CN106294743A - The recommendation method and device of application function - Google Patents
The recommendation method and device of application function Download PDFInfo
- Publication number
- CN106294743A CN106294743A CN201610652542.6A CN201610652542A CN106294743A CN 106294743 A CN106294743 A CN 106294743A CN 201610652542 A CN201610652542 A CN 201610652542A CN 106294743 A CN106294743 A CN 106294743A
- Authority
- CN
- China
- Prior art keywords
- function
- information
- training
- user
- application
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
Abstract
The invention discloses the recommendation method and device of a kind of application function.Wherein method includes: collect the training data comprising user profile and function information, extracts training characteristics from described training data;Using described training characteristics as training sample, the training algorithm preset is used to be trained obtaining recommended models;After receiving appointment user profile, the prediction of described recommended models is utilized to obtain to the function information specifying user to recommend;The function that the function information that prediction obtained is corresponding recommends described appointment user.Owing to recommended models utilizes substantial amounts of user profile and function information training to obtain, utilize this recommended models can determine should recommend which function to user according to the customized information of user, function is made to recommend more specific aim, improve the conversion ratio that function is recommended so that some New function recommend to obtain rapidly a large number of users by function.
Description
Technical field
The present invention relates to software technology field, be specifically related to the recommendation method and device of a kind of application function.
Background technology
Along with intelligent terminal's function from strength to strength, it has begun to break away from this tradition image of communication tool, increasingly
Abundant function, let us has enjoyed the advantage that intelligent terminal brings in life.The most present intelligent terminal
With former mobile phone is compared, it is simply that qualitative leap.If only paying attention to call function at mobile phone many years ago.That present intelligence
Can terminal seem, perfection be sloughed off and is become a capable helper serving life, and it even can mask the logical of basis
Telecommunication function and become a treasure case.Wherein, operating system based on intelligent terminal and the increasingly abundanter application developed to
Intelligent terminal adds diversiform function.
Apply available function the most for usual one, and along with the New function of the more new opplication offer of version is also got over
Come the most.But, being limited by intelligent terminal's display interface size, function the most frequently used for user can only be presented to use by application
Family, substantial amounts of function is the most idle, it is impossible to play its due effect.
Summary of the invention
In view of the above problems, it is proposed that the present invention in case provide one overcome the problems referred to above or at least in part solve on
State the recommendation method and device of the application function of problem.
According to an aspect of the invention, it is provided a kind of recommendation method of application function, including:
Collect the training data comprising user profile and function information, from described training data, extract training characteristics;
Using described training characteristics as training sample, the training algorithm preset is used to be trained obtaining recommended models;
After receiving appointment user profile, the prediction of described recommended models is utilized to obtain to the function specifying user to recommend
Information;
The function that the function information that prediction obtained is corresponding recommends described appointment user.
According to a further aspect in the invention, it is provided that the recommendation apparatus of a kind of application function, including:
Data collection module, is suitable to collect the training data comprising user profile and function information, from described training data
Middle extraction training characteristics;
Training module, is suitable to described training characteristics as training sample, uses the training algorithm preset to be trained
To recommended models;
Prediction module, is suitable to after receiving appointment user profile, utilizes the prediction of described recommended models to obtain to appointment
The function information that user recommends;
Recommending module, is suitable to function corresponding to the function information that prediction obtained and recommends described appointment user.
According to the recommendation method and device of the application function that the present embodiment provides, comprise user profile and function by collection
The training data of information, extracts and obtains training characteristics, uses the training algorithm preset to be trained obtaining recommended models;Receiving
After specifying user profile, recommended models prediction is utilized to obtain to the function information specifying user to recommend, merit prediction obtained
Function corresponding to information can recommend appointment user.Owing to recommended models is to utilize substantial amounts of user profile and function information training
Obtain, utilize this recommended models can determine should recommend which function to user according to the customized information of user so that merit
More specific aim can be recommended, improve the conversion ratio that function is recommended so that some New function recommend acquisition rapidly big by function
Amount user.
Described above is only the general introduction of technical solution of the present invention, in order to better understand the technological means of the present invention,
And can be practiced according to the content of description, and in order to allow above and other objects of the present invention, the feature and advantage can
Become apparent, below especially exemplified by the detailed description of the invention of the present invention.
Accompanying drawing explanation
By reading the detailed description of hereafter preferred implementation, various other advantage and benefit common for this area
Technical staff will be clear from understanding.Accompanying drawing is only used for illustrating the purpose of preferred implementation, and is not considered as the present invention
Restriction.And in whole accompanying drawing, it is denoted by the same reference numerals identical parts.In the accompanying drawings:
Fig. 1 shows the flow chart of the recommendation method of application function according to an embodiment of the invention;
Fig. 2 shows the flow chart of the recommendation method of application function in accordance with another embodiment of the present invention;
Fig. 3 shows the functional block diagram of the recommendation apparatus of application function according to an embodiment of the invention.
Detailed description of the invention
It is more fully described the exemplary embodiment of the disclosure below with reference to accompanying drawings.Although accompanying drawing shows the disclosure
Exemplary embodiment, it being understood, however, that may be realized in various forms the disclosure and should be by embodiments set forth here
Limited.On the contrary, it is provided that these embodiments are able to be best understood from the disclosure, and can be by the scope of the present disclosure
Complete conveys to those skilled in the art.
Fig. 1 shows the flow chart of the recommendation method of application function according to an embodiment of the invention.As it is shown in figure 1,
The method comprises the steps:
Step S101, collects the training data comprising user profile and function information, extracts training spy from training data
Levy.
Specify application for a, need the recommended models in advance this appointment applied to be trained in server side, profit
Would know that this kind of certain function specifying application or other associated application to provide should push away to which user by recommended models
Recommend.In the embodiment of the present invention, application refers to the application program installed in intelligent terminal, and function refers to the concrete of application offer
Function.As a example by mobile phone bodyguard, mobile phone bodyguard is a application, and the cleaning acceleration in mobile phone bodyguard, traffic monitoring, harassing and wrecking are blocked
Cut, function that virus killing etc. provides for mobile phone bodyguard.
First this step collects training data, and training data comprises user profile and function information, and user profile refers to
With user-dependent information, can specifically comprise personal attribute information, user device attribute information and/or application service condition letter
Breath.Wherein personal attribute information specifically includes the information such as sex, region, age, occupation, user preferences, and subscriber equipment attribute is believed
Breath specifically includes the facility informations such as intelligent terminal's brand, model, memory space, internal memory, and application service condition information specifically includes
The information such as the frequency of usage of application function, use time.Function information refers to the information relevant to function, can specifically comprise merit
Energy type information and/or function redirect mode information.Wherein functional type information is for example: cleaning class, virus killing class, tool-class,
The types such as amusement class, function redirects mode information for example: redirect mode, downloading mode, H5 page mode etc..
The information that user profile is a large number of users collected above, function information is also the information of a large amount of function.Wherein
Function is not limited only to be this most targeted a function specifying application to provide, it is also possible to be to specify application to be closed with this
The function that other application of connection provides.Such as, the embodiment of the present invention is directed to carry out function recommendation in mobile phone bodyguard, and it is pushed away
The function recommended is not limited only to the function that mobile phone bodyguard provides, it is also possible to comprise the function that the associated application such as mobile phone assistant provide.
Step S102, using training characteristics as training sample, uses the training algorithm preset to be trained obtaining recommending mould
Type.
The training characteristics that this step is extracted based on collected a large amount of training datas, uses the training algorithm preset to carry out
Training obtains recommended models, would know that certain money specifies certain function that application or other associated application provide by this recommended models
Should recommend to which user, specifically, would know that certain money specifies certain merit that application or other associated application provide
Can whether should recommend to certain user.
Step S103, after receiving appointment user profile, utilizes recommended models prediction to obtain to specifying user to recommend
Function information.
For specifying user, appointment user profile being input in recommended models, measurable obtaining should be recommended to this user
Function information, which function this function information describes meets and carries out, to this appointment user, the condition recommended.
Step S104, the function that the function information that prediction obtained is corresponding recommends appointment user.
Function information acquired in utilization finds qualified function, and backward appointment user recommends these functions.
The recommendation method of application function provided according to the present embodiment, comprises user profile and function information by collection
Training data, extracts and obtains training characteristics, uses the training algorithm preset to be trained obtaining recommended models;Receiving appointment
After user profile, recommended models prediction is utilized to obtain to the function information specifying user to recommend, function information prediction obtained
Corresponding function recommends appointment user.Owing to recommended models is to utilize substantial amounts of user profile and function information training to obtain
, utilize this recommended models can determine should recommend which function to user according to the customized information of user so that function pushes away
Recommend more specific aim, improve the conversion ratio that function is recommended so that some New function are recommended to obtain rapidly a large amount of using by function
Family.
Fig. 2 shows the flow chart of the recommendation method of application function in accordance with another embodiment of the present invention.Such as Fig. 2 institute
Showing, the method comprises the steps:
Step S201, collects the training data comprising user profile, function information and scene information, carries from training data
Take training characteristics.
Specify application for a, need the recommended models in advance this appointment applied to be trained in server side, profit
Would know that this kind of certain function specifying application or other associated application to provide should push away to which user by recommended models
Recommend.In the present embodiment, application refers to the application program installed in intelligent terminal, and function refers to the concrete merit that application provides
Energy.As a example by mobile phone bodyguard, mobile phone bodyguard is a application, cleaning acceleration in mobile phone bodyguard, traffic monitoring, harassing and wrecking intercept,
The function that virus killing etc. provides for mobile phone bodyguard.
First this step collects training data, and training data comprises user profile, function information and scene information, Yong Huxin
Breath refers to and user-dependent information, can specifically comprise personal attribute information, user device attribute information and/or application and use
Situation information.Wherein personal attribute information specifically includes the information such as sex, region, age, occupation, user preferences, subscriber equipment
Attribute information specifically includes the facility informations such as intelligent terminal's brand, model, memory space, internal memory, application service condition information tool
Body includes the information such as the frequency of usage of application function, use time.Function information refers to the information relevant to function, can be concrete
Comprise functional type information and/or function redirects mode information.Wherein functional type information is for example: cleaning class, virus killing class, work
The types such as tool class, amusement class, function redirects mode information for example: redirect mode, downloading mode, H5 page mode etc..Scene
Information includes recommending scene information and/or network environment information.Which kind of wherein recommend scene information to specifically describe to open up in the page
The function now recommended, alternatively, the page after the conventional func specifying application completes can be as the page of recommendation function.Lift
For example, the function items that in mobile phone bodyguard, cleaning, health check-up, virus killing etc. are conventional has one to complete the page, it is recommended that scene information can
It is described in the function completing to represent in the page recommendation of which function items.Network environment information specifically describes residing for intelligent terminal
Network environment, such as 2G, 3G, 4G or WiFi environment etc..
After collection obtains above-mentioned substantial amounts of training data, therefrom extract user characteristics, functional character and scene characteristic
Sample as follow-up training.
Step S202, carries out training characteristics manual crossover or cross processing automatically, obtains cross feature.
The training characteristics of above-mentioned three types is carried out cross processing by this step, obtains cross feature.Cross processing can be adopted
By manual mode or automated manner, the invention is not limited in this regard.
A kind of optional embodiment of the embodiment of the present invention is to use decision-tree model training characteristics automatically to be intersected
Process, obtain cross feature.Such as, GBDT (gradient promotes decision tree, Gradient Boosting Decision Tree) mould
Type is based on the boosting thought in integrated study, and each iteration all newly sets up a decision-making at the gradient direction reducing residual error
Tree, iteration how many times will generate how many decision trees.The thought of GBDT makes it have inherent advantage, has district it appeared that multiple
Dividing feature and the feature combination of property, the path of decision tree can use directly as the input feature vector of follow-up training pattern, save
Go the step that artificial searching feature, feature combine.
Step S203, uses the training algorithm preset to be trained cross feature, obtains recommended models.
A kind of optional embodiment of the embodiment of the present invention is to use binary logic to return (logistic recurrence) model
Training algorithm cross feature is trained, obtain recommended models.It is substantially linear regression that logistic returns, and simply exists
Feature adds a layer functions mapping in the mapping of result, i.e. first characteristic line is sued for peace, and then use function g (z) will be
For assuming that function is predicted.Successive value can be mapped on 0 and 1 by g (z).
Step S204, after receiving appointment user profile, utilizes recommended models prediction to obtain to specifying user to recommend
Function information and to the scene information specifying function corresponding to user's recommendation function information.
For specifying user, appointment user profile being input in recommended models, measurable obtaining should be recommended to this user
Function information, which function this function information describes meets and carries out, to this appointment user, the condition recommended.In addition, also
The measurable scene information obtained to appointment user's recommendation function, this scene information constrains to meet recommends merit to this appointment user
The scene condition of energy.
Step S205, represents configuration strategy according to default client and/or issues the merit that prediction is obtained by configuration strategy
Function corresponding to information can recommend appointment user.
This step to be accomplished that in the page specifying application, and the function that the function information that prediction obtained is corresponding pushes away
Recommend to specifying user.The function representing recommendation in which kind of page can be depending on scene information, that is, according to by scene information about
The scene condition of bundle, the function that the function information that prediction obtained is corresponding recommends appointment user.Specifically, scene information bag
Include recommendation scene information, it is recommended that scene information specifically describes the function representing recommendation in which kind of page, alternatively, answers specifying
Conventional func complete after the page can be as the page of recommendation function.As a example by mobile phone bodyguard, cleaning, health check-up, virus killing
Have one to complete the page Deng conventional function items, it is recommended that scene information can be described in which function items complete the page represents
The function recommended.
Alternatively, failing to prediction obtains not comprising in scene information, or scene information recommendation scene information, then
Scene configuration can be added so that client selects at which kind of page specifying application according to this scene configuration in issuing configuration strategy
Face represents the function of recommendation.
In the present embodiment, the functional packet that function information is corresponding is applied the function provided containing specifying and/or closes with specifying application
The function that other application of connection provides.Apply that is, the function recommended is not limited only to this appointment theing be the most targeted
The function provided, it is also possible to be the function specifying other application associated by application to provide with this.Such as, the embodiment of the present invention
Being directed to carry out function recommendation in mobile phone bodyguard, its function recommended is not limited only to the function that mobile phone bodyguard provides, and also may be used
To comprise the function that the associated application such as mobile phone assistant provide.
Further, this step also can carry out function recommendation according to default configuration strategy.Configuration strategy includes client
End represents configuration strategy and/or issues configuration strategy, and wherein issuing configuration strategy, to be that server side is pre-configured then issue
To client.Client represents configuration strategy and includes: client behavior configuration and/or the configuration of client presenting information.Client
Behavior configuration constraint function redirect mode, meeting the function of client behavior configuration just can be recommended and represent.Client
Which information is presenting information configuration constraint function present when representing.Issue configuration strategy to include: time configuration, region are joined
Put, priority configuration, scene configuration, application version configuration and/or the configuration of associated application version, these configurations constrain respectively and push away
Recommend the time of function, region, priority, scene, to the requirement of application version, requirement etc. to associated application version, function pushes away
Recommend and need to perform according to the constraint of configuration strategy.
The recommendation method of application function provided according to the present embodiment, comprises user profile and function information by collection
Training data, extracts and obtains training characteristics, uses the training algorithm preset to be trained obtaining recommended models;Receiving appointment
After user profile, recommended models prediction is utilized to obtain to the function information specifying user to recommend, function information prediction obtained
Corresponding function recommends appointment user.Owing to recommended models is to utilize substantial amounts of user profile and function information training to obtain
, utilize this recommended models can determine should recommend which function to user according to the customized information of user so that function pushes away
Recommend more specific aim, improve the conversion ratio that function is recommended so that some New function are recommended to obtain rapidly a large amount of using by function
Family.Further, utilize the most measurable scene information obtaining function recommendation of this method, select personalization for different users
Recommendation scene information, improve further user to function recommend the page clicking rate.It addition, by the page specifying application
Face is recommended the function of other associated application, substantial amounts of Adding User can be brought for other associated application, improve associated application
User coverage rate.
Fig. 3 shows the functional block diagram of the recommendation apparatus of application function according to an embodiment of the invention.Such as Fig. 3 institute
Showing, the recommendation apparatus of the application function that the present embodiment provides includes: data collection module 30, training module 31, it was predicted that module 32,
And recommending module 33.
Data collection module 30 is suitable to collect the training data comprising user profile and function information, carries from training data
Take training characteristics.
Specify application for a, need the recommended models in advance this appointment applied to be trained in server side, profit
Would know that this kind of certain function specifying application or other associated application to provide should push away to which user by recommended models
Recommend.In the present embodiment, application refers to the application program installed in intelligent terminal, and function refers to the concrete merit that application provides
Energy.As a example by mobile phone bodyguard, mobile phone bodyguard is a application, cleaning acceleration in mobile phone bodyguard, traffic monitoring, harassing and wrecking intercept,
The function that virus killing etc. provides for mobile phone bodyguard.
Data collection module 30 collects training data, and training data comprises user profile and function information, and user profile refers to
Be and user-dependent information, can specifically comprise personal attribute information, user device attribute information and/or application service condition
Information.Wherein personal attribute information specifically includes the information such as sex, region, age, occupation, user preferences, subscriber equipment attribute
Information specifically includes the facility informations such as intelligent terminal's brand, model, memory space, internal memory, and application service condition information is specifically wrapped
Include the information such as the frequency of usage of application function, use time.Function information refers to the information relevant to function, can specifically comprise
Functional type information and/or function redirect mode information.Wherein functional type information is for example: cleaning class, virus killing class, instrument
The types such as class, amusement class, function redirects mode information for example: redirect mode, downloading mode, H5 page mode etc..
Data collection module 30 is further adapted for: collect the training number comprising user profile, function information and scene information
According to;User characteristics, functional character and scene characteristic is extracted from training data.
Scene information includes recommending scene information and/or network environment information.Scene information is wherein recommended to specifically describe
Representing the function of recommendation in which kind of page, alternatively, the page after the conventional func specifying application completes can be as recommendation
The page of function.For example, the function items that in mobile phone bodyguard, cleaning, health check-up, virus killing etc. are conventional has one to complete the page, pushes away
Recommend scene information and can be described in the function completing to represent in the page recommendation of which function items.Network environment information specifically describes intelligence
Energy network environment residing for terminal, such as 2G, 3G, 4G or WiFi environment etc..
Training module 31 is suitable to training characteristics as training sample, uses the training algorithm preset to be trained being pushed away
Recommend model.
Further, training module 31 is suitable to: training characteristics carries out manual crossover or cross processing automatically, is intersected
Feature;Use the training algorithm preset that cross feature is trained, obtain recommended models.
Further, training module 31 is suitable to: uses decision-tree model that training characteristics is carried out automatic cross processing, obtains
Cross feature.
Training module 31 uses decision-tree model that training characteristics is carried out automatic cross processing, obtains cross feature.Such as,
GBDT (gradient promotes decision tree, Gradient Boosting Decision Tree) model is based in integrated study
Boosting thought, each iteration all reduce residual error gradient direction newly set up a decision tree, iteration how many times will be given birth to
Become how many decision trees.The thought of GBDT makes it have inherent advantage, it appeared that the multiple feature having distinction and feature
Combination, the path of decision tree can use directly as the input feature vector of follow-up training pattern, eliminate and manually find feature, spy
Levy the step of combination.
Further, training module 31 is suitable to: use the training algorithm of binary logic regression model to enter cross feature
Row training, obtains recommended models.It is substantially linear regression that logistic returns, and simply adds in feature to the mapping of result
One layer functions maps, and i.e. first characteristic line is sued for peace, and then uses function g (z) will assume that function is predicted the most.G (z) can
So that successive value is mapped on 0 and 1.
Prediction module 32 is suitable to after receiving appointment user profile, utilizes recommended models prediction to obtain to specifying user
The function information recommended.
Further, it was predicted that module 32 is further adapted for: utilize recommended models prediction to obtain to specifying user's recommendation function information
The scene information of corresponding function.
Recommending module 33 is suitable to function corresponding to the function information that prediction obtained and recommends appointment user.
Recommending module 33 is further adapted for: in the page specifying application, the merit that the function information that prediction obtained is corresponding
Appointment user can be recommended.Wherein, the functional packet that function information is corresponding contains the function specifying application to provide and/or applies with specifying
The function that other application of association provides.
Recommending module 33 is further adapted for: according to the scene condition retrained by scene information, function prediction obtained is believed
The function that breath is corresponding recommends appointment user.
The function representing recommendation in which kind of page can be depending on scene information, that is, recommending module 33 is according to by scene
Information constrained scene condition, the function that the function information that prediction obtained is corresponding recommends appointment user.Specifically, scene
Information includes recommending scene information, it is recommended that scene information specifically describes to represent the function of recommendation in which kind of page, alternatively,
The page after the conventional func specifying application completes can be as the page of recommendation function.As a example by mobile phone bodyguard, cleaning, body
The conventional function items such as inspection, virus killing has one to complete the page, it is recommended that what which function items scene information can be described in completes page
Face represents the function of recommendation.
Alternatively, failing to prediction obtains not comprising in scene information, or scene information recommendation scene information, then
Scene configuration can be added so that client selects at which kind of page specifying application according to this scene configuration in issuing configuration strategy
Face represents the function of recommendation.
Further, it is recommended that module 33 is suitable to: represent configuration strategy according to default client and/or issue configuration strategy
The function that the function information that prediction obtained is corresponding recommends appointment user.
Client represents configuration strategy and includes: client behavior configuration and/or the configuration of client presenting information.Client row
For configuration constraint function redirect mode, meeting the function of client behavior configuration just can be recommended and represent.Client exhibition
Which information is existing information configuration present when constraining function to represent.Issue configuration strategy to include: time configuration, region are joined
Put, priority configuration, scene configuration, application version configuration and/or the configuration of associated application version, these configurations constrain respectively and push away
Recommend the time of function, region, priority, scene, to the requirement of application version, requirement etc. to associated application version, function pushes away
Recommend and need to perform according to the constraint of configuration strategy.
The recommendation apparatus of application function provided according to the present embodiment, comprises user profile and function information by collection
Training data, extracts and obtains training characteristics, uses the training algorithm preset to be trained obtaining recommended models;Receiving appointment
After user profile, recommended models prediction is utilized to obtain to the function information specifying user to recommend, function information prediction obtained
Corresponding function recommends appointment user.Owing to recommended models is to utilize substantial amounts of user profile and function information training to obtain
, utilize this recommended models can determine should recommend which function to user according to the customized information of user so that function pushes away
Recommend more specific aim, improve the conversion ratio that function is recommended so that some New function are recommended to obtain rapidly a large amount of using by function
Family.Further, utilize the most measurable scene information obtaining function recommendation of this device, select personalization for different users
Recommendation scene information, improve further user to function recommend the page clicking rate.It addition, by the page specifying application
Face is recommended the function of other associated application, substantial amounts of Adding User can be brought for other associated application, improve associated application
User coverage rate.
Algorithm and display are not intrinsic to any certain computer, virtual system or miscellaneous equipment relevant provided herein.
Various general-purpose systems can also be used together with based on teaching in this.As described above, construct required by this kind of system
Structure be apparent from.Additionally, the present invention is also not for any certain programmed language.It is understood that, it is possible to use various
Programming language realizes the content of invention described herein, and the description done language-specific above is to disclose this
Bright preferred forms.
In description mentioned herein, illustrate a large amount of detail.It is to be appreciated, however, that the enforcement of the present invention
Example can be put into practice in the case of not having these details.In some instances, it is not shown specifically known method, structure
And technology, in order to do not obscure the understanding of this description.
Similarly, it will be appreciated that one or more in order to simplify that the disclosure helping understands in each inventive aspect, exist
Above in the description of the exemplary embodiment of the present invention, each feature of the present invention is grouped together into single enforcement sometimes
In example, figure or descriptions thereof.But, the method for the disclosure should not be construed to reflect an intention that i.e. required guarantor
The application claims feature more more than the feature being expressly recited in each claim protected.More precisely, as following
Claims reflected as, inventive aspect is all features less than single embodiment disclosed above.Therefore,
The claims following detailed description of the invention are thus expressly incorporated in this detailed description of the invention, the most each claim itself
All as the independent embodiment of the present invention.
Those skilled in the art are appreciated that and can carry out the module in the equipment in embodiment adaptively
Change and they are arranged in one or more equipment different from this embodiment.Can be the module in embodiment or list
Unit or assembly are combined into a module or unit or assembly, and can put them in addition multiple submodule or subelement or
Sub-component.In addition at least some in such feature and/or process or unit excludes each other, can use any
Combine all features disclosed in this specification (including adjoint claim, summary and accompanying drawing) and so disclosed appoint
Where method or all processes of equipment or unit are combined.Unless expressly stated otherwise, this specification (includes adjoint power
Profit requires, summary and accompanying drawing) disclosed in each feature can be carried out generation by providing identical, equivalent or the alternative features of similar purpose
Replace.
Although additionally, it will be appreciated by those of skill in the art that embodiments more described herein include other embodiments
Some feature included by rather than further feature, but the combination of the feature of different embodiment means to be in the present invention's
Within the scope of and form different embodiments.Such as, in the following claims, embodiment required for protection appoint
One of meaning can mode use in any combination.
The all parts embodiment of the present invention can realize with hardware, or to run on one or more processor
Software module realize, or with combinations thereof realize.It will be understood by those of skill in the art that and can use in practice
Microprocessor or digital signal processor (DSP) realize in the recommendation apparatus of application function according to embodiments of the present invention
The some or all functions of some or all parts.The present invention is also implemented as performing method as described herein
Part or all equipment or device program (such as, computer program and computer program).Such reality
The program of the existing present invention can store on a computer-readable medium, or can be to have the form of one or more signal.
Such signal can be downloaded from internet website and obtain, or provides on carrier signal, or with any other form
There is provided.
The present invention will be described rather than limits the invention to it should be noted above-described embodiment, and ability
Field technique personnel can design alternative embodiment without departing from the scope of the appended claims.In the claims,
Any reference marks that should not will be located between bracket is configured to limitations on claims.Word " comprises " and does not excludes the presence of not
Arrange element in the claims or step.Word "a" or "an" before being positioned at element does not excludes the presence of multiple such
Element.The present invention and can come real by means of including the hardware of some different elements by means of properly programmed computer
Existing.If in the unit claim listing equipment for drying, several in these devices can be by same hardware branch
Specifically embody.Word first, second and third use do not indicate that any order.These word explanations can be run after fame
Claim.
The invention discloses: A1, a kind of recommendation method of application function, including:
Collect the training data comprising user profile and function information, from described training data, extract training characteristics;
Using described training characteristics as training sample, the training algorithm preset is used to be trained obtaining recommended models;
After receiving appointment user profile, the prediction of described recommended models is utilized to obtain to the function specifying user to recommend
Information;
The function that the function information that prediction obtained is corresponding recommends described appointment user.
A2, according to the method described in A1, the function that described function information prediction obtained is corresponding recommends described appointment
User farther includes: in the page specifying application, and the function that the function information that prediction obtained is corresponding recommends described finger
Determine user.
A3, according to the method described in A2, functional packet corresponding to described function information is containing the described function specifying application to provide
And/or the function of other application offer with described appointment association.
A4, according to the method described in any one of A1-A3, collect training data and farther include: collect comprise user profile,
Function information and the training data of scene information;
Described training characteristics of extracting from training data farther includes: extract user characteristics, function from training data
Feature and scene characteristic.
A5, according to the method described in any one of A1-A4, described user profile includes: personal attribute information, subscriber equipment belong to
Property information and/or application service condition information;
Described function information includes: functional type information and/or function redirect mode information.
A6, according to the method described in A4, described scene information includes: recommend scene information and/or network environment information.
A7, according to the method described in A4 or A6, after receiving appointment user profile, described method also includes: utilize
The prediction of described recommended models obtains to the scene information specifying function corresponding to user's recommendation function information;
The function that described function information prediction obtained is corresponding is recommended described appointment user and is farther included: according to being subject to
The scene condition of described scene information constraint, the function that the function information that prediction obtained is corresponding recommends described appointment user.
A8, according to the method described in any one of A1-A7, described using training characteristics as training sample, use the instruction preset
White silk algorithm is trained obtaining recommended models and farther includes:
Training characteristics is carried out manual crossover or cross processing automatically, obtains cross feature;
Use the training algorithm preset that cross feature is trained, obtain recommended models.
A9, according to the method described in A8, described training characteristics carried out automatic cross processing farther include: use decision-making
Training characteristics is carried out automatic cross processing by tree-model, obtains cross feature.
A10, according to the method described in A8, cross feature is trained by the training algorithm that described employing is preset, and is pushed away
Recommend model to farther include: use the training algorithm of binary logic regression model that cross feature is trained, recommended
Model.
A11, according to the method described in any one of A1-A10, function corresponding to described function information prediction obtained is recommended
Farther include to described appointment user: represent according to default client and configure strategy and/or issue configuration strategy and will predict
The function that the function information that obtains is corresponding recommends described appointment user.
A12, according to the method described in A11, described client represent configuration strategy include: client behavior configuration and/or
Client presenting information configures.
A13, according to the method described in A11, described in issue configuration strategy include: the time configuration, region configuration, preferential grating
Put, application version configures and/or the configuration of associated application version.
The invention also discloses: B14, the recommendation apparatus of a kind of application function, including:
Data collection module, is suitable to collect the training data comprising user profile and function information, from described training data
Middle extraction training characteristics;
Training module, is suitable to described training characteristics as training sample, uses the training algorithm preset to be trained
To recommended models;
Prediction module, is suitable to after receiving appointment user profile, utilizes the prediction of described recommended models to obtain to appointment
The function information that user recommends;
Recommending module, is suitable to function corresponding to the function information that prediction obtained and recommends described appointment user.
B15, according to the device described in B14, described recommending module is further adapted for: in the page specifying application, will be pre-
The function that the function information that records is corresponding recommends described appointment user.
B16, according to the device described in B15, functional packet corresponding to described function information is containing the described merit specifying application to provide
The function that energy and/or other application with described appointment association provide.
B17, according to the device described in any one of B14-B16, described data collection module is further adapted for: collect comprise use
The training data of family information, function information and scene information;User characteristics, functional character and scene is extracted special from training data
Levy.
B18, according to the device described in any one of B14-B17, described user profile includes: personal attribute information, Yong Hushe
Standby attribute information and/or application service condition information;
Described function information includes: functional type information and/or function redirect mode information.
B19, according to the device described in B17, described scene information includes: recommend scene information and/or network environment information.
B20, according to the device described in B17 or B19, described prediction module is further adapted for: utilize described recommended models pre-
Record to the scene information specifying function corresponding to user's recommendation function information;
Described recommending module is further adapted for: according to the scene condition retrained by described scene information, prediction obtained
The function that function information is corresponding recommends described appointment user.
B21, according to the device described in any one of B14-B20, described training module is further adapted for:
Training characteristics is carried out manual crossover or cross processing automatically, obtains cross feature;
Use the training algorithm preset that cross feature is trained, obtain recommended models.
B22, according to the device described in B21, described training module is further adapted for: use decision-tree model by training characteristics
Carry out automatic cross processing, obtain cross feature.
B23, according to the device described in B21, described training module is further adapted for: use binary logic regression model
Cross feature is trained by training algorithm, obtains recommended models.
B24, according to the device described in any one of B14-B23, described recommending module is further adapted for: according to default client
Hold to represent to configure strategy and/or issue function corresponding to tactful function information prediction obtained of configuration and recommend described appointment use
Family.
B25, according to the device described in B24, described client represent configuration strategy include: client behavior configuration and/or
Client presenting information configures.
B26, according to the device described in B24, described in issue configuration strategy include: the time configuration, region configuration, preferential grating
Put, application version configures and/or the configuration of associated application version.
Claims (10)
1. a recommendation method for application function, including:
Collect the training data comprising user profile and function information, from described training data, extract training characteristics;
Using described training characteristics as training sample, the training algorithm preset is used to be trained obtaining recommended models;
After receiving appointment user profile, the prediction of described recommended models is utilized to obtain to the function letter specifying user to recommend
Breath;
The function that the function information that prediction obtained is corresponding recommends described appointment user.
Method the most according to claim 1, the function that described function information prediction obtained is corresponding recommends described finger
Determining user to farther include: in the page specifying application, the function that the function information that prediction obtained is corresponding is recommended described
Specify user.
Method the most according to claim 2, the functional packet that described function information is corresponding applies, containing described appointment, the merit provided
The function that energy and/or other application with described appointment association provide.
4. according to the method described in any one of claim 1-3, collect training data and farther include: collect and comprise user's letter
Breath, function information and the training data of scene information;
Described training characteristics of extracting from training data farther includes: extract user characteristics, functional character from training data
And scene characteristic.
5., according to the method described in any one of claim 1-4, described user profile includes: personal attribute information, subscriber equipment
Attribute information and/or application service condition information;
Described function information includes: functional type information and/or function redirect mode information.
Method the most according to claim 4, described scene information includes: recommend scene information and/or network environment information.
7., according to the method described in claim 4 or 6, after receiving appointment user profile, described method also includes: utilize
The prediction of described recommended models obtains to the scene information specifying function corresponding to user's recommendation function information;
The function that described function information prediction obtained is corresponding is recommended described appointment user and is farther included: according to by described
The scene condition of scene information constraint, the function that the function information that prediction obtained is corresponding recommends described appointment user.
8. according to the method described in any one of claim 1-7, described using training characteristics as training sample, use the instruction preset
White silk algorithm is trained obtaining recommended models and farther includes:
Training characteristics is carried out manual crossover or cross processing automatically, obtains cross feature;
Use the training algorithm preset that cross feature is trained, obtain recommended models.
Method the most according to claim 8, described carry out automatic cross processing by training characteristics and farther includes: use certainly
Training characteristics is carried out automatic cross processing by plan tree-model, obtains cross feature.
10. a recommendation apparatus for application function, including:
Data collection module, is suitable to collect the training data comprising user profile and function information, carries from described training data
Take training characteristics;
Training module, is suitable to described training characteristics as training sample, uses the training algorithm preset to be trained being pushed away
Recommend model;
Prediction module, is suitable to after receiving appointment user profile, utilizes the prediction of described recommended models to obtain to specifying user
The function information recommended;
Recommending module, is suitable to function corresponding to the function information that prediction obtained and recommends described appointment user.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610652542.6A CN106294743A (en) | 2016-08-10 | 2016-08-10 | The recommendation method and device of application function |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610652542.6A CN106294743A (en) | 2016-08-10 | 2016-08-10 | The recommendation method and device of application function |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106294743A true CN106294743A (en) | 2017-01-04 |
Family
ID=57668035
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610652542.6A Pending CN106294743A (en) | 2016-08-10 | 2016-08-10 | The recommendation method and device of application function |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106294743A (en) |
Cited By (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107766086A (en) * | 2017-10-19 | 2018-03-06 | 北京金山安全软件有限公司 | Function option display method and device, electronic equipment and readable storage medium |
CN107977072A (en) * | 2017-07-28 | 2018-05-01 | 北京物灵智能科技有限公司 | What a kind of robot used form method, forms expert system and electronic equipment |
CN108958826A (en) * | 2017-05-22 | 2018-12-07 | 北京京东尚科信息技术有限公司 | The method and apparatus of dynamic configuration application installation package |
CN109725957A (en) * | 2017-10-27 | 2019-05-07 | 钉钉控股(开曼)有限公司 | Group's configuration method and device, content share system |
CN109951318A (en) * | 2019-02-22 | 2019-06-28 | 珠海天燕科技有限公司 | The function configuration method and device of application |
WO2019128598A1 (en) * | 2017-12-29 | 2019-07-04 | Oppo广东移动通信有限公司 | Application processing method, electronic device, and computer readable storage medium |
CN110119474A (en) * | 2018-05-16 | 2019-08-13 | 华为技术有限公司 | Recommended models training method, the prediction technique based on recommended models and device |
CN110134040A (en) * | 2019-05-07 | 2019-08-16 | 上海电气集团股份有限公司 | The processing method and system of the operation data of industrial equipment |
CN110471735A (en) * | 2019-08-21 | 2019-11-19 | 中国工商银行股份有限公司 | Operational view interface method for pushing and device |
CN110522289A (en) * | 2018-05-25 | 2019-12-03 | 潘通 | A kind of control method of intelligent electric cooker, device and intelligent electric cooker |
CN110569428A (en) * | 2019-08-08 | 2019-12-13 | 阿里巴巴集团控股有限公司 | recommendation model construction method, device and equipment |
CN110895721A (en) * | 2018-09-12 | 2020-03-20 | 珠海格力电器股份有限公司 | Method and device for predicting electric appliance function |
WO2020082855A1 (en) * | 2018-10-25 | 2020-04-30 | 阿里巴巴集团控股有限公司 | Service recommendation method, apparatus and device |
CN111798259A (en) * | 2019-04-09 | 2020-10-20 | Oppo广东移动通信有限公司 | Application recommendation method and device, storage medium and electronic equipment |
CN112506406A (en) * | 2020-12-03 | 2021-03-16 | 中国地质大学(北京) | Target control method and system based on user habit self-learning control |
Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103593434A (en) * | 2013-11-12 | 2014-02-19 | 北京奇虎科技有限公司 | Application recommendation method and device and server equipment |
CN103941853A (en) * | 2013-01-22 | 2014-07-23 | 三星电子株式会社 | Electronic device for determining emotion of user and method for determining emotion of user |
CN103995704A (en) * | 2014-06-03 | 2014-08-20 | 北京金山安全软件有限公司 | Function providing method and device for application program |
CN104750807A (en) * | 2015-03-25 | 2015-07-01 | 百度在线网络技术(北京)有限公司 | Application recommending method and device |
US9086783B2 (en) * | 2011-09-30 | 2015-07-21 | Oracle International Corporation | Methods and systems for using search and select forms on a mobile device |
CN104933075A (en) * | 2014-03-20 | 2015-09-23 | 百度在线网络技术(北京)有限公司 | User attribute predicting platform and method |
CN105160548A (en) * | 2015-08-20 | 2015-12-16 | 北京奇虎科技有限公司 | Method and apparatus for predicting advertisement click-through rate |
CN105631707A (en) * | 2015-12-23 | 2016-06-01 | 北京奇虎科技有限公司 | Advertisement click rate estimation method based on decision tree, application recommendation method and device |
CN105786993A (en) * | 2016-02-17 | 2016-07-20 | 北京奇虎科技有限公司 | Function plug-in recommending method and device of application |
-
2016
- 2016-08-10 CN CN201610652542.6A patent/CN106294743A/en active Pending
Patent Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9086783B2 (en) * | 2011-09-30 | 2015-07-21 | Oracle International Corporation | Methods and systems for using search and select forms on a mobile device |
CN103941853A (en) * | 2013-01-22 | 2014-07-23 | 三星电子株式会社 | Electronic device for determining emotion of user and method for determining emotion of user |
CN103593434A (en) * | 2013-11-12 | 2014-02-19 | 北京奇虎科技有限公司 | Application recommendation method and device and server equipment |
CN104933075A (en) * | 2014-03-20 | 2015-09-23 | 百度在线网络技术(北京)有限公司 | User attribute predicting platform and method |
CN103995704A (en) * | 2014-06-03 | 2014-08-20 | 北京金山安全软件有限公司 | Function providing method and device for application program |
CN104750807A (en) * | 2015-03-25 | 2015-07-01 | 百度在线网络技术(北京)有限公司 | Application recommending method and device |
CN105160548A (en) * | 2015-08-20 | 2015-12-16 | 北京奇虎科技有限公司 | Method and apparatus for predicting advertisement click-through rate |
CN105631707A (en) * | 2015-12-23 | 2016-06-01 | 北京奇虎科技有限公司 | Advertisement click rate estimation method based on decision tree, application recommendation method and device |
CN105786993A (en) * | 2016-02-17 | 2016-07-20 | 北京奇虎科技有限公司 | Function plug-in recommending method and device of application |
Cited By (22)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108958826A (en) * | 2017-05-22 | 2018-12-07 | 北京京东尚科信息技术有限公司 | The method and apparatus of dynamic configuration application installation package |
CN108958826B (en) * | 2017-05-22 | 2022-06-07 | 北京京东尚科信息技术有限公司 | Method and device for dynamically configuring application installation package |
CN107977072A (en) * | 2017-07-28 | 2018-05-01 | 北京物灵智能科技有限公司 | What a kind of robot used form method, forms expert system and electronic equipment |
CN107977072B (en) * | 2017-07-28 | 2021-06-08 | 北京物灵智能科技有限公司 | Formation method for robot, formation expert system and electronic equipment |
CN107766086A (en) * | 2017-10-19 | 2018-03-06 | 北京金山安全软件有限公司 | Function option display method and device, electronic equipment and readable storage medium |
CN107766086B (en) * | 2017-10-19 | 2020-11-13 | 北京金山安全软件有限公司 | Function option display method and device, electronic equipment and readable storage medium |
CN109725957A (en) * | 2017-10-27 | 2019-05-07 | 钉钉控股(开曼)有限公司 | Group's configuration method and device, content share system |
WO2019128598A1 (en) * | 2017-12-29 | 2019-07-04 | Oppo广东移动通信有限公司 | Application processing method, electronic device, and computer readable storage medium |
CN110119474A (en) * | 2018-05-16 | 2019-08-13 | 华为技术有限公司 | Recommended models training method, the prediction technique based on recommended models and device |
CN110119474B (en) * | 2018-05-16 | 2022-10-18 | 华为技术有限公司 | Recommendation model training method, prediction method and device based on recommendation model |
CN110522289A (en) * | 2018-05-25 | 2019-12-03 | 潘通 | A kind of control method of intelligent electric cooker, device and intelligent electric cooker |
CN110895721A (en) * | 2018-09-12 | 2020-03-20 | 珠海格力电器股份有限公司 | Method and device for predicting electric appliance function |
CN110895721B (en) * | 2018-09-12 | 2021-11-16 | 珠海格力电器股份有限公司 | Method and device for predicting electric appliance function |
WO2020082855A1 (en) * | 2018-10-25 | 2020-04-30 | 阿里巴巴集团控股有限公司 | Service recommendation method, apparatus and device |
CN109951318A (en) * | 2019-02-22 | 2019-06-28 | 珠海天燕科技有限公司 | The function configuration method and device of application |
CN111798259A (en) * | 2019-04-09 | 2020-10-20 | Oppo广东移动通信有限公司 | Application recommendation method and device, storage medium and electronic equipment |
CN110134040A (en) * | 2019-05-07 | 2019-08-16 | 上海电气集团股份有限公司 | The processing method and system of the operation data of industrial equipment |
CN110134040B (en) * | 2019-05-07 | 2021-09-24 | 上海电气集团股份有限公司 | Method and system for processing operation data of industrial equipment |
CN110569428A (en) * | 2019-08-08 | 2019-12-13 | 阿里巴巴集团控股有限公司 | recommendation model construction method, device and equipment |
CN110569428B (en) * | 2019-08-08 | 2023-10-20 | 创新先进技术有限公司 | Recommendation model construction method, device and equipment |
CN110471735A (en) * | 2019-08-21 | 2019-11-19 | 中国工商银行股份有限公司 | Operational view interface method for pushing and device |
CN112506406A (en) * | 2020-12-03 | 2021-03-16 | 中国地质大学(北京) | Target control method and system based on user habit self-learning control |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106294743A (en) | The recommendation method and device of application function | |
CN104270429B (en) | A kind of method and apparatus for pushing VAS application -to-terminal service | |
US9996531B1 (en) | Conversational understanding | |
CN109074514A (en) | Pass through the deep learning of the robot of example and experience | |
CN106997357A (en) | A kind of message treatment method, apparatus and system | |
CN112036577B (en) | Method and device for applying machine learning based on data form and electronic equipment | |
CN110782318A (en) | Marketing method and device based on audio interaction and storage medium | |
CN104679532B (en) | kernel module loading method and device | |
CN108595497A (en) | Data screening method, apparatus and terminal | |
CN110442737A (en) | The twin method and system of number based on chart database | |
CN108804188A (en) | Changing an interface skin method and device | |
CN105975537A (en) | Sorting method and device of application program | |
CN109828906A (en) | UI automated testing method, device, electronic equipment and storage medium | |
JP2018014765A5 (en) | ||
CN109947399A (en) | Code structure generation method, device, computer installation and readable storage medium storing program for executing | |
CN108171528A (en) | A kind of attribution method and attribution system | |
CN109815154A (en) | A kind of test method, device, system and medium | |
CN109543734A (en) | User portrait method and device, storage medium | |
CN110515896A (en) | Model resource management method, model file production method, device and system | |
CN110069676A (en) | Keyword recommendation method and device | |
CN110515986A (en) | A kind of processing method of social network diagram, device and storage medium | |
US10762089B2 (en) | Open ended question identification for investigations | |
CN106843984A (en) | The update method and device of application database | |
CN110858160A (en) | Resource scheduling method and device, storage medium and processor | |
CN106878562A (en) | Control the method and device of wifi functional switches |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170104 |
|
RJ01 | Rejection of invention patent application after publication |