be explicit relation, it is also possible to implicit relationship.Explicit relation can
Be such as it is as described above " can ", "Yes".Implicit relationship can be A is related to B, and A and B occurs simultaneously etc., this
Relationship can be indicated by statistical model (for example, neural network model).For example, by void between " dog " 363 and " performer " 365
The relationship that line indicates can be implicit relationship, because usually dog will not be performer.Implicit relationship indicates " dog " 363 and " performer "
The association for having certain implicit between 365, this association possibly can not or be not suitable for indicating using above-mentioned explicit relation.
In " hypergraph " expression figure when can be super, super side can connect multiple nodes.For example, Zhang San, Li Si and king
Fifth is that friend, can be linked together these three people by super side, to indicate the correlation between three people.
Recurrence vehicle mixing can by various types of instructions by unified form expression in a figure, consequently facilitating
To operating for knowledge.
Continue introduce how knowledge based map obtains reasoning below in conjunction with Fig. 3 and Fig. 4.It in some implementations, can be with
Determine multiple nodes relevant to the multiple difference extracted from natural language sentence in the knowledge mapping.For example, for
Problem " why Ai Er Bert Einstein can think deeply? ", node " Ai Erbai can be determined in knowledge mapping 300
Special Einstein " and " thinking ".The path comprising these nodes, the road are determined from knowledge mapping 300 (for example, datagram)
Diameter includes a part of side in knowledge mapping 300 (for example, datagram).
In some implementations, it can determine relevant to extracted item difference in knowledge mapping 300 (for example, datagram)
Node.Then, the path comprising these nodes is determined from knowledge mapping 300 (for example, datagram), path includes knowledge mapping
A part of side in 300.Then, it is determined based on logical relation represented by the side that path includes corresponding with natural language sentence
Reasoning.
In some implementations, determine that matched path can be by by the subgraph and data of logic rules in knowledge mapping
Subgraph carries out pattern match to obtain.For example, can determine subgraph corresponding with predefined sentence template in logic rules layer.
As shown in figure 3, for natural language problem " why B can be with C? ", which includes node 320, node 330 and node 340,
Define node 320 can C, node 320 is a part of B.
Based on the node extracted from natural language sentence, the data Layer with the subgraph match of logic rules layer can be determined
Subgraph.The subgraph of data Layer includes above-mentioned node.Then, the subgraph based on matched data Layer can determine and natural language
The corresponding path of speech sentence.
During pattern match, due to being stored in the form of chart database, knowledge mapping is supported to grasp graph traversal
Make, thus, pattern match can be by realizing graph traversal.In such a case, it is possible to the efficiency of parallel computation is improved,
Correspondingly, the capacity of figure can be increased.It will be appreciated, however, that the pattern match between figure can pass through any existing or general
It is realized the method developed, the disclosure is unrestricted herein.
As described above, at least part in knowledge mapping when indicating these represented by relationship probability.This
In the case of, multiple subgraphs with the subgraph match of logic rules layer can be determined from datagram by pattern match.Then, may be used
Probability with the relationship indicated based on side, determines the matching degree of the subgraph of these subgraphs and logic rules layer, and by matching degree height
It is determined as the subgraph of data Layer in the subgraph of predetermined threshold.For example, the subgraph with highest matching degree can be determined as data
The subgraph of layer.
Fig. 4 shows a part of one or more example knowledge mappings 400 realized according to the disclosure.Knowledge mapping
400 show for natural language sentence " why Einstein can think deeply, and computer cannot be thought deeply? " it is obtained
Subgraph.
Specifically, receiving natural language problem, " why Einstein can think deeply, and computer cannot be thought
Examine? " later, figure engine 220 by the natural language sentence in knowledge mapping predefined template or rule match, that is, Fig. 3
Shown in logic rules layer 370 rule " why B can be with C? " " why A cannot C? ".Then, by rules layer
Subgraph corresponding with above-mentioned rule and data Layer in 370 carry out pattern match, to determine the subgraph to match, as shown in Figure 4.
In knowledge mapping 400, in computer 410 and logic rules figure "? A " 310 matchings, " people " 430 and logic are advised
Then in figure "? B " 330 matchings, and in " thinking " 440 and logic rules layer "? C " 340 matchings.In addition, " brain " 422 with "?
X " 322 is matched, " cerebral cortex " 424 with "? Y " 324 matchings, " neuron " 426 with "? Z " 326 matchings.
Based on the matched knowledge mapping 400 of logic rules, the reasoning for natural language sentence can be obtained.Specifically
Ground, as shown in figure 4, " brain " 422 " can think deeply " 440, " people " 430 has " brain " 422, " Ai Er Bert Einstein "
450 be " people " 430.Therefore, " Ai Er Bert Einstein " 450 can " thinking " 440.Similarly, " thinking " 440 needs " big
Brain " 422, " computer " 410 do not have " brain " 422.Therefore, " computer " 410 " cannot think deeply " 440.
In some implementations, since the relationship between node has transitivity, multiple nodes can about be turned to two sections
Point and its between relationship.Fig. 5 a and Figure 5b shows that one or more example knowledge mappings 500 realized according to the disclosure
With 550 schematic diagram.As shown in Figure 5 a, " Paar " 510 is " dog " 520, and " dog " 520 is " animal " 530.Due to "Yes"
Relationship have transitivity, Figure 50 0 can be converted into Figure 55 0, be " animal " 530 there is shown with " Paar " 510.This can be real
The logical deduction of existing transitive relation.
Such as, if it is determined that include between two nodes (for example, " Paar " 510 and " animal " 530) in datagram 500
Intermediate node (for example, " dog " 520) can determine relationship and " dog " 520 between " Paar " 510 and " animal " 530 and " move
Whether the relationship between object " 530 has transitivity.Since relationship between the two is "Yes" (" isA "), then this can be determined
Two relationships have transitivity.In such a case, it is possible to knowledge mapping 500 is converted into knowledge mapping 550, it correspondingly, can
That will be " Paar " 510 to the path of " animal " 530 to the path integration of " animal " 530 from " Paar " 510 to " dog " 520.
Fig. 6 shows the flow chart of one or more logic reasonings 600 realized according to the disclosure.Method 600
It can be executed by figure engine 220 or processing unit 110, the disclosure is unrestricted herein.
602, in response to receiving natural language sentence, the determining predefined sentence mould with natural language statement matching
Plate.In some implementations, natural language sentence can be natural language problem, such as " why Einstein can think deeply? ".
604, multiple items are extracted from natural language sentence based on predefined sentence template.
606, by obtaining the reasoning for natural language sentence, knowledge mapping using multiple inquiry knowledge mappings
The side of logical relation between node and expression node including presentation-entity or concept.Knowledge mapping can be with chart database
Formation be stored at such as memory 120 or storage device 150.
In some implementations, it obtains reasoning and comprises determining that multiple nodes relevant to multiple difference in knowledge mapping；From
Determined in knowledge mapping include multiple nodes path, path by knowledge mapping while it is a part of while constitute；And it is based on
Logical relation represented by the side that path includes determines reasoning.
In some implementations, knowledge mapping includes logic rules layer and data Layer, and data Layer includes presentation-entity or concept
Node, and logic rules layer include indicate logic rules subgraph, and determine path include: logic rules layer determine
Subgraph corresponding with predefined sentence template；Based on multiple nodes, the determining data Layer with the subgraph match of logic rules layer
Subgraph；And the subgraph based on data Layer, determine path.
In some implementations, at least part in path while instruction while represented by relationship probability, and determine number
Multiple subgraphs with the data Layer of the subgraph match of logic rules layer are comprised determining that according to the subgraph of layer；Based on probability, determine multiple
The matching degree of the subgraph of subgraph and logic rules layer；And the subgraph that matching degree is higher than predetermined threshold is determined as to the son of data Layer
Figure.
In some implementations, the subgraph for determining data Layer includes: in response to determining the first node and second in datagram
Include intermediate node between node, determines the first relationship and intermediate node and second node between first node and intermediate node
Between the second relationship whether there is transitivity；And in response to determining that the first relationship and the second relationship have transitivity, by the
One node is first node to the path of second node via the path integration of intermediate node to second node.
Function described herein can be executed at least partly by one or more hardware logic components.Such as but
It is not limited to, the exemplary types for the hardware logic component that can be used include field programmable gate array (FPGA), dedicated integrated
Circuit (ASIC), Application Specific Standard Product (ASSP), system on chip (SOC), Complex Programmable Logic Devices (CPLD) etc..
For implement disclosed method program code can using any combination of one or more programming languages come
It writes.These program codes can be supplied to the place of general purpose computer, special purpose computer or other programmable data processing units
Device or controller are managed, so that program code makes defined in flowchart and or block diagram when by processor or controller execution
Function/operation is carried out.Program code can be executed completely on machine, partly be executed on machine, as stand alone software
Is executed on machine and partly execute or executed on remote machine or server completely on the remote machine to packet portion.
In the context of present disclosure, machine readable media can be tangible medium, may include or stores
The program for using or being used in combination with instruction execution system, device or equipment for instruction execution system, device or equipment.Machine
Device readable medium can be machine-readable signal medium or machine-readable storage medium.Machine readable media may include but unlimited
In times of electronics, magnetic, optical, electromagnetism, infrared or semiconductor system, device or equipment or above content
What appropriate combination.The more specific example of machine readable storage medium will include the electrical connection of line based on one or more, portable
Formula computer disks, hard disk, random access memory (RAM), read-only memory (ROM), Erasable Programmable Read Only Memory EPROM
(EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage are set
Standby or above content any appropriate combination.
Although this should be understood as requiring operating in this way with shown in addition, depicting each operation using certain order
Certain order out executes in sequential order, or requires the operation of all diagrams that should be performed to obtain desired result.
Under certain environment, multitask and parallel processing be may be advantageous.Similarly, although containing several tools in being discussed above
Body realization thanks to elder sister, but these are not construed as the limitation to the scope of the present disclosure.In the context individually realized
Certain features of description can also be realized in combination in single realize.On the contrary, described in the context individually realized
Various features can also be realized individually or in any suitable subcombination in multiple realizations.
It is listed below some example implementations of theme described herein.
According to some realizations, a method of computer implementation is provided.This method comprises: in response to receiving nature language
Speech sentence, the determining predefined sentence template with the natural language statement matching；Based on the predefined sentence template from institute
Predicate sentence extracts multiple items；The reasoning for the natural language sentence is obtained using the multiple item inquiry knowledge mapping,
The semantic correctness of the problem or the verifying sentence related with the sentence is answered in the reasoning, and the knowledge mapping includes
The side of logical relation between presentation-entity or the node and expression node of concept.
In some implementations, obtain the reasoning comprise determining that in the knowledge mapping to the multiple item respectively it is related
Multiple nodes；The path comprising the multiple node is determined from the knowledge mapping, the path is by the knowledge mapping
In it is described while it is a part of while constitute；And based on the logical relation represented by the side that path includes to determine
State reasoning.
In some implementations, the knowledge mapping includes logic rules layer and data Layer, and the data Layer includes indicating real
The node of body or concept, and the logic rules layer includes the subgraph for indicating logic rules, and determines that the path includes:
In determining the first subgraph corresponding with the predefined sentence template of the logic rules layer；Based on the multiple node,
Determining the second subgraph with first subgraph match in the data Layer；And the road is determined based on second subgraph
Diameter.
In some implementations, at least part in the path when indicating described represented by relationship probability, and
And determine that second subgraph comprises determining that multiple subgraphs with the data Layer of first subgraph match；Based on described
Probability determines the matching degree of the multiple subgraph Yu first subgraph；And it is matching degree is true higher than the subgraph of predetermined threshold
It is set to second subgraph.
In some implementations, determine second subgraph include: in response in the determination datagram first node and
Include intermediate node between second node, determine the first relationship between the first node and the intermediate node and it is described in
Whether the second relationship between intermediate node and the second node has transitivity；And in response to determination first relationship with
Second relationship has transitivity, by the first node via the intermediate node to the path integration of the second node
For the first node to the path of the second node.
In some implementations, the knowledge mapping is stored in the form of chart database.
In some implementations, the natural language sentence is natural language problem.
According to some realizations, a kind of electronic equipment is provided.The electronic equipment includes: processing unit；And memory, coupling
It is bonded to the processing unit and is stored with instruction, described instruction executes following movement when being executed by the processing unit: ringing
Ying Yu receives natural language sentence, the determining predefined sentence template with the natural language statement matching；Based on described pre-
Definition statement template extracts multiple items from the natural language sentence；Knowledge mapping is inquired using the multiple item to be directed to
The semanteme of the problem or the verifying sentence related with the sentence is answered in the reasoning of the natural language sentence, the reasoning
Correctness, the knowledge mapping include the node of presentation-entity or concept and the side for indicating the logical relation between node.
In some implementations, obtain the reasoning comprise determining that in the knowledge mapping to the multiple item respectively it is related
Multiple nodes；The path comprising the multiple node is determined from the knowledge mapping, the path is by the knowledge mapping
In it is described while it is a part of while constitute；And based on the logical relation represented by the side that path includes to determine
State reasoning.
In some implementations, the knowledge mapping includes logic rules layer and data Layer, and the data Layer includes indicating real
The node of body or concept, and the logic rules layer includes the subgraph for indicating logic rules, and determines that the path includes:
In determining the first subgraph corresponding with the predefined sentence template of the logic rules layer；Based on the multiple node,
Determining the second subgraph with first subgraph match in the data Layer；And it is based on second subgraph, determine the road
Diameter.
In some implementations, at least part in the path when indicating described represented by relationship probability, and
And determine that second subgraph comprises determining that multiple subgraphs with the data Layer of first subgraph match；Based on described
Probability determines the matching degree of the multiple subgraph Yu first subgraph；And it is matching degree is true higher than the subgraph of predetermined threshold
It is set to second subgraph.
In some implementations, determine second subgraph include: in response in the determination datagram first node and
Include intermediate node between second node, determine the first relationship between the first node and the intermediate node and it is described in
Whether the second relationship between intermediate node and the second node has transitivity；And in response to determination first relationship with
Second relationship has transitivity, by the first node via the intermediate node to the path integration of the second node
For the first node to the path of the second node.
In some implementations, the memory stores the knowledge mapping in the form of chart database.
In some implementations, the natural language sentence is natural language problem.
According to some realizations, a kind of computer program product is provided.The computer program product is visibly stored
In non-transitory, computer storage medium and including machine-executable instruction.The machine-executable instruction is executed by equipment
When make the equipment: in response to receiving natural language sentence, the determining predefined sentence with the natural language statement matching
Template；Multiple items are extracted from the natural language sentence based on the predefined sentence template；Known using the inquiry of the multiple item
Map is known to obtain the reasoning for the natural language sentence, and the reasoning is answered the problem related with the sentence or tested
The semantic correctness of the sentence is demonstrate,proved, the knowledge mapping includes the node of presentation-entity or concept and indicates between node
The side of logical relation.
In some implementations, obtain the reasoning comprise determining that in the knowledge mapping to the multiple item respectively it is related
Multiple nodes；The path comprising the multiple node is determined from the knowledge mapping, the path is by the knowledge mapping
In it is described while it is a part of while constitute；And based on the logical relation represented by the side that path includes to determine
State reasoning.
In some implementations, the knowledge mapping includes logic rules layer and data Layer, and the data Layer includes indicating real
The node of body or concept, and the logic rules layer includes the subgraph for indicating logic rules, and determines that the path includes:
In determining the first subgraph corresponding with the predefined sentence template of the logic rules layer；Based on the multiple node,
Determining the second subgraph with first subgraph match in the data Layer；And it is based on second subgraph, determine the road
Diameter.
In some implementations, at least part in the path when indicating described represented by relationship probability, and
And determine that second subgraph comprises determining that multiple subgraphs with the data Layer of first subgraph match；Based on described
Probability determines the matching degree of the multiple subgraph Yu first subgraph；And it is matching degree is true higher than the subgraph of predetermined threshold
It is set to second subgraph.
In some implementations, determine second subgraph include: in response in the determination datagram first node and
Include intermediate node between second node, determine the first relationship between the first node and the intermediate node and it is described in
Whether the second relationship between intermediate node and the second node has transitivity；And in response to determination first relationship with
Second relationship has transitivity, by the first node via the intermediate node to the path integration of the second node
For the first node to the path of the second node.
In some implementations, the knowledge mapping is stored in the form of chart database.
In some implementations, the natural language sentence is natural language problem.
Although having used specific to this theme of the language description of structure feature and/or method logical action, answer
When understanding that theme defined in the appended claims is not necessarily limited to special characteristic described above or movement.On on the contrary,
Special characteristic described in face and movement are only to realize the exemplary forms of claims.
Claims (20)
1. a method of computer implementation, comprising:
In response to receiving natural language sentence, the determining predefined sentence template with the natural language statement matching；
Multiple items are extracted from the natural language sentence based on the predefined sentence template；
Inquire knowledge mapping using the multiple item to obtain the reasoning for the natural language sentence, the reasoning answer with
The semantic correctness of the related problem of the sentence or the verifying sentence, the knowledge mapping includes presentation-entity or concept
Node and indicate node between logical relation side.
2. according to the method described in claim 1, wherein obtaining the reasoning and including:
Determine multiple nodes relevant to the multiple item difference in the knowledge mapping；
The path comprising the multiple node is determined from the knowledge mapping, the path is as described in the knowledge mapping
While it is a part of while constitute；And
The reasoning is determined based on the logical relation represented by the side that path includes.
3. according to the method described in claim 2, wherein the knowledge mapping includes logic rules layer and data Layer, the data
Layer includes the node of presentation-entity or concept, and the logic rules layer includes the subgraph for indicating logic rules, and is determined
The path includes:
In determining the first subgraph corresponding with the predefined sentence template of the logic rules layer；
Based on the multiple node, determining the second subgraph with first subgraph match in the data Layer；And
The path is determined based on second subgraph.
4. according to the method described in claim 3, wherein at least part in the path when indicating described represented by
The probability of relationship, and determine that second subgraph includes:
Determining multiple subgraphs with the data Layer of first subgraph match；
Based on the probability, the matching degree of the multiple subgraph Yu first subgraph is determined；And
The subgraph that matching degree is higher than predetermined threshold is determined as second subgraph.
5. according to the method described in claim 3, wherein determining that second subgraph includes:
In response to including intermediate node between the first node and second node in the determination datagram, the first segment is determined
Whether the second relationship between the first relationship and the intermediate node and the second node between point and the intermediate node
With transitivity；And
There is transitivity in response to determination first relationship and second relationship, by the first node via the centre
The path integration of node to the second node is the first node to the path of the second node.
6. according to the method described in claim 1, wherein the knowledge mapping is stored in the form of chart database.
7. according to the method described in claim 1, wherein the natural language sentence is natural language problem.
8. a kind of electronic equipment, comprising:
Processing unit；And
Memory is coupled to the processing unit and is stored with instruction, and described instruction is held when being executed by the processing unit
The following movement of row:
In response to receiving natural language sentence, the determining predefined sentence template with the natural language statement matching；
Multiple items are extracted from the natural language sentence based on the predefined sentence template；
Inquire knowledge mapping using the multiple item to obtain the reasoning for the natural language sentence, the reasoning answer with
The semantic correctness of the related problem of the sentence or the verifying sentence, the knowledge mapping includes presentation-entity or concept
Node and indicate node between logical relation side.
9. equipment according to claim 8, wherein obtaining the reasoning and including:
Determine multiple nodes relevant to the multiple item difference in the knowledge mapping；
The path comprising the multiple node is determined from the knowledge mapping, the path is as described in the knowledge mapping
While it is a part of while constitute；And
The reasoning is determined based on the logical relation represented by the side that path includes.
10. equipment according to claim 9, wherein the knowledge mapping includes logic rules layer and data Layer, the number
It include the node of presentation-entity or concept according to layer, and the logic rules layer includes the subgraph for indicating logic rules, and really
The fixed path includes:
In determining the first subgraph corresponding with the predefined sentence template of the logic rules layer；
Based on the multiple node, determining the second subgraph with first subgraph match in the data Layer；And
Based on second subgraph, the path is determined.
11. equipment according to claim 10, wherein at least part in the path when indicating described represented by
Relationship probability, and determine that second subgraph includes:
Determining multiple subgraphs with the data Layer of first subgraph match；
Based on the probability, the matching degree of the multiple subgraph Yu first subgraph is determined；And
The subgraph that matching degree is higher than predetermined threshold is determined as second subgraph.
12. equipment according to claim 10, wherein determining that second subgraph includes:
In response to including intermediate node between the first node and second node in the determination datagram, the first segment is determined
Whether the second relationship between the first relationship and the intermediate node and the second node between point and the intermediate node
With transitivity；And
There is transitivity in response to determination first relationship and second relationship, by the first node via the centre
The path integration of node to the second node is the first node to the path of the second node.
13. equipment according to claim 8, wherein the memory stores the knowledge graph in the form of chart database
Spectrum.
14. equipment according to claim 8, wherein the natural language sentence is natural language problem.
15. a kind of computer program product, the computer program product is tangibly stored in non-transient computer storage and is situated between
In matter and including machine-executable instruction, the machine-executable instruction makes the equipment when being executed by equipment:
In response to receiving natural language sentence, the determining predefined sentence template with the natural language statement matching；
Multiple items are extracted from the natural language sentence based on the predefined sentence template；
Inquire knowledge mapping using the multiple item to obtain the reasoning for the natural language sentence, the reasoning answer with
The semantic correctness of the related problem of the sentence or the verifying sentence, the knowledge mapping includes presentation-entity or concept
Node and indicate node between logical relation side.
16. computer program product according to claim 15, wherein obtaining the reasoning and including:
Determine multiple nodes relevant to the multiple item difference in the knowledge mapping；
The path comprising the multiple node is determined from the knowledge mapping, the path is as described in the knowledge mapping
While it is a part of while constitute；And
The reasoning is determined based on the logical relation represented by the side that path includes.
17. computer program product according to claim 16, wherein the knowledge mapping includes logic rules layer sum number
According to layer, the data Layer includes the node of presentation-entity or concept, and the logic rules layer includes indicating logic rules
Subgraph, and determine that the path includes:
In determining the first subgraph corresponding with the predefined sentence template of the logic rules layer；
Based on the multiple node, determining the second subgraph with first subgraph match in the data Layer；And
Based on second subgraph, the path is determined.
18. computer program product according to claim 17, wherein at least part side in the path indicates institute
The probability of relationship represented by side is stated, and determines that second subgraph includes:
Determining multiple subgraphs with the data Layer of first subgraph match；
Based on the probability, the matching degree of the multiple subgraph Yu first subgraph is determined；And
The subgraph that matching degree is higher than predetermined threshold is determined as second subgraph.
19. computer program product according to claim 17, wherein determining that second subgraph includes:
In response to including intermediate node between the first node and second node in the determination datagram, the first segment is determined
Whether the second relationship between the first relationship and the intermediate node and the second node between point and the intermediate node
With transitivity；And
There is transitivity in response to determination first relationship and second relationship, by the first node via the centre
The path integration of node to the second node is the first node to the path of the second node.
20. computer program product according to claim 15, wherein the knowledge mapping is deposited in the form of chart database
Storage.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710433308.9A CN109033063B (en) | 2017-06-09 | 2017-06-09 | Machine inference method based on knowledge graph, electronic device and computer readable storage medium |
| PCT/US2018/034017 WO2018226404A1 (en) | 2017-06-09 | 2018-05-23 | Machine reasoning based on knowledge graph |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710433308.9A CN109033063B (en) | 2017-06-09 | 2017-06-09 | Machine inference method based on knowledge graph, electronic device and computer readable storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109033063A true CN109033063A (en) | 2018-12-18 |
| CN109033063B CN109033063B (en) | 2022-02-25 |
Family
ID=63643046
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710433308.9A Active CN109033063B (en) | 2017-06-09 | 2017-06-09 | Machine inference method based on knowledge graph, electronic device and computer readable storage medium |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN109033063B (en) |
| WO (1) | WO2018226404A1 (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109766453A (en) * | 2019-01-18 | 2019-05-17 | 广东小天才科技有限公司 | A kind of method and system of user's corpus semantic understanding |
| CN110008413A (en) * | 2019-03-14 | 2019-07-12 | 海信集团有限公司 | A kind of traffic trip problem querying method and device |
| CN110245253A (en) * | 2019-05-21 | 2019-09-17 | 华中师范大学 | A kind of Semantic interaction method and system based on environmental information |
| CN110717025A (en) * | 2019-10-08 | 2020-01-21 | 北京百度网讯科技有限公司 | Question answering method and device, electronic equipment and storage medium |
| CN111898004A (en) * | 2020-06-20 | 2020-11-06 | 中国建设银行股份有限公司 | Data mining method and device, electronic equipment and readable storage medium thereof |
| CN112580357A (en) * | 2019-09-29 | 2021-03-30 | 微软技术许可有限责任公司 | Semantic parsing of natural language queries |
| CN112581955A (en) * | 2020-11-30 | 2021-03-30 | 广州橙行智动汽车科技有限公司 | Voice control method, server, voice control system and readable storage medium |
| CN112818092A (en) * | 2020-04-20 | 2021-05-18 | 腾讯科技（深圳）有限公司 | Knowledge graph query statement generation method, device, equipment and storage medium |
| CN114238648A (en) * | 2021-11-17 | 2022-03-25 | 中国人民解放军军事科学院国防科技创新研究院 | Game countermeasure behavior decision method and device based on knowledge graph |
| WO2022194086A1 (en) * | 2021-03-16 | 2022-09-22 | International Business Machines Corporation | A neuro-symbolic approach for entity linking |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109840255B (en) * | 2019-01-09 | 2023-09-19 | 平安科技（深圳）有限公司 | Reply text generation method, device, equipment and storage medium |
| CN111831797B (en) * | 2019-04-19 | 2024-06-14 | 广东省智能制造研究所 | Management and recommendation system for manufacturing industry processing equipment model |
| CN110175226B (en) * | 2019-05-09 | 2021-06-08 | 厦门邑通软件科技有限公司 | Dialogue decision method based on multi-dimensional scene analysis |
| US11227018B2 (en) | 2019-06-27 | 2022-01-18 | International Business Machines Corporation | Auto generating reasoning query on a knowledge graph |
| CN113282720A (en) * | 2020-02-20 | 2021-08-20 | 清华大学 | Visual reasoning method and device |
| CN112541043A (en) * | 2020-12-24 | 2021-03-23 | 北京明略软件系统有限公司 | Method, device and equipment for detecting connectivity of nodes of knowledge graph |
| CN114297411B (en) * | 2022-03-08 | 2023-05-02 | 支付宝(杭州)信息技术有限公司 | Configuration method, analysis method and device of entities in knowledge graph |
| CN115827935B (en) * | 2023-02-09 | 2023-05-23 | 支付宝(杭州)信息技术有限公司 | Data processing method, device and equipment |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7013308B1 (en) * | 2000-11-28 | 2006-03-14 | Semscript Ltd. | Knowledge storage and retrieval system and method |
| CN104750795A (en) * | 2015-03-12 | 2015-07-01 | 北京云知声信息技术有限公司 | Intelligent semantic searching system and method |
| CN105787105A (en) * | 2016-03-21 | 2016-07-20 | 浙江大学 | Iterative-model-based establishment method of Chinese encyclopedic knowledge graph classification system |
| CN105868313A (en) * | 2016-03-25 | 2016-08-17 | 浙江大学 | Mapping knowledge domain questioning and answering system and method based on template matching technique |
-
2017
- 2017-06-09 CN CN201710433308.9A patent/CN109033063B/en active Active
-
2018
- 2018-05-23 WO PCT/US2018/034017 patent/WO2018226404A1/en active Application Filing
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7013308B1 (en) * | 2000-11-28 | 2006-03-14 | Semscript Ltd. | Knowledge storage and retrieval system and method |
| CN104750795A (en) * | 2015-03-12 | 2015-07-01 | 北京云知声信息技术有限公司 | Intelligent semantic searching system and method |
| CN105787105A (en) * | 2016-03-21 | 2016-07-20 | 浙江大学 | Iterative-model-based establishment method of Chinese encyclopedic knowledge graph classification system |
| CN105868313A (en) * | 2016-03-25 | 2016-08-17 | 浙江大学 | Mapping knowledge domain questioning and answering system and method based on template matching technique |
Non-Patent Citations (2)
| Title |
|---|
| ABDALGHANI ABUJABAL 等: "Automated Template Generation for Question Answering over Knowledge Graphs", 《PROCEEDINGS OF THE 26TH INTERNATIONAL CONFERENCE ON WORLD WIDE WEB》 * |
| 刘峤 等: "知识图谱构建技术综述", 《计算机研究与发展》 * |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109766453A (en) * | 2019-01-18 | 2019-05-17 | 广东小天才科技有限公司 | A kind of method and system of user's corpus semantic understanding |
| CN110008413A (en) * | 2019-03-14 | 2019-07-12 | 海信集团有限公司 | A kind of traffic trip problem querying method and device |
| CN110245253A (en) * | 2019-05-21 | 2019-09-17 | 华中师范大学 | A kind of Semantic interaction method and system based on environmental information |
| CN110245253B (en) * | 2019-05-21 | 2021-11-23 | 华中师范大学 | Semantic interaction method and system based on environmental information |
| CN112580357A (en) * | 2019-09-29 | 2021-03-30 | 微软技术许可有限责任公司 | Semantic parsing of natural language queries |
| CN110717025A (en) * | 2019-10-08 | 2020-01-21 | 北京百度网讯科技有限公司 | Question answering method and device, electronic equipment and storage medium |
| CN112818092B (en) * | 2020-04-20 | 2023-08-11 | 腾讯科技（深圳）有限公司 | Knowledge graph query statement generation method, device, equipment and storage medium |
| CN112818092A (en) * | 2020-04-20 | 2021-05-18 | 腾讯科技（深圳）有限公司 | Knowledge graph query statement generation method, device, equipment and storage medium |
| CN111898004A (en) * | 2020-06-20 | 2020-11-06 | 中国建设银行股份有限公司 | Data mining method and device, electronic equipment and readable storage medium thereof |
| CN112581955A (en) * | 2020-11-30 | 2021-03-30 | 广州橙行智动汽车科技有限公司 | Voice control method, server, voice control system and readable storage medium |
| CN112581955B (en) * | 2020-11-30 | 2024-03-08 | 广州橙行智动汽车科技有限公司 | Voice control method, server, voice control system, and readable storage medium |
| WO2022194086A1 (en) * | 2021-03-16 | 2022-09-22 | International Business Machines Corporation | A neuro-symbolic approach for entity linking |
| CN114238648A (en) * | 2021-11-17 | 2022-03-25 | 中国人民解放军军事科学院国防科技创新研究院 | Game countermeasure behavior decision method and device based on knowledge graph |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2018226404A1 (en) | 2018-12-13 |
| CN109033063B (en) | 2022-02-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109033063A (en) | The machine inference of knowledge based map | |
| Liu et al. | Federaser: Enabling efficient client-level data removal from federated learning models | |
| Geman et al. | Visual turing test for computer vision systems | |
| CN104346366B (en) | Extend the method and apparatus of test data | |
| CN110109940A (en) | By the figure traversal language conversion of functional expression at the structured query language of extension | |
| EP2795481A1 (en) | Detecting cases with conflicting rules | |
| CN108875057A (en) | For determining the method, apparatus, equipment and computer-readable medium of Data subject | |
| Yan et al. | A clustering algorithm for multi-modal heterogeneous big data with abnormal data | |
| US20230402125A1 (en) | Drug screening model construction method, a drug screening model construction device, a drug screening method, apparatus and a medium | |
| CN107818501A (en) | Calculating method and device | |
| CN104598216B (en) | The method and system write back for realizing the virtual data of business intelligence report | |
| Xu et al. | Understanding human mobility: A multi-modal and intelligent moving objects database | |
| Badu-Marfo et al. | A differentially private multi-output deep generative networks approach for activity diary synthesis | |
| Mahmud et al. | REVEALE: Reward Verification and Learning Using Explanations. | |
| Riesco et al. | A natural implementation of plural semantics in Maude | |
| CN113221578B (en) | Disease entity retrieval method, device, equipment and medium | |
| Porto et al. | A scientific hypothesis conceptual model | |
| US11829735B2 (en) | Artificial intelligence (AI) framework to identify object-relational mapping issues in real-time | |
| CN107633870A (en) | Data extraction method and device, storage medium, electronic equipment | |
| Rao | Effficient Graph-based Computation and Analytics | |
| Wen | Parallelizing Path Exploration and Optimizing Constraint Solving for Efficient Symbolic Execution | |
| Roopa et al. | Aadhaar Data Analysis Comparison in MapReduce, Hive and Spark | |
| Bylander et al. | Structured matching: A task-specific technique for making decisions | |
| Alwani | Fused Convolutional Neural Network Accelerators | |
| Savkovic et al. | Query Stability in Monotonic Data-Aware Business Processes [Extended Version] |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |