CN109960759A - Recommender system clicking rate prediction technique based on deep neural network - Google Patents
Recommender system clicking rate prediction technique based on deep neural network Download PDFInfo
- Publication number
- CN109960759A CN109960759A CN201910223510.8A CN201910223510A CN109960759A CN 109960759 A CN109960759 A CN 109960759A CN 201910223510 A CN201910223510 A CN 201910223510A CN 109960759 A CN109960759 A CN 109960759A
- Authority
- CN
- China
- Prior art keywords
- user
- article
- neural network
- matrix
- model
- 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
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 24
- 238000000034 method Methods 0.000 title claims abstract description 14
- 239000011159 matrix material Substances 0.000 claims abstract description 35
- 238000012549 training Methods 0.000 claims abstract description 31
- 239000013598 vector Substances 0.000 claims abstract description 29
- 238000003062 neural network model Methods 0.000 claims abstract description 10
- 230000002452 interceptive effect Effects 0.000 claims abstract description 8
- 230000003993 interaction Effects 0.000 claims abstract description 4
- 239000000284 extract Substances 0.000 claims abstract 2
- 230000004913 activation Effects 0.000 claims description 14
- 210000002569 neuron Anatomy 0.000 claims description 8
- 210000005036 nerve Anatomy 0.000 claims description 5
- 239000013604 expression vector Substances 0.000 claims description 3
- 239000000203 mixture Substances 0.000 claims 1
- 238000000354 decomposition reaction Methods 0.000 abstract description 3
- 230000006399 behavior Effects 0.000 description 29
- 230000006870 function Effects 0.000 description 8
- 210000004218 nerve net Anatomy 0.000 description 3
- 230000003542 behavioural effect Effects 0.000 description 2
- 238000013527 convolutional neural network Methods 0.000 description 2
- 238000003066 decision tree Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 241001269238 Data Species 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Classifications
-
- 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
- G06N3/06—Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons
-
- 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
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Computational Linguistics (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- General Health & Medical Sciences (AREA)
- Economics (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- Entrepreneurship & Innovation (AREA)
- Neurology (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Quality & Reliability (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Operations Research (AREA)
- Tourism & Hospitality (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The present invention discloses a kind of recommender system clicking rate prediction technique based on deep neural network, including acquisition user clicks behavior as sample, extracts the numerical characteristics input GBDT tree-model training with numerical values recited relationship of sample, obtains GBDT leaf segment dot matrix E1；The behavior sequence that constitutes of article will be clicked by all users in sample and input Attention network, obtain in sample all users to the key training matrix E2 of article；The article characteristics vector for clicking interaction to user carries out sum-average arithmetic, obtains the corresponding click Interactive matrix E3 of user, splices E1, E2 and E3, and inputting, there is the deep neural network model of three layers of hidden layer and one layer of output layer to export prediction result.User is clicked behavior decomposition attributive character by the present invention, by GBDT tree-model, Attention network and deep neural network model nonlinear fitting, construct recommender system clicking rate prediction model, model training obtains prediction result, this method depth excavates the recent interest of user, and extensive degree is high and scalability is strong.
Description
Technical field
The present invention relates to information recommendation system field, in particular to a kind of recommender system based on deep neural network is clicked
Rate prediction technique.
Background technique
With the development of information technology and internet industry, information overload handles the challenge of information at people.For with
For family, it is one extremely important that the content of oneself needs how is quickly and accurately navigated in the resource with exponential increase
And the thing of great challenge.For businessman, how appropriate article is presented to the user in time, thus promote trading volume and
Economic growth and one have much the thing of difficulty.The birth of recommender system greatly alleviates this difficulty.
Recommender system is a kind of information filtering system, main search, point using family portrait, Item Information and user
It the behavioral datas such as hits, collect, recommending its may interested article for user.Wherein, the clicking rate of article is to measure user couple
The important indicator of article fancy grade.The recommender system algorithm that needs to design a model goes prediction user to the clicking rate of article, and will
Article sorts by the clicking rate size predicted, chooses the maximum K article of clicking rate and recommends user.Therefore, recommending system
It is a very important task that clicking rate, which estimates (CTR), in system, and accurately estimating for clicking rate is conducive to improve recommender system
Performance and bring maximized business earnings.
The data that recommender system can utilize are mainly derived from user's portrait, Item Information and user behavior, however
In practical application scene, these data are often the categorical data of discrete type.For example, shaped like [time=Wednesday, gender
=male, city=Beijing] as discrete data, we would generally use onehot coded treatment special at the sparse two-value of higher-dimension
Sign, since model data to be treated are mostly the sparse features of higher-dimension, this is for traditional machine learning model
Bring certain challenge.Carry out clicking rate prediction using linear Logic Regression Models, although linear model have it is simple, can
It is explanatory it is strong, can large-scale parallel training the advantages that, but linear model can not learn the cross feature to high-order, model automatically
Effect too depend on artificial Feature Engineering, when in face of higher-dimension, multi-field sparse features, artificial Feature Engineering is very
Hardly possible finds strong high-order feature.It is some to do the non-thread of feature combination automatically in view of the limitation of linear model itself
Property model be also applied to during CTR estimates, as gradient promotes decision-tree model (GBDT) and Factorization machine model (FM).It utilizes
GBDT model finds strong feature and carries out feature combination, and way is to input the leaf node output of decision tree as the feature of LR,
This is the thinking of early sign combination, but GBDT is ineffective in large-scale sparse id data.The FM model of proposition is very
Good solves the problems, such as the combination of sparse data feature, significantly reduces computation complexity, but model is confined to only
The intersection for considering second order feature is unable to get the feature combination of higher order.As it can be seen that be ctr pre- for traditional machine learning model
Some problems can be had by estimating:
(1) a large amount of manual features are depended on
(2) cross feature of high-order is hardly resulted in
(3) for the sparse id data of higher-dimension, generalization ability is inadequate,
Recently, deep neural network (DNN) has been applied in classification and returning for task, in computer vision, nature
There is good performance in the fields such as Language Processing.Since deep neural network can automatically carry out high-order feature combination, have
Stronger model expressive force, DNN are also successfully applied to CTR and estimate in task.Wide&Deep model is attempted DNN and LR
The common training pattern parameter of models coupling, while low order feature and high-order feature are considered, but LR model part needs to compare
More complex data prediction.In addition, there are also some CTR prediction models based on CNN and RNN, CNN model is confined to adjacent
Feature combination is done between feature, RNN is more suitable for sequential click data in contrast, but both lays particular emphasis on high order cross
The extraction of feature has ignored some useful low order latent structures.DeepFM is equivalent to the LR model optimization in Wide&Deep
For FM model, to construct a clicking rate prediction model end to end.Although DeepFM model has become the master that CTR is estimated
Flow model, but find in practical applications, the scalability of DeepFM is inadequate.There are also the important companies in part in recommendation task
Continuous feature, still may require that sometimes and carries out a small amount of Feature Engineering, such as number of clicks, the Yong Hudian of counting user these last few days
The difference etc. of time is hit, if directly continuous feature is input in neural network model, not only needs to carry out complicated return
One changes, and good effect can not be played to prediction.
Summary of the invention
The main object of the present invention is that one kind is proposed based on recommender system clicking rate prediction model based on depth nerve net
The recommender system clicking rate prediction technique of network, it is intended to overcome problem above.
To achieve the above object, a kind of recommender system clicking rate prediction side based on deep neural network proposed by the present invention
Method includes the following steps:
S10 acquires the click behavior of user as sample, and click behavior includes User ID, click time t, the object that is clicked
Product id, according to the attributive character of User ID splicing correspondence user, it is special to extract attribute for the attributive character of article id splicing correspondence article
With the numerical characteristics of numerical values recited relationship in sign, the training of GBDT tree-model is inputted, GBDT leaf segment dot matrix E1 is obtained；By sample
In click the behavior sequence that constitutes of article by all users and input Attention network, obtain in sample all users to article
Key training matrix E2；The article characteristics vector for clicking interaction to user carries out sum-average arithmetic, obtains the corresponding click of user
Interactive matrix E3；
S20 splices GBDT leaf segment dot matrix E1, key training matrix E2 and click Interactive matrix E3, inputs depth
Neural network model, wherein deep neural network model has three layers of hidden layer and one layer of output layer；
The neuron number that deep neural network model first layer is arranged in S30 is 1024, its second neuron is arranged
Number is 512, and it is 256 that its third layer neuron number, which is arranged, and the activation primitive of hidden layer uses ReLU as activation primitive, formula
It is as follows:
a(l+1)=ReLU (W(l+1)a(l)+b(l+1))
a(l)It indicates l layers of output, while being also l+1 layers of input, b(l+1)Indicate l layers of biasing, W(l+1)Table
Show that l+1 layers of weight matrix, output layer are activation primitive using Sigmoid, formula is as follows；
H indicates the number of plies of hidden layer, a(H)It is H layers of output, w(H+1)And b(H+1)It is all parameter to be asked in output layer,It is the final predicted value of model,
The loss function of model uses negative log-likelihood function:
S is training set sample, and x is the input of neural network, and y ∈ (0,1) is the true tag of sample, and p (x) is nerve net
The output predicted value of network exports predicted value.
Preferably, the columns of the GBDT leaf segment dot matrix E1, key training matrix E2 and click Interactive matrix E3 are homogeneous
Together, the GBDT leaf segment dot matrix E1, key training matrix E2 and click Interactive matrix E3 be spliced into it is longitudinal spliced.
Preferably, the different user clicks the behavior sequence that different articles are constituted and includes user's history behavior sequence and call together
The vector for returning article, activates associated user behavior sequence using article vector is recalled, user is calculated to the article
Key training matrix 2, calculation formula is as follows:
Wherein, { e1...eHBe user behavior sequence, vAIt is the expression vector of article A, a () is preceding to nerve
The activation primitive of network, WjIt is the parameter in neural network, passes through the correlation between each behavior of network query function article A and user
Weight, finally export it is after the weighted sum of user behavior sequence as a result, key training vector as user, all users'
Key training vector forms user to the key training matrix 2 of the article.
User is clicked behavior decomposition attributive character by the present invention, by GBDT tree-model, Attention network and depth nerve
Network model nonlinear fitting, constructs recommender system clicking rate prediction model, and model training obtains prediction result, this method depth
The recent interest of user is excavated, extensive degree is high and scalability is strong.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of invention for those of ordinary skill in the art without creative efforts, can be with
The structure shown according to these attached drawings obtains other attached drawings.
Fig. 1 is the structural schematic diagram of one embodiment of recommender system clicking rate prediction model framework；
Fig. 2 is the Attention network structure；
The embodiments will be further described with reference to the accompanying drawings for the realization, the function and the advantages of the object of the present invention.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiment is only a part of the embodiments of the present invention, instead of all the embodiments.Base
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts it is all its
His embodiment, shall fall within the protection scope of the present invention.
It is to be appreciated that if relating to directionality instruction (such as up, down, left, right, before and after ...) in the embodiment of the present invention,
Then directionality instruction be only used for explain under a certain particular pose (as shown in the picture) between each component relative positional relationship,
Motion conditions etc., if the particular pose changes, directionality instruction is also correspondingly changed correspondingly.
In addition, being somebody's turn to do " first ", " second " etc. if relating to the description of " first ", " second " etc. in the embodiment of the present invention
Description be used for description purposes only, be not understood to indicate or imply its relative importance or implicitly indicate indicated skill
The quantity of art feature." first " is defined as a result, the feature of " second " can explicitly or implicitly include at least one spy
Sign.It in addition, the technical solution between each embodiment can be combined with each other, but must be with those of ordinary skill in the art's energy
It is enough realize based on, will be understood that the knot of this technical solution when conflicting or cannot achieve when occurs in the combination of technical solution
Conjunction is not present, also not the present invention claims protection scope within.
Present invention consideration combines GBDT model with deep neural network, and it is strong, general to construct a nonlinear fitting ability
Change ability height and expansible recommender system clicking rate prediction model, propose one kind based on recommender system clicking rate prediction model
Recommender system clicking rate prediction technique based on deep neural network, the model framework is as shown in Figure 1, the model includes three big portions
Point:
1. importation
It is estimated in scene in clicking rate, feature can be divided into category feature and numerical characteristics, and category feature is in physical meaning
On be that there is no size relations such as user's gender: male or female, and numerical characteristics are that there are size relations such as age of user.
A GBDT tree-model is individually trained using the numerical characteristics in sample, each sample can be assigned to leaf node by GBDT tree-model
On, when a sample by certain tree finally fall in this tree a leaf node on, then in new feature vector this
The corresponding element value of leaf node is 1, and the corresponding element value of other leaf nodes of this tree is 0, such as the leaf of GBDT tree
Child node quantity is 5, and sample 1 has been assigned in the 2nd leaf node, is [0,1,0,0,0] in the corresponding vector of sample 1, with
All samples of this mode can obtain corresponding new feature vector.Eigenmatrix is obtained after the progress of new feature vector is longitudinal spliced
E1.The design of this part is primarily due to GBDT tree-model and is generally better at processing numerical characteristics, by successional numerical characteristics weight
The newly encoded leaf node at discretization facilitates the training for carrying out neural network, and the feature manually extracted can be added in this part,
It is the embodiment of model scalability.
For the behavioural characteristic of user, such as exposure, click, collection behavior, the article id interacted using different behaviors
The behavior sequence of user is constituted by the time sequencing that behavior generates.The behavior sequence of these users is input to Attention net
In network, the user interest profile of a regular length is calculated for each user, obtains spy after the interest characteristics splicing of all users
Levy matrix E2.
For article characteristics, due to can have a variety of different classes of attributes, we are encoded an article using onehot,
The attribute is indicated with 1, and 0 indicates without the attribute, then the also article vector of available each article corresponding one-dimensional 0/1.
For a user, click behavior may be generated with multiple articles whithin a period of time, so, will have and the user produces
The raw article characteristics vector for clicking interaction carries out sum-average arithmetic, and after average, each user corresponds to one-dimensional article characteristics vector,
The article vector of all users obtains eigenmatrix E3 by splicing.
After E1, E2, E2 are spliced, it is input in deep neural network model.
2.Attention network portion
Attention network mechanism is derived from when it should be noted that when some target or some scene, the target internal and
Attention distribution in the scene on the spatial position of everywhere is different.It amplifies on user behavior, user is every time to object
The click behavior of product only can be related with several behaviors in behavior sequence.The structure of Attention is as shown in Fig. 2, user is gone through
History behavior sequence and the vector for recalling article are inputted into Attention network, and using recalling, the activation of article vector is associated
User behavior sequence, calculate user to the key training feature of the article.
Expression formula are as follows:
Wherein, { e1...eHBe user behavior sequence, vAIt is the expression vector of article A, a () is preceding to nerve
The activation primitive of network, WjIt is the parameter in neural network, passes through the correlation between each behavior of network query function article A and user
Weight, finally export after the weighted sum of user behavior sequence as a result, interest vector as user.
3. deep neural network part
Deep neural network (DNN) is designed as three layers of hidden layer structure, every layer of neuron number is respectively 1024,512,
256.The activation primitive of hidden layer uses ReLU as activation primitive, the output of every layer of hidden layer are as follows:
a(l+1)=ReLU (W(l+1)a(l)+b(l+1))
a(l)It indicates l layers of output, while being also l+1 layers of input, b(l+1)Indicate l layers of biasing, W(l+1)Table
Show l+1 layers of weight matrix.It is the output layer of activation primitive finally by a Sigmoid, exports user to candidate item
Click probability, that is, prediction result, expression formula are as follows:
H indicates the number of plies of hidden layer, a(H)It is H layers of output, w(H+1)And b(H+1)It is all parameter to be asked in output layer,It is the final predicted value of model.
The loss function of model uses negative log-likelihood function:
S is training set sample, and x is the input of neural network, and y ∈ (0,1) is the true tag of sample, and p (x) is nerve net
The output predicted value of network.
Generate prediction model practical operation step:
S1: continuous type numerical characteristics are as input one GBDT tree-model of training；
GBDT tree-model output leaf node vector in S2:S1 is denoted as feature vector v1；
S3: user's history behavior sequence and the vector for recalling article are inputted into Attent i on network, using recalling object
Product vector activates associated user behavior sequence, and user is calculated to the key training feature of the article, is denoted as V2；
S4: for remaining article characteristics using user as grouping, the article characteristics vector in every group carries out arithmetic average, obtains
The corresponding feature vector of a user form, this category feature is denoted as v3；
S5:v1, v2, v3 carry out vector splicing by user, and are input in the neural network model of three-decker；
S6: the neuron number of every layer of neural network of setting is respectively 1024,512,256, and every layer of hidden layer uses ReLU
As activation primitive, output layer uses Sigmoid function, the prediction result of output layer final output model.
User is clicked behavior decomposition attributive character by the present invention, by GBDT tree-model, Attention network and depth nerve
Network model nonlinear fitting, constructs recommender system clicking rate prediction model, and model training obtains prediction result, this method depth
The recent interest of user is excavated, extensive degree is high and scalability is strong.
The above description is only a preferred embodiment of the present invention, is not intended to limit the scope of the invention, all at this
Under the inventive concept of invention, using equivalent structure transformation made by description of the invention and accompanying drawing content, or directly/use indirectly
It is included in other related technical areas in scope of patent protection of the invention.
Claims (3)
1. a kind of recommender system clicking rate prediction technique based on deep neural network, which comprises the steps of:
S10 acquires the click behavior of user as sample, and click behavior includes User ID, click time t, the article that is clicked
Id, according to the attributive character of User ID splicing correspondence user, the attributive character of article id splicing correspondence article extracts attributive character
In with numerical values recited relationship numerical characteristics, input GBDT tree-model training, obtain GBDT leaf segment dot matrix E1；It will be in sample
The behavior sequence that constitutes of article is clicked by all users and inputs Attention network, obtains in sample all users to article
Key training matrix E2；The article characteristics vector for clicking interaction to user carries out sum-average arithmetic, obtains the corresponding click of user and hands over
Mutual matrix E3；
S20 splices GBDT leaf segment dot matrix E1, key training matrix E2 and click Interactive matrix E3, input depth nerve
Network model, wherein deep neural network model has three layers of hidden layer and one layer of output layer；
The neuron number that deep neural network model first layer is arranged in S30 is 1024, its second neuron number, which is arranged, is
512, it is 256 that its third layer neuron number, which is arranged, and the activation primitive of hidden layer uses ReLU as activation primitive, and formula is such as
Under:
a(l+1)=ReLU (W(l+1)a(l)+b(l+1))
a(l)It indicates l layers of output, while being also l+1 layers of input, b(l+1)Indicate l layers of biasing, W(l+1)Indicate l+
1 layer of weight matrix, output layer are activation primitive using Sigmoid, and formula is as follows；
H indicates the number of plies of hidden layer, a(H)It is H layers of output, w(H+1)And b(H+1)It is all parameter to be asked in output layer,It is
The final predicted value of model,
The loss function of model uses negative log-likelihood function:
S is training set sample, and x is the input of neural network, and y ∈ (0,1) is the true tag of sample, and p (x) is neural network
Predicted value is exported, predicted value is exported.
2. the recommender system clicking rate prediction technique based on deep neural network as described in claim 1, which is characterized in that institute
The columns for stating GBDT leaf segment dot matrix E1, key training matrix E2 and click Interactive matrix E3 is all the same, the GBDT leaf node
Matrix E1, key training matrix E2 and click Interactive matrix E3 be spliced into it is longitudinal spliced.
3. the recommender system clicking rate prediction technique based on deep neural network as described in claim 1, which is characterized in that institute
Stating different user and clicking the behavior sequence that different articles are constituted includes user's history behavior sequence and the vector for recalling article, is used
It recalls article vector and activates associated user behavior sequence, user is calculated to the key training matrix 2 of the article, meter
It is as follows to calculate formula:
Wherein, { e1...eHBe user behavior sequence, vAIt is the expression vector of article A, a () is feedforward neural network
Activation primitive, WjIt is the parameter in neural network, is weighed by the correlation between each behavior of network query function article A and user
Weight, finally export after the weighted sum of user behavior sequence as a result, key training vector as user, all users' is emerging
Key training matrix 2 of the interesting intensity vector composition user to the article.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910223510.8A CN109960759B (en) | 2019-03-22 | 2019-03-22 | Recommendation system click rate prediction method based on deep neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910223510.8A CN109960759B (en) | 2019-03-22 | 2019-03-22 | Recommendation system click rate prediction method based on deep neural network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109960759A true CN109960759A (en) | 2019-07-02 |
| CN109960759B CN109960759B (en) | 2022-07-12 |
Family
ID=67024795
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910223510.8A Active CN109960759B (en) | 2019-03-22 | 2019-03-22 | Recommendation system click rate prediction method based on deep neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109960759B (en) |
Cited By (49)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110442810A (en) * | 2019-08-08 | 2019-11-12 | 广州华建工智慧科技有限公司 | A kind of mobile terminal BIM model intelligent buffer method based on DeepFM proposed algorithm |
| CN110619540A (en) * | 2019-08-13 | 2019-12-27 | 浙江工业大学 | Click stream estimation method of neural network |
| CN110689376A (en) * | 2019-09-27 | 2020-01-14 | 支付宝(杭州)信息技术有限公司 | Click rate prediction method and device and electronic equipment |
| CN110689368A (en) * | 2019-08-22 | 2020-01-14 | 北京大学（天津滨海）新一代信息技术研究院 | Method for designing advertisement click rate prediction system in mobile application |
| CN110837596A (en) * | 2019-09-16 | 2020-02-25 | 中国平安人寿保险股份有限公司 | Intelligent recommendation method and device, computer equipment and storage medium |
| CN110889063A (en) * | 2019-11-29 | 2020-03-17 | 中山大学 | Video pre-caching method based on Hox process and matrix decomposition |
| CN111045716A (en) * | 2019-11-04 | 2020-04-21 | 中山大学 | Related patch recommendation method based on heterogeneous data |
| CN111046076A (en) * | 2019-12-05 | 2020-04-21 | 深圳索信达数据技术有限公司 | Efficient and accurate analysis method for bank user behavior data |
| CN111062775A (en) * | 2019-12-03 | 2020-04-24 | 中山大学 | Recommendation system recall method based on attention mechanism |
| CN111079014A (en) * | 2019-12-17 | 2020-04-28 | 携程计算机技术（上海）有限公司 | Recommendation method, system, medium and electronic device based on tree structure |
| CN111125521A (en) * | 2019-12-13 | 2020-05-08 | 上海喜马拉雅科技有限公司 | Information recommendation method, device, equipment and storage medium |
| CN111125530A (en) * | 2019-12-24 | 2020-05-08 | 中山大学 | Information flow recommendation method based on multi-type feature deep learning |
| CN111209386A (en) * | 2020-01-07 | 2020-05-29 | 重庆邮电大学 | Personalized text recommendation method based on deep learning |
| CN111242283A (en) * | 2020-01-09 | 2020-06-05 | 支付宝(杭州)信息技术有限公司 | Training method and device for evaluating self-encoder of interaction event |
| CN111339415A (en) * | 2020-02-25 | 2020-06-26 | 中国科学技术大学 | Click rate prediction method and device based on multi-interactive attention network |
| CN111340537A (en) * | 2020-02-20 | 2020-06-26 | 大连理工大学 | Method for establishing advertisement click rate estimation model |
| CN111369278A (en) * | 2020-02-19 | 2020-07-03 | 杭州电子科技大学 | Click rate prediction method based on long-term interest modeling of user |
| CN111400592A (en) * | 2020-03-12 | 2020-07-10 | 山东师范大学 | Personalized course recommendation method and system based on eye movement technology and deep learning |
| CN111538761A (en) * | 2020-04-21 | 2020-08-14 | 中南大学 | Click rate prediction method based on attention mechanism |
| CN111737578A (en) * | 2020-06-22 | 2020-10-02 | 陕西师范大学 | Recommendation method and system |
| CN111784062A (en) * | 2020-07-07 | 2020-10-16 | 上海携旅信息技术有限公司 | Click probability prediction method, system, electronic device, and medium |
| CN111815064A (en) * | 2020-07-21 | 2020-10-23 | 上海数鸣人工智能科技有限公司 | Method for classifying and predicting demands of financial products of users |
| CN111898904A (en) * | 2020-07-28 | 2020-11-06 | 拉扎斯网络科技（上海）有限公司 | Data processing method and device |
| CN111950195A (en) * | 2020-07-28 | 2020-11-17 | 兰笺(苏州)科技有限公司 | Engineering progress prediction method based on portrait system and depth regression model |
| CN112053188A (en) * | 2020-08-25 | 2020-12-08 | 杭州电子科技大学 | Internet advertisement recommendation method based on hybrid deep neural network model |
| CN112084450A (en) * | 2020-09-09 | 2020-12-15 | 长沙理工大学 | Click rate prediction method and system based on convolutional attention network deep session sequence |
| CN112102004A (en) * | 2020-09-18 | 2020-12-18 | 合肥工业大学 | Click rate prediction fusion method based on residual error learning |
| CN112200316A (en) * | 2020-10-30 | 2021-01-08 | 福州大学 | GBDT learning method for online prediction task |
| CN112307257A (en) * | 2020-11-25 | 2021-02-02 | 中国计量大学 | Short video click rate prediction method based on multi-information node graph network |
| CN112418402A (en) * | 2020-11-24 | 2021-02-26 | 百度在线网络技术（北京）有限公司 | Method for recommending object, neural network and training method thereof, and computing device |
| CN112465226A (en) * | 2020-11-27 | 2021-03-09 | 上海交通大学 | User behavior prediction method based on feature interaction and graph neural network |
| CN112492396A (en) * | 2020-12-08 | 2021-03-12 | 中国计量大学 | Short video click rate prediction method based on fine-grained multi-aspect analysis |
| CN112528161A (en) * | 2021-02-07 | 2021-03-19 | 电子科技大学 | Conversation recommendation method based on item click sequence optimization |
| CN112633937A (en) * | 2020-12-30 | 2021-04-09 | 上海数鸣人工智能科技有限公司 | Marketing prediction method based on dimension reduction of depth automatic encoder and gradient lifting decision tree |
| CN112950291A (en) * | 2021-03-31 | 2021-06-11 | 北京奇艺世纪科技有限公司 | Model deviation optimization method, device, equipment and computer readable medium |
| CN113344615A (en) * | 2021-05-27 | 2021-09-03 | 上海数鸣人工智能科技有限公司 | Marketing activity prediction method based on GBDT and DL fusion model |
| CN113342868A (en) * | 2021-08-05 | 2021-09-03 | 腾讯科技（深圳）有限公司 | Information recommendation method, device, equipment and computer readable storage medium |
| CN113360816A (en) * | 2020-03-05 | 2021-09-07 | 北京沃东天骏信息技术有限公司 | Click rate prediction method and device |
| CN113407579A (en) * | 2021-07-15 | 2021-09-17 | 北京百度网讯科技有限公司 | Group query method and device, electronic equipment and readable storage medium |
| CN113409090A (en) * | 2021-07-05 | 2021-09-17 | 中国工商银行股份有限公司 | Training method, prediction method and device of advertisement click rate prediction model |
| CN113420227A (en) * | 2021-07-21 | 2021-09-21 | 北京百度网讯科技有限公司 | Training method of click rate estimation model, and method and device for estimating click rate |
| CN113704441A (en) * | 2021-09-06 | 2021-11-26 | 中国计量大学 | Conversation recommendation method considering importance of item and item attribute feature level |
| CN113722578A (en) * | 2020-05-25 | 2021-11-30 | 北京沃东天骏信息技术有限公司 | Data processing method and device |
| CN113761392A (en) * | 2021-09-14 | 2021-12-07 | 上海任意门科技有限公司 | Content recall method, computing device, and computer-readable storage medium |
| CN113761378A (en) * | 2021-09-14 | 2021-12-07 | 上海任意门科技有限公司 | Content ordering method, computing device and computer-readable storage medium |
| CN113888238A (en) * | 2021-10-25 | 2022-01-04 | 重庆邮电大学 | Advertisement click rate prediction method and device and computer equipment |
| CN113887694A (en) * | 2020-07-01 | 2022-01-04 | 复旦大学 | Click rate estimation model based on characteristic representation under attention mechanism |
| CN114168845A (en) * | 2021-11-24 | 2022-03-11 | 电子科技大学 | Serialization recommendation method based on multi-task learning |
| CN115062220A (en) * | 2022-06-16 | 2022-09-16 | 成都集致生活科技有限公司 | Attention merging-based recruitment recommendation system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110131166A1 (en) * | 2009-12-01 | 2011-06-02 | Hulu Llc | Fuzzy users' attributes prediction based on users' behaviors |
| CN104090919A (en) * | 2014-06-16 | 2014-10-08 | 华为技术有限公司 | Advertisement recommending method and advertisement recommending server |
| CN108416625A (en) * | 2018-02-28 | 2018-08-17 | 阿里巴巴集团控股有限公司 | The recommendation method and apparatus of marketing product |
| CN109299976A (en) * | 2018-09-07 | 2019-02-01 | 深圳大学 | Clicking rate prediction technique, electronic device and computer readable storage medium |
-
2019
- 2019-03-22 CN CN201910223510.8A patent/CN109960759B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110131166A1 (en) * | 2009-12-01 | 2011-06-02 | Hulu Llc | Fuzzy users' attributes prediction based on users' behaviors |
| CN104090919A (en) * | 2014-06-16 | 2014-10-08 | 华为技术有限公司 | Advertisement recommending method and advertisement recommending server |
| CN108416625A (en) * | 2018-02-28 | 2018-08-17 | 阿里巴巴集团控股有限公司 | The recommendation method and apparatus of marketing product |
| CN109299976A (en) * | 2018-09-07 | 2019-02-01 | 深圳大学 | Clicking rate prediction technique, electronic device and computer readable storage medium |
Cited By (78)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110442810B (en) * | 2019-08-08 | 2023-06-13 | 广州华建工智慧科技有限公司 | Mobile terminal BIM model intelligent caching method based on deep FM recommendation algorithm |
| CN110442810A (en) * | 2019-08-08 | 2019-11-12 | 广州华建工智慧科技有限公司 | A kind of mobile terminal BIM model intelligent buffer method based on DeepFM proposed algorithm |
| CN110619540A (en) * | 2019-08-13 | 2019-12-27 | 浙江工业大学 | Click stream estimation method of neural network |
| CN110689368B (en) * | 2019-08-22 | 2022-07-19 | 北京大学（天津滨海）新一代信息技术研究院 | Method for designing advertisement click rate prediction system in mobile application |
| CN110689368A (en) * | 2019-08-22 | 2020-01-14 | 北京大学（天津滨海）新一代信息技术研究院 | Method for designing advertisement click rate prediction system in mobile application |
| CN110837596A (en) * | 2019-09-16 | 2020-02-25 | 中国平安人寿保险股份有限公司 | Intelligent recommendation method and device, computer equipment and storage medium |
| CN110689376A (en) * | 2019-09-27 | 2020-01-14 | 支付宝(杭州)信息技术有限公司 | Click rate prediction method and device and electronic equipment |
| CN111045716A (en) * | 2019-11-04 | 2020-04-21 | 中山大学 | Related patch recommendation method based on heterogeneous data |
| CN111045716B (en) * | 2019-11-04 | 2022-02-22 | 中山大学 | Related patch recommendation method based on heterogeneous data |
| CN110889063A (en) * | 2019-11-29 | 2020-03-17 | 中山大学 | Video pre-caching method based on Hox process and matrix decomposition |
| CN110889063B (en) * | 2019-11-29 | 2023-05-02 | 中山大学 | Video pre-caching method based on Hox process and matrix decomposition |
| CN111062775A (en) * | 2019-12-03 | 2020-04-24 | 中山大学 | Recommendation system recall method based on attention mechanism |
| CN111062775B (en) * | 2019-12-03 | 2023-05-05 | 中山大学 | Recommendation system recall method based on attention mechanism |
| CN111046076A (en) * | 2019-12-05 | 2020-04-21 | 深圳索信达数据技术有限公司 | Efficient and accurate analysis method for bank user behavior data |
| CN111125521A (en) * | 2019-12-13 | 2020-05-08 | 上海喜马拉雅科技有限公司 | Information recommendation method, device, equipment and storage medium |
| CN111079014A (en) * | 2019-12-17 | 2020-04-28 | 携程计算机技术（上海）有限公司 | Recommendation method, system, medium and electronic device based on tree structure |
| CN111079014B (en) * | 2019-12-17 | 2024-02-09 | 携程计算机技术（上海）有限公司 | Recommendation method, system, medium and electronic equipment based on tree structure |
| CN111125530A (en) * | 2019-12-24 | 2020-05-08 | 中山大学 | Information flow recommendation method based on multi-type feature deep learning |
| CN111125530B (en) * | 2019-12-24 | 2023-04-28 | 中山大学 | Information flow recommendation method based on multi-type feature deep learning |
| CN111209386B (en) * | 2020-01-07 | 2022-04-12 | 重庆邮电大学 | Personalized text recommendation method based on deep learning |
| CN111209386A (en) * | 2020-01-07 | 2020-05-29 | 重庆邮电大学 | Personalized text recommendation method based on deep learning |
| CN111242283A (en) * | 2020-01-09 | 2020-06-05 | 支付宝(杭州)信息技术有限公司 | Training method and device for evaluating self-encoder of interaction event |
| CN111242283B (en) * | 2020-01-09 | 2021-06-25 | 支付宝(杭州)信息技术有限公司 | Training method and device for evaluating self-encoder of interaction event |
| CN111369278A (en) * | 2020-02-19 | 2020-07-03 | 杭州电子科技大学 | Click rate prediction method based on long-term interest modeling of user |
| CN111340537B (en) * | 2020-02-20 | 2022-10-04 | 大连理工大学 | Method for establishing advertisement click rate estimation model |
| CN111340537A (en) * | 2020-02-20 | 2020-06-26 | 大连理工大学 | Method for establishing advertisement click rate estimation model |
| CN111339415A (en) * | 2020-02-25 | 2020-06-26 | 中国科学技术大学 | Click rate prediction method and device based on multi-interactive attention network |
| CN111339415B (en) * | 2020-02-25 | 2023-06-20 | 中国科学技术大学 | Click rate prediction method and device based on multi-interactive attention network |
| CN113360816A (en) * | 2020-03-05 | 2021-09-07 | 北京沃东天骏信息技术有限公司 | Click rate prediction method and device |
| CN111400592B (en) * | 2020-03-12 | 2023-06-20 | 山东师范大学 | Personalized course recommendation method and system based on eye movement technology and deep learning |
| CN111400592A (en) * | 2020-03-12 | 2020-07-10 | 山东师范大学 | Personalized course recommendation method and system based on eye movement technology and deep learning |
| CN111538761A (en) * | 2020-04-21 | 2020-08-14 | 中南大学 | Click rate prediction method based on attention mechanism |
| CN113722578A (en) * | 2020-05-25 | 2021-11-30 | 北京沃东天骏信息技术有限公司 | Data processing method and device |
| CN111737578B (en) * | 2020-06-22 | 2024-04-02 | 陕西师范大学 | Recommendation method and system |
| CN111737578A (en) * | 2020-06-22 | 2020-10-02 | 陕西师范大学 | Recommendation method and system |
| CN113887694A (en) * | 2020-07-01 | 2022-01-04 | 复旦大学 | Click rate estimation model based on characteristic representation under attention mechanism |
| CN111784062A (en) * | 2020-07-07 | 2020-10-16 | 上海携旅信息技术有限公司 | Click probability prediction method, system, electronic device, and medium |
| CN111815064A (en) * | 2020-07-21 | 2020-10-23 | 上海数鸣人工智能科技有限公司 | Method for classifying and predicting demands of financial products of users |
| CN111950195B (en) * | 2020-07-28 | 2024-01-19 | 兰笺(苏州)科技有限公司 | Project progress prediction method based on portrait system and depth regression model |
| CN111898904B (en) * | 2020-07-28 | 2024-03-22 | 拉扎斯网络科技（上海）有限公司 | Data processing method and device |
| CN111950195A (en) * | 2020-07-28 | 2020-11-17 | 兰笺(苏州)科技有限公司 | Engineering progress prediction method based on portrait system and depth regression model |
| CN111898904A (en) * | 2020-07-28 | 2020-11-06 | 拉扎斯网络科技（上海）有限公司 | Data processing method and device |
| CN112053188A (en) * | 2020-08-25 | 2020-12-08 | 杭州电子科技大学 | Internet advertisement recommendation method based on hybrid deep neural network model |
| CN112053188B (en) * | 2020-08-25 | 2024-01-16 | 杭州电子科技大学 | Internet advertisement recommendation method based on hybrid deep neural network model |
| CN112084450A (en) * | 2020-09-09 | 2020-12-15 | 长沙理工大学 | Click rate prediction method and system based on convolutional attention network deep session sequence |
| CN112102004B (en) * | 2020-09-18 | 2022-08-30 | 合肥工业大学 | Click rate prediction fusion method based on residual error learning |
| CN112102004A (en) * | 2020-09-18 | 2020-12-18 | 合肥工业大学 | Click rate prediction fusion method based on residual error learning |
| CN112200316A (en) * | 2020-10-30 | 2021-01-08 | 福州大学 | GBDT learning method for online prediction task |
| CN112200316B (en) * | 2020-10-30 | 2022-07-15 | 福州大学 | GBDT learning method for online prediction task |
| CN112418402A (en) * | 2020-11-24 | 2021-02-26 | 百度在线网络技术（北京）有限公司 | Method for recommending object, neural network and training method thereof, and computing device |
| CN112418402B (en) * | 2020-11-24 | 2023-08-11 | 百度在线网络技术（北京）有限公司 | Method for recommending object, neural network, training method of neural network and computing device |
| CN112307257B (en) * | 2020-11-25 | 2021-06-15 | 中国计量大学 | Short video click rate prediction method based on multi-information node graph network |
| CN112307257A (en) * | 2020-11-25 | 2021-02-02 | 中国计量大学 | Short video click rate prediction method based on multi-information node graph network |
| CN112465226A (en) * | 2020-11-27 | 2021-03-09 | 上海交通大学 | User behavior prediction method based on feature interaction and graph neural network |
| CN112492396A (en) * | 2020-12-08 | 2021-03-12 | 中国计量大学 | Short video click rate prediction method based on fine-grained multi-aspect analysis |
| CN112633937B (en) * | 2020-12-30 | 2023-10-20 | 上海数鸣人工智能科技有限公司 | Marketing prediction method based on dimension reduction and GBDT (Global positioning System) of depth automatic encoder |
| CN112633937A (en) * | 2020-12-30 | 2021-04-09 | 上海数鸣人工智能科技有限公司 | Marketing prediction method based on dimension reduction of depth automatic encoder and gradient lifting decision tree |
| CN112528161B (en) * | 2021-02-07 | 2021-04-30 | 电子科技大学 | Conversation recommendation method based on item click sequence optimization |
| CN112528161A (en) * | 2021-02-07 | 2021-03-19 | 电子科技大学 | Conversation recommendation method based on item click sequence optimization |
| CN112950291A (en) * | 2021-03-31 | 2021-06-11 | 北京奇艺世纪科技有限公司 | Model deviation optimization method, device, equipment and computer readable medium |
| CN113344615A (en) * | 2021-05-27 | 2021-09-03 | 上海数鸣人工智能科技有限公司 | Marketing activity prediction method based on GBDT and DL fusion model |
| CN113344615B (en) * | 2021-05-27 | 2023-12-05 | 上海数鸣人工智能科技有限公司 | Marketing campaign prediction method based on GBDT and DL fusion model |
| CN113409090A (en) * | 2021-07-05 | 2021-09-17 | 中国工商银行股份有限公司 | Training method, prediction method and device of advertisement click rate prediction model |
| CN113407579A (en) * | 2021-07-15 | 2021-09-17 | 北京百度网讯科技有限公司 | Group query method and device, electronic equipment and readable storage medium |
| CN113407579B (en) * | 2021-07-15 | 2024-01-19 | 北京百度网讯科技有限公司 | Group query method, device, electronic equipment and readable storage medium |
| CN113420227B (en) * | 2021-07-21 | 2024-05-14 | 北京百度网讯科技有限公司 | Training method of click rate estimation model, click rate estimation method and device |
| CN113420227A (en) * | 2021-07-21 | 2021-09-21 | 北京百度网讯科技有限公司 | Training method of click rate estimation model, and method and device for estimating click rate |
| CN113342868A (en) * | 2021-08-05 | 2021-09-03 | 腾讯科技（深圳）有限公司 | Information recommendation method, device, equipment and computer readable storage medium |
| CN113342868B (en) * | 2021-08-05 | 2021-11-02 | 腾讯科技（深圳）有限公司 | Information recommendation method, device, equipment and computer readable storage medium |
| CN113704441A (en) * | 2021-09-06 | 2021-11-26 | 中国计量大学 | Conversation recommendation method considering importance of item and item attribute feature level |
| CN113761392B (en) * | 2021-09-14 | 2022-04-12 | 上海任意门科技有限公司 | Content recall method, computing device, and computer-readable storage medium |
| CN113761392A (en) * | 2021-09-14 | 2021-12-07 | 上海任意门科技有限公司 | Content recall method, computing device, and computer-readable storage medium |
| CN113761378A (en) * | 2021-09-14 | 2021-12-07 | 上海任意门科技有限公司 | Content ordering method, computing device and computer-readable storage medium |
| CN113888238A (en) * | 2021-10-25 | 2022-01-04 | 重庆邮电大学 | Advertisement click rate prediction method and device and computer equipment |
| CN113888238B (en) * | 2021-10-25 | 2024-04-09 | 重庆邮电大学 | Advertisement click rate prediction method and device and computer equipment |
| CN114168845A (en) * | 2021-11-24 | 2022-03-11 | 电子科技大学 | Serialization recommendation method based on multi-task learning |
| CN114168845B (en) * | 2021-11-24 | 2023-08-15 | 电子科技大学 | Serialized recommendation method based on multitask learning |
| CN115062220A (en) * | 2022-06-16 | 2022-09-16 | 成都集致生活科技有限公司 | Attention merging-based recruitment recommendation system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109960759B (en) | 2022-07-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109960759A (en) | Recommender system clicking rate prediction technique based on deep neural network | |
| Zhou | Product advertising recommendation in e-commerce based on deep learning and distributed expression | |
| CN109299396A (en) | Merge the convolutional neural networks collaborative filtering recommending method and system of attention model | |
| Hsu et al. | Incremental clustering of mixed data based on distance hierarchy | |
| CN109299976A (en) | Clicking rate prediction technique, electronic device and computer readable storage medium | |
| CN107122809A (en) | Neural network characteristics learning method based on image own coding | |
| CN110083770A (en) | It is a kind of based on feature level deeper from the sequence of recommendation method of attention network | |
| CN110674407A (en) | Hybrid recommendation method based on graph convolution neural network | |
| Wang et al. | Case-based reasoning for product style construction and fuzzy analytic hierarchy process evaluation modeling using consumers linguistic variables | |
| CN109214562A (en) | A kind of power grid scientific research hotspot prediction and method for pushing based on RNN | |
| CN111563770A (en) | Click rate estimation method based on feature differentiation learning | |
| CN112765480A (en) | Information pushing method and device and computer readable storage medium | |
| CN108984555A (en) | User Status is excavated and information recommendation method, device and equipment | |
| CN106897776A (en) | A kind of continuous type latent structure method based on nominal attribute | |
| Lin et al. | Intelligent decision support for new product development: a consumer-oriented approach | |
| CN115481325A (en) | Personalized news recommendation method and system based on user global interest migration perception | |
| Upadhya et al. | Deep neural network models for question classification in community question-answering forums | |
| CN106600347A (en) | Method for constructing sequence forecasting model based on multi-visual-angle data and cycle network | |
| CN111597428A (en) | Recommendation method for splicing user and article with q-separation k sparsity | |
| CN111241372B (en) | Method for predicting color harmony degree according to user preference learning | |
| CN110472108A (en) | Garment fabric sample retrieving method based on text profile matching | |
| JP2023550510A (en) | Recommendation methods, devices, electronic devices and storage media | |
| Wang et al. | Research on GRU and CNN combined recommendation model with self-attention mechanism | |
| Wu | Evolutionary design method for transforming ceramic cultural elements into cultural and creative products. | |
| Liu et al. | An enterprise operation management method based on mobile edge computing and data mining |
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 |