CN108076036A - It is a kind of to threaten vector extraction and the system and method for label for labelling - Google Patents
It is a kind of to threaten vector extraction and the system and method for label for labelling Download PDFInfo
- Publication number
- CN108076036A CN108076036A CN201710010774.6A CN201710010774A CN108076036A CN 108076036 A CN108076036 A CN 108076036A CN 201710010774 A CN201710010774 A CN 201710010774A CN 108076036 A CN108076036 A CN 108076036A
- Authority
- CN
- China
- Prior art keywords
- matrix
- metadata
- vector
- label
- calculating
- 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.)
- Withdrawn
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L63/00—Network architectures or network communication protocols for network security
- H04L63/14—Network architectures or network communication protocols for network security for detecting or protecting against malicious traffic
- H04L63/1441—Countermeasures against malicious traffic
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/10—File systems; File servers
- G06F16/16—File or folder operations, e.g. details of user interfaces specifically adapted to file systems
- G06F16/164—File meta data generation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
Abstract
Vector extraction and the system and method for label for labelling are threatened the invention discloses a kind of, including：Extraction module, for extracting metadata vector from data source, the whole metadata vector extracted in data source forms metadata matrix；Analysis module for being based on metadata matrix, is calculated with calculating matrix, analysis structure matrix of consequence；Comparing module is compared with pattern matrix for being based on matrix of consequence, obtains label-vector；Update module for updating label-vector to metadata matrix, and carries out label-vector corresponding metadata vector label for labelling, while is setting the value in recalculating metadata matrix and matrix of consequence in number.The extraction for solving the problem of to threaten vector reaches the purpose for threatening vector extraction and label for labelling to vector is threatened to carry out label for labelling.
Description
Technical field
The present invention relates to computer network security technology fields, relate more specifically to a kind of threat vector extraction and label mark
The system and method for note.
Background technology
At present, application of the label in community network information is increasingly extensive, by using label by data resource and user
It links together.The research of the degree of association helps to carry out systematic searching and browsing to information between label, while can also excavate
Go out the similitude between user, so as to carry out personalized recommendation to user.In information security field, to known threat or dive
All it is a kind of producer side ability all the time in the threat vector extraction of threat, the more educated of threat vector to extraction is also
It is short of very much.This causes user to rely only on the scheme of producer's offer, and can not the participation of active come in, and form oneself
Exclusive safe value system and security knowledge, this causes label exists in user individual application aspect to be short of.
The content of the invention
In order to solve the above technical problem, the present invention provides a kind of threat vector extractions according to the present invention and label mark
The system and method for note solves the problem of that the extraction of vector is threatened to reach and vector is threatened to carry to vector is threatened to carry out label for labelling
Take the purpose with label for labelling.
According to the first aspect of the invention, a kind of system for threatening vector extraction and label for labelling is provided.The system bag
It includes：Extraction module, for extracting metadata vector from data source, the whole metadata vector extracted in data source forms first number
According to matrix；Analysis module for being based on metadata matrix, is calculated with calculating matrix, analysis structure matrix of consequence；Compare mould
Block is compared with pattern matrix for being based on matrix of consequence, obtains label-vector；Update module, for by label-vector more
Newly to metadata matrix, and label for labelling is carried out to label-vector corresponding metadata vector, while in setting number again
Calculate the value in metadata matrix and matrix of consequence.
In some embodiments, further include：Index module is indexed metadata for being based on label.
In some embodiments, the analysis module, is additionally operable to：Define the calculation expression of the metadata and calculating matrix
Formula.
In some embodiments, the data source includes initial data, metadata, vector, wherein, the raw data packets
It includes binary file, binary system message, dynamic apis and performs records series, security log, dynamic apis monitored results.
In some embodiments, the element of the calculating matrix includes function, operator, numerical value, and the result of calculating includes cloth
Value of, character string, numerical value, matrix.
According to the second aspect of the invention, a kind of method for threatening vector extraction and label for labelling is provided, including：From data
Metadata vector is extracted in source, the whole metadata vector extracted in data source forms metadata matrix；Based on metadata matrix,
It is calculated with calculating matrix, analysis structure matrix of consequence；Be compared based on matrix of consequence with pattern matrix, obtain marking to
Amount；Label-vector is updated to metadata matrix, and label for labelling is carried out to label-vector corresponding metadata vector, while
The value in metadata matrix and matrix of consequence is recalculated in setting number.
In some embodiments, further include：Metadata is indexed based on label.
In some embodiments, it is described based on metadata matrix, it is calculated with calculating matrix, analysis structure result square
Battle array, further includes：
Define the calculation expression of the metadata and calculating matrix.
In some embodiments, the data source includes initial data, metadata, vector, wherein, the raw data packets
It includes binary file, binary system message, dynamic apis and performs records series, security log, dynamic apis monitored results.
In some embodiments, the element of the calculating matrix includes function, operator, numerical value, and the result of calculating includes cloth
Value of, character string, numerical value, matrix.
Technical solution provided by the present invention carries out label for labelling by label clustering algorithm to metadata vector, according to
The calculating matrix and metadata matrix that service condition defines are calculated, then the area that matrix of consequence and pattern matrix are compared
Vector update is not to metadata matrix, if difference vector carries out label for labelling to difference vector, come without corresponding mark
It realizes the problem of threatening vector label mark, reaches the purpose for threatening vector extraction and label for labelling.The present invention can establish one kind
Flexible label for labelling mode not only provided and ability of the information as label knowledge is extracted from metadata, but also can establish more flexible
Customized label system.
Description of the drawings
In order to illustrate more clearly of technical scheme, letter will be made to attached drawing needed in the embodiment below
Singly introduce, it should be apparent that, the accompanying drawings in the following description is only some embodiments described in the present invention, for this field
For those of ordinary skill, without creative efforts, other attached drawings are can also be obtained according to these attached drawings.
Fig. 1 is according to a kind of threat vector extraction of the embodiment of the present invention and the block diagram of the system of label for labelling；
Fig. 2 is according to a kind of threat vector extraction of the embodiment of the present invention and the flow chart of the method for label for labelling.
Specific embodiment
With reference to the accompanying drawings to a preferred embodiment of the present invention will be described in detail, it is omitted in the course of the description for this
It is unnecessary details and function for invention, to prevent the understanding of the present invention from causing to obscure.Show although being shown in attached drawing
Example property embodiment, it being understood, however, that may be realized in various forms the present invention without should be limited by embodiments set forth here
System.On the contrary, these embodiments are provided to facilitate a more thoroughly understanding of the present invention, and can be complete by the scope of the present invention
Be communicated to those skilled in the art.
Fig. 1 is according to a kind of threat vector extraction of the embodiment of the present invention and the block diagram of the system of label for labelling.Such as Fig. 1 institutes
It states, system can include：Extraction module 110, analysis module 120, comparing module 130, update module 140, index module 150.
Extraction module 110, for extracting metadata vector from data source, the whole metadata vector extracted in data source
Form metadata matrix.
Extraction module 110 is used to extract metadata from data source.Data source includes initial data, metadata, vector,
In, the initial data includes binary file, binary system message, dynamic apis and performs records series, security log, dynamic apis
Monitored results.
Data source can be described as a vector, and according to the difference of data type, the metadata vector can be one
A binary file, wherein, binary file includes the networks such as IP address, domain name, the URL that can be extracted in binary file and visits
Ask information, digital signature information, the potential ability of binary executable, metadata vector can also be a network equipment
The dynamic apis of the binary system message obtained whithin a period of time or a file performs records series, dynamic apis prison
Security log controlled result or obtained etc..
For flow, extraction is similar to flow and parses, and flow parsing includes：Data flow restructuring, protocol identification, protocol analysis
Deng；For file, extraction content includes EXIF information, executable program file code sequence and API Calls sequence, the word of file
Symbol string information etc..If flow transmits file, then the metadata that this document extracts can be the metadata of flow extraction
A part.File is carried out the executions API sequential extraction procedures that are extracted of sandbox analysis to metadata can be that file extracts
Metadata a part.
A plurality of dynamic apis performs records series and can be formed by a sparse matrix in itself, can in extraction process above
To form the item of new sparse matrix.
The metadata of extraction can be described as a sparse matrix, common, and each column vector of matrix is to single
File, data flow, the description of the metadata of session, each row vector of matrix is a dimension of data source metadata.Extraction
The work of module 110 can be abstracted as and carry out matrixing to metadata vector, obtain the sparse matrix of metadata composition.Wall scroll
Metadata is a metadata vector, often marks a label, then vectorial（Sparse matrix）In corresponding Xiang Ze be assigned, other
The not vector of the label, then be not assigned.
Analysis module 120 for being based on metadata matrix, is calculated with calculating matrix, analysis structure matrix of consequence.
Specifically, the working method of analysis module 120, it is to do metadata sparse matrix with calculating matrix that can be abstracted as
"×" calculates.The element of the calculating matrix can be function, operator, numerical value.Each of which column vector is all nonidentity operation to same
The operation of one metadata, each of which row vector all represent the different calculating to each metadata.One possible lambda is calculated
Sub- example is as follows：
In some embodiments, the analysis module 120 is additionally operable to：
Define the calculation expression of the metadata and calculating matrix.
Define the multiplication of metadata and calculating matrix for metadata computation matrix lambda expression formulas, function table
Up to formula or numerical value.The result of calculating can be Boolean, character string, numerical value, matrix etc..As above a first number defined in formula
According to sparse matrix and a calculating matrix.The result of its "×" computing is a matrix, and the row vector expression one of matrix is single
File, data flow, whole result of calculations of session, matrix column vector are represented to a result of calculation, all files, data
Stream, the result of calculation of session.
As it can be seen that if desired increase more computings, it is only necessary to which the row for increasing calculating matrix can or define multiple
Calculating matrix.All it is a rule per a line, each row are all this regular metadata items for calculating matrix.If
Any one metadata item is calculated, then in the corresponding calculating metadata item of this in the form of lambda expression formulas
It writes, if need not calculate this, leaves blank this.Between rule and rule, between each metadata be "AND"
Relation, if necessary to the relation of "or", then need to be split as two or more rules.
It has been observed that user policy can be added to after producer's rule.Producer's rule can be in equipment R＆D process, generally
The many algorithms in neutral net, convolutional calculation, graph theory can be used, carry out effective approximate calculation, are calculated with reaching to reduce
The effect of amount.The result of approximate calculation can be exported directly as result, while splice the calculating to user side rule, be tied
Fruit matrix.
Comparing module 130 is compared with pattern matrix for being based on matrix of consequence, obtains label-vector.
Pattern matrix be the matrix of a similar calculating matrix, simply this matrix and matrix of consequence × operation result is complete
Portion is all Boolean.Wherein, True represents that corresponding metadata needs to be demarcated by the corresponding label of the row, False represent this to
Amount is not demarcated by the corresponding label of the row.Matrix of consequence and pattern matrix are compared into the calculating of row matrix multiplication cross, the difference of generation to
Amount and the corresponding readable data mark of difference vector that may be present can be used as metadata label-vector more new metadata dilute
Dredge matrix.
Due to the update of metadata sparse matrix may result in result generate variation, correlated results may need by
It recalculates.
On why wanting Comparative result：
When producer's rule, user policy split into calculating matrix, the compound conditions information such as "AND", "or", " non-" will be split
For a plurality of simple condition.The computings such as its "AND", "or", " non-" will lose in matrix computations.The place of these compound condition information
Reason is completed in this module.
In producer's rule, by multiple matrix of consequence column vectors（Single file, data flow, the result of session）Into traveling
The analysis of one step can support computing by vector, evaluate more complicated vector.Such as the whole network situation.
Update module 140, for by label-vector update to metadata matrix, and to the corresponding metadata of label-vector to
Amount carries out label for labelling, while is setting the value in recalculating metadata matrix and matrix of consequence in number.
Result of calculation may be updated by way of more new metadata sparse matrix, data-triggered can be based on and updated, only
The item for being corresponded in matrix and having value in row is recalculated according to above-mentioned calculation expression, and updates matrix of consequence.In general, this
Process should terminate in number is taken turns, therefore can enumerate maximum upper limit（Such as：Number is set as 3 times）If the update wheel number upper limit
It is reached, then terminates.
Further, the system also includes：
Index module 150 is indexed metadata for being based on label.
Index module 150 is used to be indexed metadata and newly generated human-readable label.With effectively by label,
Metadata is associated event or load, for subsequently showing, analyzing use.
Fig. 2 shows a kind of threat vector extraction according to embodiments of the present invention and the flow chart of the method for label for labelling.
As shown in Fig. 2, method includes the following steps：
S210 extracts metadata vector from data source, and the whole metadata vector extracted in data source forms metadata matrix.
Metadata is extracted from data source, data source includes initial data, metadata, vector, wherein, initial data includes
Binary file, binary system message, dynamic apis perform records series, security log, dynamic apis monitored results.
Data source can be described as a vector, and according to the difference of data type, the metadata vector can be one
A binary file, wherein, binary file includes the networks such as IP address, domain name, the URL that can be extracted in binary file and visits
Ask information, digital signature information, the potential ability of binary executable, metadata vector can also be a network equipment
The dynamic apis of the binary system message obtained whithin a period of time or a file performs records series, dynamic apis prison
Security log controlled result or obtained etc..
For flow, extraction is similar to flow and parses, and flow parsing includes：Data flow restructuring, protocol identification, protocol analysis
Deng；For file, extraction content includes EXIF information, executable program file code sequence and API Calls sequence, the word of file
Symbol string information etc..If flow transmits file, then the metadata that this document extracts can be the metadata of flow extraction
A part.File is carried out the executions API sequential extraction procedures that are extracted of sandbox analysis to metadata can be that file extracts
Metadata a part.
A plurality of dynamic apis performs records series and can be formed by a sparse matrix in itself, can in extraction process above
To form the item of new sparse matrix.
The metadata of extraction can be described as a sparse matrix, common, and each column vector of matrix is to single
File, data flow, the description of the metadata of session, each row vector of matrix is a dimension of data source metadata.Extraction
Work can be abstracted as carries out matrixing to metadata vector, obtains the sparse matrix of metadata composition.Wall scroll metadata is one
A metadata vector, often marks a label, then vectorial（Sparse matrix）In corresponding Xiang Ze be assigned, other not no labels
Vector, then be not assigned.
S220 based on metadata matrix, is calculated with calculating matrix, analysis structure matrix of consequence.
It is that metadata sparse matrix and calculating matrix are done "×" calculating that calculating, which can be abstracted as,.The member of the calculating matrix
Element can be function, operator, numerical value.Each of which column vector is all operation of the nonidentity operation to same metadata, each of which row to
Amount all represents the different calculating to each metadata.For example, lambda operators expression formula can be：
In some embodiments, further included before being calculated：
Define the calculation expression of the metadata and calculating matrix.
Define the multiplication of metadata and calculating matrix for metadata computation matrix lambda expression formulas, function table
Up to formula or numerical value.The result of calculating can be Boolean, character string, numerical value, matrix etc..As above a first number defined in formula
According to sparse matrix and a calculating matrix.The result of its "×" computing is a matrix, and the row vector expression one of matrix is single
File, data flow, whole result of calculations of session, matrix column vector are represented to a result of calculation, all files, data
Stream, the result of calculation of session.
As it can be seen that if desired increase more computings, it is only necessary to which the row for increasing calculating matrix can or define multiple
Calculating matrix.All it is a rule per a line, each row are all this regular metadata items for calculating matrix.If
Any one metadata item is calculated, then in the corresponding calculating metadata item of this in the form of lambda expression formulas
It writes, if need not calculate this, leaves blank this.Between rule and rule, between each metadata be "AND"
Relation, if necessary to the relation of "or", then need to be split as two or more rules.It has been observed that user policy can be added to
After producer's rule.
Producer's rule can generally use more in neutral net, convolutional calculation, graph theory in equipment R＆D process
Kind algorithm, carries out effective approximate calculation, to have the function that reduce calculation amount.The result of approximate calculation can be directly as result
Output, while splice the calculating to user side rule, obtain matrix of consequence.
S230 is compared with pattern matrix based on matrix of consequence, obtains label-vector.
Pattern matrix be the matrix of a similar calculating matrix, simply this matrix and matrix of consequence × operation result is complete
Portion is all Boolean.Wherein, True represents that corresponding metadata needs to be demarcated by the corresponding label of the row, False represent this to
Amount is not demarcated by the corresponding label of the row.Matrix of consequence and pattern matrix are compared into the calculating of row matrix multiplication cross, the difference of generation to
Amount and the corresponding readable data mark of difference vector that may be present can be used as metadata label-vector more new metadata dilute
Dredge matrix.Due to the update of metadata sparse matrix may result in result generate variation, correlated results may need by
It recalculates.
S240 updates label-vector to metadata matrix, and to the corresponding metadata vector of label-vector into row label
Mark, while setting the value in recalculating metadata matrix and matrix of consequence in number.
Result of calculation may be updated by way of more new metadata sparse matrix, data-triggered can be based on and updated, only
The item for being corresponded in matrix and having value in row is recalculated according to above-mentioned calculation expression, and updates matrix of consequence.In general, this
Process should terminate in number is taken turns, therefore can enumerate maximum upper limit（Such as：Number is set as 3 times）If the update wheel number upper limit
It is reached, then terminates.
In some embodiments, following step is further included：
S250 is indexed metadata based on label.
Metadata and newly generated human-readable label are indexed, with effectively by label, metadata to event or
Load is associated, for subsequently showing, analyzing use.
Above-described embodiment carries out label for labelling by label clustering algorithm to metadata vector, is defined according to service condition
Calculating matrix and metadata matrix are calculated, then member is arrived in the difference vector update that matrix of consequence and pattern matrix are compared
Data matrix, if difference vector carries out label for labelling to difference vector, vector mark is threatened to realize without corresponding mark
The problem of label mark, reaches the purpose for threatening vector extraction and label for labelling.
So far having been combined preferred embodiment, invention has been described.It should be understood that those skilled in the art are not
In the case of departing from the spirit and scope of the present invention, various other changes, replacement and addition can be carried out.Therefore, it is of the invention
Scope be not limited to above-mentioned specific embodiment, and should be defined by the appended claims.
Claims (10)
1. a kind of threaten vector extraction and the system of label for labelling, which is characterized in that including：
Extraction module, for extracting metadata vector from data source, the whole metadata vector extracted in data source forms member
Data matrix；
Analysis module for being based on metadata matrix, is calculated with calculating matrix, analysis structure matrix of consequence；
Comparing module is compared with pattern matrix for being based on matrix of consequence, obtains label-vector；
Update module for updating label-vector to metadata matrix, and carries out label-vector corresponding metadata vector
Label for labelling, while setting the value in recalculating metadata matrix and matrix of consequence in number.
2. system according to claim 1, which is characterized in that further include：
Index module is indexed metadata for being based on label.
3. system according to claim 1, which is characterized in that the analysis module is additionally operable to：
Define the calculation expression of the metadata and calculating matrix.
4. system according to claim 1, which is characterized in that the data source includes initial data, metadata, vector,
Wherein, the initial data includes binary file, binary system message, dynamic apis execution records series, security log, dynamic
API monitored results.
5. system according to claim 1, which is characterized in that the element of the calculating matrix includes function, operator, number
Value, the result of calculating include Boolean, character string, numerical value, matrix.
6. a kind of threaten vector extraction and the method for label for labelling, which is characterized in that including：
Metadata vector is extracted from data source, the whole metadata vector extracted in data source forms metadata matrix；
It based on metadata matrix, is calculated with calculating matrix, analysis structure matrix of consequence；
It is compared based on matrix of consequence with pattern matrix, obtains label-vector；
Label-vector is updated to metadata matrix, and label for labelling is carried out to label-vector corresponding metadata vector, simultaneously
Value in recalculating metadata matrix and matrix of consequence in setting number.
7. it according to the method described in claim 6, it is characterized in that, further includes：Metadata is indexed based on label.
8. according to the method described in claim 6, it is characterized in that, described based on metadata matrix, counted with calculating matrix
It calculates, analysis structure matrix of consequence further includes：
Define the calculation expression of the metadata and calculating matrix.
9. according to the method described in claim 6, it is characterized in that, the data source include initial data, metadata, vector,
Wherein, the initial data includes binary file, binary system message, dynamic apis execution records series, security log, dynamic
API monitored results.
10. according to the method described in claim 6, it is characterized in that, the element of the calculating matrix includes function, operator, number
Value, the result of calculating include Boolean, character string, numerical value, matrix.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710010774.6A CN108076036A (en) | 2017-01-06 | 2017-01-06 | It is a kind of to threaten vector extraction and the system and method for label for labelling |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710010774.6A CN108076036A (en) | 2017-01-06 | 2017-01-06 | It is a kind of to threaten vector extraction and the system and method for label for labelling |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108076036A true CN108076036A (en) | 2018-05-25 |
Family
ID=62159112
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710010774.6A Withdrawn CN108076036A (en) | 2017-01-06 | 2017-01-06 | It is a kind of to threaten vector extraction and the system and method for label for labelling |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108076036A (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103049479A (en) * | 2012-11-26 | 2013-04-17 | 北京奇虎科技有限公司 | Method and system for generating online video label |
| US20140279834A1 (en) * | 2013-03-15 | 2014-09-18 | Amiato, Inc. | Scalable Analysis Platform For Semi-Structured Data |
| CN106095966A (en) * | 2016-06-15 | 2016-11-09 | 成都品果科技有限公司 | A kind of user's extendible label for labelling method and system |
-
2017
- 2017-01-06 CN CN201710010774.6A patent/CN108076036A/en not_active Withdrawn
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103049479A (en) * | 2012-11-26 | 2013-04-17 | 北京奇虎科技有限公司 | Method and system for generating online video label |
| US20140279834A1 (en) * | 2013-03-15 | 2014-09-18 | Amiato, Inc. | Scalable Analysis Platform For Semi-Structured Data |
| CN106095966A (en) * | 2016-06-15 | 2016-11-09 | 成都品果科技有限公司 | A kind of user's extendible label for labelling method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Byskov | Enumerating maximal independent sets with applications to graph colouring | |
| US20100332476A1 (en) | Web graph compression through scalable pattern mining | |
| CN108363686A (en) | A kind of character string segmenting method, device, terminal device and storage medium | |
| CN109582289B (en) | Method, system, storage medium and processor for processing rule flow in rule engine | |
| CN111756706A (en) | Abnormal flow detection method and device and storage medium | |
| CN110309368A (en) | Determination method, apparatus, storage medium and the electronic device of data address | |
| CN111443901B (en) | Java reflection-based service expansion method and device | |
| CN107330094A (en) | The Bloom Filter tree construction and key-value pair storage method of dynamic memory key-value pair | |
| CN103309893A (en) | Character string comparing method and device | |
| CN108259195A (en) | The determining method and system of the coverage of anomalous event | |
| Sun et al. | Finding group steiner trees in graphs with both vertex and edge weights | |
| Irving et al. | The suffix binary search tree and suffix AVL tree | |
| CN115905633A (en) | Image similarity retrieval method and system with privacy protection function | |
| van Iersel et al. | Approximation algorithms for nonbinary agreement forests | |
| CN107766036A (en) | A kind of construction method of module, construction device and terminal device | |
| CN114662108A (en) | Software detection method and device and electronic equipment | |
| CN102708285B (en) | Coremedicine excavation method based on complex network model parallelizing PageRank algorithm | |
| CN114254353A (en) | Data processing method and device based on privacy protection and server | |
| WO2018135515A1 (en) | Information processing device, neural network design method, and recording medium | |
| CN106778352B (en) | Multisource privacy protection method for combined release of set value data and social network data | |
| CN108076036A (en) | It is a kind of to threaten vector extraction and the system and method for label for labelling | |
| CN112084500A (en) | Method and device for clustering virus samples, electronic equipment and storage medium | |
| CN103124273A (en) | Method and system for building and matching path inverted list based on user behavior analysis | |
| CN107220702B (en) | Computer vision processing method and device of low-computing-capacity processing equipment | |
| Bonsma et al. | Extremal graphs having no matching cuts |
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 | ||
| CB02 | Change of applicant information | ||
| CB02 | Change of applicant information |
Address after: 150028 Building 7, Innovation Plaza, Science and Technology Innovation City, Harbin High-tech Industrial Development Zone, Heilongjiang Province (838 Shikun Road) Applicant after: Harbin antiy Technology Group Limited by Share Ltd Address before: 506 room 162, Hongqi Avenue, Nangang District, Harbin Development Zone, Heilongjiang, 150090 Applicant before: Harbin Antiy Technology Co., Ltd. |
|
| WW01 | Invention patent application withdrawn after publication | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20180525 |