CN108694211A - Using distribution method and device - Google Patents
Using distribution method and device Download PDFInfo
- Publication number
- CN108694211A CN108694211A CN201710234047.8A CN201710234047A CN108694211A CN 108694211 A CN108694211 A CN 108694211A CN 201710234047 A CN201710234047 A CN 201710234047A CN 108694211 A CN108694211 A CN 108694211A
- Authority
- CN
- China
- Prior art keywords
- information
- application
- terminal
- level
- user
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/01—Input arrangements or combined input and output arrangements for interaction between user and computer
- G06F3/048—Interaction techniques based on graphical user interfaces [GUI]
- G06F3/0481—Interaction techniques based on graphical user interfaces [GUI] based on specific properties of the displayed interaction object or a metaphor-based environment, e.g. interaction with desktop elements like windows or icons, or assisted by a cursor's changing behaviour or appearance
- G06F3/04812—Interaction techniques based on cursor appearance or behaviour, e.g. being affected by the presence of displayed objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/01—Input arrangements or combined input and output arrangements for interaction between user and computer
- G06F3/048—Interaction techniques based on graphical user interfaces [GUI]
- G06F3/0481—Interaction techniques based on graphical user interfaces [GUI] based on specific properties of the displayed interaction object or a metaphor-based environment, e.g. interaction with desktop elements like windows or icons, or assisted by a cursor's changing behaviour or appearance
- G06F3/04817—Interaction techniques based on graphical user interfaces [GUI] based on specific properties of the displayed interaction object or a metaphor-based environment, e.g. interaction with desktop elements like windows or icons, or assisted by a cursor's changing behaviour or appearance using icons
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04M—TELEPHONIC COMMUNICATION
- H04M1/00—Substation equipment, e.g. for use by subscribers
- H04M1/72—Mobile telephones; Cordless telephones, i.e. devices for establishing wireless links to base stations without route selection
- H04M1/724—User interfaces specially adapted for cordless or mobile telephones
- H04M1/72448—User interfaces specially adapted for cordless or mobile telephones with means for adapting the functionality of the device according to specific conditions
- H04M1/72454—User interfaces specially adapted for cordless or mobile telephones with means for adapting the functionality of the device according to specific conditions according to context-related or environment-related conditions
-
- 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
Abstract
The embodiment of the invention discloses a kind of application distribution method and device, the method includes:Obtain the current application scenarios information of terminal, wherein the application scenarios information, the application demand current for reflecting the user of the terminal;Determine the intended application for meeting presupposed information matching relationship with the application scenarios information;Determine that intended application distributes the display parameters at interface in application, wherein the display parameters include at least display location parameter;According to the display parameters, the application message of the intended application is exported at application distribution interface.
Description
Technical field
The present invention relates to information technology field more particularly to a kind of application distribution methods and device.
Background technology
With the development of electronic technology, it is more and more to develop the application that can be applicable in electronic equipment.User is if desired
The service provided using some service or some specific merchant, may all need to arrive first the application shop page and go to download and correspond to
Application installation package, then installed again.Many application messages are all generally shown in the distribution interface of distribution application, it should
Application message can be used for terminal and be connected to corresponding application installation package.But the application sequence at present in application shop is if it were not for root
It is arranged according to the type of application, if it were not for being arranged according to the download of each application installation package.If user will download pair
The application installation package answered if it were not for searching corresponding application message one by one by page turning, otherwise is exactly manually to search
The application installation package oneself to be searched for is inputted in rope frame.Using first method, it is clear that user needs to find answering of oneself wanting
It is low with the cumbersome and efficiency of installation kit.By the way of manual search, it is clear that sometimes user is difficult to expect suitably searching
Rope keyword, therefore there is also cumbersome, operation difficulty be big and the phenomenons such as efficiency is low.
Invention content
In view of this, an embodiment of the present invention is intended to provide a kind of application distribution method and device, at least partly solve above-mentioned
Problem.
In order to achieve the above objectives, the technical proposal of the invention is realized in this way:
The offer of first aspect of the embodiment of the present invention is a kind of using distribution method, including:
Obtain the current application scenarios information of terminal, wherein the application scenarios information, the use for reflecting the terminal
The current application demand in family;
Determine the intended application for meeting presupposed information matching relationship with the application scenarios information;
Determine that intended application distributes the display parameters at interface in application, wherein the display parameters include at least display
Location parameter;
According to the display parameters, the application message of the intended application is exported at application distribution interface.
Based on said program, the current application scenarios information of the acquisition terminal, including:
According to the current location information of the terminal, the schedule information of the terminal storage, the terminal it is currently used
At least one in status information and user's operation behavioral data, determines level-one scene information, wherein the level-one field
Scape information, including:It is carried out using the scene information divided according to default partition dimension, the level-one scene information is used to indicate
The current desired corresponding level-one dimension of terminal.
It is described to obtain the current application scenarios information of terminal based on said program, further include:
According to the current location information of the terminal, the schedule information of the terminal storage, the terminal it is currently used
At least one in status information and user's operation behavioral data, determines two level scene information;Wherein, the two level field
Scape information includes:The two level dimension of the corresponding level-one dimension of the level-one scene information of the two level scene information description is retouched
State information;The two level scene information is used to indicate the corresponding two level dimension of the current desired application of the terminal.
It is described to obtain the current application scenarios information of terminal based on said program, further include:
According to the current location information of the terminal, the schedule information of the terminal storage, the terminal it is currently used
At least one in status information and user's operation behavioral data, determines three-level scene information, wherein the three-level field
Scape information includes:The description information of three-level dimension under the two level dimension is described.
Based on said program, the three-level scene information, including:Describe main demand main demand information and with description and institute
State the back-up needs information of the relevant back-up needs of main demand.
Based on said program, the method further includes:
Obtain predetermined information, wherein the predetermined information includes:The historical behavior data of user's operation terminal, current line
For at least one of data and user's portrait attribute;
The intended application for determining to meet presupposed information matching relationship with the application scenarios information, including:
Determine that the target that presupposed information matching relationship is satisfied by with the application scenarios information and the predetermined information is answered
With.
Based on said program, the method further includes:
When determining to be satisfied by the alternative of presupposed information matching relationship with the application scenarios information and the predetermined information
Using at least one for the conversion ratio CVR and/or recommended parameter for later, obtaining the alternative application;
According to the CVR and the recommended parameter, intended application is selected from the alternative application.
Based on said program, the current application scenarios information of the acquisition terminal, including:
Obtain the user behavior data that user applies the terminal;
The user behavior data is generated into behavior five-tuple, wherein the five-tuple includes:Navigation patterns information, point
Behavioural information, search behavior information are hit, behavioural information is downloaded and uses switching behavior information.
It is described to determine that the target for meeting presupposed information matching relationship with the application scenarios information is answered based on said program
With, including:
The input layer for the neural network that the application scenarios information input is trained using sample data;
The hidden layer of the neural network handles the scene information inputted from the input layer, from the nerve net
The output layer of network exports the identification information of the intended application.
The offer of second aspect of the embodiment of the present invention is a kind of using diostribution device, including:
Acquiring unit, the application scenarios information current for obtaining terminal, wherein the application scenarios information, for anti-
Reflect the current application demand of the user of the terminal;
First determination unit, for determining that the target for meeting presupposed information matching relationship with the application scenarios information is answered
With;
Second determination unit, for determining intended application in the display parameters using distribution interface, wherein the display
Parameter includes at least display location parameter;
Output unit, for according to the display parameters, answering for the intended application to be exported at application distribution interface
Use information.
Based on said program, the acquiring unit is specifically used for according to the current location information of the terminal, the terminal
The schedule information of storage, the currently used status information of the terminal and at least one in user's operation behavioral data,
Determine level-one scene information, wherein the level-one scene information, including:It is carried out using division according to default partition dimension
Scene information, the level-one scene information are used to indicate the current desired corresponding level-one dimension of the terminal.
Based on said program, the acquiring unit is specifically used for according to the current location information of the terminal, the terminal
The schedule information of storage, the currently used status information of the terminal and at least one in user's operation behavioral data,
Determine two level scene information;Wherein, the two level scene information includes:The level-one field of the two level scene information description
The description information of the two level dimension of the corresponding level-one dimension of scape information;The two level scene information is used to indicate the terminal and works as
The corresponding two level dimension of preceding required application.
Based on said program, the acquiring unit is specifically used for according to the current location information of the terminal, the terminal
The schedule information of storage, the currently used status information of the terminal and at least one in user's operation behavioral data,
Determine three-level scene information, wherein the three-level scene information includes:Retouching for three-level dimension under the two level dimension is described
State information.
Based on said program, the three-level scene information, including:Describe main demand main demand information and with description and institute
State the back-up needs information of the relevant back-up needs of main demand.
Based on said program, the acquiring unit is additionally operable to obtain predetermined information, wherein the predetermined information includes:With
The historical behavior data of family operating terminal, current behavior data and user draw a portrait attribute at least one;
First determination unit, specifically for determining to be satisfied by with the application scenarios information and the predetermined information
The intended application of presupposed information matching relationship.
Based on said program, first determination unit, also particularly useful for when determine with the application scenarios information and
The predetermined information is satisfied by after the alternative application of presupposed information matching relationship, obtains the conversion ratio CVR of the alternative application
And/or at least one of recommended parameter;According to the CVR and the recommended parameter, selected from the alternative application
Intended application.
Based on said program, the acquiring unit is additionally operable to obtain the user behavior data that user applies the terminal;It will
The user behavior data generates behavior five-tuple, wherein the five-tuple includes:Navigation patterns information, click behavioural information,
Search behavior information downloads behavioural information and uses switching behavior information.
Based on said program, first determination unit is specifically used for the application scenarios information input using sample
The input layer for the neural network that data are trained;The hidden layer of the neural network believes the scene inputted from the input layer
Breath is handled, and the identification information of the intended application is exported from the output layer of the neural network.
Application distribution method and device provided in an embodiment of the present invention can obtain the current application scenarios information of terminal, root
Intended application is determined according to application scenarios information, and further determines the display parameters of intended application;The display parameters are at least
Including display location parameter, goal application would generally so that intended application is aobvious by the setting of the display location parameter
Show the conspicuous position at application distribution interface, to reduce the operation that user searched or searched for current desired application, simplifies user
It searches and oneself thinks operation to be applied, promote application distribution efficiency, promote the intelligent and users' satisfaction degree of application distribution.
Description of the drawings
Fig. 1 be it is provided in an embodiment of the present invention the first apply distribution method flow diagram;
Fig. 2 is a kind of structural schematic diagram of dissemination system provided in an embodiment of the present invention;
Fig. 3 is the display effect schematic diagram that interface is distributed in a kind of application provided in an embodiment of the present invention;
Fig. 4 is the flow diagram of second provided in an embodiment of the present invention application distribution method;
Fig. 5 be it is provided in an embodiment of the present invention the third apply distribution method flow diagram;
Fig. 6 is the flow diagram of the training method of neural network provided in an embodiment of the present invention;
Fig. 7 is a kind of structural schematic diagram using diostribution device provided in an embodiment of the present invention;
Fig. 8 is the flow diagram of 4th kind provided in an embodiment of the present invention application distribution method;
Fig. 9 is another structural schematic diagram for applying diostribution device provided in an embodiment of the present invention.
Specific implementation mode
Technical scheme of the present invention is further elaborated below in conjunction with Figure of description and specific embodiment.
As shown in Figure 1, the present embodiment provides it is a kind of apply distribution method, including
Step S110:Obtain the current application scenarios information of terminal, wherein the application scenarios information, for reflecting
State the current application demand of the user of terminal;
Step S120:Determine the intended application for meeting presupposed information matching relationship with the application scenarios information;
Step S130:Determine that intended application distributes the display parameters at interface in application, wherein the display parameters are at least
Including display location parameter;
Step S140:According to the display parameters, the application letter of the intended application is exported at application distribution interface
Breath.
It can be for applied to the information processing method using distribution platform or terminal using distribution method described in this implementation.Figure
2 show using the attachment structure schematic diagram between distribution platform and terminal.The distribution platform can be applied to network side
One or more application distribution server.The terminal can be the electronics such as mobile phone, tablet computer or the wearable device of user
Equipment etc., terminal can be connected through the internet to distribution platform.
The processing using distribution platform S110 to step S140 through the above steps, when particular terminal occurs in determination
Intended application can then be exported the application message of intended application to the terminal by step S140, facilitate terminal in its installation
Distribution application application distribution interface in feature position show the application message of the intended application.Here application distribution
Interface is the application interface of the distribution application.Fig. 3 show a kind of application distribution interface of the distribution application, in application point
Hair interface display has the application icon of various applications to be distributed, user to click the application icon, then terminal can download the application
Application installation package.
Specifically, as shown in figure 4, including applied to the application process in application platform:
Step S1:Application platform receives the application scenarios information that terminal reports;
Step S2:Application platform is determined to need the mesh to the terminal keypoint recommendation according to the application scenarios information
Mark application;
Step S3:According to default Generalization bounds, determine that the intended application distributes the aobvious of interface display in the application of terminal
Show parameter, display parameters here include at least display location parameter;
Step S4:The application message of the intended application and the display parameters are exported to terminal, to control terminal
According to the application message of display parameters intended application described in application distribution interface display.
Certainly, it is described in the present embodiment using distribution method may be be applied to the terminal in method.This
If sample, the distribution platform needs the application message for the various applications distributed to be issued in the terminal.The terminal is logical
Information processings of the step S110 to step S140 is crossed, is determined described in the application distribution interface output of the distribution application of its installation
The application message of intended application.Specifically such as, as shown in figure 5, may include applied to the application distribution method in terminal:
Step S11:Obtain the application scenarios information;
Step S12:According to the application scenarios information, selects from the various applications that application platform pushes and answered with described
Meet the intended application of the presupposed information matching relationship with scene information;
Step S13:According to default Generalization bounds, the display parameters of the intended application are determined, display parameters here
Including at least the display location parameter;
Step S14:According to the display parameters, the intended application described in application distribution interface display.
The current application scene information of the terminal in the present embodiment, it may include:The position that the terminal is currently located
Information, the purposes information of present position, current time, the operating status of user terminal, currently detected user's operation.
The terminal can be mobile terminal in the present embodiment, and being can be by the portable equipment of user, such words,
The position where position, that is, user where terminal.If the terminal is fixed terminal, if user is in using terminal, terminal institute
Position namely user where position.
The purposes information of present position, the purposes information may indicate that the purposes of present position, for example, meeting
The various uses information such as room, classroom, market or dining room.
The current time can be the various information such as current at the time of point, season or month.
The operating status of the user terminal may include:Whether the terminal is currently in open state, terminal
The various information such as application, the foreground application in terminal and the background application of open state.
User's operation can be the various operations that control terminal input by user executes, and indicate to open some using user
Using or close some application, the various application operatings etc. that carry out in the application interface applied at one.
Above- mentioned information can directly or indirectly reflect application demand of the user using application.
For example, if the terminal is currently located in user family, possible user has the application using leisure, amusement and shopping to need
It asks.For example, if the terminal, in meeting room, user may need the application of minutes.
For another example user is when market is done shopping, it may be necessary to shopping application or shopping payment class application.
After determining the application scenarios information in the step S120 in the present embodiment, it can determine whether out to answer with described
With the compatible intended application of scene demand.
Each its specific purposes of application provides specific service;If purposes or service that the application provides can
Meet current application demand, then the application can be used as the intended application.Obvious goal application may be application
The application of demand.
When the service of the purposes and/or offer applied when one can meet the application demand, then it is believed that the application
Meet the presupposed information matching relationship with the application scenarios information.
If user may currently need to use mother and baby's product shopping application in some embodiments, then the user is also possible to
The potential user of toiletries shopping application, if at this point, one application purposes and/or offer service, with according to described in
The application demand that application scenarios information is determined has incidence relation, it is also assumed that being to meet the presupposed information matching relationship
One of intended application.
Meet presupposed information matching relationship, including two kinds with the application scenarios information described in the present embodiment:
The first:
Meet positive match relationship with the application scenarios information, i.e., the application, which can be provided just, meets the application need
The service asked or purposes.
Second:
Meeting proximity matching relationship with the application scenarios information, the machine application can partly meet the application demand,
Or it can provide and the associated service of the application demand and/or purposes.
In short, the mode for meeting presupposed information matching relationship with the application scenarios information in the present embodiment has very much
Kind, it is not limited to any one of the above.
In step s 130 if it is determined that after intended application, join it is determined that going out its display at application distribution interface
Number.The display parameters include at least display location parameter in this embodiment.
It is usual one application distribution interface in is provided with multiple display positions, some display positions are forward, some display positions rearward,
Some shows that position is very easy to cause the attention of user, and some display positions would not.The display position in the present embodiment
Set the identification parameter that parameter can be the display position.
The display parameters may also include in some implementations:Other display parameters, for example, display color, display area
Parameter etc..In short, the display parameters are to highlight parameter in this embodiment.Here the parameter that highlights is:So that institute
Intended application is stated than being shown in around it or neighbouring other application more quotes the display parameters that family pays attention to.For example, target is answered
Show that area can be more than the display area of non-targeted application, the non-targeted application of display background color ratio of intended application is shown
Show that background colour is more prominent etc..In short, the default Generalization bounds described in the present embodiment can be by highlighting intended application,
Reached to the display parameters for the application message applied needed for user's express user determination strategy with attracting the user's attention power.
Then directly according to the display parameters determined in step S140, in the application distribution interface display institute of distribution application
State the application message of intended application.
Here the application message of intended application may include:The acquisition link of intended application, the application icon of intended application, should
Application icon and the chained address for the application installation package for obtaining intended application are bound, when detecting the behaviour for acting on the application icon
When making, the electronic equipment can obtain the application installation package of the intended application automatically.
In some embodiments, the application message may also include the relevant information of the application, relevant information here
It may include:The description information of the purposes of the intended application, the industry of ownership or available service is described, or recommends the mesh
Mark the recommendation information of application.When user being facilitated to see the application message in this way, determine whether the intended application is exactly that itself is current
Required intended application.
In short, the application distribution flow before carrying out obtaining the output of the application message of application installation package in the present embodiment
When, no longer it is to show into row stochastic application message in application distribution interface or be distributed based on download, but can obtain
The application scenarios information, according to application scenarios information sifting go out disclosure satisfy that the intended application of user's current demand preferentially into
Row distribution reduces the operation searched or searched for one by one in magnanimity application message of the user on the interface of distribution application, letter
The operation that user obtains application installation package is changed, the intelligent and user for improving electronic equipment searches application installation needed for oneself
The rate of the application message of packet.
In some embodiments, the step S110 may include:
According to the current location information of the terminal, the schedule information of the terminal storage, the terminal it is currently used
At least one in status information and user's operation behavioral data, determines level-one scene information, wherein the level-one field
Scape information, including:It is carried out using the scene information divided according to default partition dimension, the level-one scene information is used to indicate
The current desired corresponding level-one dimension of terminal.
The current location information of terminal can be obtained in the present embodiment, specifically such as, pass through global positioning system (Global
Position System, GPS), base station auxiliary positioning or the location information for waiting various positioning methods.
The schedule information of the terminal storage can be the schedule that the user of terminal acquisition is manually entered, pre- based on user
Determine the schedule information that the operating terminals such as air ticket voluntarily generate.For example, user can arrange the work of oneself, and it is input in terminal,
When handling different schedule things, substantial required application is different user.
For example, the plan that user has arranged to travel outdoors by application of travelling, then before user goes on a tour or during going on a tour,
User may generate interest to the cuisines and culture on ground of going on a tour, then user applies there is cuisines and/or culture recommends class to answer
Application demand;Therefore the scene information can be extracted according to schedule information.
The currently used state of the terminal.For example, terminal is that booting uses for the first time, then terminal may be downloaded in batches
Many common applications.If present terminal is just in one game application of front stage operation, the user using the terminal may be one
A game lover, then the user may also generate interest to other game, then game application is the application demand of active user.It is aobvious
So, the currently used state of terminal is also that can characterize the current application demand of user.
The user's operation behavioral data may include:Historical operation behavioral data and/or current application operation behavior data.
Here user's operation behavior can be the data that the various operations that user inputs from human-computer interaction interface generate, for example, clicking row
For data, download behavioral data, reading behavior data etc..In some embodiments, historical operation behavioral data and current behavior
Operation data, can be with the long-term hobby and/or preference of characterized user;The current application behavioral data can characterize user and work as
Preceding hobby and/or preference, therefore the user's operation behavioral data can reflect the application demand of the user.
In the present embodiment first according to above-mentioned current location information, schedule information, terminal use state information and
At least one of user's operation behavioral data determines the level-one scene information.In the present embodiment, the step S110
Can be obtained according to prefixed time interval the above-mentioned current location information, schedule information, terminal use state information and
At least one of user's operation behavioral data, then according to the correspondence of these information and application scenarios information, and/
Or, weighing factor of these information to application scenarios information, it can be in conjunction with the use of current location information, schedule information, terminal
Status information and user's operation behavioral data it is multinomial, synthesis determines the level-one scene information.
The partition dimension of the scene information in the present embodiment, including a variety of dimensions, for example, industry dimension or function dimension
Degree etc..
The industry dimension, can be divided into:Shopping, social, travel and amusement etc..
The function dimension, can be divided into:Multiple dimensions such as multimedia messages output, multimedia messages input and communication.
In this case, according to the current location information, it may be determined which current desired dimension application of user gone out.
For example, determining that user has plan of travel according to schedule information, then user may need travel class application.
The scene information includes level-one scene information in the present embodiment, in the step s 120 can be according to the level-one field
Scape information, it may be determined that go out the place dimension of the intended application.If in this case, can in step S130 and step S140
To include facilitating user to check at the front row at application distribution interface by the application message for corresponding to the application of dimension.For example, for having
The user of plan of travel can be then shown in the front row using distribution interface with travel class application, facilitate that user checks and acquisition is answered
Use installation kit.
The level-one application scenarios information in the present embodiment at least can be used to determine the class of the intended application.
In step s 130, it may be determined which kind of application is arranged as intended application in the notable of application distribution interface
Position.Here notable position can be according to statistical data or emulation or practice, and that determines is easier attraction user
The display position of attention.
Optionally, the step S110 further includes:
According to the current location information of the terminal, the schedule information of the terminal storage, the terminal it is currently used
At least one in status information and user's operation behavioral data, determines two level scene information;Wherein, the two level field
Scape information includes:The two level dimension of the corresponding level-one dimension of the level-one scene information of the two level scene information description is retouched
State information;The two level scene information is used to indicate the corresponding two level dimension of the current desired application of the terminal.
It is located at the level-one scene if the level-one scene information is divided according to industry, in the present embodiment to believe
Two level scene information under breath can be the sub-industry described under some industry.For example, shopping class is applied and is carried out according to sub-industry
It divides, may include:The shopping of mother and baby's class, toiletries shopping, clothing shopping, electric type shopping and comprehensive shopping etc..
If the level-one scene information description is the trade information of the affiliated industry of intended application, the two level scene information then describes
Intended application ownership sub-industry sub-industry information.In the present embodiment, if the level-one dimension is industry, described two
Grade scene information may be used to indicate described then to describe the description information of the sub-industry under the industry where the intended application
The corresponding sub-industry of the current desired application of terminal.
It is located at the level-one scene if the level-one scene information is divided according to purposes, in the present embodiment to believe
Two level scene information under breath can be the relevant information for describing a certain purposes or the subdivision purposes under function.Such as, it currently determines
The corresponding communication class of the level-one scene information, then communication class can be subdivided into instant messaging and non-instant communication again.According to current
The various data such as user's operation behavioral data may further determine that out that user is currently needed instant messaging application, still
Non-instant communications applications.If in this case, when determining intended application, can determine that target is answered according to level-one scene information
With the major class at place the group where intended application is determined then according to the two level application scenarios information.
As further improvement of this embodiment, the step S110 further includes:
According to the current location information of the terminal, the schedule information of the terminal storage, the terminal it is currently used
At least one in status information and user's operation behavioral data, determines three-level scene information, wherein the three-level field
Scape information includes:The description information of three-level dimension under the two level dimension is described.
The three-level dimension is the three-level dimension further refined to said second dimension in the present embodiment.For example, institute
It is industry to state level-one dimension, and the two level dimension is sub-industry, then three-level dimension here can be the corresponding product of the sub-industry
Category information.Here category information can be to provide provider and/or brand message of the application needed for the sub-industry etc..
For example, being based on the level-one scene information, determine that the current user demand of user is shopping class application, and according to
Two level scene information is it is determined that need mother and baby's class shopping application;It further can be according to the history Shopping Behaviors data of user
And/or level of consumption information and/or hobby, further orient user it is desirable that mother and baby's class net purchase application overseas, still
Domestic shopping application.
If the level-one scene information can be used for determining that the level-one purposes of intended application, two level scene information can determine that
The two level purposes of intended application, then three-level application scenarios information can be used for determining the current desired three-level purposes of user.For example,
Instant messaging application can be divided into application such as QQ, wechat, facebook and footpath between fields footpath between fields etc. again.It in the present embodiment can be according to the three-level field
Scape information further determines that out the specific purposes category needed for the target user.
In short, the scene information is divided into level-one scene information, two level scene information and three-level scene in the present embodiment
In this case information can be accurately positioned out the current desired intended application of user.
The method can be in conjunction with above-mentioned three-level scene information altogether one or more progress intended applications therein
It determines.In some embodiments, the method may also include:If determining the intended application according to above application scene information
Later, the mount message that application has been installed in the terminal is obtained;
According to the mount message, the application message that terminal described in multiple intended applications has installed application is filtered;
The application of uninstalled intended application is believed in the terminal described in application distribution interface display in step S140
Breath, reduces the distribution or recommendation of unnecessary application message.
In some embodiments, the three-level scene information, including:Describe main demand main demand information and with description and institute
State the back-up needs information of the relevant back-up needs of main demand.
For example, the three-level scene information, including:Describe main demand main category information and with description with the main demand
The supplement category information of relevant back-up needs.
It plays when user fights in complete sports class, then the main category information can be:Instruction user currently needs to fight class
The description information of the A game applications of game application.The supplement category information can be then to belong to the same exploitation with A game applications
Team is outstanding or belongs to the information using B of the same brand.
For example, current A game is the game application of king's honor, then it is described to be read using the news that B be king's honor
Using or game strategy application etc..
Accordingly, in the step S140, can the application distribute interface can with the first predeterminated position can be with
The application message for meeting the main demand is highlighted, then the second predeterminated position that can distribute interface in the application can be more
It is outstanding to show the application message for meeting the back-up needs.Second predeterminated position can be the phase of first predeterminated position
It sets at ortho position.
For example, showing the application message of game A applications in first row, the application message using B is shown in second row.
In some embodiments, the method further includes:
Obtain predetermined information, wherein the predetermined information includes:The historical behavior data of user's operation terminal, current line
For at least one of data and user's portrait attribute;
The step S120 may include:
Determine that the target that presupposed information matching relationship is satisfied by with the application scenarios information and the predetermined information is answered
With.
The historical behavior data in the present embodiment, it may include the various actions data and other rows of user's using terminal
For data.The various actions data of the using terminal may include:The click behavior that is executed to terminal, various is answered at download behavior
Frequency of usage information etc..Other described behavioral datas may include:User recently for child go to school look for school etc. other
Approach obtains and the data of the direct using terminal of non-user.
User's portrait attribute can be the information of various description individual subscriber features;Here individual subscriber feature packet
It includes:Change rate is not less than the behavioral characteristics of predeterminable level less than the static nature and change rate of predeterminable level.The static nature
It may include:The information such as the gender of user, the age bracket, schooling, permanent residence and/or occupation.The behavioral characteristics
It may include:The information such as user interest hobby, the level of consumption of user, user mood.
Not only it can determine the intended application according to application scenarios information in the present embodiment, it can also be according to history row
Go out the target user for information, comprehensive selections such as data, current behavior data and user portrait attributes.
When determining the intended application, following formula may be used and carry out comprehensive score, then the row for passing through comprehensive score
The comparison of sequence and scoring threshold value, selects the intended application:
Wherein, the yiFor the comprehensive score of i-th of alternative application;The xjFor the individual scores of j-th of reference factor;
The wjFor the weight of j-th of reference factor;The J is the total number of reference factor.The i, j, J are positive integer,
And the j is the positive integer no more than the J.
The reference factor includes at least in the present embodiment:The application scenarios information of the terminal.In some embodiments
In may also include:User's portrait attribute, historical behavior data and/or current behavior data etc..
The comprehensive score can be calculated for each application, multiple alternative applications of comprehensive score can be selected
Or comprehensive score is more than the alternative application of predetermined threshold value as the intended application.
Certainly, it the above is only the specific example for providing a selection intended application, specific implementation is not limited to above-mentioned
Functional relation.
In some embodiments, the method further includes:
When determining to be satisfied by the alternative of presupposed information matching relationship with the application scenarios information and the predetermined information
Using at least one for the conversion ratio CVR and recommended parameter for later, obtaining the alternative application;
According to the CVR and the recommended parameter, intended application is selected from the alternative application.
In the present embodiment when carrying out using distribution, corresponding application can be also considered towards all users or corresponding area
The CVR and recommended parameter of intra domain user.It is applied in the Chinese CVR etc. for being clicked download for example, some will be considered.It is described
Recommended parameter can be according to whether be the reference factors such as application provider's payment request recommendation, the application of emphasis distribution.
The determination of the intended application is participated according to the CVR in the present embodiment, in this case, can be recommended to user
The relatively high application of other application favorable comment or the high intended application of other users frequency of usage, to ensure that it is corresponding that user downloads
Using there is preferable use feeling later.
Further include in the present embodiment:Recommended parameter, the recommended parameter can be it is reserved it is various think keypoint recommendation some
Using when for making the application of keypoint recommendation has larger probability to be displayed on the application as the intended application to distribute boundary
One parameter in face.
In the present embodiment, the step S110 may include:
Obtain the user behavior data that user applies the terminal;
The user behavior data is generated into behavior five-tuple, wherein the five-tuple includes:Navigation patterns information, point
Behavioural information, search behavior information are hit, behavioural information is downloaded and uses switching behavior information.
The user behavior data is converted into five-tuple in the present embodiment.Here five-tuple includes 5 elements;This
In each element describe a kind of user behavior information.The operation that each element describes in five-tuple in the present embodiment
Behavioural information can be the behavioural information that corresponding user behavior is executed in the pervious predetermined amount of time of current time.Here behavior
Information may include:The temporal information for executing corresponding behavior, executes the information such as place at the application for executing number, being directed to.
Each element can correspond to an array or an array, array here in five-tuple in the present embodiment
And/or each element has its corresponding meaning again in array.
For another example the user behavior information of each element in the five-tuple can be user behavior sequence.Array
User behavior sequence is to execute time series, the location sequence etc. of the user behavior.It may include for example, downloading behavioural information:Under
Carry time series, download location sequence (downloads the sequence of moment terminal position, download application identities sequence (indicates to download
Using) etc. information.
The user behavior data is accurately described using five-tuple in the present embodiment, facilitate the application platform and/or
The terminal, user behavior data described in convenient disposal, to quick and be accurately positioned out and disclosure satisfy that current application demand
Intended application.
It in the present embodiment can be by above application scene information, user's Figure Characteristics of individual subscriber, historical behavior number
According to, one or more of current behavior data, CVR and recommended parameter be input in pre-determined grader, pass through classification
The classification of device exports the intended application.Here grader may include:The various graders such as the vector machine that can be used for classifying.
In the present embodiment, the step S130 may include:
The input layer for the neural network that the application scenarios information input is trained using sample data;
The hidden layer of the neural network handles the scene information inputted from the input layer, from the nerve net
The output layer of network exports the identification information of the intended application.
In the present embodiment the screening of the intended application and the identification information of intended application are carried out using neural network
Output.The neural network can be a kind of grader being trained in advance using sample data in the present embodiment.
Therefore further include in this implementation method:
Before determining the intended application, the training of the neural network is carried out using sample data.
Here the training for carrying out the neural network may include:
Sample data is built, sample data here includes:It the behavior sequences of the various applications being directed to for behavior and answers
Identification information;
Neural metwork training is carried out using the sample data;When carrying out the training of the neural network, may be used
Back propagation (Back Propagation Algorithm, BPA) method is trained.It is trained using BPA methods, then
When carrying out neural metwork training, the output result trained and the error of the legitimate reading in sample data can be returned to
Neural network, god can be each according to the node of the error transfer factor neural network hidden layer, the weight of node and symbol etc. by network
Kind parameter so that the neural network after adjustment inputs the output result obtained later and sample using the behavior sequence of sample data
As a result the god that the difference smaller between legitimate reading or without differences, in this way training obtain positions network with intended application
Accurate high feature.
As shown in fig. 6, user behavior data is converted into five-tuple, the application scenarios that the five-tuple of user behavior is generated are believed
The sequence for ceasing composition, the correspondence that the application sequence of the application that user behavior operates in five-tuple is formed, as neural network
Sample data, be input to untrained neural network or not training complete neural network, neural network is trained, with
Being obtained by one or many training can be with the neural network of subsequent applications.
As further improvement of this embodiment, the method further includes:
The application scenarios information is cleaned into row information, it is ensured that be input to the god by the data of the input floor height of network
It is true, effective data, so that it is guaranteed that the accuracy of intended application positioning.For example, the current location information of user terminal,
Previous minute is also in China, and the current location information of present terminal is then shown in the U.S., it is clear that it is fixed to be then likely to occur at this time
Position is abnormal;This abnormal data can be removed by data cleansing, to be accurately positioned the intended application.
When carrying out the neural metwork training, sample data is also needed by data cleansing, it is ensured that nerve net
The reasonability of network itself, to further ensure that the determine accuracy to intended application.
The accurate and easy distribution of the intended application, the step S140 may include for convenience in the present embodiment:
Pop-up is shown on application distribution interface;Here pop-up is shown in using the display surface on distribution interface
Product is less than the wicket at application distribution interface;The application message of the intended application is shown in wicket.
In order to simplify the distribution of the intended application, may also include:
It is applied in the not actuated distribution application of terminal or the distribution when backstage starts, distribution platform is to the terminal
Send the PUSH message for the application message for including the intended application;Or the terminal voluntarily generates and recommends the intended application
Recommendation information.Here PUSH message or recommendation information can be the corresponding system message of operating system or the distribution application
Application message.
In some embodiments, using the pop-up and/or PUSH message or recommendation information, the intended application of keypoint recommendation
It can be the intended application for meeting the main demand, can also be the intended application for meeting the back-up needs.
As shown in fig. 7, the present embodiment provides it is a kind of apply diostribution device, including:
Acquiring unit 110, the application scenarios information current for obtaining terminal, wherein the application scenarios information is used for
Reflect the current application demand of the user of the terminal;
First determination unit 120, the mesh for determining to meet presupposed information matching relationship with the application scenarios information
Mark application;
Second determination unit 130, for determining intended application in the display parameters using distribution interface, wherein described
Display parameters include at least display location parameter;
Output unit 140, for according to the display parameters, the intended application to be exported at application distribution interface
Application message.
Diostribution device is applied described in the present embodiment, is applied in distribution platform, then the acquiring unit 110 can correspond to
In communication interface, the application scenarios information can be received from the terminal, or with the relevant information of application scenarios information, then profit
The application scenarios information is determined with processor or processing circuit.First determination unit 120 and second described in the present embodiment
Determination unit 130 can correspond to the processor in application platform or processing circuit.The processor can be central processing unit, number
Signal processor, microprocessor, application processor or programmable array etc..The processing circuit can be application-specific integrated circuit.Institute
It states the execution of processor or processing circuit by executable code, determination and the display parameters of above-mentioned intended application may be implemented
It determines.When described device is total applied to distribution platform, then the output unit 140 can correspond to communication interface, output here
The application message can be to export application message and the display parameters to terminal, to which control terminal is aobvious according to display parameters
Show the application message.
If the application diostribution device is applied in terminal, the acquiring unit 110, the first determination unit 120 and the
Two determination units 130 may both correspond to the processor or processing circuit of terminal.Here processor or the detailed of processing circuit is retouched
It states and may refer to preceding sections, not just being repeated here.The output unit 140 can correspond to display screen, display here
Screen can be the display that the various defeated displays such as liquid crystal display, electric ink display screen or projection display screen export the application message
Structure.
The associated description of the application scenarios information, intended application and the application message may refer in the present embodiment
The associated description of preceding method embodiment is not just unfolded to describe here.
Optionally, the acquiring unit 110, specifically for being deposited according to the current location information of the terminal, the terminal
The schedule information of storage, the currently used status information of the terminal and at least one in user's operation behavioral data, really
Make level-one scene information, wherein the level-one scene information, including:It is carried out using the field divided according to default partition dimension
Scape information, the level-one scene information are used to indicate the current desired corresponding level-one dimension of the terminal.
Further, the acquiring unit, it may also be used for according to the current location information of the terminal, the terminal storage
Schedule information, the terminal currently used status information and at least one in user's operation behavioral data, determine
Go out two level scene information;Wherein, the two level scene information includes:The level-one scene letter of the two level scene information description
Cease the description information of the two level dimension of corresponding level-one dimension;The two level scene information is used to indicate the current institute of the terminal
The corresponding two level dimension of application needed.
In addition, the acquiring unit 110, is specifically used for according to the current location information of the terminal, the terminal storage
Schedule information, the terminal currently used status information and at least one in user's operation behavioral data, determine
Go out three-level scene information, wherein the three-level scene information includes:The description letter of three-level dimension under the two level dimension is described
Breath.
The application scenarios information can be divided into many levels in the present embodiment, specifically may include level-one scene information, two
Grade scene information and three-level scene information can facilitate simplicity accurately to determine target by successively orienting scene information
Using.
In some embodiments, the three-level scene information, including:Describe main demand main demand information and with description with
The back-up needs information of the relevant back-up needs of main demand.
Main demand information described in the present embodiment is the information for reflecting the main demand of active user, the back-up needs letter
Breath can be the demand information for reflecting active user's potential demand.In the present embodiment, if according to such three-level scene information
The determination of the intended application is carried out, the intended application met with householder's demand can be not only oriented, may further determine that out satisfaction
The intended application of user's back-up needs, to the intelligent and users' satisfaction degree of lifting device.
Optionally, the acquiring unit 110 is additionally operable to obtain predetermined information, wherein the predetermined information includes:User
The historical behavior data of operating terminal, current behavior data and user draw a portrait attribute at least one;Accordingly, described
One determination unit 120, specifically for determining to be satisfied by presupposed information with the application scenarios information and the predetermined information
Intended application with relationship.
Acquiring unit 110, which is also re-used, in the present embodiment carries out historical behavior data, current behavior data, going through here
History behavioral data and current behavior data, are not limited to the behavioral data of this terminal of user's operation, can also be not operation terminal
Data or the behavioral datas of other electronic equipments is operated under line.In this way, the first determination unit 120 is positioning the intended application
Reference factor is more, so as to more accurately provide the intended application, to which the aobvious of interface is distributed in application as far as possible
Eye position (being easiest to attract the user's attention the position of power) shows the application message of corresponding intended application, reduce user check and
It searches for and oneself thinks operation to be applied, promote the acquisition efficiency of application installation package, promote the efficiency of application distribution, lifting device
Intelligent and users' satisfaction degree.
In further embodiments, first determination unit 120 is determined and the applied field also particularly useful for working as
Scape information and the predetermined information are satisfied by after the alternative application of presupposed information matching relationship, obtain turning for the alternative application
At least one of rate CVR and/or recommended parameter;According to the CVR and the recommended parameter, from the alternative application
Select intended application.
In the present embodiment not only can be in conjunction with userspersonal information, personal information here and individual subscriber behavior and attribute
The information such as relevant application scenarios information, behavioral data and/or user's portrait attribute can also combine the CVR of each application and push away
The non-personal information such as parameter is recommended, to carry out the selection of the intended application, to which another level is in the aobvious of application distribution interface
Eye position shows the intended application.
In addition, the acquiring unit 110, is additionally operable to obtain the user behavior data that user applies the terminal;It will be described
User behavior data generates behavior five-tuple, wherein the five-tuple includes:Navigation patterns information clicks behavioural information, search
Behavioural information downloads behavioural information and uses switching behavior information.
User behavior data is converted into being classified the five-tuple that device is handled in the present embodiment, five-tuple is basic
On include user's operation terminal a few major class operation behaviors, there is information and grader to be facilitated to be handled in detail and comprehensively
The characteristics of.
In also some embodiments, first determination unit 120, specifically for adopting the application scenarios information input
With the input layer for the neural network that sample data is trained;The hidden layer of the neural network is to inputting from the input layer
Scene information is handled, and the identification information of the intended application is exported from the output layer of the neural network.
The application scenarios information is carried out using neural network in the present embodiment and be output to neural network, by neural network
The identification information of intended application is exported, it is thus easy and be accurately obtained the intended application.
Two specific examples are provided below in conjunction with above-mentioned any embodiment:
Example one:
This example provides a kind of distribution application, which includes using distribution interface;At the application distribution interface
On be provided with multiple display positions, different display positions attracts the ability of the attention of user different.Usually got at application distribution interface
Forward position, to the attraction of the attention of user more before, the suction in a row closer to the display position in centre position to user
Gravitation is stronger.
In this example using preceding method, the display position of intended application and intended application is determined, in order to further
Attract the attention of user, can also increase in display position that target is answered and highlight icon, this highlights icon and can wrap
Include the red point for attracting the user's attention power, or the Dynamic Graph etc. of switching so that intended application comes with other non-targeted application divisions.
Application scenarios are divided by this in this example:Level-one scene, two level scene and three-level scene.Here description level-one
The information of scene is aforementioned level-one scene information, the description two level scene is said second scene information, description described three
Grade scene is aforementioned three-level scene information;Following one level scene, two level scene and three-level scene are introduced individually below.
Level-one scene is divided into two kinds of the clear demand scene of user and user's Fuzzy Demand scene.Demand is specified, is had clear
The demand purpose of one industry, such as:It has been ready to travelling, has prepared necessary application (Application, App) of travelling;Prepare
It holds wedding party, prepares relevant App.Fuzzy Demand, without the purpose of specific industry:It is newly mounted with mobile phone, some indispensabilities is installed
Application;It is boring, want to have a look at either with or without some joyful applications;Have a look the activity of distribution application and preferential etc..
Two level scene is finely divided under level-one scene.For clear demand class scene, it is sub-divided into each sub-industry.It is right
In user's Fuzzy Demand scene, some specific scenes for being intended to definite conception are sub-divided into, are such as newly mounted with mobile phone, it is boring to stroll one
Distribution application is strolled, looks at activity etc..
Three-level scene:In each two level scene in the following, can be divided into:Browsing/query process, downloading process are left point
Hair application action, leave distribution application after etc. scenes.
The technology point of division:User distribution application session behavior include:{ click Xu Lie [a1], Sou Suoxulie [a2],
Xia Zaixulie [a3]}.Here click sequence can be the user's operation behavior for checking details page of clicking.By by session behavior into
Row cluster, gathers the type of the classification two level scene.Each type includes that (application scenarios here can wrap for the application scenarios of user
It includes:One or more of level-one scene, two level scene and three-level scene), including main demand and back-up needs two parts are for example
Main demand is king's honor, and back-up needs are the App, strategy App etc. that the news of king's honor is read.
After user begins to use heavy burden, into the acquisition of application scenarios information, specifically it may include:
Initialization procedure:User initially enters the disconnected equipment whether newly to come into operation when distribution application, e.g. newly
The mobile phone to come into operation.New equipment enters indispensable application scenarios, otherwise, scene is strolled into boring.
Scene behavior is mainly made of five-tuple:It is browsing respectively, clicks, search for, downloads, handoff scenario.
The each tuple of scene has corresponding behavior sequence, has such as browsed application sequence, the application sequence clicked, search
Application sequence, the application sequence downloaded, theme or function switch sequence.There may be multiple application masters under an application
Topic or multiple and different functions.For example, news category application, has news reading page, there are review pages, there is Community Page, it is different
The page corresponded to different themes and/or function.Here theme or function switch sequence can be user behavior in application
Theme switching and/or function switch behavior.Using these behavior sequences, carries out characterizing abstract work, obtain feature vector.
Characterize method of the work by deep learning, use at present method for:
The first step:
Behavior sequence is reflected by training using the more common text depth representing model (word2vec) of industry at present
Penetrate into K dimension real vectors.
Second step:
5 layers of convolutional neural networks of construction are classified, and (5 layers of convolutional neural networks are divided into three parts, input layer, output layer
And hidden layer, wherein the neural network that hidden layer is three layers counts 5 layers altogether of input in, and input layer 120+K nodes, 120 be row
For the input of sequence, k inputs for real vector, 120 node of intermediate hidden layer, and neuron is convolutional Neural member, output layer number etc.
In two level scene quantity)
Artificial constructed training sample:The behavior sequence and corresponding scene, labeled data amount of artificial mark user reaches pre-
Determine item number (for example, 10000 or 20000), is trained.
Training method uses the backpropagation training method of neural network.
Third walks:The corresponding feature vector of application scenarios information is estimated, the real-time of 5 layers of neural network row user behavior is passed through
Prediction obtains the estimate sequence information that result is scene.
3, it is carried out using distribution in conjunction with customized information and application scenarios information.
It is first random to recommend back-up needs in cold start-up (terminal initial start-up) stage after the identification of each application scenarios
App, accumulate training sample.After equal cold start-ups terminate, under each application scenarios, there is showing as positive sample for download,
That does not download shows as negative sample, using user's portrait information of user, utilizes and returns a point (Logistic
Regression) model be trained obtain the current application scenarios of instruction terminal application scenarios information you.Specific training
Flow is as follows:
Using the characteristic of user's portrait attribute and application scenarios information as the basis of training data, it is clear to carry out data
After a series of processing such as washing, it is trained using the Logistic regression models under spark, obtains applicable god
Through network.
On-line prediction part is divided into:Candidate App probably has certain number (for example, 1000);Then primary election is carried out, is obtained
Go out 200 App (200 applications can be alternative application);Then with train come Logistic regression models into
Row essence row, finally according to CVR*price into rearrangement (price is advertisement bid, it is therefore an objective to maximum revenue) from alternative application
In select intended application.
It can also classify in some embodiments to intended application, which can be the first of highest priority
Grade intended application, n-th grade of intended application of n-th grade of priority.Here n is positive integer less than N, the n and N;The N can
For total priority series.N is smaller in this embodiment, then priority is higher.The N can be the positive integer not less than 2.By upper
Stating operation realizes the combination of customized information and application scenarios information, obtains being best able to meet the current user demand of user,
And it is capable of the intended application of the distribution requirements of full marks hair application.
Example two:
As shown in figure 8, this example is based on above application distribution method, a specific example is provided, including:Processed offline portion
Divide and online processing part.
The processed offline part, including:
Data Layer obtains user's portrait attribute and the corresponding sample data of application scenarios information;
Model layer carries out data cleansing to user's portrait attribute and application scenarios information.Here data cleansing included
Filter the sample data that Characteristic Number is less than predetermined number.The data cleansing may also include:It filters out in all sample datas
Sample number including a feature is less than the feature of specified number.The confidence in sample data can be filtered out by data cleansing
Low dirty data is spent, to which training for promotion obtains the recognition accuracy of neural network.
Further include in model layer:Obtain characteristic module;The standardization statement shape of feature is stored in the feature templates
Original user feature in sample data can be passed through Feature Mapping, changed by formula according to the feature templates of acquisition in this example
At the standardized feature of the special statement of table conversion of the numeralization or stringification feature of characterization primitive character.
Further include in model layer:Sample form is obtained, record has how to arrange or store the sample in the sample form
The information format of notebook data.For example, providing a kind of training data format, the data format in the present embodiment, in sample form
Including:One or more data cells;Each data cell includes:Sample label, feature:Feature weight.Sample label pair
Answer recognition result;The features such as attribute and behavioral data that feature corresponds to the application scenarios information of front, user draws a portrait;Feature is weighed
Weight, is used to indicate disturbance degree of each feature to recognition result.Multiple data cells constitute feature vector successively.
Then, the standardized feature after obtaining sample characteristics and being turned according to feature is input to common interface, this is public to connect
Mouth is for by the sample form of sample data device to control, forming sample data file.
Further, the sample data file is subjected to Logistic regression processing, obtains can be used for knowing
Do not go out the model of current desired intended application.
Before carrying out model and reaching the standard grade, need through monitoring module using verify data to Logistic regression
The model that model and training in training obtain is verified and is calibrated, to ensure to train the accuracy of obtained model.
When carrying out the Logistic regression processing, Logistic can be provided by configuration module
Parameter needed for regression processing.
Final mask is reached the standard grade.
Online processing part includes:
Obtain the application message of candidate APP;
Primary election, it may include:Candidate APP is screened for the first time using application scenarios information, obtains alternative application
It is selected, it may include:The selected of alternative application is carried out using the methods of regression analysis, obtains intended application;
It retakes, it may include:Based on CVR and recommended parameter etc., minor sort again is carried out to the sequence of intended application, is determined excellent
The intended application first distributed or recommended;
In short, in online processing part, real-time keyword (Key Value, KV) storage, stream calculation and score online
To the comprehensive score of each alternative application, the intended application is finally determined according to comprehensive score.
The KV is stored as the various correlations for determining intended application such as storage user's portrait attribute, application scenarios information
The keyword of information, to reduce information storage.And real-time storage is used in this example, real-time storage here can be to use
Sufficiently small predetermined time interval carries out the storage of the keyword.
Here stream calculation may include:First data are stored, the analyzing processings such as analysis inquiry then are carried out to data,
Obtain handling result;The mode of this stream calculation has the characteristics that processing rhythm is compact and time delay is small.
Example three:
This example provides a kind of using diostribution device, it may include:
Fig. 9 is the structural schematic diagram of the mobile terminal of another embodiment of the present invention.Application diostribution device packet shown in Fig. 9
It includes:At least one processor 201, memory 202 and at least one network interface 204.Various components in mobile terminal 200 are logical
Bus system 205 is crossed to be coupled.It is understood that bus system 205 is for realizing the connection communication between these components.Bus
System 205 further includes power bus, controlling bus and status signal bus in addition in addition to including data/address bus.But in order to clear
For the sake of explanation, various buses are all designated as bus system 205 in the figure 7.
It is appreciated that the memory 202 in the embodiment of the present invention can be volatile memory or nonvolatile memory,
Both or may include volatile and non-volatile memory.Wherein, nonvolatile memory can be read-only memory (Read-
Only Memory, ROM), programmable read only memory (Programmable ROM, PROM), the read-only storage of erasable programmable
Device (Erasable PROM, EPROM), electrically erasable programmable read-only memory (Electrically EPROM, EEPROM) or
Flash memory.Volatile memory can be random access memory (Random Access Memory, RAM), be used as external high
Speed caching.By exemplary but be not restricted explanation, the RAM of many forms is available, such as static RAM
(Static RAM, SRAM), dynamic random access memory (Dynamic RAM, DRAM), Synchronous Dynamic Random Access Memory
(Synchronous DRAM, SDRAM), double data speed synchronous dynamic RAM (Double Data Rate
SDRAM, DDRSDRAM), enhanced Synchronous Dynamic Random Access Memory (Enhanced SDRAM, ESDRAM), synchronized links
Dynamic random access memory (Synchlink DRAM, SLDRAM) and direct rambus random access memory (Direct
Rambus RAM, DRRAM).The memory 202 of system and method described herein be intended to including but not limited to these and it is arbitrary its
It is suitble to the memory of type.
In some embodiments, memory 202 stores following element, executable modules or data structures, or
Their subset of person or their superset:Operating system 2021 and application program 2022.
Wherein, operating system 2021, including various system programs, such as ccf layer, core library layer, driving layer etc., are used for
Realize various basic businesses and the hardware based task of processing.Application program 2022, including various application programs, such as media
Player (Media Player), browser (Browser) etc., for realizing various applied business.Realize the embodiment of the present invention
The program of method may be embodied in application program 2022.
In embodiments of the present invention, by the program for calling memory 202 to store or instruction, specifically, can be application
The program stored in program 2022 or instruction, processor 201 can be used for executing answering for aforementioned any one embodiment of the method offer
With distribution method, specifically such as, at least executes Fig. 1, Fig. 4 and shown in fig. 5 apply distribution method.
In several embodiments provided herein, it should be understood that disclosed device and method can pass through it
Its mode is realized.Apparatus embodiments described above are merely indicative, for example, the division of the unit, only
A kind of division of logic function, formula that in actual implementation, there may be another division manner, such as:Multiple units or component can combine, or
It is desirably integrated into another system, or some features can be ignored or not executed.In addition, shown or discussed each composition portion
It can be the INDIRECT COUPLING by some interfaces, equipment or unit to divide mutual coupling or direct-coupling or communication connection
Or communication connection, can be electrical, mechanical or other forms.
The above-mentioned unit illustrated as separating component can be or may not be and be physically separated, aobvious as unit
The component shown can be or may not be physical unit, you can be located at a place, may be distributed over multiple network lists
In member;Some or all of wherein unit can be selected according to the actual needs to achieve the purpose of the solution of this embodiment.
In addition, each functional unit in various embodiments of the present invention can be fully integrated into a processing module, also may be used
It, can also be during two or more units be integrated in one unit to be each unit individually as a unit;It is above-mentioned
The form that hardware had both may be used in integrated unit is realized, can also be realized in the form of hardware adds SFU software functional unit.
One of ordinary skill in the art will appreciate that:Realize that all or part of step of above method embodiment can pass through
The relevant hardware of program instruction is completed, and program above-mentioned can be stored in a computer read/write memory medium, the program
When being executed, step including the steps of the foregoing method embodiments is executed;And storage medium above-mentioned includes:Movable storage device, ROM,
The various media that can store program code such as RAM, magnetic disc or CD.
The above description is merely a specific embodiment, but scope of protection of the present invention is not limited thereto, any
Those familiar with the art in the technical scope disclosed by the present invention, can easily think of the change or the replacement, and should all contain
Lid is within protection scope of the present invention.Therefore, protection scope of the present invention should be based on the protection scope of the described claims.
Claims (15)
1. a kind of applying distribution method, which is characterized in that including:
Obtain the current application scenarios information of terminal, wherein the application scenarios information, for reflecting that the user of the terminal works as
Preceding application demand;
Determine the intended application for meeting presupposed information matching relationship with the application scenarios information;
Determine that intended application distributes the display parameters at interface in application, wherein the display parameters include at least display location
Parameter;
According to the display parameters, the application message of the intended application is exported at application distribution interface.
2. according to the method described in claim 1, it is characterized in that,
The current application scenarios information of the acquisition terminal, including:
According to the currently used state of the current location information of the terminal, the schedule information of the terminal storage, the terminal
At least one in information and user's operation behavioral data, determines level-one scene information, wherein the level-one scene letter
Breath, including:According to default partition dimension using the scene information divided, the level-one scene information, be used to indicate described
The current desired corresponding level-one dimension of terminal.
3. according to the method described in claim 2, it is characterized in that,
It is described to obtain the current application scenarios information of terminal, further include:
According to the currently used state of the current location information of the terminal, the schedule information of the terminal storage, the terminal
At least one in information and user's operation behavioral data, determines two level scene information;Wherein, the two level scene letter
Breath includes:The description letter of the two level dimension of the corresponding level-one dimension of the level-one scene information of the two level scene information description
Breath;The two level scene information is used to indicate the corresponding two level dimension of the current desired application of the terminal.
4. according to the method described in claim 3, it is characterized in that,
It is described to obtain the current application scenarios information of terminal, further include:
According to the currently used state of the current location information of the terminal, the schedule information of the terminal storage, the terminal
At least one in information and user's operation behavioral data, determines three-level scene information, wherein the three-level scene letter
Breath includes:The description information of three-level dimension under the two level dimension is described.
5. method according to any one of claims 1 to 4, which is characterized in that
The method further includes:
Obtain predetermined information, wherein the predetermined information includes:The historical behavior data of user's operation terminal, current behavior number
According to and user draw a portrait attribute at least one;
The intended application for determining to meet presupposed information matching relationship with the application scenarios information, including:
Determine the intended application that presupposed information matching relationship is satisfied by with the application scenarios information and the predetermined information.
6. according to the method described in claim 5, it is characterized in that,
The method further includes:
When the alternative application for determining to be satisfied by presupposed information matching relationship with the application scenarios information and the predetermined information
Later, obtain the alternative application conversion ratio CVR and/or recommended parameter at least one;
According to the CVR and the recommended parameter, intended application is selected from the alternative application.
7. method according to any one of claims 1 to 4, which is characterized in that
The current application scenarios information of the acquisition terminal, including:
Obtain the user behavior data that user applies the terminal;
The user behavior data is generated into behavior five-tuple, wherein the five-tuple includes:Navigation patterns information clicks row
For information, search behavior information, downloads behavioural information and use switching behavior information.
8. method according to any one of claims 1 to 4, which is characterized in that
The intended application for determining to meet presupposed information matching relationship with the application scenarios information, including:
The input layer for the neural network that the application scenarios information input is trained using sample data;
The hidden layer of the neural network handles the scene information inputted from the input layer, from the neural network
Output layer exports the identification information of the intended application.
9. a kind of applying diostribution device, which is characterized in that including:
Acquiring unit, the application scenarios information current for obtaining terminal, wherein the application scenarios information, for reflecting
State the current application demand of the user of terminal;
First determination unit, the intended application for determining to meet presupposed information matching relationship with the application scenarios information;
Second determination unit, for determining intended application in the display parameters using distribution interface, wherein the display parameters
Including at least display location parameter;
Output unit, for according to the display parameters, the application letter of the intended application to be exported at application distribution interface
Breath.
10. device according to claim 9, which is characterized in that
The acquiring unit is specifically used for according to the current location information of the terminal, the schedule information of the terminal storage, institute
The currently used status information of terminal and at least one in user's operation behavioral data are stated, determines that level-one scene is believed
Breath, wherein the level-one scene information, including:It is carried out using the scene information divided, the level-one according to default partition dimension
Scene information is used to indicate the current desired corresponding level-one dimension of the terminal.
11. device according to claim 10, which is characterized in that
The acquiring unit is specifically used for according to the current location information of the terminal, the schedule information of the terminal storage, institute
The currently used status information of terminal and at least one in user's operation behavioral data are stated, determines that two level scene is believed
Breath;Wherein, the two level scene information includes:The corresponding level-one of the level-one scene information of the two level scene information description
The description information of the two level dimension of dimension;The two level scene information is used to indicate the current desired application of the terminal and corresponds to
Two level dimension.
12. according to the devices described in claim 11, which is characterized in that
The acquiring unit is specifically used for according to the current location information of the terminal, the schedule information of the terminal storage, institute
The currently used status information of terminal and at least one in user's operation behavioral data are stated, determines that three-level scene is believed
Breath, wherein the three-level scene information includes:The description information of three-level dimension under the two level dimension is described.
13. according to claim 10 to 12 any one of them device, which is characterized in that
The acquiring unit is additionally operable to obtain predetermined information, wherein the predetermined information includes:The history of user's operation terminal
At least one of behavioral data, current behavior data and user's portrait attribute;
First determination unit, it is default specifically for determining to be satisfied by with the application scenarios information and the predetermined information
The intended application of information matches relationship.
14. device according to claim 13, which is characterized in that
First determination unit is determined to be satisfied by with the application scenarios information and the predetermined information also particularly useful for working as
After the alternative application of presupposed information matching relationship, obtain the alternative application conversion ratio CVR and/or recommended parameter at least
One of them;According to the CVR and the recommended parameter, intended application is selected from the alternative application.
15. according to claim 9 to 12 any one of them device, which is characterized in that
First determination unit, specifically for the application scenarios information input is used the nerve that sample data is trained
The input layer of network;The hidden layer of the neural network handles the scene information inputted from the input layer, from described
The output layer of neural network exports the identification information of the intended application.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710234047.8A CN108694211B (en) | 2017-04-11 | 2017-04-11 | Application distribution method and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710234047.8A CN108694211B (en) | 2017-04-11 | 2017-04-11 | Application distribution method and device |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108694211A true CN108694211A (en) | 2018-10-23 |
CN108694211B CN108694211B (en) | 2023-05-12 |
Family
ID=63843381
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710234047.8A Active CN108694211B (en) | 2017-04-11 | 2017-04-11 | Application distribution method and device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108694211B (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109710853A (en) * | 2018-12-27 | 2019-05-03 | 深圳铁盒子文化科技发展有限公司 | A kind of artificial intelligence classified matching method and system |
CN110515901A (en) * | 2019-08-23 | 2019-11-29 | 掌阅科技股份有限公司 | Information-pushing method calculates equipment and computer storage medium |
CN111143424A (en) * | 2018-11-05 | 2020-05-12 | 百度在线网络技术(北京)有限公司 | Characteristic scene data mining method and device and terminal |
CN111339291A (en) * | 2018-12-18 | 2020-06-26 | 腾讯科技(深圳)有限公司 | Information display method and device and storage medium |
Citations (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1853212A (en) * | 2003-09-18 | 2006-10-25 | 株式会社查纳位资讯情报 | On-vehicle information terminal, summarized map creation device, summarized map creation method |
CN102300205A (en) * | 2010-08-10 | 2011-12-28 | 微软公司 | Publicity and transfer of mobile application program based on position and context |
CN103455559A (en) * | 2011-12-27 | 2013-12-18 | 北京奇虎科技有限公司 | Method and device for automatically recommending application |
CN103714084A (en) * | 2012-10-08 | 2014-04-09 | 腾讯科技(深圳)有限公司 | Method and device for recommending information |
CN103810030A (en) * | 2014-02-20 | 2014-05-21 | 北京奇虎科技有限公司 | Application recommendation method, device and system based on mobile terminal application market |
CN104158885A (en) * | 2014-08-21 | 2014-11-19 | 中南大学 | Method and system of streaming loading of application based on location information |
CN104221403A (en) * | 2012-03-24 | 2014-12-17 | 英特尔公司 | Location-based application recommendation |
CN104573054A (en) * | 2015-01-21 | 2015-04-29 | 杭州朗和科技有限公司 | Information pushing method and equipment |
CN105068869A (en) * | 2015-09-29 | 2015-11-18 | 北京网诺星云科技有限公司 | Method and device for pushing information in mobile terminal |
CN106021507A (en) * | 2016-05-20 | 2016-10-12 | 广东欧珀移动通信有限公司 | Application recommendation method and apparatus, and mobile terminal |
CN106126210A (en) * | 2016-06-15 | 2016-11-16 | 维沃移动通信有限公司 | A kind of application icon display packing and mobile terminal |
CN106484392A (en) * | 2016-09-08 | 2017-03-08 | 北京小米移动软件有限公司 | icon display method and device |
US20170215024A1 (en) * | 2016-01-25 | 2017-07-27 | Baidu Online Network Technology (Beijing) Co., Ltd . | Methods and apparatuses for recommending application and presenting recommended applications |
-
2017
- 2017-04-11 CN CN201710234047.8A patent/CN108694211B/en active Active
Patent Citations (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1853212A (en) * | 2003-09-18 | 2006-10-25 | 株式会社查纳位资讯情报 | On-vehicle information terminal, summarized map creation device, summarized map creation method |
CN102300205A (en) * | 2010-08-10 | 2011-12-28 | 微软公司 | Publicity and transfer of mobile application program based on position and context |
CN103455559A (en) * | 2011-12-27 | 2013-12-18 | 北京奇虎科技有限公司 | Method and device for automatically recommending application |
CN104221403A (en) * | 2012-03-24 | 2014-12-17 | 英特尔公司 | Location-based application recommendation |
CN103714084A (en) * | 2012-10-08 | 2014-04-09 | 腾讯科技(深圳)有限公司 | Method and device for recommending information |
CN103810030A (en) * | 2014-02-20 | 2014-05-21 | 北京奇虎科技有限公司 | Application recommendation method, device and system based on mobile terminal application market |
CN104158885A (en) * | 2014-08-21 | 2014-11-19 | 中南大学 | Method and system of streaming loading of application based on location information |
CN104573054A (en) * | 2015-01-21 | 2015-04-29 | 杭州朗和科技有限公司 | Information pushing method and equipment |
CN105068869A (en) * | 2015-09-29 | 2015-11-18 | 北京网诺星云科技有限公司 | Method and device for pushing information in mobile terminal |
US20170215024A1 (en) * | 2016-01-25 | 2017-07-27 | Baidu Online Network Technology (Beijing) Co., Ltd . | Methods and apparatuses for recommending application and presenting recommended applications |
CN106021507A (en) * | 2016-05-20 | 2016-10-12 | 广东欧珀移动通信有限公司 | Application recommendation method and apparatus, and mobile terminal |
CN106126210A (en) * | 2016-06-15 | 2016-11-16 | 维沃移动通信有限公司 | A kind of application icon display packing and mobile terminal |
CN106484392A (en) * | 2016-09-08 | 2017-03-08 | 北京小米移动软件有限公司 | icon display method and device |
Non-Patent Citations (1)
Title |
---|
周正: "基于LBS的移动环境中POI推送方法研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111143424A (en) * | 2018-11-05 | 2020-05-12 | 百度在线网络技术(北京)有限公司 | Characteristic scene data mining method and device and terminal |
CN111339291A (en) * | 2018-12-18 | 2020-06-26 | 腾讯科技(深圳)有限公司 | Information display method and device and storage medium |
CN109710853A (en) * | 2018-12-27 | 2019-05-03 | 深圳铁盒子文化科技发展有限公司 | A kind of artificial intelligence classified matching method and system |
CN109710853B (en) * | 2018-12-27 | 2021-05-04 | 深圳铁盒子文化科技发展有限公司 | Artificial intelligence classification matching method and system |
CN110515901A (en) * | 2019-08-23 | 2019-11-29 | 掌阅科技股份有限公司 | Information-pushing method calculates equipment and computer storage medium |
CN110515901B (en) * | 2019-08-23 | 2023-03-14 | 掌阅科技股份有限公司 | Information pushing method, computing device and computer storage medium |
Also Published As
Publication number | Publication date |
---|---|
CN108694211B (en) | 2023-05-12 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110046952A (en) | A kind of training method and device, a kind of recommended method and device of recommended models | |
CN109902708A (en) | A kind of recommended models training method and relevant apparatus | |
CN110162700A (en) | The training method of information recommendation and model, device, equipment and storage medium | |
CN110008399A (en) | A kind of training method and device, a kind of recommended method and device of recommended models | |
US20160342902A1 (en) | Interestingness recommendations in a computing advice facility | |
CN108694211A (en) | Using distribution method and device | |
CN108780518A (en) | Web interface generation and test are carried out using artificial neural network | |
CN108604222A (en) | System and method for deployment customized machine learning service | |
CN107958317A (en) | A kind of method and apparatus that crowdsourcing participant is chosen in crowdsourcing project | |
CN110378434A (en) | Training method, recommended method, device and the electronic equipment of clicking rate prediction model | |
CN107800801A (en) | A kind of pushing learning resource method and system for learning preference based on user | |
WO2012094519A1 (en) | Geographically localized recommendations in a computing advice facility | |
WO2015106353A1 (en) | Item classification method and selection system for electronic solicitation | |
WO2012094516A1 (en) | Interestingness recommendations in a computing advice facility | |
CN109903103B (en) | Method and device for recommending articles | |
CN108021708B (en) | Content recommendation method and device and computer readable storage medium | |
CN110147821A (en) | Targeted user population determines method, apparatus, computer equipment and storage medium | |
CN110033342A (en) | A kind of training method and device, a kind of recommended method and device of recommended models | |
CN109903087A (en) | The method, apparatus and storage medium of Behavior-based control feature prediction user property value | |
CN110263235A (en) | Information pushes object updating method, device and computer equipment | |
CN111949886B (en) | Sample data generation method and related device for information recommendation | |
CN105975537A (en) | Sorting method and device of application program | |
CN108564393A (en) | Potential customers' methods of marking, device and system | |
CN110163703A (en) | A kind of disaggregated model method for building up, official documents and correspondence method for pushing and server | |
CN108769440A (en) | Preposition shunt method and device |
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 |