CN110083778A - The figure convolutional neural networks construction method and device of study separation characterization - Google Patents
The figure convolutional neural networks construction method and device of study separation characterization Download PDFInfo
- Publication number
- CN110083778A CN110083778A CN201910277434.9A CN201910277434A CN110083778A CN 110083778 A CN110083778 A CN 110083778A CN 201910277434 A CN201910277434 A CN 201910277434A CN 110083778 A CN110083778 A CN 110083778A
- Authority
- CN
- China
- Prior art keywords
- characterization
- neural networks
- factor
- convolutional neural
- node
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000012512 characterization method Methods 0.000 title claims abstract description 55
- 238000000926 separation method Methods 0.000 title claims abstract description 36
- 238000013527 convolutional neural network Methods 0.000 title claims abstract description 31
- 238000010276 construction Methods 0.000 title claims abstract description 30
- 238000000034 method Methods 0.000 claims abstract description 18
- 230000003447 ipsilateral effect Effects 0.000 claims abstract description 7
- 230000015572 biosynthetic process Effects 0.000 claims abstract description 6
- 238000010030 laminating Methods 0.000 claims description 4
- 230000008901 benefit Effects 0.000 description 6
- 238000003475 lamination Methods 0.000 description 5
- 238000013528 artificial neural network Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 230000001939 inductive effect Effects 0.000 description 3
- 238000002955 isolation Methods 0.000 description 3
- 210000004218 nerve net Anatomy 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 230000001537 neural effect Effects 0.000 description 2
- 238000009795 derivation Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000003012 network analysis Methods 0.000 description 1
- 238000000547 structure data Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9536—Search customisation based on social or collaborative filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/047—Probabilistic or stochastic networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/01—Social networking
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Health & Medical Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Biophysics (AREA)
- Evolutionary Computation (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Biomedical Technology (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Databases & Information Systems (AREA)
- Business, Economics & Management (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Probability & Statistics with Applications (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses the figure convolutional neural networks construction methods and device of a kind of study separation characterization, wherein method includes: to carry out probabilistic Modeling to the forming process of input figure, generates the generative probabilistic model of the multiple latent factors that may cause a line formation of description；It is made inferences in each convolutional layer using guidable dynamic EM algorithm by generative probabilistic model, obtains the factor corresponding to each neighbours of each node, neighbor node is separated；In each convolutional layer, the characterization for describing each node not ipsilateral is constructed according to the neighbor node of the different factors.This method can generate the characterization that can accurately describe multiple sides of each data point in figure comprehensively according to each factor.
Description
Technical field
The present invention relates to social network analysis technical field, in particular to the picture scroll product nerve net of a kind of study separation characterization
Network construction method and device.
Background technique
Currently, being for handling the complexity such as social networks, information network to scheme figure neural network of the convolutional network as representative
The end-to-end depth learning technology of a new generation of graph structure data.However, the formation on the side in existing figure neural network default figure
It is all the diversified origin cause of formation for being pushed by the same monofactor, therefore real data behind can not being captured.
Summary of the invention
The present invention is directed to solve at least some of the technical problems in related technologies.
For this purpose, an object of the present invention is to provide a kind of figure convolutional neural networks building sides of study separation characterization
The characterization that can accurately describe multiple sides of each data point in figure comprehensively can be generated in method, this method.
It is another object of the present invention to the figure convolutional neural networks construction devices for proposing a kind of study separation characterization.
In order to achieve the above objectives, one aspect of the present invention embodiment proposes a kind of picture scroll product nerve net of study separation characterization
Network construction method, comprising: probabilistic Modeling is carried out to the forming process of input figure, the multiple a lines that may cause of description is generated and is formed
Latent factor generative probabilistic model；Guidable dynamic EM is used in each convolutional layer by the generative probabilistic model
Algorithm (Expectation-Maximization, EM algorithm) makes inferences, and obtains each neighbours institute of each node
The corresponding factor, neighbor node is separated；In each described convolutional layer, according to the neighbor node structure of the different factors
Build out the characterization for describing each node not ipsilateral.
The figure convolutional neural networks construction method of the study separation characterization of the embodiment of the present invention, considers a figure at behind
Multiple factors, these factors are separated, obtain more accurate comprehensive characterization, and each because of the period of the day from 11 p.m. to 1 a.m, still reserved graph mind separating
The advantages of through the end-to-end study of network support, inductive learning, after separating each factor, can be generated according to each factor can be complete
Face accurately describes the characterization of multiple sides of each data point in figure.
In addition, the figure convolutional neural networks construction method of study separation characterization according to the above embodiment of the present invention can be with
With following additional technical characteristic:
Further, in one embodiment of the invention, further includes: each multiple described convolutional layer of superposition, with benefit
With preset higher order topology structure.
Further, in one embodiment of the invention, corresponding one of each side is by the isolated factor.
Further, in one embodiment of the invention, the factor of the input figure is that plural number is multiple.
In order to achieve the above objectives, another aspect of the present invention embodiment proposes a kind of figure convolutional Neural of study separation characterization
Network struction device, comprising: modeling module carries out probabilistic Modeling for the forming process to input figure, and generation description is multiple can
The generative probabilistic model for the latent factor that a line can be caused to be formed；Reasoning module, for being existed by the generative probabilistic model
Made inferences in each convolutional layer using guidable dynamic EM algorithm, obtain corresponding to each neighbours of each node because
Son separates neighbor node；Module is constructed, is used in each described convolutional layer, according to the neighbours of the different factors
Node constructs the characterization for describing each node not ipsilateral.
The figure convolutional neural networks construction device of the study separation characterization of the embodiment of the present invention, considers a figure at behind
Multiple factors, these factors are separated, obtain more accurate comprehensive characterization, and each because of the period of the day from 11 p.m. to 1 a.m, still reserved graph mind separating
The advantages of through the end-to-end study of network support, inductive learning, after separating each factor, can be generated according to each factor can be complete
Face accurately describes the characterization of multiple sides of each data point in figure.
In addition, the figure convolutional neural networks construction device of study separation characterization according to the above embodiment of the present invention can be with
With following additional technical characteristic:
Further, in one embodiment of the invention, further includes: laminating module, it is multiple described each for being superimposed
A convolutional layer, to utilize preset higher order topology structure.
Further, in one embodiment of the invention, corresponding one of each side is by the isolated factor.
Further, in one embodiment of the invention, the factor of the input figure is that plural number is multiple.
The additional aspect of the present invention and advantage will be set forth in part in the description, and will partially become from the following description
Obviously, or practice through the invention is recognized.
Detailed description of the invention
Above-mentioned and/or additional aspect and advantage of the invention will become from the following description of the accompanying drawings of embodiments
Obviously and it is readily appreciated that, in which:
Fig. 1 is the process according to the figure convolutional neural networks construction method of the study separation characterization of one embodiment of the invention
Figure；
Fig. 2 is the figure convolutional neural networks construction method according to the study separation characterization of one specific embodiment of the present invention
Flow chart；
Fig. 3 is the structure according to the figure convolutional neural networks construction device of the study separation characterization of one embodiment of the invention
Schematic diagram.
Specific embodiment
The embodiment of the present invention is described below in detail, examples of the embodiments are shown in the accompanying drawings, wherein from beginning to end
Same or similar label indicates same or similar element or element with the same or similar functions.Below with reference to attached
The embodiment of figure description is exemplary, it is intended to is used to explain the present invention, and is not considered as limiting the invention.
The figure convolutional neural networks structure of study separation characterization proposed according to embodiments of the present invention is described with reference to the accompanying drawings
Construction method and device describe the figure convolutional Neural of study separation characterization proposed according to embodiments of the present invention with reference to the accompanying drawings first
Network establishing method.
Fig. 1 is the flow chart of the figure convolutional neural networks construction method of the study separation characterization of one embodiment of the invention.
As shown in Figure 1, the study separation characterization figure convolutional neural networks construction method the following steps are included:
In step s101, probabilistic Modeling is carried out to the forming process of input figure, generation description is multiple to may cause one
The generative probabilistic model for the latent factor that side is formed.
It is understood that as shown in Fig. 2, firstly, carry out probabilistic Modeling to the forming process of the figure of input, foundation it is general
Rate generates model and describes multiple latent factors that may cause a line formation.
It specifically,, can in given figure after a node and its neighbours based on the reasoning module of generative probabilistic model
The latent factor for pushing each side to be formed is found unsupervisedly and neighbours are sorted out or divided according to its corresponding factor
From.
Wherein, in one embodiment of the invention, it is multiple for plural number to input the factor of figure.
In step s 102, it is carried out in each convolutional layer using guidable dynamic EM algorithm by generative probabilistic model
Reasoning obtains the factor corresponding to each neighbours of each node, neighbor node is separated.
It is understood that as shown in Fig. 2, in each convolutional layer, according to the generative probabilistic model of foundation, using can
The dynamic EM algorithm led makes inferences, and the factor corresponding to each neighbours of one node of reasoning accordingly separates neighbours.
In step s 103, in each convolutional layer, each section of description is constructed according to the neighbor node of the different factors
Put the characterization of not ipsilateral.
It is understood that as shown in Fig. 2, in each convolutional layer, the difference factor pair according to obtained in previous step
The neighbours answered construct the characterization for describing the node not ipsilateral, and corresponding one of each side is by the isolated factor.
Specifically, the embodiment of the present invention proposes a kind of new picture scroll lamination for applying factor isolation technics, the picture scroll
Lamination accurately can comprehensively describe the characterization of its multiple side to the output of each node.That is, the embodiment of the present invention
Picture scroll lamination applies factor isolation technics, after carrying out factor separation, parallel, to be independently located in using multiple figure convolution operations
Manage information corresponding with each factor.
Wherein, factor isolation technics is in given figure after a node and its neighbours, and one kind can be sent out unsupervisedly
It now pushes the latent factor of each side formation and neighbours is subjected to the technology sorted out/separated according to its corresponding factor.
In specific application, in recommender system, more fully accurate user's portrait etc. is automatically generated；And recommending system
In system, the interaction between the individual such as user, article naturally forms a figure, method energy through the embodiment of the present invention
Enough more accurate multiple points of interest of plural number for comprehensively capturing user or demand point.
Further, in one embodiment of the invention, the method for the embodiment of the present invention further include: superposition is multiple each
A convolutional layer, to utilize preset higher order topology structure.
It is understood that the embodiment of the present invention, which passes through, is superimposed multiple above-mentioned convolutional layers, to efficiently use the height in figure
Rank topological structure.
Specifically, the embodiment of the present invention proposes a kind of picture scroll product nerve net for being superimposed multiple above-mentioned new picture scroll laminations
Network can further utilize the additional informations such as the higher order topology structure in figure.That is, the picture scroll product of the embodiment of the present invention
Neural network has been superimposed multiple above-mentioned new picture scroll laminations, further to utilize the additional letter such as higher order topology structure in figure
Breath.
To sum up, the embodiment of the present invention tries to find out and separates brought by multiple factors mainly for when carrying out picture scroll product
Challenge proposes targetedly measure, to improved figure convolutional neural networks can export can it is more accurate, data point is described comprehensively
Characterization:
(1) challenge one: diagram data will not usually mark out the specific factor for pushing a line to be formed.The embodiment of the present invention is
This proposes a kind of unsupervised technology based on generative probabilistic model, to infer the corresponding latent factor of each edge.
(2) two are challenged: how to keep two big advantages of figure neural network while carrying out complicated inference --- support end
To end study, support inductive learning (result is extrapolated to the new data point that do not see).The embodiment of the present invention thus pushes away probability
Manage process description at it is a kind of can (end-to-end to support) of derivation, Dynamic Execution (support to conclude) EM algorithm.
The figure convolutional neural networks construction method of the study separation characterization proposed according to embodiments of the present invention, it is contemplated that facilitate
One figure at the factor may be to have plural number multiple, can infer potential multiple factors unsupervisedly, and by they point
From, and after separating each factor, can accordingly generate can accurately describe multiple sides of each data point in figure comprehensively
Characterization, thus consider push a figure at factor quantity may be plural number it is multiple, by carry out picture scroll product the time-division
The factor different from these, and then obtaining can the more accurate multiple and different side for comprehensively describing each data point in figure
Characterization.
Referring next to the figure convolutional neural networks structure for the study separation characterization that attached drawing description proposes according to embodiments of the present invention
Build device.
Fig. 3 is the structural representation of the figure convolutional neural networks construction device of the study separation characterization of one embodiment of the invention
Figure.
As shown in figure 3, the figure convolutional neural networks construction device 10 of study separation characterization includes: modeling module 100, pushes away
Manage module 200 and building module 300.
Wherein, modeling module 100 is used to carry out the forming process of input figure probabilistic Modeling, and generation description is multiple to be led
The generative probabilistic model for the latent factor for causing a line to be formed.Reasoning module 200 is used for through generative probabilistic model at each
It is made inferences in convolutional layer using guidable dynamic EM algorithm, obtains the factor corresponding to each neighbours of each node, it will
Neighbor node separation.It constructs module 300 to be used in each convolutional layer, description is constructed according to the neighbor node of the different factors
The characterization of each node not ipsilateral.The device 10 of the embodiment of the present invention can be generated according to each factor and can accurately be retouched comprehensively
State the characterization of multiple sides of each data point in figure.
Further, in one embodiment of the invention, the device 10 of the embodiment of the present invention further include: laminating module.
Wherein, laminating module, for being superimposed each multiple convolutional layer, to utilize preset higher order topology structure.
Further, in one embodiment of the invention, corresponding one of each side is by the isolated factor.
Further, in one embodiment of the invention, it is multiple for plural number to input the factor of figure.
It should be noted that the aforementioned figure convolutional neural networks construction method embodiment to study separation characterization is explained
The figure convolutional neural networks construction device of the bright study separation characterization for being also applied for the embodiment, details are not described herein again.
The figure convolutional neural networks construction device of the study separation characterization proposed according to embodiments of the present invention, it is contemplated that facilitate
One figure at the factor may be to have plural number multiple, can infer potential multiple factors unsupervisedly, and by they point
From, and after separating each factor, can accordingly generate can accurately describe multiple sides of each data point in figure comprehensively
Characterization, thus consider push a figure at factor quantity may be plural number it is multiple, by carry out picture scroll product the time-division
The factor different from these, and then obtaining can the more accurate multiple and different side for comprehensively describing each data point in figure
Characterization.
In the description of the present invention, it is to be understood that, term " center ", " longitudinal direction ", " transverse direction ", " length ", " width ",
" thickness ", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom" "inner", "outside", " up time
The orientation or positional relationship of the instructions such as needle ", " counterclockwise ", " axial direction ", " radial direction ", " circumferential direction " be orientation based on the figure or
Positional relationship is merely for convenience of description of the present invention and simplification of the description, rather than the device or element of indication or suggestion meaning must
There must be specific orientation, be constructed and operated in a specific orientation, therefore be not considered as limiting the invention.
In addition, term " first ", " second " are used for descriptive purposes only and cannot be understood as indicating or suggesting relative importance
Or implicitly indicate the quantity of indicated technical characteristic.Define " first " as a result, the feature of " second " can be expressed or
Implicitly include at least one this feature.In the description of the present invention, the meaning of " plurality " is at least two, such as two, three
It is a etc., unless otherwise specifically defined.
In the present invention unless specifically defined or limited otherwise, term " installation ", " connected ", " connection ", " fixation " etc.
Term shall be understood in a broad sense, for example, it may be being fixedly connected, may be a detachable connection, or integral；It can be mechanical connect
It connects, is also possible to be electrically connected；It can be directly connected, can also can be in two elements indirectly connected through an intermediary
The interaction relationship of the connection in portion or two elements, unless otherwise restricted clearly.For those of ordinary skill in the art
For, the specific meanings of the above terms in the present invention can be understood according to specific conditions.
In the present invention unless specifically defined or limited otherwise, fisrt feature in the second feature " on " or " down " can be with
It is that the first and second features directly contact or the first and second features pass through intermediary mediate contact.Moreover, fisrt feature exists
Second feature " on ", " top " and " above " but fisrt feature be directly above or diagonally above the second feature, or be merely representative of
First feature horizontal height is higher than second feature.Fisrt feature can be under the second feature " below ", " below " and " below "
One feature is directly under or diagonally below the second feature, or is merely representative of first feature horizontal height less than second feature.
In the description of this specification, reference term " one embodiment ", " some embodiments ", " example ", " specifically show
The description of example " or " some examples " etc. means specific features, structure, material or spy described in conjunction with this embodiment or example
Point is included at least one embodiment or example of the invention.In the present specification, schematic expression of the above terms are not
It must be directed to identical embodiment or example.Moreover, particular features, structures, materials, or characteristics described can be in office
It can be combined in any suitable manner in one or more embodiment or examples.In addition, without conflicting with each other, the skill of this field
Art personnel can tie the feature of different embodiments or examples described in this specification and different embodiments or examples
It closes and combines.
Although the embodiments of the present invention has been shown and described above, it is to be understood that above-described embodiment is example
Property, it is not considered as limiting the invention, those skilled in the art within the scope of the invention can be to above-mentioned
Embodiment is changed, modifies, replacement and variant.
Claims (8)
1. a kind of figure convolutional neural networks construction method of study separation characterization characterized by comprising
Probabilistic Modeling is carried out to the forming process of input figure, generates the multiple latent factors that may cause a line formation of description
Generative probabilistic model；
It is made inferences, is obtained each using guidable dynamic EM algorithm in each convolutional layer by the generative probabilistic model
The factor corresponding to each neighbours of node, neighbor node is separated；
In each described convolutional layer, it is different that description each node is constructed according to the neighbor node of the different factors
The characterization of side.
2. the figure convolutional neural networks construction method of study separation characterization according to claim 1, which is characterized in that also wrap
It includes:
Each multiple described convolutional layer are superimposed, to utilize preset higher order topology structure.
3. the figure convolutional neural networks construction method of study separation characterization according to claim 1, which is characterized in that each
Corresponding one of side is by the isolated factor.
4. the figure convolutional neural networks construction method of study separation characterization according to claim 1, which is characterized in that described
The factor for inputting figure is that plural number is multiple.
5. a kind of figure convolutional neural networks construction device of study separation characterization characterized by comprising
Modeling module carries out probabilistic Modeling for the forming process to input figure, and generation description is multiple to may cause a line shape
At latent factor generative probabilistic model；
Reasoning module, for being carried out in each convolutional layer using guidable dynamic EM algorithm by the generative probabilistic model
Reasoning obtains the factor corresponding to each neighbours of each node, neighbor node is separated；
Module is constructed, for constructing description institute according to the neighbor node of the different factors in each described convolutional layer
State the characterization of each node not ipsilateral.
6. the figure convolutional neural networks construction device of study separation characterization according to claim 5, which is characterized in that also wrap
It includes:
Laminating module, for being superimposed each multiple described convolutional layer, to utilize preset higher order topology structure.
7. the figure convolutional neural networks construction device of study separation characterization according to claim 5, which is characterized in that each
Corresponding one of side is by the isolated factor.
8. the figure convolutional neural networks construction device of study separation characterization according to claim 5, which is characterized in that described
The factor for inputting figure is that plural number is multiple.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910277434.9A CN110083778A (en) | 2019-04-08 | 2019-04-08 | The figure convolutional neural networks construction method and device of study separation characterization |
| PCT/CN2019/098236 WO2020206876A1 (en) | 2019-04-08 | 2019-07-29 | Method and device for constructing graph convolutional neural network for learning disentangled representation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910277434.9A CN110083778A (en) | 2019-04-08 | 2019-04-08 | The figure convolutional neural networks construction method and device of study separation characterization |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110083778A true CN110083778A (en) | 2019-08-02 |
Family
ID=67414479
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910277434.9A Pending CN110083778A (en) | 2019-04-08 | 2019-04-08 | The figure convolutional neural networks construction method and device of study separation characterization |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN110083778A (en) |
| WO (1) | WO2020206876A1 (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110889015A (en) * | 2019-10-31 | 2020-03-17 | 天津工业大学 | Independent decoupling convolutional neural network characterization algorithm for graph data |
| CN112148998A (en) * | 2020-09-08 | 2020-12-29 | 浙江工业大学 | Online social platform user friend recommendation method based on multi-core graph convolutional network |
| CN116127204A (en) * | 2023-04-17 | 2023-05-16 | 中国科学技术大学 | Multi-view user portrayal method, multi-view user portrayal system, apparatus, and medium |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113299079B (en) * | 2021-03-29 | 2022-06-10 | 东南大学 | Regional intersection signal control method based on PPO and graph convolution neural network |
| CN113722603A (en) * | 2021-11-02 | 2021-11-30 | 阿里巴巴达摩院(杭州)科技有限公司 | Object pushing method, product pushing method, computer terminal and storage medium |
| CN115883147B (en) * | 2022-11-22 | 2023-10-13 | 浙江御安信息技术有限公司 | Attacker portrait method based on graphic neural network |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104866558A (en) * | 2015-05-18 | 2015-08-26 | 中国科学院计算技术研究所 | Training method of social networking account mapping model, mapping method and system |
| CN106559290A (en) * | 2016-11-29 | 2017-04-05 | 北京邮电大学 | Method and system based on the link prediction of community structure |
| CN106649659A (en) * | 2016-12-13 | 2017-05-10 | 重庆邮电大学 | Link prediction system and method for social network |
| CN106959967A (en) * | 2016-01-12 | 2017-07-18 | 中国科学院声学研究所 | A kind of training of link prediction model and link prediction method |
| CN107332687A (en) * | 2017-05-23 | 2017-11-07 | 浙江工业大学 | Link prediction method based on Bayesian estimation and common neighbor |
| CN107451703A (en) * | 2017-08-31 | 2017-12-08 | 杭州师范大学 | A kind of social networks multitask Forecasting Methodology based on factor graph model |
| CN109347697A (en) * | 2018-10-10 | 2019-02-15 | 南昌航空大学 | Opportunistic network link prediction method, apparatus and readable storage medium storing program for executing |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107203511B (en) * | 2017-05-27 | 2020-07-17 | 中国矿业大学 | Network text named entity identification method based on neural network probability disambiguation |
| CN108681775B (en) * | 2018-05-25 | 2021-04-13 | 厦门大学 | Tree network method for testing and updating through WordNet embedding |
| CN109063841B (en) * | 2018-08-27 | 2020-07-03 | 北京航空航天大学 | Fault mechanism intelligent analysis method based on Bayesian network and deep learning algorithm |
| CN109376769A (en) * | 2018-09-21 | 2019-02-22 | 广东技术师范学院 | The information transfer method of multitask classification is used for based on production confrontation neural network |
| CN109582960B (en) * | 2018-11-27 | 2020-11-24 | 上海交通大学 | Zero example learning method based on structured association semantic embedding |
-
2019
- 2019-04-08 CN CN201910277434.9A patent/CN110083778A/en active Pending
- 2019-07-29 WO PCT/CN2019/098236 patent/WO2020206876A1/en active Application Filing
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104866558A (en) * | 2015-05-18 | 2015-08-26 | 中国科学院计算技术研究所 | Training method of social networking account mapping model, mapping method and system |
| CN106959967A (en) * | 2016-01-12 | 2017-07-18 | 中国科学院声学研究所 | A kind of training of link prediction model and link prediction method |
| CN106559290A (en) * | 2016-11-29 | 2017-04-05 | 北京邮电大学 | Method and system based on the link prediction of community structure |
| CN106649659A (en) * | 2016-12-13 | 2017-05-10 | 重庆邮电大学 | Link prediction system and method for social network |
| CN107332687A (en) * | 2017-05-23 | 2017-11-07 | 浙江工业大学 | Link prediction method based on Bayesian estimation and common neighbor |
| CN107451703A (en) * | 2017-08-31 | 2017-12-08 | 杭州师范大学 | A kind of social networks multitask Forecasting Methodology based on factor graph model |
| CN109347697A (en) * | 2018-10-10 | 2019-02-15 | 南昌航空大学 | Opportunistic network link prediction method, apparatus and readable storage medium storing program for executing |
Non-Patent Citations (1)
| Title |
|---|
| 黄立威 等: "一种基于因子图模型的半监督社区发现方法", 《自动化学报》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110889015A (en) * | 2019-10-31 | 2020-03-17 | 天津工业大学 | Independent decoupling convolutional neural network characterization algorithm for graph data |
| CN110889015B (en) * | 2019-10-31 | 2024-01-30 | 天津工业大学 | Independent decoupling convolutional neural network characterization method for graph data |
| CN112148998A (en) * | 2020-09-08 | 2020-12-29 | 浙江工业大学 | Online social platform user friend recommendation method based on multi-core graph convolutional network |
| CN112148998B (en) * | 2020-09-08 | 2021-10-26 | 浙江工业大学 | Online social platform user friend recommendation method based on multi-core graph convolutional network |
| CN116127204A (en) * | 2023-04-17 | 2023-05-16 | 中国科学技术大学 | Multi-view user portrayal method, multi-view user portrayal system, apparatus, and medium |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2020206876A1 (en) | 2020-10-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110083778A (en) | The figure convolutional neural networks construction method and device of study separation characterization | |
| CN102300051B (en) | Camera layout determination support device | |
| Dong et al. | A hybrid group decision making framework for achieving agreed solutions based on stable opinions | |
| Hadlak et al. | A Survey of Multi-faceted Graph Visualization. | |
| EP3928473B1 (en) | Systems and methods for communications node upgrade and selection | |
| CN105677648B (en) | A kind of Combo discovering method and system based on label propagation algorithm | |
| CN104217073B (en) | A kind of visual layout's method of network community gravitation guiding | |
| CN104573915B (en) | Bullet train product structure tree constructing method and device | |
| CN106548124A (en) | Theme deduction system, theme presumption method | |
| CN110390548A (en) | The selection method and device of coupon distribution strategy | |
| Alonso et al. | Deep reinforcement learning for the management of software-defined networks and network function virtualization in an edge-IoT architecture | |
| CN111160954A (en) | Recommendation method facing group object based on graph convolution network model | |
| CN107391542A (en) | A kind of open source software community expert recommendation method based on document knowledge collection of illustrative plates | |
| CN105814582A (en) | Method and system for recognizing faces | |
| CN109165787A (en) | Bus routes method for customizing, device and equipment | |
| Kojić et al. | A neural networks-based hybrid routing protocol for wireless mesh networks | |
| CN106294299A (en) | A kind of report form generation method and device | |
| CN112085171B (en) | Recommendation method based on clustered multi-entity graph neural network | |
| CN103198432B (en) | Detection method and detection system of network groups in online social network | |
| Wang et al. | A computing perspective on smart city [guest editorial] | |
| CN109120669A (en) | Method, medium and system based on the internet P2P building block chain | |
| CN104363612B (en) | Piecemeal wireless sensor network restorative procedure based on mobile relay scheduling | |
| CN111126607B (en) | Data processing method, device and system for model training | |
| CN109299849B (en) | Group demand level calculation method in social network | |
| CN107220742A (en) | A kind of development of information system common support method analyzed based on system vulnerability and platform |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190802 |