CN107451832A - The method and apparatus of pushed information - Google Patents
The method and apparatus of pushed information Download PDFInfo
- Publication number
- CN107451832A CN107451832A CN201610371042.5A CN201610371042A CN107451832A CN 107451832 A CN107451832 A CN 107451832A CN 201610371042 A CN201610371042 A CN 201610371042A CN 107451832 A CN107451832 A CN 107451832A
- Authority
- CN
- China
- Prior art keywords
- data
- user
- characteristic
- dimension
- cycle
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- 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
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0251—Targeted advertisements
-
- 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
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0251—Targeted advertisements
- G06Q30/0253—During e-commerce, i.e. online transactions
-
- 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
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0251—Targeted advertisements
- G06Q30/0255—Targeted advertisements based on user history
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Business, Economics & Management (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Engineering & Computer Science (AREA)
- Development Economics (AREA)
- Strategic Management (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Economics (AREA)
- Marketing (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
This application discloses the method and apparatus of pushed information.One embodiment of methods described includes:Based on the preferential dimension in preferential dimension data storehouse, the characteristic of the various dimensions of user in predetermined amount of time is obtained;The characteristic of various dimensions is divided to multiple cycles;Obtain the user activity value in each cycle;Within the predetermined cycle, the Euclidean distance between the user activity value in cycle two-by-two is calculated;According to Euclidean distance, cluster user population data;According to user group's data, to user terminal pushed information.This embodiment improves the degree of accuracy of the utilization rate to user data and cluster user colony, so that the information pushed to user has more specific aim.
Description
Technical field
The application is related to field of computer technology, and in particular to Internet technical field, especially relates to
And the method and apparatus of generation pushed information.
Background technology
The method of pushed information, refer to be based on objective data to user's pushed information.At present more
The method of common pushed information, it is to be drawn a portrait based on user to user's pushed information.Here use
Family is drawn a portrait, and is the virtual representations of real user, is built upon a series of True Data (order numbers
According to useful data etc.) on targeted customer's model.User, which draws a portrait, includes the basis of user
Data (such as age, sex), interest (such as books, dress ornament) and consumption habit are (when such as
Between, quantity and the amount of money) etc., therefore, businessman can draw a portrait to provide according to user has more specific aim
Service.
The method of current pushed information, first by carrying out artificial point to existing user data
Analysis, the rule of user group is determined, and then build the rule of user's portrait model, then manually
The modeling of user's portrait is carried out, after generation model, regeneration performs sentence, finally in big data
Platform performs and obtains user group's data.
However, the method for current pushed information by manual analysis, it is necessary to determine user group
Rule and user base data utilization rate it is relatively low, therefore obtain user group's data efficiency
Relatively low and to user's pushed information the degree of accuracy is relatively low.
The content of the invention
The purpose of the application is to propose a kind of method and apparatus of improved pushed information, to solve
The technical problem that certainly background section above is mentioned.
In a first aspect, this application provides a kind of method of pushed information, methods described includes:
Based on the preferential dimension in preferential dimension data storehouse, the various dimensions of user in predetermined amount of time are obtained
Characteristic;The characteristic of the various dimensions is divided to multiple cycles;Obtain each week
The user activity value of phase;Within the predetermined cycle, the user activity value in cycle two-by-two is calculated
Between Euclidean distance;According to the Euclidean distance, cluster user population data;According to described
User group's data, to user terminal pushed information.
Second aspect, this application provides a kind of device of pushed information, described device includes:
Described device includes:First acquisition unit, for based on the preferential dimension in preferential dimension data storehouse
Degree, obtain the characteristic of the various dimensions of user in predetermined amount of time;Division unit, for inciting somebody to action
The characteristic of the various dimensions is divided to multiple cycles;Second acquisition unit, it is every for obtaining
The user activity value in individual cycle;Computing unit, within the predetermined cycle, calculating two-by-two
Euclidean distance between the user activity value in cycle;Cluster cell, for according to described European
Distance, cluster user population data;Push unit, for according to user group's data,
To user terminal pushed information.
The method and apparatus for the pushed information that the application provides, by based on preferential dimension data storehouse
In preferential dimension, obtain predetermined amount of time in user various dimensions characteristic, then will
The characteristic of various dimensions is divided to multiple cycles, obtains the user activity in each cycle afterwards
Value, afterwards within the predetermined cycle, calculate European between the user activity value in cycle two-by-two
Distance, afterwards according to Euclidean distance, cluster user population data, afterwards according to user group's number
According to user terminal pushed information, so as to which the rule of user group be determined by computer, carrying
It is high to determine the efficiency of user group, and improve the utilization rate of user base data so that
The information pushed to user has more specific aim.
Brief description of the drawings
Retouched by reading with reference to the detailed of being made to non-limiting example of being made of the following drawings
State, other features, objects and advantages will become more apparent upon:
Fig. 1 is that the application can apply to exemplary system architecture figure therein;
Fig. 2 is the flow chart according to one embodiment of the method for the pushed information of the application;
Fig. 3 is the flow chart according to another embodiment of the method for the pushed information of the application;
Fig. 4 is the schematic diagram according to an application scenarios of the method for the pushed information of the application;
Fig. 5 is the structural representation according to one embodiment of the device of the pushed information of the application
Figure;
Fig. 6 is the structural representation according to another embodiment of the device of the pushed information of the application
Figure;
Fig. 7 is adapted for for realizing the terminal device of the embodiment of the present application or the computer of server
The structural representation of system.
Embodiment
The application is described in further detail with reference to the accompanying drawings and examples.It is appreciated that
, specific embodiment described herein is used only for explaining related invention, rather than to the hair
Bright restriction.It also should be noted that for the ease of description, illustrate only in accompanying drawing with
About the related part of invention.
It should be noted that in the case where not conflicting, embodiment and embodiment in the application
In feature can be mutually combined.Describe this in detail below with reference to the accompanying drawings and in conjunction with the embodiments
Application.
Fig. 1 shows the method for pushed information or the device of pushed information that can apply the application
Embodiment exemplary system architecture 100.
As shown in figure 1, system architecture 100 can include terminal device 101,102,103,
Network 104 and server 105,106.Network 104 is in terminal device 101,102,103
The medium of communication link is provided between server 105,106.Network 104 can include various
Connection type, such as wired, wireless communication link or fiber optic cables etc..
User 110 can pass through network 104 and service with using terminal equipment 101,102,103
Device 105,106 interacts, to receive or send message etc..Terminal device 101,102,103
On various telecommunication customer end applications can be installed, such as web browser applications, shopping class should
With, searching class application, JICQ, mailbox client, social platform software etc..
Terminal device 101,102,103 can have display screen and support man-machine interaction
Various electronic equipments, including but not limited to smart mobile phone, tablet personal computer, MP3 player (Moving
Picture Experts Group Audio Layer III, dynamic image expert's compression standard audio layer
Face 3), (Moving Picture Experts Group Audio Layer IV, dynamic image are special by MP4
Family's compression standard audio aspect 4) player, pocket computer on knee and desktop computer etc..
Server 105,106 can be to provide the server of various services, such as to terminal device
101st, the information shown on 102,103 provides the background server supported.Background server can
Carry out the processing such as analyzing with the characteristic of the various dimensions to the user received, and processing is tied
Fruit (such as pushed information) feeds back to terminal device.
It should be noted that the method for the pushed information that the embodiment of the present application is provided is typically by taking
Device 105,106 of being engaged in performs, correspondingly, the device of pushed information be generally positioned at server 105,
In 106.
It should be understood that the number of the terminal device, network and server in Fig. 1 is only to illustrate
Property.According to needs are realized, can have any number of terminal device, network and server.
With continued reference to Fig. 2, show that one of the method for the pushed information according to the application is implemented
The flow 200 of example.The method of described pushed information, comprises the following steps:
Step 201, based on the preferential dimension in preferential dimension data storehouse, predetermined amount of time is obtained
The characteristic of the various dimensions of interior user.
In the present embodiment, the electronic equipment of the method operation of pushed information thereon (such as is schemed
Server shown in 1) can be whole to user by wired connection mode or radio connection
Hold pushed information.It is pointed out that above-mentioned radio connection can include but is not limited to
3G/4G connections, WiFi connections, bluetooth connection, WiMAX connections, Zigbee connections, UWB
(ultra wideband) is connected and other currently known or exploitation in the future wireless connection sides
Formula.Generally, user is applied using web browser applications, the shopping class installed in terminal, searched
The application of rope class, JICQ etc. browse webpage or shopping.
It should be appreciated that in above-mentioned preferential dimension data storehouse, technological development personnel can be included
The quantity of default preferential dimension and preferential dimension.For example, technological development personnel can preset it is excellent
First dimension is order data and shopping cart data the two dimensions, then based on preferential dimension data
Preferential dimension in storehouse, can when obtaining the characteristic of the various dimensions of user in predetermined amount of time
Preferentially to obtain the order data of user and shopping cart data.
Here the characteristic of various dimensions, can include multinomial in following dimension:Order numbers
According to, shopping cart data, collection data, advisory data, comment data, focused data, forwarding
Data, record data is browsed, record data is searched for, delivers data, gender data, age number
According to, income data, occupation data, psychological characteristics data, values data, consumer behavior it is inclined
Good data, attitude data and custom data.For example, order data can be chosen and browse data
The two dimensions obtain each user the two dimensions as the dimension for preparing to obtain afterwards
Characteristic.
Here predetermined amount of time refers to that basis is actually needed time dimension set in advance.For example,
The data before choosing pushed information in 1 year can be set as needed as in predetermined amount of time
The characteristic of the various dimensions of user.
Step 202, the characteristic of various dimensions is divided to multiple cycles.
In the present embodiment, the characteristic based on the various dimensions obtained in step 201, can be with
The characteristic of various dimensions is divided to multiple cycles as required.Used for example, if desired obtaining
Characteristic of the family in 10 days, then it can be 10 by each cycle set, so that will be upper
The characteristic for stating the various dimensions in 1 year of acquisition is divided to 37 cycles.
Step 203, the user activity value in each cycle is obtained.
In the present embodiment, can be to each when obtaining the user activity value in each cycle
The characteristic of the various dimensions in cycle is analyzed and processed, and according to default user activity meter
Calculate rule and the characteristic of the various dimensions in each cycle is converted into user activity value.Here
User activity rule can be the user activity in the prior art or in the technology of future development
Computation rule, the application are not construed as limiting to this.
In the optional implementation of the present embodiment, the user activity value for obtaining each cycle can
With including:The weighted average of the characteristic of various dimensions is obtained in each cycle;Normalization
Weighted average, the characteristic after being normalized;Characteristic after fitting normalization,
Obtain user activity curve;According to user activity curve, the user for obtaining each cycle lives
Jerk value.
In this implementation, the characteristic of corresponding different dimensions sets weight in advance, this
In weight refer to relatively important journey of the characteristic of the dimension in the characteristic of various dimensions
Degree, therefore can be according to the weighted average of the characteristic of Weight Acquisition various dimensions in each cycle
Value;Weighted average is normalized afterwards so that the absolute value of weighted average turns
The relative value relation being changed in same section, so as to the characteristic after being normalized;Afterwards
Characteristic after fitting normalization, obtains user activity curve, fitting here refers to
Know some discrete function values { f1, f2 ..., fn } of certain function, it is some in the function by adjusting
Undetermined coefficient f (λ 1, λ 2 ..., λ n) so that the difference of the function and known point set (such as
The difference of least square meaning) it is minimum;Finally according to user activity curve, each cycle is obtained
User activity value, it will be appreciated that user activity value now is corresponding with each cycle
User activity curve on value.
It should be appreciated that approximating method here can use prior art or the technology of future development
In approximating method, the application is not construed as limiting to this.Exemplary, the spy after fitting normalization
Data are levied, obtaining user activity curve can include:Characteristic after normalization is inputted
Support vector regression model (SVR) is fitted, so as to obtain user activity curve.
Using support vector regression models fitting normalize after characteristic when, above-mentioned pushed information
Method can also include:Using the characteristic after normalization, support vector regression mould is trained
Type.Here the method for training support vector regression model can use in the prior art or future
Existing training method, will not be repeated here in the technology of development.
Step 204, within the predetermined cycle, calculating is two-by-two between the user activity value in cycle
Euclidean distance.
In the present embodiment, Euclidean distance namely euclidean metric (euclidean metric),
It is the distance definition of a generally use, refers to the actual distance between two points in m-dimensional space,
Or the natural length (i.e. the distance of the point to origin) of vector.In the two-dimensional space of the application
Euclidean distance be exactly actual range between 2 points, therefore, within the predetermined cycle, namely
Within according to the cycle set in advance paid close attention to, the user activity value in cycle two-by-two is calculated
Between Euclidean distance, namely calculate the actual range between the user activity value in cycle two-by-two.
Step 205, according to Euclidean distance, cluster user population data.
In the present embodiment, can basis on the basis of the Euclidean distance that step 204 obtains
Euclidean distance, user group's data similar in Euclidean distance are clustered, namely obtained in weight
The similar user group of the variation tendency of user activity value in the cycle of point concern.Here poly-
Class, refer to the process of for user group's data to be divided into the multiple classes being made up of similar Euclidean distance.
Step 206, according to user group's data, to user terminal pushed information.
In the present embodiment, based on the user group's data obtained in step 205, can to
The similar user group's pushed information of the variation tendency of angle value is enlivened in the cycle paid close attention to.
The method for the pushed information that above-described embodiment of the application provides, can be to determining user
User's pushed information of colony's rule, improve the utilization rate to user data and cluster user group
The degree of accuracy of body, so that the information pushed to user has more specific aim.
With further reference to Fig. 3, the method for pushed information according to the application another is shown
The flow 300 of embodiment.The method of described pushed information, comprises the following steps:
In step 301, based on the preferential dimension in preferential dimension data storehouse, pre- timing is obtained
Between in section the various dimensions of user characteristic, perform step 302 afterwards;
In step 302, the characteristic of various dimensions is divided to multiple cycles, performed afterwards
Step 303;
In step 303, the user activity value in each cycle is obtained, performs step 304 afterwards;
In step 304, within the predetermined cycle, the user activity value in cycle two-by-two is calculated
Between Euclidean distance, afterwards perform step 305;
In step 305, according to Euclidean distance, cluster user population data, step is performed afterwards
Rapid 306;
Within step 306, detect what is be consistent in user group's data with historic user order data
User's individual data items, step 307 is performed afterwards;
In step 307, obtain user's individual data items acquired in characteristic dimension and
The quantity of the dimension of characteristic acquired in each, step 308 is performed afterwards;
In step 308, according to quantity from big to small, the feature acquired in predetermined number is obtained
The dimension of data, step 309 is performed afterwards;
In a step 309, using the dimension of the characteristic acquired in predetermined number as preferential dimension
Degree, updates preferential dimension data storehouse, performs step 301 afterwards.
It should be appreciated that step 201 in above-mentioned steps 301 to step 305 and above-mentioned Fig. 2 to
Step 205 corresponds, therefore, the operation of the description in step 201 to step 205 and step
Suddenly step 301 is equally applicable to step 305, will not be repeated here.
In step 306 to step 309, user group's data are the characteristic according to various dimensions
According to the cluster data obtained after analyzing and processing, and historic user order data is what is truly occurred
The data of order, according to the order data truly occurred, it can detect in user group's data
The characteristic of the user correctly bought, obtain the dimension of the characteristic of the user correctly bought
(namely dimension of the characteristic acquired in above-mentioned user's individual data items).Specifically, can be with
The class of user group's data corresponding to historic user order data is found by way of recursive lookup
Not, after lookup, the dimension for the characteristic that user itself is included is counted side by side
Sequence, the dimension of the characteristic of predetermined number is found out, by the dimension of the characteristic of predetermined number
, should if including in preferential dimension data storehouse as preferential dimensions updating to preferential dimension data storehouse
Dimension, then the renewal step is skipped, will if not including the dimension in preferential dimension data storehouse
The dimensions updating performs step 301 afterwards to preferential dimension data storehouse.
The method for the pushed information that above-described embodiment of the application provides, can obtain the scheduled time
The characteristic of the preferential dimension of user is analyzed as user base data in section, is improved
To the utilization rate of user base data, and efficiency and the degree of accuracy of cluster user colony are improved,
So that the information pushed to user has more specific aim.
With continued reference to Fig. 4, Fig. 4 is the applied field according to the method for the pushed information of the present embodiment
One indicative flowchart of scape.
As shown in figure 4, in the application scenarios of the method for the pushed information, following four are devised
Individual module realizes the method for pushed information:Data extraction module 410, category of model module 420,
Intensified learning module 430 and pushed information module 440.
Wherein, the major function of data extraction module 410 was tieed up in big data platform extraction time
The total data of degree, and after elapsed time periodic transformation and the normalization of data magnitude change, make
The mixed and disorderly data of user are obtained, become to carry out calculating unified data format using model.
Data extraction module can obtain unified data format by following steps:
First, in step 411, the characteristic of various dimensions is extracted.
, it is necessary to extract the Back ground Information of user before model is established, first in big data platform
According to History Order data, the historical viewings data of user in time format extraction predetermined amount of time
Characteristic Deng various dimensions (pays attention to:It is to choose by rule of thumb when the dimension of characteristic most starts
(such as order, browse), it is excellent when the preferential dimension data storehouse in intensified learning module 430
It after first dimension is continuously increased, will be called automatically by characteristic extracting module, ensure preferential dimension quilt
It is preferential to choose), it is assumed that the characteristic of the various dimensions within being extracted 1 year by following formatted data:
Afterwards, in step 412, the time cycle of the characteristic of transforming multidimensional degree.
After it have chosen the characteristic for the chronological various dimensions specified, it is necessary to according to
The different cycles is converted using the time cycle, and it also requires each user's different dimensions (are ordered
Forms data, browse data) characteristic be all put into a unified time life cycle (with
10 days is exemplified by a cycles) it is inner.Form after change is:
testname 0.0 0.0 4 6 6 0.0 0.0 4 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
0.0 0.0
Explanation:Testname therein is test user's name used, here
4=(4 × 0.9+0.1 × 4), the sum for referring to order numbers in one of them cycle is 4, and it is total to browse number
Number is 4, and the weight of order is 0.9, and the weight browsed is 0.1, so as to obtain order data,
The weighted average for browsing the characteristic of two dimensions of data is 4, similarly, in other cycles
Numerical value be also order data, browse two dimensions of data characteristic weighted average.
Afterwards, in step 413, the characteristic after the conversion time cycle is normalized.
After the characteristic after obtaining the conversion time cycle, the data between different users,
It there may be the difference of the order of magnitude, it is assumed that party A-subscriber, in one month, order numbers are 10
Effectively single, number of visits 1 time, party B-subscriber, in one month, order numbers are 100 effective
Order, number of visits 10, if at this moment directly carrying out periodic transformation, party B-subscriber finally calculates
Active value out is certainly higher than party A-subscriber, but is so inaccurate, it is necessary to enter line number in fact
Change on, the data of party A-subscriber and the data of party B-subscriber are all divided into same section,
Such as between [0,4], it is assumed in the following that the data taken at random in data pool:
testname 0.0 0.0 0.5 1.0 1.0 0.0 0.0 0.5 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
0.0 0.0
This is arrived, user's portrait of a time cycle just is so depicted, and is with one group
The data of unified form represent.
The major function of category of model module 420 is the data inputted by data extraction module,
Model is established by SVR algorithms, and calculates the user based on time life cycle and enlivens
Degree, and come cluster user colony according to user activity.
Category of model module 420 passes through following steps cluster user colony:
Obtaining after the normalized data of data extraction module input, in step 421,
Begin setting up SVR models and matched curve, Zhi Hou are generated according to the SVR models established
In step 422, value of each user in the liveness in each cycle is calculated, afterwards in step 423
In, according to each user in the value of the liveness in each cycle, cluster user colony.
Currently employed method is:Train obtain cycle data, come calculate SVR return (and
Be worth contrast) afterwards with SVR prediction (unknown data contrast) periodic quantity, for example, testname
User it is present have 37 periodic quantities of 15 days as follows:
Testname 0.00.0 0.5 1.0 1.0 0.0 0.0 0.5 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
0.0 0.0
Earliest extraction time is 2014-1-1, and the time the latest is 2015-1-1, then we
It is right after being calculated by SVR when wondering the value of January 1 to diurnal periodicity in January 14 in 2015
The value answered is:
Wherein, 0.0996304954031 is active user's state in which, 0.156469039467
May state in which for the next cycle.
And at present after being counted to all data, the frequency occurred to data is ranked up,
Frequency is taken 5 sections of highest occur, so active period division is as follows:
Less than 0.11:Without active
0.11 to 0.5:It is low active
0.5 to 1.5:Normal active
1.5 to 2.5:Height is active
More than 2.5:Superelevation is enlivened
This is arrived, obtain all users enlivens angle value.
Next, starting to be included in user group's classification submodule, the input of this module is as follows:
Citing:
Then according to the cycle of delimitation, corresponding time series, such as last 4 15 are found
It cycle:
0.0999411429443 0.100038003852 0.0996304954031
0.156469039467, difference between any two is calculated, obtains three differences, similarly, is calculated
The value in all user's last 4 cycles, finally calculates the Euclidean distance between them two-by-two again,
Stop until calculating, export user group.
The major function of intensified learning module 430 is classified by input model sort module
User group's data are carried out with the real user purchase order data in the true marketing cycle are inputted
Evaluation is compared, finds out the dimension of the characteristic of the user correctly bought, statistics is all these just
The dimension of the characteristic of true user, is passed to independent preferential dimension data storehouse, and carry in data
When modulus block is using the characteristic for extracting various dimensions, there is provided preferential dimension is to characteristic extracting module
Use.
Intensified learning module 430 generates preferential dimension data storehouse by following steps:
In step S431, the effect of the user group of cluster is assessed, uses the side of recursive lookup
Formula finds the classification where the user group of corresponding real user purchase order data, afterwards,
In step S432 after lookup, the dimension for the characteristic that this certain customers is included itself
Degree carries out quantity statistics and is ranked up, and finds out the dimension of first three data feature, afterwards,
In step S433, by these three dimensions updatings into preferential dimension data storehouse, if preferential dimension
There is this dimension in database, then skipped, do not updated.So that next time, modeling could
Automatically complete, rather than the analysis and modeling of artificial carry out data.
The major function of pushed information module 440 is gathered according to what category of model module 420 exported
Class user group is to user's pushed information.
Pushed information module 440 passes through step S441 propelling user informations.In step S441,
The user group's inquired about using the mode of inquiry in the information pushed to user after meeting cluster
The information of classification, rear line push meet cluster after user group classification information.
In the application scenarios of above-described embodiment of the application, the method for pushed information is by poly-
User group's pushed information after class, and obtain the spy of the preferential dimension of user in predetermined amount of time
Sign data are analyzed as user base data, are improved to the utilization rate of user data and are gathered
The degree of accuracy of class user group, so that the information pushed to user has more specific aim.
With further reference to Fig. 5, as the realization to method shown in above-mentioned each figure, the application provides
A kind of one embodiment of the device of pushed information, the device embodiment and the side shown in Fig. 2
Method embodiment is corresponding, and the device specifically can apply in various electronic equipments.
As shown in figure 5, the device 500 of the pushed information described in the present embodiment includes:First obtains
Take unit 510, division unit 520, second acquisition unit 530, computing unit 540, cluster
Unit 550 and push unit 560.
First acquisition unit 510, for based on the preferential dimension in preferential dimension data storehouse, obtaining
Take the characteristic of the various dimensions of user in predetermined amount of time.
Division unit 520, for the characteristic of various dimensions to be divided into multiple cycles.
Second acquisition unit 530, for obtaining the user activity value in each cycle.
Computing unit 540, within the predetermined cycle, the user for calculating the cycle two-by-two to enliven
Euclidean distance between angle value.
Cluster cell 550, for according to Euclidean distance, cluster user population data.
Push unit 560, for according to user group's data, to user terminal pushed information.
It will be understood by those skilled in the art that the device 500 of above-mentioned pushed information also includes
Other known features, such as processor, memory etc., in order to unnecessarily obscure the disclosure
Embodiment, these known structures are not shown in Figure 5.
It should be appreciated that in all units described in device 500 and the method with reference to the description of figure 2
Each step is corresponding.Thus, the operation described above with respect to the method for pushed information and feature
Device 500 and the unit wherein included are equally applicable to, will not be repeated here.In device 500
Corresponding units can be cooperated with the unit in terminal device and/or server to realize this
Apply for the scheme of embodiment.
With further reference to Fig. 6, as the realization to method shown in above-mentioned each figure, the application provides
A kind of one embodiment of the device of pushed information, the device embodiment and the side shown in Fig. 2
Method embodiment is corresponding, and the device specifically can apply in various electronic equipments.
As shown in fig. 6, the device 600 of the pushed information described in the present embodiment includes:First obtains
Take unit 610, division unit 620, second acquisition unit 630, computing unit 640, cluster
Unit 660 and push unit 660.
First acquisition unit 610, for based on the preferential dimension in preferential dimension data storehouse, obtaining
Take the characteristic of the various dimensions of user in predetermined amount of time.
Division unit 620, for the characteristic of various dimensions to be divided into multiple cycles.
Second acquisition unit 630, for obtaining the user activity value in each cycle.
Computing unit 640, within the predetermined cycle, the user for calculating the cycle two-by-two to enliven
Euclidean distance between angle value.
Cluster cell 650, for according to Euclidean distance, cluster user population data.
Detection unit 660, for detecting the History Order data in user group's data with user
The user's individual data items being consistent.
3rd acquiring unit 670, for obtaining the characteristic acquired in user's individual data items
The quantity of the dimension of dimension and the characteristic acquired in each.
4th acquiring unit 680, obtained for from big to small, obtaining predetermined number according to quantity
The dimension of the characteristic taken.
Setup unit 690, for using the dimension of the characteristic acquired in predetermined number as excellent
First dimension, update preferential dimension data storehouse.
It will be understood by those skilled in the art that the device 600 of above-mentioned pushed information also includes
Other known features, such as processor, memory etc., in order to unnecessarily obscure the disclosure
Embodiment, these known structures are not shown in figure 6.
It should be appreciated that in all units described in device 600 and the method with reference to the description of figure 3
Each step is corresponding.Thus, the operation described above with respect to the method for pushed information and feature
Device 600 and the unit wherein included are equally applicable to, will not be repeated here.In device 600
Corresponding units can be cooperated with the unit in terminal device and/or server to realize this
Apply for the scheme of embodiment.
In above-described embodiment of the application, first acquisition unit, second acquisition unit, the 3rd
Acquiring unit and the 4th acquiring unit, which only represent, obtains four different units of object.Ability
Field technique personnel should be appreciated that therein first, second, third or the 4th not form to obtaining
Take the particular determination of unit.
Similarly, the first acquisition subelement, the second acquisition subelement, which only represent, obtains each not phase of object
Two same subelements.It will be appreciated by those skilled in the art that therein first or second not
Form the particular determination to obtaining subelement.
Below with reference to Fig. 7, it illustrates suitable for for realizing the terminal device of the embodiment of the present application
Or the structural representation of the computer system 700 of server.
As shown in fig. 7, computer system 700 includes CPU (CPU) 701, its
Can according to the program being stored in read-only storage (ROM) 702 or from storage part 708
The program that is loaded into random access storage device (RAM) 703 and perform various appropriate actions
And processing.In RAM 703, also it is stored with system 700 and operates required various program sums
According to.CPU 701, ROM 702 and RAM 703 are connected with each other by bus 704.Input
/ output (I/O) interface 705 is also connected to bus 704.
I/O interfaces 705 are connected to lower component:Importation 706 including keyboard, mouse etc.;
Including cathode-ray tube (CRT), liquid crystal display (LCD) etc. and loudspeaker etc.
Output par, c 707;Storage part 708 including hard disk etc.;And including such as LAN card,
The communications portion 709 of the NIC of modem etc..Communications portion 709 is via such as
The network of internet performs communication process.Driver 710 is also according to needing to be connected to I/O interfaces
705.Detachable media 711, such as disk, CD, magneto-optic disk, semiconductor memory etc.,
Be arranged on as needed on driver 710, in order to the computer program that reads from it according to
Need to be mounted into storage part 708.
Especially, in accordance with an embodiment of the present disclosure, can be with above with reference to the process of flow chart description
It is implemented as computer software programs.For example, embodiment of the disclosure includes a kind of computer journey
Sequence product, it includes being tangibly embodied in the computer program on machine readable media, the meter
Calculation machine program bag contains the program code for being used for the method shown in execution flow chart.In such implementation
In example, the computer program can be downloaded and installed by communications portion 709 from network,
And/or it is mounted from detachable media 711.
Flow chart and block diagram in accompanying drawing, it is illustrated that according to the system of the various embodiments of the application,
Architectural framework in the cards, function and the operation of method and computer program product.This point
On, each square frame in flow chart or block diagram can represent a unit, program segment or code
A part, a part for the unit, program segment or code is used for comprising one or more
The executable instruction of logic function as defined in realization.It should also be noted that at some as replacement
In realization, the function of being marked in square frame can also be with different from the order marked in accompanying drawing hair
It is raw.For example, two square frames succeedingly represented can essentially perform substantially in parallel, they
Sometimes can also perform in the opposite order, this is depending on involved function.It is also noted that
It is, each square frame and block diagram in block diagram and/or flow chart and/or the square frame in flow chart
Combination, function or the special hardware based system of operation it can be realized as defined in execution,
Or it can be realized with the combination of specialized hardware and computer instruction.
Being described in unit involved in the embodiment of the present application can be real by way of software
It is existing, it can also be realized by way of hardware.Described unit can also be arranged on processing
In device, for example, can be described as:A kind of processor include first acquisition unit, division unit,
Second acquisition unit, computing unit, cluster cell and push unit.Wherein, these units
Title does not form the restriction to the unit in itself under certain conditions, for example, semantic receive list
Member is also described as " based on the preferential dimension in preferential dimension data storehouse, obtaining pre- timing
Between in section the characteristic of the various dimensions of user unit ".
As on the other hand, present invention also provides a kind of nonvolatile computer storage media,
The nonvolatile computer storage media can be described in above-described embodiment included in device
Nonvolatile computer storage media;Can also be individualism, without non-in supplying terminal
Volatile computer storage medium.Above-mentioned nonvolatile computer storage media be stored with one or
The multiple programs of person, when one or more of programs are performed by an equipment so that described
Equipment:Based on the preferential dimension in preferential dimension data storehouse, user in predetermined amount of time is obtained
The characteristic of various dimensions;The characteristic of various dimensions is divided to multiple cycles;Obtain each
The user activity value in cycle;Within the predetermined cycle, the user activity in cycle two-by-two is calculated
Euclidean distance between value;According to Euclidean distance, cluster user population data;According to customer group
Volume data, to user terminal pushed information.
Above description is only the preferred embodiment of the application and saying to institute's application technology principle
It is bright.It will be appreciated by those skilled in the art that invention scope involved in the application, and it is unlimited
In the technical scheme that the particular combination of above-mentioned technical characteristic forms, while it should also cover and not depart from
In the case of the inventive concept, it is combined by above-mentioned technical characteristic or its equivalent feature
And the other technical schemes formed.Such as features described above and (but not limited to) disclosed herein
The technical scheme that technical characteristic with similar functions is replaced mutually and formed.
Claims (11)
- A kind of 1. method of pushed information, it is characterised in that methods described includes:Based on the preferential dimension in preferential dimension data storehouse, user's is more in acquisition predetermined amount of time The characteristic of dimension;The characteristic of the various dimensions is divided to multiple cycles;Obtain the user activity value in each cycle;Within the predetermined cycle, the Euclidean distance between the user activity value in cycle two-by-two is calculated;According to the Euclidean distance, cluster user population data;According to user group's data, to user terminal pushed information.
- 2. the method for pushed information according to claim 1, it is characterised in that described to obtain Taking the user activity value in each cycle includes:The weighted average of the characteristic of various dimensions is obtained in each cycle;The weighted average is normalized, the characteristic after being normalized;The characteristic being fitted after the normalization, obtain user activity curve;According to the user activity curve, the user activity value in each cycle is obtained.
- 3. the method for the pushed information according to any one of claim 1 or 2, its feature It is, the characteristic after the fitting normalization, obtaining user activity curve includes:Characteristic input support vector regression model after the normalization is fitted, obtained To user activity curve.
- 4. the method for pushed information according to claim 3, it is characterised in that described to obtain Taking the user activity value in each cycle also includes:Using the characteristic after the normalization, the support vector regression model is trained.
- 5. the method for the pushed information according to any one of claim 1 or 2, its feature It is, methods described also includes:Detect the user's individual being consistent in user group's data with the History Order data of user Data;Obtain the characteristic acquired in user's individual data items dimension and each obtained The quantity of the dimension of the characteristic taken;According to the quantity from big to small, the dimension of the characteristic acquired in predetermined number is obtained;Using the dimension of the characteristic acquired in the predetermined number as preferential dimension, institute is updated State preferential dimension data storehouse.
- 6. the method for pushed information according to claim 1, it is characterised in that the spy Data are levied including multinomial in following dimension:Order data, shopping cart data, collection data, Advisory data, comment data, focused data, forward data, browse record data, search note Record data, deliver data, gender data, age data, income data, occupation data, the heart Manage characteristic, values data, consumer behavior preference data, attitude data and custom data.
- 7. a kind of device of pushed information, it is characterised in that described device includes:First acquisition unit, for based on the preferential dimension in preferential dimension data storehouse, obtaining pre- The characteristic of the various dimensions of user in section of fixing time;Division unit, for the characteristic of the various dimensions to be divided into multiple cycles;Second acquisition unit, for obtaining the user activity value in each cycle;Computing unit, within the predetermined cycle, calculating the user activity value in cycle two-by-two Between Euclidean distance;Cluster cell, for according to the Euclidean distance, cluster user population data;Push unit, for according to user group's data, to user terminal pushed information.
- 8. the device of pushed information according to claim 7, it is characterised in that described Two acquiring units include:First obtain subelement, in each cycle obtain various dimensions characteristic plus Weight average value;Normalize subelement, for normalizing the weighted average, the spy after being normalized Levy data;Subelement is fitted, for being fitted the characteristic after the normalization, user is obtained and enlivens Write music line;Second obtains subelement, for according to the user activity curve, obtaining each cycle User activity value.
- 9. the device of the pushed information according to any one of claim 7 or 8, its feature It is, the fitting subelement is further used for:Characteristic after the normalization is inputted Support vector regression model is fitted, and obtains user activity curve.
- 10. the device of pushed information according to claim 9, it is characterised in that described Second acquisition unit also includes:Subelement is trained, for using the characteristic after the normalization, training the support Vector regression model.
- 11. the device of the pushed information according to any one of claim 7 or 8, it is special Sign is that described device also includes:Detection unit, for detecting the History Order data in user group's data with user The user's individual data items being consistent;3rd acquiring unit, for obtaining the characteristic acquired in user's individual data items The quantity of the dimension of dimension and the characteristic acquired in each;4th acquiring unit, obtained for from big to small, obtaining predetermined number according to the quantity The dimension of the characteristic taken;AndSetup unit, for using the dimension of the characteristic acquired in the predetermined number as excellent First dimension, update the preferential dimension data storehouse.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610371042.5A CN107451832B (en) | 2016-05-30 | 2016-05-30 | Method and device for pushing information |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610371042.5A CN107451832B (en) | 2016-05-30 | 2016-05-30 | Method and device for pushing information |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107451832A true CN107451832A (en) | 2017-12-08 |
CN107451832B CN107451832B (en) | 2023-09-05 |
Family
ID=60485643
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610371042.5A Active CN107451832B (en) | 2016-05-30 | 2016-05-30 | Method and device for pushing information |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107451832B (en) |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108648000A (en) * | 2018-04-24 | 2018-10-12 | 腾讯科技(深圳)有限公司 | Method and device, the electronic equipment that life cycle is assessed are retained to user |
CN109241133A (en) * | 2018-08-14 | 2019-01-18 | 北京粉笔未来科技有限公司 | Data monitoring method, calculates equipment and storage medium at device |
CN109711871A (en) * | 2018-12-13 | 2019-05-03 | 北京达佳互联信息技术有限公司 | A kind of potential customers determine method, apparatus, server and readable storage medium storing program for executing |
WO2019169756A1 (en) * | 2018-03-06 | 2019-09-12 | 平安科技(深圳)有限公司 | Product recommendation method and apparatus, and storage medium |
CN111062824A (en) * | 2019-12-04 | 2020-04-24 | 腾讯科技(深圳)有限公司 | Group member processing method and device, computer equipment and storage medium |
CN111400764A (en) * | 2020-03-25 | 2020-07-10 | 支付宝(杭州)信息技术有限公司 | Personal information protection wind control model training method, risk identification method and hardware |
CN111612499A (en) * | 2020-04-03 | 2020-09-01 | 浙江口碑网络技术有限公司 | Information pushing method and device, storage medium and terminal |
CN111797848A (en) * | 2019-04-09 | 2020-10-20 | 成都鼎桥通信技术有限公司 | User classification method, device, equipment and storage medium |
CN112053212A (en) * | 2020-09-30 | 2020-12-08 | 杭州拼便宜网络科技有限公司 | Commodity group purchase method, commodity group purchase system and electronic equipment |
CN112150179A (en) * | 2019-06-28 | 2020-12-29 | 京东数字科技控股有限公司 | Information pushing method and device |
CN117808473A (en) * | 2024-03-01 | 2024-04-02 | 深圳迅策科技股份有限公司 | Privacy calculation method and system for transaction data |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102339296A (en) * | 2010-07-26 | 2012-02-01 | 阿里巴巴集团控股有限公司 | Method and device for sorting query results |
CN102346894A (en) * | 2010-08-03 | 2012-02-08 | 阿里巴巴集团控股有限公司 | Output method, system and server of recommendation information |
CN102411596A (en) * | 2010-09-21 | 2012-04-11 | 阿里巴巴集团控股有限公司 | Information recommendation method and system |
US8533144B1 (en) * | 2012-11-12 | 2013-09-10 | State Farm Mutual Automobile Insurance Company | Automation and security application store suggestions based on usage data |
CN103377242A (en) * | 2012-04-25 | 2013-10-30 | Tcl集团股份有限公司 | User behavior analysis method, user behavior analytical prediction method and television program push system |
CN104077332A (en) * | 2013-03-29 | 2014-10-01 | 上海城际互通通信有限公司 | User behavior analysis method based on charging information |
CN104111946A (en) * | 2013-04-19 | 2014-10-22 | 腾讯科技(深圳)有限公司 | Clustering method and device based on user interests |
CN104866540A (en) * | 2015-05-04 | 2015-08-26 | 华中科技大学 | Personalized recommendation method based on group user behavior analysis |
-
2016
- 2016-05-30 CN CN201610371042.5A patent/CN107451832B/en active Active
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102339296A (en) * | 2010-07-26 | 2012-02-01 | 阿里巴巴集团控股有限公司 | Method and device for sorting query results |
CN102346894A (en) * | 2010-08-03 | 2012-02-08 | 阿里巴巴集团控股有限公司 | Output method, system and server of recommendation information |
CN102411596A (en) * | 2010-09-21 | 2012-04-11 | 阿里巴巴集团控股有限公司 | Information recommendation method and system |
CN103377242A (en) * | 2012-04-25 | 2013-10-30 | Tcl集团股份有限公司 | User behavior analysis method, user behavior analytical prediction method and television program push system |
US8533144B1 (en) * | 2012-11-12 | 2013-09-10 | State Farm Mutual Automobile Insurance Company | Automation and security application store suggestions based on usage data |
CN104077332A (en) * | 2013-03-29 | 2014-10-01 | 上海城际互通通信有限公司 | User behavior analysis method based on charging information |
CN104111946A (en) * | 2013-04-19 | 2014-10-22 | 腾讯科技(深圳)有限公司 | Clustering method and device based on user interests |
CN104866540A (en) * | 2015-05-04 | 2015-08-26 | 华中科技大学 | Personalized recommendation method based on group user behavior analysis |
Cited By (17)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2019169756A1 (en) * | 2018-03-06 | 2019-09-12 | 平安科技(深圳)有限公司 | Product recommendation method and apparatus, and storage medium |
CN108648000A (en) * | 2018-04-24 | 2018-10-12 | 腾讯科技(深圳)有限公司 | Method and device, the electronic equipment that life cycle is assessed are retained to user |
CN108648000B (en) * | 2018-04-24 | 2022-10-28 | 腾讯科技(深圳)有限公司 | Method and device for evaluating user retention life cycle and electronic equipment |
CN109241133A (en) * | 2018-08-14 | 2019-01-18 | 北京粉笔未来科技有限公司 | Data monitoring method, calculates equipment and storage medium at device |
CN109711871A (en) * | 2018-12-13 | 2019-05-03 | 北京达佳互联信息技术有限公司 | A kind of potential customers determine method, apparatus, server and readable storage medium storing program for executing |
CN111797848B (en) * | 2019-04-09 | 2023-10-24 | 成都鼎桥通信技术有限公司 | User classification method, device, equipment and storage medium |
CN111797848A (en) * | 2019-04-09 | 2020-10-20 | 成都鼎桥通信技术有限公司 | User classification method, device, equipment and storage medium |
CN112150179A (en) * | 2019-06-28 | 2020-12-29 | 京东数字科技控股有限公司 | Information pushing method and device |
CN112150179B (en) * | 2019-06-28 | 2024-04-09 | 京东科技控股股份有限公司 | Information pushing method and device |
CN111062824B (en) * | 2019-12-04 | 2023-08-18 | 腾讯科技(深圳)有限公司 | Group member processing method, device, computer equipment and storage medium |
CN111062824A (en) * | 2019-12-04 | 2020-04-24 | 腾讯科技(深圳)有限公司 | Group member processing method and device, computer equipment and storage medium |
CN111400764A (en) * | 2020-03-25 | 2020-07-10 | 支付宝(杭州)信息技术有限公司 | Personal information protection wind control model training method, risk identification method and hardware |
CN111612499B (en) * | 2020-04-03 | 2023-07-28 | 浙江口碑网络技术有限公司 | Information pushing method and device, storage medium and terminal |
CN111612499A (en) * | 2020-04-03 | 2020-09-01 | 浙江口碑网络技术有限公司 | Information pushing method and device, storage medium and terminal |
CN112053212A (en) * | 2020-09-30 | 2020-12-08 | 杭州拼便宜网络科技有限公司 | Commodity group purchase method, commodity group purchase system and electronic equipment |
CN117808473A (en) * | 2024-03-01 | 2024-04-02 | 深圳迅策科技股份有限公司 | Privacy calculation method and system for transaction data |
CN117808473B (en) * | 2024-03-01 | 2024-05-31 | 深圳迅策科技股份有限公司 | Privacy calculation method and system for transaction data |
Also Published As
Publication number | Publication date |
---|---|
CN107451832B (en) | 2023-09-05 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107451832A (en) | The method and apparatus of pushed information | |
CN105320766B (en) | Information-pushing method and device | |
CN104102648B (en) | Interest based on user behavior data recommends method and device | |
CN106227786B (en) | Method and apparatus for pushed information | |
CN110060090A (en) | Method, apparatus, electronic equipment and the readable storage medium storing program for executing of Recommendations combination | |
CN107908740A (en) | Information output method and device | |
CN107679217B (en) | Associated content extraction method and device based on data mining | |
CN107172151A (en) | Method and apparatus for pushed information | |
CN106651544B (en) | Conversational recommendation system with minimal user interaction | |
CN105976212B (en) | Commodity display method and device and electronic commerce platform | |
CN109829108B (en) | Information recommendation method and device, electronic equipment and readable storage medium | |
CN107784390A (en) | Recognition methods, device, electronic equipment and the storage medium of subscriber lifecycle | |
CN108510373A (en) | Paintings recommend method, paintings recommendation apparatus, equipment and storage medium | |
CN105718184A (en) | Data processing method and apparatus | |
CN110363604B (en) | Page generation method and device | |
CN107797982A (en) | For identifying the method, apparatus and equipment of text type | |
WO2019080662A1 (en) | Information recommendation method, device and apparatus | |
CN107145485A (en) | Method and apparatus for compressing topic model | |
CN109064265A (en) | Purchase vehicle recommended method and system based on the network platform | |
CN107944481A (en) | Method and apparatus for generating information | |
CN108632311A (en) | Information-pushing method and device | |
CN110111167A (en) | A kind of method and apparatus of determining recommended | |
CN111291191B (en) | Broadcast television knowledge graph construction method and device | |
CN107885784A (en) | The method and apparatus for extracting user characteristic data | |
CN112116426A (en) | Method and device for pushing article information |
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 |