CN108460489A - A kind of user behavior analysis based on big data technology and service recommendation frame - Google Patents
A kind of user behavior analysis based on big data technology and service recommendation frame Download PDFInfo
- Publication number
- CN108460489A CN108460489A CN201810212948.1A CN201810212948A CN108460489A CN 108460489 A CN108460489 A CN 108460489A CN 201810212948 A CN201810212948 A CN 201810212948A CN 108460489 A CN108460489 A CN 108460489A
- Authority
- CN
- China
- Prior art keywords
- user
- behavior
- service
- service recommendation
- user behavior
- 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
- 238000004458 analytical method Methods 0.000 title claims abstract description 22
- 238000005516 engineering process Methods 0.000 title claims abstract description 17
- 230000006399 behavior Effects 0.000 claims abstract description 79
- 238000013528 artificial neural network Methods 0.000 claims abstract description 12
- 238000007418 data mining Methods 0.000 claims abstract description 5
- 238000000034 method Methods 0.000 claims description 30
- 238000003066 decision tree Methods 0.000 claims description 21
- 238000000605 extraction Methods 0.000 claims description 12
- 230000003542 behavioural effect Effects 0.000 claims description 9
- 230000033001 locomotion Effects 0.000 claims description 8
- 230000000306 recurrent effect Effects 0.000 claims description 8
- 238000013135 deep learning Methods 0.000 claims description 5
- 238000013461 design Methods 0.000 claims description 4
- 230000011218 segmentation Effects 0.000 claims description 3
- 230000006870 function Effects 0.000 claims 1
- 238000013507 mapping Methods 0.000 claims 1
- 238000003062 neural network model Methods 0.000 claims 1
- 238000004364 calculation method Methods 0.000 abstract description 7
- 238000010801 machine learning Methods 0.000 abstract description 5
- 238000010223 real-time analysis Methods 0.000 abstract 1
- 238000012549 training Methods 0.000 description 8
- 238000012545 processing Methods 0.000 description 5
- 239000011159 matrix material Substances 0.000 description 4
- 238000012544 monitoring process Methods 0.000 description 3
- 210000000988 bone and bone Anatomy 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 238000000714 time series forecasting Methods 0.000 description 2
- 241001269238 Data Species 0.000 description 1
- 241000270708 Testudinidae Species 0.000 description 1
- 230000001133 acceleration Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 239000002131 composite material Substances 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 210000005036 nerve Anatomy 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000013138 pruning Methods 0.000 description 1
- 230000003252 repetitive effect Effects 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
Classifications
-
- 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"
-
- 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
- G06Q50/10—Services
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- General Physics & Mathematics (AREA)
- Tourism & Hospitality (AREA)
- Marketing (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- General Business, Economics & Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Biophysics (AREA)
- Game Theory and Decision Science (AREA)
- Primary Health Care (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Development Economics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
A kind of user behavior analysis based on big data technology and service recommendation mechanism is claimed in the present invention, is related to the users' behavior model based on data mining and service recommendation two parts.Real-time Research on Calculation based on big data provides accurate and real-time analysis result for data service layer.Modeling analysis and service recommendation are carried out to user behavior based on machine learning algorithm, users' behavior model is accurately generated using machine learning algorithm.On the basis of users' behavior model, service recommendation is carried out using deep neural network algorithm, achieve the purpose that quick and precisely to identify user behavior and carries out service recommendation.
Description
Technical field
The present invention relates to a kind of user behavior analysis based on big data technology and service recommendation mechanism.Pass through intelligent video
Monitoring system obtains user behavior characteristics, carries out modeling analysis and service recommendation to user behavior based on machine learning algorithm, adopts
Users' behavior model is accurately generated with machine learning algorithm.On the basis of users' behavior model, depth nerve is utilized
Network algorithm carries out service recommendation, belongs to the crossing domain of data mining and Internet of Things.
Background technology
Big data is inextricably linked with cloud computing platform.Data are assets, are indispensable basic resources,
Bottom of the cloud computing as computing resource supports the big data processing on upper layer, realizes the search efficiency of real-time interactive and divides
Analysis ability.The solution providers such as IBM, the inscriptions on bones or tortoise shells and HP, equipment vendor are based primarily upon Hadoop framework and are provided for corporate client
Big data application product and solution.Wherein, the big data product that IBM is provided includes being developed based on Hadoop Open Source Platforms
The number greatly such as cloud platform system, flow data processing software Streams, analysis tool BigIn-sights, data warehouse Warehouse
According to analysis product.
The existing technology in intelligent Community has been fallen behind, the data processing method for being still website formula of use.And it is existing
External famous enterprise IT core calculations process layers are mostly Hadoop, MapReduce, Spark, herein on, also application has different
Calculation paradigm, such as batch processing, stream process and figure calculating etc..Famous Amazon exactly establishes flexible MapReduce programs
On Hadoop frames top, personal behavior model is efficiently accurately analyzed and establishes, and fast and accurately to billions of people
Group carries out service push.
China " intelligence community " achieves positive progress, has benefited from the fast development of big data, cloud computing technology, based on big
The service cloud platform of data is also at the primary development object of many " intelligence communities ".So under current technology, user behavior mould
Formula is portrayed inaccurately, and not in time, the service so as to cause service and the user's actual need of system recommendation tries to go south by driving the chariot north.
Invention content
In order to overcome above-mentioned defect existing in the prior art, the present invention provides a kind of users based on big data technology
Behavioural analysis and service recommendation frame, the behavior pattern for accurately portraying user, and carry out accurately service recommendation.In this base
On plinth, the present invention quickly and accurately generates user's behavior prediction mould using machine algorithms such as star skeleton technique, decision tree C5.0
Type provides accurate personalized recommendation to the user in service recommendation based on deep neural network learning algorithm.
The frame that above-mentioned purpose to realize the present invention is proposed, it includes two portions to we provide specific solution
Point；
1, the quick service recommendation method of tensor resolution based on deep learning.
For accurate, the rapid feature of the following community service cloud platform analysis engine, propose a kind of based on data mining
Users' behavior model, the model by extract user behavior characteristics generate user draw a portrait, using one kind be based on resident's row
User's behavior prediction method for prognostic chart and Recognition with Recurrent Neural Network predicts user behavior.Use tensor under deep learning
Decompose quick proposed algorithm：It is based on user's portrait likelihood segmentation region using user, service, time and is split processing, then
Service recommendation is carried out using the quick proposed algorithm of tensor resolution, is reached and is quick and precisely identified user behavior and carry out service recommendation
Purpose provides support for big data analysis engine.
2, the real-time computing technique based on parallel computation frame.
The Large Scale Neural Networks training parallel computing platform of a customization is realized, platform is excellent using distributed storage
Change strategy, the design philosophy that memory calculates, realizes the efficient neural network repetitive exercise parallelization calculation block calculated based on memory
Frame ensures the quick execution of neural network BP training algorithm.Uniform programming model and interface are realized based on matrix model, are answered for system
It is analyzed to calculate with layer data and transparent acceleration platform and ease for use interface is provided.
Description of the drawings
Fig. 1 is the system construction drawing of the present invention；
Fig. 2 is decision Tree algorithms flow chart
Fig. 3 is decision tree implementation procedure flow chart
Fig. 4 is user behavior Time series forecasting model figure
Specific implementation mode
The specific implementation of the present invention is further explained in detail below in conjunction with the accompanying drawings.
Fig. 1 is the system construction drawing of the present invention.It is related to users' behavior model and service based on data mining to push away
Recommend two parts.User behavior pattern is predicted to establish prediction model for the behavior of community users with service recommendation, be generated accurately
Service recommendation.The feature of user is extracted using star-like skeleton technique, and feature is stored in resident's Behavioral training number
According to library.User's behavior prediction method tentative prediction user behavior based on behaviour prognostic chart and Recognition with Recurrent Neural Network.For solution
Certainly the problem of recommendation service accuracy rate, the quick proposed algorithm of tensor resolution under deep learning is used：Utilize user, service, time
It is split based on user's portrait likelihood segmentation region and is processed into 3 dimension datas, recycle instantly popular tensor resolution label
Calculating of the quick proposed algorithm into row label.It is finally introducing depth learning technology, is non-convex optimization problem, nothing for tensor resolution
Method is ensured to generate the problems such as three hidden eigenmatrixes, algorithmic statement process are highly dependent upon initial value design, can be instructed using deep learning
Accurate initial hidden eigenmatrix is practised, and then is more accurately predicted by the acquisition of tensor resolution algorithm.Steps are as follows：
1, user behavior characteristics are extracted
User behavior characteristics are obtained by intelligent video monitoring system.The extraction user behavior of user from video,
It is the key that user behavior extraction.Movement is made of a series of human body attitude sequences, during gesture recognition, if
The characteristics of human motion is described using frontal outline, is shot due to camera, it is difficult to the variation characteristic of movement is extracted, because
This, the description of human motion is carried out using side profile.The behavioural characteristic for describing human motion simultaneously is retouched if too simple
Inaccuracy is stated, it is difficult to accurately identify human body behavior, if behavioural characteristic is excessively complicated, calculation amount is too big, influences system operation
Efficiency.This project using the motion feature of star-like skeleton technique extraction is made with the angular deviation of human body head and four limbs to barycenter
It needs to be come out moving target recognition with contours extract method before extraction for the behavioural characteristic of moving target in video pictures.
Step 1：Mixed Gauss model background modeling.According to video monitoring, the video data of needs is acquired, using mixing
Gauss model background modeling method, by moving target from video pictures extraction process.Image background is obtained by background modeling
Model, then difference is carried out by background model, obtained image carries out binary conversion treatment again, obtains a pixel for being less than threshold value
Point, the pixel as the moving target.
Step 2：Contours extract algorithm based on OpenCV.The contours extract algorithm that this project uses is in OpenCV letters
Number is changed on the basis of library, is mainly used imgproc modules and is carried out image procossing, video modules carry out video point
Analysis, calib3d modules are carried out 3D modeling and camera Calibration, are finally carried using cvFindContours to carry out contour feature
It takes.
Step 3：Star skeleton technique.After having obtained the integrity profile of moving object attitude, so that it may to use star bone
Frame method obtains the star skeleton of moving object.It is specifically divided into the following steps：1. the posture profile based on acquisition obtains profile matter
The heart；2. calculate profile on point to profile barycenter Euclidean distance；IDFT, LPF, DFT processing are carried out 3. adjusting the distance, reduction is made an uproar
Sound, the output sequence after obtaining smoothly；4. finding the local maximum of smooth output sequence, then constitutes and move with profile barycenter
The star skeleton of target.The offset of 5 maximum points and barycenter that are obtained according to star skeleton technique is special as human body behavior
Sign.
Platform prepares for user behavior analysis by extracting user behavior characteristics, the user behavior characteristics of extraction is deposited
Access customer Behavioral training database, user behavior are trained library as user behavior characteristics matching library, are provided for user's behavior prediction
Basis.
2, users' behavior model
Fig. 2 is decision Tree algorithms flow chart.By the user behavior characteristics of extraction, behavior prediction is carried out to user.User
Behavioural analysis prediction model is designed based on decision Tree algorithms C5.0.Decision Tree algorithms are a kind of typical machine learning
Sorting algorithm is generally applied in various prediction models and grader, and application decision tree realizes the behavior to resident
It is predicted.
Entire decision tree implementation procedure includes mainly following two steps：
Step 1：Training sample is extracted from resident's Behavioral training database and establishes decision-tree model, that is, carries out machine
Device learns, and realization generally includes two parts of structure decision tree and decision tree beta pruning.
Step 2：Classification processing is carried out to the training sample of extraction based on the decision-tree model built above, and then is determined
The direction of each branch in next step.
Structure decision tree mainly determines the topological structure between different characteristic attribute by Attributions selection measurement, to complete
The structure of decision tree.Node N is created by sample set S first, then it is judged, finally chooses best attribute pair
S carries out splitting operation, finally obtains complete decision tree, further generates user's behavior prediction vector and indicates.It is determined based on above
Plan tree principle, decision tree during user behavior analysis are obtained a user's behavior prediction figure to structure, are indicated using Fig. 3
Decision tree predict the behavior of some user.
3, the user's behavior prediction method based on user's behavior prediction figure and Recognition with Recurrent Neural Network
Fig. 4 is user behavior Time series forecasting model figure.After obtaining the user's behavior prediction vector based on decision tree and indicating,
The selection record of service models the sequential behavior of each user according to user, to which the acts and efforts for expediency for portraying user are special
Sign.
Modeling and forecasting is carried out to the behavior of user using Recognition with Recurrent Neural Network.Entire modeling implementation procedure includes mainly following
Three steps：
Step 1：For each user u, sequence can be obtained after being ranked up according to the time to the usage record of service
v1,v2,...vt-1,vt..., corresponding result is y1,y2...,yt-1,yt...。
Step 2：In current time t, it is based on usage history y1...,yt-1, to the parameter W in Recognition with Recurrent Neural Network4,W5With
β is learnt, and then predicts select probability of the user to all services.
Step 3：In t moment, with the historical information v of user u1,...,vt-1And y1,...,yt-1It is training data to each
User establishes temporal model.When carrying out service recommendation, to each service v ∈ V prediction users to the selection rate of the service, and will
The high service of selection rate carries out service accuracy rate and calculates.
4, data prediction
Pretreatment is intended to implement cutting to user-service-time three-dimensional tensor global view, to isolate data scale more
Suitable several small data sets so as to proposed algorithm parallel execution.User-service-time three-dimensional tensor global view is enabled to be
GRM×N×C, wherein M, N, C respectively represents number of users in global view, service number, time points.Obviously, this is a super large
Scale three-dimensional tensor, preprocessing process are as follows.
Step 1：Service dimension cutting.In field of service calculation, any given user task can be divided into two classes：Atom
Demand for services type task, composite services demand type task.The former need to only be based on single candidate service collection and implement preferably, and the latter only needs
A series of relevant candidate service collection in subtasks is implemented preferably, to be both not required to run proposed algorithm based on global view.It borrows
Help Relevant Service Discovery Technologies can be extracted from global view only with the relevant Services Subset of given user task.
By semantic-based Web service discovery technique process, a large amount of unrelated services can be filtered, by task related service
It is cut into from global view, and several Services Subsets is separated into according to subtask.Each Services Subset corresponds to one respectively
User-service that a data scale is greatly reduced-time three-dimensional tensor RM×N×C(wherein, n represents the service number in tensor, and n
≤N)。
Step 2：User's dimension is cut.In view of user-service-time tensor R of gained after service dimension cuttingM×N×C
It is still large-scale, userbase is extremely huge in especially practical SOA system environments, therefore still needs to cut user's dimension
It cuts.User's context feature can help to existing user-service-time tensor RM×N×CCutting.
Step 3：Time dimension stipulations.Tensor RM×N×CIn, C represents most fine-grained time point, this recommends operation
Complexity is higher for algorithm, must carry out stipulations.Stipulations key is the division of timeslice, and c is enabled to represent shape after division time point
At time the piece number, then the tensor R after time stipulationsM×N×CIt is converted into Ri×j×k.It is final that data are advised by data prediction
Mould is RM×N×CSmall data set.
5, tensor resolution label recommendations algorithm
After data prediction, several user-service-time decimals will be obtained according to collection, enable Ri×j×kRepresent arbitrary small number evidence
Collection, can be based on Ri×j×kThe tensor resolution algorithm of design.It is defined as follows：
Xijk≈Yijk (1)
Wherein Y represents different tensor resolution methods.
The principle of tensor resolution is by high-dimensional data space Rm×n×cIt is mapped to by the hidden eigenmatrix U of userij, the hidden spy of service
Levy matrix Vjk, time hidden eigenmatrix Wik(1≤ijk of dimension) 1≤ijk in the low-dimensional feature space of composition：In view of tripartite graph
Correlation two-by-two between middle triple element, it is proposed that following tensor resolution model：
Yijk=UijVjk+VjkWik+UijWik (2)
The relationship of user node i and service node j are as follows：
Wherein, UijIt is I × J matrixes, VjkIt is J × K matrix, WikIt is I × K matrix, two formulas above comparison expression.(3) model in
Show the direct relation between user and service, other direct relations are similarly.
Claims (6)
1. a kind of user behavior analysis based on big data technology and service recommendation mechanism, it is characterised in that：Main includes being based on
The users' behavior model and service recommendation two parts of data mining.Users' behavior model is by extracting user behavior
Feature generates user's portrait, uses a kind of user's behavior prediction method pair based on behaviour prognostic chart and Recognition with Recurrent Neural Network
User behavior is predicted.To achieve the purpose that quick and precisely to identify user behavior and carrying out service recommendation, deep learning is used
The lower quick proposed algorithm of tensor resolution：It is based on user's portrait likelihood segmentation region using user, service, time and is split place
Reason recycles the quick proposed algorithm of tensor resolution to carry out service recommendation.
2. a kind of user behavior analysis based on big data technology and service recommendation mechanism according to claim 1, feature
It is：User behavior characteristics extraction is as follows：
S21：Moving target is extracted using contours extract method.It is made of a series of human body attitude sequences due to moving, knows in posture
During other, it is difficult to extract the variation characteristic of movement, the description of human motion is carried out using side profile.
S22：Mixed Gauss model background modeling：By moving target from video pictures extraction process.
S23：Contours extract algorithm based on OpenCV：It is changed, and then is used on the basis of OpenCV function libraries
CvFindContours carries out contour feature extraction.
S24：Star-like skeleton technique：The offset of 5 maximum points and barycenter that are obtained according to star skeleton technique is as human body
Behavioural characteristic.
3. a kind of user behavior analysis based on big data technology and service recommendation mechanism according to claim 1, feature
It is：Users' behavior model is as follows：
By the user behavior characteristics of extraction, behavior prediction is carried out to user.User behavior analysis prediction model is to be based on decision
Tree algorithm C5.0 is designed, and application decision tree realizes that the behavior to user is predicted, and then it is pre- to obtain user behavior
Mapping.
4. a kind of user behavior analysis based on big data technology and service recommendation mechanism according to claim 1, feature
It is：User's behavior prediction method based on user's behavior prediction figure and Recognition with Recurrent Neural Network is as follows：
User's behavior prediction vector based on decision tree, the sequential behavior according to user to the selection record of service to each user
It is modeled, to portray the acts and efforts for expediency feature of user.Modeling and forecasting is carried out to the behavior of user using Recognition with Recurrent Neural Network
Model.
5. a kind of user behavior analysis based on big data technology and service recommendation mechanism according to claim 1, feature
It is：Data prediction is as follows：
S51：Service dimension cutting：By semantic-based Web service discovery technique process, a large amount of unrelated services can be filtered,
Task related service is cut into from global view, and several Services Subsets are separated into according to subtask.
S52：User's latitude is cut：Since userbase is extremely huge in practical SOA system environments, user's dimension need to be cut
It cuts
S53：Time latitude stipulations：Tensor RM×N×CIn, C represents most fine-grained time point, this is for running proposed algorithm
Complexity is higher, must carry out stipulations.
6. a kind of user behavior analysis based on big data technology and service recommendation mechanism according to claim 1, feature
It is：Tensor resolution label recommendations algorithm is as follows：
After data prediction, several user-service-time decimals will be obtained according to collection, enable Ri×j×kArbitrarily small data set is represented,
Based on Ri×j×kDesign tensor resolution algorithm.Definition：Xijk≈Yijk。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810212948.1A CN108460489A (en) | 2018-03-15 | 2018-03-15 | A kind of user behavior analysis based on big data technology and service recommendation frame |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810212948.1A CN108460489A (en) | 2018-03-15 | 2018-03-15 | A kind of user behavior analysis based on big data technology and service recommendation frame |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108460489A true CN108460489A (en) | 2018-08-28 |
Family
ID=63236431
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810212948.1A Pending CN108460489A (en) | 2018-03-15 | 2018-03-15 | A kind of user behavior analysis based on big data technology and service recommendation frame |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108460489A (en) |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109525665A (en) * | 2018-11-16 | 2019-03-26 | 济南浪潮高新科技投资发展有限公司 | A kind of wound visitor's cloud center configuration recommended method based on crowdsourcing |
| CN109522742A (en) * | 2018-10-26 | 2019-03-26 | 贵州斯曼特信息技术开发有限责任公司 | A kind of batch processing method of computer big data |
| CN110020228A (en) * | 2019-04-08 | 2019-07-16 | 浙江大学城市学院 | A kind of relevance evaluation method for Internet of Things Item Information searching order |
| CN110362780A (en) * | 2019-07-17 | 2019-10-22 | 北京航空航天大学 | A kind of big data tensor canonical decomposition calculation method based on Shen prestige many-core processor |
| CN111476202A (en) * | 2020-04-30 | 2020-07-31 | 杨九妹 | User behavior analysis method and system of financial institution security system and robot |
| CN111797874A (en) * | 2019-04-09 | 2020-10-20 | Oppo广东移动通信有限公司 | Behavior prediction method, behavior prediction device, storage medium and electronic equipment |
| CN112182395A (en) * | 2020-10-10 | 2021-01-05 | 深圳市万佳安物联科技股份有限公司 | Financial service personalized recommendation device and method based on time sequence |
| CN112307352A (en) * | 2020-11-26 | 2021-02-02 | 腾讯科技（深圳）有限公司 | Content recommendation method, system, device and storage medium |
| CN112417310A (en) * | 2019-08-21 | 2021-02-26 | 上海掌门科技有限公司 | Method for establishing intelligent service index and recommending intelligent service |
| CN112541407A (en) * | 2020-08-20 | 2021-03-23 | 同济大学 | Visual service recommendation method based on user service operation flow |
| CN113095084A (en) * | 2021-03-16 | 2021-07-09 | 重庆邮电大学 | Semantic service matching method and device in Internet of things and storage medium |
| US11250322B2 (en) | 2019-07-15 | 2022-02-15 | Sony Interactive Entertainment LLC | Self-healing machine learning system for transformed data |
| WO2022121705A1 (en) * | 2020-12-10 | 2022-06-16 | 株式会社日立制作所 | Information processing method, apparatus and device |
| CN116485587A (en) * | 2023-04-21 | 2023-07-25 | 深圳润高智慧产业有限公司 | Community service acquisition method, community service providing method, electronic device and storage medium |
| US11983609B2 (en) | 2019-07-10 | 2024-05-14 | Sony Interactive Entertainment LLC | Dual machine learning pipelines for transforming data and optimizing data transformation |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105550950A (en) * | 2015-11-20 | 2016-05-04 | 广东工业大学 | Location-based service travel recommendation method |
| CN106022231A (en) * | 2016-05-11 | 2016-10-12 | 浙江理工大学 | Multi-feature-fusion-based technical method for rapid detection of pedestrian |
| CN106649658A (en) * | 2016-12-13 | 2017-05-10 | 重庆邮电大学 | Recommendation system and method for improving user role undifferentiated treatment and data sparseness |
| CN107423442A (en) * | 2017-08-07 | 2017-12-01 | 火烈鸟网络(广州)股份有限公司 | Method and system, storage medium and computer equipment are recommended in application based on user's portrait behavioural analysis |
| CN107507054A (en) * | 2017-07-24 | 2017-12-22 | 哈尔滨工程大学 | A kind of proposed algorithm based on Recognition with Recurrent Neural Network |
-
2018
- 2018-03-15 CN CN201810212948.1A patent/CN108460489A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105550950A (en) * | 2015-11-20 | 2016-05-04 | 广东工业大学 | Location-based service travel recommendation method |
| CN106022231A (en) * | 2016-05-11 | 2016-10-12 | 浙江理工大学 | Multi-feature-fusion-based technical method for rapid detection of pedestrian |
| CN106649658A (en) * | 2016-12-13 | 2017-05-10 | 重庆邮电大学 | Recommendation system and method for improving user role undifferentiated treatment and data sparseness |
| CN107507054A (en) * | 2017-07-24 | 2017-12-22 | 哈尔滨工程大学 | A kind of proposed algorithm based on Recognition with Recurrent Neural Network |
| CN107423442A (en) * | 2017-08-07 | 2017-12-01 | 火烈鸟网络(广州)股份有限公司 | Method and system, storage medium and computer equipment are recommended in application based on user's portrait behavioural analysis |
Cited By (22)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109522742A (en) * | 2018-10-26 | 2019-03-26 | 贵州斯曼特信息技术开发有限责任公司 | A kind of batch processing method of computer big data |
| CN109525665A (en) * | 2018-11-16 | 2019-03-26 | 济南浪潮高新科技投资发展有限公司 | A kind of wound visitor's cloud center configuration recommended method based on crowdsourcing |
| CN110020228A (en) * | 2019-04-08 | 2019-07-16 | 浙江大学城市学院 | A kind of relevance evaluation method for Internet of Things Item Information searching order |
| CN111797874B (en) * | 2019-04-09 | 2024-04-09 | Oppo广东移动通信有限公司 | Behavior prediction method and device, storage medium and electronic equipment |
| CN111797874A (en) * | 2019-04-09 | 2020-10-20 | Oppo广东移动通信有限公司 | Behavior prediction method, behavior prediction device, storage medium and electronic equipment |
| US11983609B2 (en) | 2019-07-10 | 2024-05-14 | Sony Interactive Entertainment LLC | Dual machine learning pipelines for transforming data and optimizing data transformation |
| TWI755778B (en) * | 2019-07-15 | 2022-02-21 | 美商索尼互動娛樂有限責任公司 | Self-healing machine learning system for transformed data |
| US11250322B2 (en) | 2019-07-15 | 2022-02-15 | Sony Interactive Entertainment LLC | Self-healing machine learning system for transformed data |
| CN110362780B (en) * | 2019-07-17 | 2021-03-23 | 北京航空航天大学 | Large data tensor canonical decomposition calculation method based on Shenwei many-core processor |
| CN110362780A (en) * | 2019-07-17 | 2019-10-22 | 北京航空航天大学 | A kind of big data tensor canonical decomposition calculation method based on Shen prestige many-core processor |
| CN112417310A (en) * | 2019-08-21 | 2021-02-26 | 上海掌门科技有限公司 | Method for establishing intelligent service index and recommending intelligent service |
| CN112417310B (en) * | 2019-08-21 | 2023-10-03 | 上海掌门科技有限公司 | Method for establishing intelligent service index and recommending intelligent service |
| CN111476202A (en) * | 2020-04-30 | 2020-07-31 | 杨九妹 | User behavior analysis method and system of financial institution security system and robot |
| CN112541407A (en) * | 2020-08-20 | 2021-03-23 | 同济大学 | Visual service recommendation method based on user service operation flow |
| CN112182395B (en) * | 2020-10-10 | 2023-08-29 | 深圳市万佳安物联科技股份有限公司 | Financial service personalized recommendation device and method based on time sequence |
| CN112182395A (en) * | 2020-10-10 | 2021-01-05 | 深圳市万佳安物联科技股份有限公司 | Financial service personalized recommendation device and method based on time sequence |
| CN112307352A (en) * | 2020-11-26 | 2021-02-02 | 腾讯科技（深圳）有限公司 | Content recommendation method, system, device and storage medium |
| CN112307352B (en) * | 2020-11-26 | 2023-05-26 | 腾讯科技（深圳）有限公司 | Content recommendation method, system, device and storage medium |
| WO2022121705A1 (en) * | 2020-12-10 | 2022-06-16 | 株式会社日立制作所 | Information processing method, apparatus and device |
| CN113095084A (en) * | 2021-03-16 | 2021-07-09 | 重庆邮电大学 | Semantic service matching method and device in Internet of things and storage medium |
| CN116485587A (en) * | 2023-04-21 | 2023-07-25 | 深圳润高智慧产业有限公司 | Community service acquisition method, community service providing method, electronic device and storage medium |
| CN116485587B (en) * | 2023-04-21 | 2024-04-09 | 深圳润高智慧产业有限公司 | Community service acquisition method, community service providing method, electronic device and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108460489A (en) | A kind of user behavior analysis based on big data technology and service recommendation frame | |
| Nabil et al. | Prediction of students’ academic performance based on courses’ grades using deep neural networks | |
| CN108334605B (en) | Text classification method and device, computer equipment and storage medium | |
| CN110532996B (en) | Video classification method, information processing method and server | |
| CN109993102B (en) | Similar face retrieval method, device and storage medium | |
| CN110866140A (en) | Image feature extraction model training method, image searching method and computer equipment | |
| CN104899561A (en) | Parallelized human body behavior identification method | |
| CN108765383B (en) | Video description method based on deep migration learning | |
| CN110659678B (en) | User behavior classification method, system and storage medium | |
| CN113139141B (en) | User tag expansion labeling method, device, equipment and storage medium | |
| CN115797606B (en) | 3D virtual digital human interaction action generation method and system based on deep learning | |
| CN110399895A (en) | The method and apparatus of image recognition | |
| Gaier et al. | Aerodynamic design exploration through surrogate-assisted illumination | |
| CN113204988A (en) | Small sample viewpoint estimation | |
| CN110110663A (en) | A kind of age recognition methods and system based on face character | |
| KR20200010672A (en) | Smart merchandise searching method and system using deep learning | |
| CN113128478A (en) | Model training method, pedestrian analysis method, device, equipment and storage medium | |
| CN112015902A (en) | Least-order text classification method under metric-based meta-learning framework | |
| CN115526246A (en) | Self-supervision molecular classification method based on deep learning model | |
| Cai et al. | Training deep convolution network with synthetic data for architectural morphological prototype classification | |
| CN112115744A (en) | Point cloud data processing method and device, computer storage medium and electronic equipment | |
| Sun et al. | Improving and evaluating deep learning models of cellular organization | |
| CN109784404A (en) | A kind of the multi-tag classification prototype system and method for fusion tag information | |
| KR20080097753A (en) | Method for region based on image retrieval using multi-class support vector machine | |
| CN115937516B (en) | Image semantic segmentation method and device, storage medium and terminal |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20180828 |
|
| WD01 | Invention patent application deemed withdrawn after publication |