CN108683734A - Category method for pushing, device and storage device, computer equipment - Google Patents
Category method for pushing, device and storage device, computer equipment Download PDFInfo
- Publication number
- CN108683734A CN108683734A CN201810460322.2A CN201810460322A CN108683734A CN 108683734 A CN108683734 A CN 108683734A CN 201810460322 A CN201810460322 A CN 201810460322A CN 108683734 A CN108683734 A CN 108683734A
- Authority
- CN
- China
- Prior art keywords
- category
- pushed
- score value
- value
- live streaming
- 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
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/50—Network services
- H04L67/55—Push-based network services
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Information Transfer Between Computers (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
A kind of category method for pushing of present invention offer, device and storage device, computer equipment.This method includes:Obtain the characteristic information data of category to be pushed;The characteristic information data is subjected to Feature Conversion, obtains feature vector;According to default score value operational model, the corresponding score value of described eigenvector is calculated, using the score value of score value category to be pushed as described in;Wherein, the default score value operational model is used to characterize the incidence relation of described eigenvector and the category score value to be pushed;Category is pushed according to the score value of the category to be pushed.Above-mentioned category method for pushing can match corresponding score value according to the hobby of user, the corresponding category to be pushed of matching score value be pushed to user, to improve the competitiveness of product in market.
Description
Technical field
The present invention relates to Internet technical fields, specifically, the present invention relates to a kind of category method for pushing, device and depositing
Store up equipment, computer equipment.
Background technology
With the rapid development of Internet, various categories emerge one after another.User can be by downloading corresponding APP client
End, searches for corresponding category.In order to improve the market competitiveness, category operator pushes various product to subscription client
Class.For example, the category of live streaming type.Live streaming operator can constantly push various users according to live content to subscription client can
The interested live streaming of energy improves product competitiveness in the market to attract more users.
However, with category increase and the variation of user demand, how to subscription client push meet user's need
The category asked, the problem of becoming urgent need to resolve at present.Such as, the category that type is broadcast live includes game live streaming category, and product are broadcast live in diet
Class, amusement live streaming category and outdoor activities live streaming category etc..How could attract more to live streaming client push category
User becomes the big problem that live streaming market competitiveness needs solve that improves.
Invention content
The purpose of the present invention is intended to provide a kind of category method for pushing, device and storage device, computer equipment, with push
The interested category of user improves the market competitiveness.
The present invention provides following scheme:
A kind of category method for pushing, includes the following steps:Obtain the characteristic information data of category to be pushed;By the feature
Information data carries out Feature Conversion, obtains feature vector;According to default score value operational model, it is corresponding to calculate described eigenvector
Score value, using the score value of score value category to be pushed as described in;Wherein, the default score value operational model is described for characterizing
The incidence relation of feature vector and the category score value to be pushed;Category is pushed according to the score value of the category to be pushed.
The characteristic information data is included in the precedence data using store, creation time in one of the embodiments,
With the Temporal-distance value and volumes of searches of current time;It is described that multiple characteristic information datas are subjected to Feature Conversion respectively,
Corresponding multidimensional characteristic vectors are obtained, including:Respectively by the precedence data in application store, creation time and current time
Temporal-distance value and volumes of searches be normalized, obtain the data after normalized;According to least square method pair
Described search amount is handled, and the variation tendency value of described search amount is obtained;According to after the normalized data and institute
Volumes of searches variation tendency value is stated, the corresponding multidimensional characteristic vectors are obtained.
The category to be pushed is live streaming category in one of the embodiments,;The characteristic information data, which is included in, answers
With the precedence data in store, the Temporal-distance value and volumes of searches of creation time and current time, start broadcasting main broadcaster's number;Described point
Multiple characteristic information datas are not subjected to Feature Conversion, obtain corresponding multidimensional characteristic vectors, including:Respectively will it is described
Using the precedence data in store, the Temporal-distance value and volumes of searches of creation time and current time, the main broadcaster's number that starts broadcasting are returned
One change is handled, and obtains the data after normalized;According to least square method respectively to described search amount and the master that starts broadcasting
It broadcasts number to be handled, obtains the variation tendency value of the variation tendency value and the main broadcaster's number that starts broadcasting of described search amount;According to described
The variation tendency value of the variation tendency value of data, described search amount after normalized and the main broadcaster's number that starts broadcasting, obtains institute
State corresponding multidimensional characteristic vectors.
The score value of category to be pushed described in the basis pushes category in one of the embodiments, including:Obtain first
In preset time period, the score value of multiple categories to be pushed;Push score value reaches the category to be pushed in default score range;Or it presses
The multiple category to be pushed is ranked up from big to small according to score value, obtains the preceding preset quantity category to be pushed that sorts,
Push preset quantity category to be pushed.
The push score value reaches the category to be pushed in default score range in one of the embodiments, including:It obtains
Take the quantity for the category to be pushed that score value reaches in default score range;Judge whether the quantity is less than preset quantity;If so,
The difference for calculating preset quantity and the quantity, using difference as quantity to be obtained;It is obtained from the first preset category list and waits obtaining
The category of quantity;The score value is pushed to reach the category to be pushed in default score range and arrange from the described first preset category
The category of the quantity to be obtained obtained in table;If it is not, pushing the score value reaches category to be pushed in default score range.
The push score value reaches the category to be pushed in default score range in one of the embodiments, including:It obtains
Score value in the second preset category list is taken to reach the category in default score range;Push score value reaches in default score range
Score value reaches the category in default score range in category to be pushed and the second preset category list.
The category to be pushed is live streaming category in one of the embodiments,;The push score value reaches default score value
Category to be pushed in range, including:It obtains score value and reaches the live streaming category in default score range in the second preset time period
Interior minimum main broadcaster's number and/or the minimum popularity number of starting broadcasting;It pushes the minimum main broadcaster's number that starts broadcasting and is more than the first preset value and/or minimum people
Destiny is more than the live streaming category of the second preset value.
The category to be pushed is live streaming category in one of the embodiments,;Category to be pushed described in the basis
Score value pushes category, including:Obtain the score value of multiple live streaming categories;It is issued according to the score value determination of the multiple live streaming category
Category is broadcast live;The viewing number of users of the live streaming category issued described in acquisition;According to the viewing user of the live streaming category issued
Number push live streaming category.
The score value of category to be pushed described in the basis pushes category in one of the embodiments, including:According to described
The score value of category to be pushed obtains category;Obtain position, title and figure that the category of the acquisition is shown in APP client homepages
Mark;The position, title and icon are sent to APP clients.
A kind of category pusher, including:First acquisition module, the characteristic information data for obtaining category to be pushed;
Second acquisition module obtains feature vector for the characteristic information data to be carried out Feature Conversion;Computing module is used for root
According to default score value operational model, the corresponding score value of described eigenvector is calculated, using score value category to be pushed as described in
Score value;Wherein, the default score value operational model is used to characterize being associated with for described eigenvector and the category score value to be pushed
Relationship;Pushing module, the score value for the category to be pushed according to push category.
A kind of storage device is stored thereon with a plurality of instruction;Described instruction is suitable for being loaded by processor and executing above-mentioned
Category method for pushing described in one embodiment.
A kind of computer equipment comprising:One or more processors;Memory;One or more application program, wherein
One or more of application programs are stored in the memory and are configured as being held by one or more of processors
Row, one or more of application programs are configured to carry out the category method for pushing according to any of the above-described embodiment.
Compared with prior art, the solution of the present invention has the following advantages:
A kind of category method for pushing provided by the invention, obtains the characteristic information data of category to be pushed, and will wait pushing
The characteristic information data of category is converted into feature vector, and point of the category to be pushed can be obtained according to default score value operational model
Value, to push category to subscription client according to the score value of category to be pushed.Wherein, default score value operational model, which characterizes, waits for
Push the incidence relation of the feature vector of category and the score value of final output.It that is to say, believed by the feature of each category to be pushed
The feature vector of breath data conversion can obtain one and the category characteristic information pair to be pushed according to default score value operational model
The score value answered.Therefore, corresponding score value can be matched according to the hobby of user, pushing matching score value to user corresponds to
Category to be pushed, to improve the competitiveness of product in market.
Further, the characteristic information data of category to be pushed is multiple, by the feature feature that obtains later of conversion to
Amount is multidimensional characteristic vectors, and the score value for the category to be pushed that final operation obtains is joined with reference to multiple correlations of category to be pushed
Number.This improves the reference values of the score value of category to be pushed so that the category of push is more accurate, more meets user's
Demand.
Preferably, category to be pushed is live streaming category, and characteristic information data is included in the precedence data using store, creates
The Temporal-distance value and volumes of searches of time and current time, start broadcasting main broadcaster's number.The present invention will be respectively by aforementioned four reference number
According to being normalized, the data after normalized are obtained.Meanwhile according to least square method respectively to above-mentioned volumes of searches and
The main broadcaster's number that starts broadcasting is handled, and obtains corresponding variation tendency value.Further according to after normalized data and variation become
Gesture value obtains the corresponding multidimensional characteristic vectors.It that is to say, be pushed to the live streaming category of live streaming client with reference to live streaming
Precedence data, the main broadcaster's number that starts broadcasting, volumes of searches and the creation time that category is broadcast live, to ensure that the live streaming category of push more accords with
Be fated the demand of lower user, to improve the competitiveness of product in market.
Description of the drawings
Above-mentioned and/or additional aspect and advantage of the invention will become from the following description of the accompanying drawings of embodiments
Obviously and it is readily appreciated that, wherein:
Interaction schematic diagrames of the Fig. 1 between one embodiment of the invention server provided and client;
Fig. 2 is the flow chart in an a kind of embodiment of category method for pushing provided by the invention;
Fig. 3 is the flow chart in an embodiment of step S200 provided by the invention;
Fig. 4 is the flow chart in a kind of another embodiment of category method for pushing provided by the invention;
Fig. 5 is the flow chart in a kind of another embodiment of category method for pushing provided by the invention;
Fig. 6 is the flow chart in a kind of another embodiment of category method for pushing provided by the invention;
Fig. 7 is the flow chart in a kind of another embodiment of category method for pushing provided by the invention;
Fig. 8 is the flow chart in a kind of another embodiment of category method for pushing provided by the invention;
Fig. 9 is a kind of implementing procedure frame diagram of the category method for pushing provided by the invention in an application environment;
Figure 10 is configured list figure in operation module 113 provided by the invention;
Figure 11 is the structure diagram in an a kind of embodiment of category pusher provided by the invention;
Figure 12 is the schematic diagram in one embodiment of computer equipment structure provided by the invention.
Specific implementation mode
The embodiment of the present invention is described below in detail, examples of the embodiments are shown in the accompanying drawings, wherein from beginning to end
Same or similar label indicates same or similar element or element with the same or similar functions.Below with reference to attached
The embodiment of figure description is exemplary, and is only used for explaining the present invention, and is not construed as limiting the claims.
Those skilled in the art of the present technique are appreciated that unless expressly stated, singulative " one " used herein, " one
It is a ", " described " and "the" may also comprise plural form.It is to be further understood that is used in the specification of the present invention arranges
Diction " comprising " refer to there are the feature, integer, step, operation, but it is not excluded that presence or addition it is one or more other
Feature, integer, step, operation.
Those skilled in the art of the present technique are appreciated that unless otherwise defined, all terms used herein (including technology art
Language and scientific terminology), there is meaning identical with the general understanding of the those of ordinary skill in fields of the present invention.Should also
Understand, those terms such as defined in the general dictionary, it should be understood that have in the context of the prior art
The consistent meaning of meaning, and unless by specific definitions as here, the meaning of idealization or too formal otherwise will not be used
To explain.
It will be appreciated by those skilled in the art that so-called " application " of the invention, " application program ", " application software " and class
It is the same concept well known to those skilled in the art like the concept of statement, refers to being instructed by series of computation machine and related data
The computer software for being suitable for electronics operation of the organic construction of resource.Unless specified, this name itself is not by programming language
Type, rank, the operating system of operation of also not rely by it or platform are limited.In the nature of things, this genus also not by appoint
The terminal of what form is limited.
A kind of category method for pushing provided by the invention is suitable for various types of content platforms with category push.
In one embodiment, which is applied in application environment as shown in Figure 1.
As shown in Figure 1, server 100 is located at user terminal 300 in 200 environment of the same network, server 100 and use
Family terminal 300 carries out the interaction of data information by network 200.Server 100 and the quantity of user terminal 300 are not construed as limiting,
It is only used as illustrating shown in Fig. 1.Client is installed, client is third-party application software, is such as broadcast live in user terminal 300
Client etc..User can carry out information exchange by the client end AP P in user terminal 300 with corresponding server 100.Visitor
Family end is corresponding with server (Server) end, follows same set of data protocol jointly so that server end can with client
The data for parsing other side mutually push category for user.
Server 100 may be, but not limited to, network server, management server, apps server, database
Server, cloud server etc..User terminal 300 may be, but not limited to, smart mobile phone, PC (personal
Computer, PC), tablet computer, personal digital assistant (personal digital assistant, PDA), mobile Internet access set
Standby (mobile Internet device, MID) etc..The operating system of user terminal 300 may be, but not limited to, Android
(Android) system, IOS (iPhone operating system) system, Windows phone systems, Windows systems
Deng.
In embodiments of the present invention, server 100 can wait in the client by the acquisition user terminal 300 of network 200
The correlated characteristic information data for pushing category can also be directly to obtain category to be pushed in being locally stored of server 100
Correlated characteristic information data.Meanwhile the category of push is issued to the visitor of user terminal 300 by server 100 by network 200
Family end.
A kind of category method for pushing provided by the invention is applied to server end.In one embodiment, as shown in Fig. 2, originally
A kind of category method for pushing of invention includes the following steps:
S100 obtains the characteristic information data of category to be pushed.
In the present embodiment, server obtains the characteristic information data of category to be pushed.Category to be pushed can be live streaming
Category, short-sighted frequency category, APP categories etc..The characteristic information data of category to be pushed can be:When pushing the establishment of category
Between, precedence data of the category to be pushed on operation platform, the popularity of category to be pushed, the volumes of searches of category to be pushed
Deng.Wherein, the characteristic information data of category to be pushed can be one or more.In order to improve the accuracy of push category,
Multiple characteristic information datas of category to be pushed can be acquired.Multiple characteristic information datas herein can be and category to be pushed
Characteristic and the relevant characteristic information data of interest-degree with user, so that the subsequently characteristic information number based on category to be pushed
The category can be more embodied according to the score value of acquisition to be worth the push of user group.
The characteristic information data is carried out Feature Conversion, obtains feature vector by S200.
Further, the characteristic information data of category to be pushed is carried out Feature Conversion by server, waits pushing away to obtain this
Send the feature vector of category.The technical implementation way of Feature Conversion can be, to characteristic information data carry out linear process or
Nonlinear Processing, or linear process is carried out to partial data in characteristic information data, other part data carry out non-linear place
Reason.Specific Feature Conversion mode can be characteristic information data is normalized, and/or, according to least square method
Characteristic information data is handled.
In the present embodiment, the characteristic information data of category to be pushed is multiple.Multiple characteristic information datas are carried out special
After sign conversion, multidimensional characteristic vectors are obtained.Specifically, server carries out Feature Conversion to each characteristic information data respectively, obtains
To corresponding multidimensional characteristic vectors.Wherein, in multidimensional characteristic vectors, each element corresponds to a characteristic information data and carries out feature
Numerical value after conversion.
In one embodiment, the characteristic information data of category to be pushed is included in the precedence data using store, when establishment
Between Temporal-distance value and volumes of searches with current time.As shown in figure 3, step S200 includes the following steps:
S210, respectively by it is described application store precedence data, the Temporal-distance value of creation time and current time, with
And volumes of searches is normalized, and obtains the data after normalized.
In the present embodiment, server obtains the category searching times in preset time period.It specifically can be according to preset time
The search term of user, the volumes of searches of the category to be recommended obtained from the background with Keywords matching in section.Server treats push product
The normalized mode of the volumes of searches of class can be:The category for obtaining preset quantity through the above way (including waits for push product
Class) in each category volumes of searches, then the average value of volumes of searches in the category of the preset quantity is calculated, according to being averaged for volumes of searches
The volumes of searches that value, maximum value and minimum value treat push category is normalized, after obtaining volumes of searches normalized
Numerical value.
Server obtains the creation time of category to be pushed, and calculates time difference of the creation time apart from current time, then
The time difference is normalized.Server can be to the mode of the time difference normalized:By above-mentioned
Mode obtains each time difference of the category creation time apart from current time in the category (including category to be pushed) of preset quantity
Value, further according to maximum value and minimum of the category creation time in the category of preset quantity in the time difference of current time
Value is treated push category and is normalized, and the Temporal-distance value normalizing of category creation time and current time to be pushed is obtained
Change treated numerical value.
Server obtains the ranking of category to be pushed other categories in the precedence data using store, and application store
Data, and treat precedence data of the push category in application store and be normalized.Server is treated push category and is being answered
Mode with the precedence data in store can be:Obtain the row using the category (including category to be pushed) of preset quantity in store
Name data, using in the precedence data maximum value and minimum value treat push category application store precedence data returned
One change is handled, and obtains the numerical value after precedence data normalized of the push category in application store.Specific normalized
It is represented by:X=(x-min)/(max-min).Wherein X indicates that category to be pushed carries out normalizing in the precedence data in application store
Change treated corresponding numerical value, x indicates category push before applying the precedence data in store to be normalized
Rank value, min indicate that minimum value described above, max indicate maximum value described above.Wherein it is possible to be to pass through reptile
Technology is applied the precedence data of the category (including category to be pushed) of preset quantity in store.
It should be noted that treating precedence data of the push category in application store, creation time and current time
It is previously mentioned in the step of Temporal-distance value and volumes of searches are normalized【The category of preset quantity (including waits pushing
Category)】Refer to the same meaning.It that is to say, server treats the row in application store of push category using same category packet
Name data, the Temporal-distance value and volumes of searches of creation time and current time are normalized.Therefore, it just can ensure that
Each relevant parameter of push category is treated when being handled, using same reference standard.It that is to say in normalized
In the process, other categories of reference are to determine, after ensureing that the relevant parameter for treating push category is normalized
The accuracy of obtained numerical value.
S230 handles described search amount according to least square method, obtains the variation tendency value of described search amount.
In the present embodiment, server according to least square method treat push category volumes of searches handled, waited for
Push the variation tendency value of category volumes of searches.It can be specifically the preset time calculated with least square method described in step S210
The variation tendency value of the volumes of searches of category to be pushed in section.
Least square method (also known as least squares method) is a kind of mathematical optimization techniques.It by minimize error quadratic sum
Find the optimal function matching of data.Unknown data can be easily acquired using least square method, and these are acquired
Data and real data between error quadratic sum be minimum.Least square method can be additionally used in curve matching.It is some other excellent
Change problem can also be expressed by minimizing energy or maximizing entropy with least square method.
S250, according to after the normalized data and the variation tendency value, it is special to obtain the corresponding multidimensional
Sign vector.
In the present embodiment, the precedence data by the category to be pushed obtained in step S210 in application store carries out normalizing
Change treated numerical value, the number after the Temporal-distance value of push category creation time and current time is normalized
Value is obtained according to least square method in numerical value and step S230 that the volumes of searches of category to be pushed is normalized
To category volumes of searches to be pushed variation tendency value, obtain corresponding multidimensional characteristic vectors.Each element pair in feature vector
Answer the numerical value obtained after above-mentioned processing.
S300 calculates the corresponding score value of described eigenvector, using the score value as institute according to default score value operational model
State the score value of category to be pushed;Wherein, the default score value operational model waits pushing for characterizing described eigenvector with described
The incidence relation of category score value.
In the present embodiment, described eigenvector is inputted and is preset in score value operational model by server, to obtain waiting pushing
The score value of category.Wherein, the incidence relation of score value operational model characteristic feature vector and category score value to be pushed is preset.Default point
Value operational model can be the score value operational model trained according to sample data, can also be existing score value operation mould
Type.Such as the score value operational model or multiple parameters trained according to neural network model (can be characteristic information datas
In parameter) product sum function value formed score value operational model.
In one embodiment, category to be pushed is live streaming category.The characteristic information data of live streaming category is included in application
The Temporal-distance value and volumes of searches of the precedence data in store, creation time and current time, start broadcasting main broadcaster's number.Step S200
Include the following steps:
Respectively by the precedence data in application store, the Temporal-distance value of creation time and current time, and search
Suo Liang, the main broadcaster's number that starts broadcasting are normalized, and obtain the data after normalized;According to least square method respectively to described
Volumes of searches and the main broadcaster's number that starts broadcasting are handled, and obtain corresponding variation tendency value;After the normalized
It is special to obtain the corresponding multidimensional for the variation tendency value of the variation tendency value of data and described search amount, the main broadcaster's number that starts broadcasting
Sign vector.
In the present embodiment, server obtains the live streaming category searching times in preset time period.It specifically can be according to default
The search term of user in period, the volumes of searches of the live streaming category obtained from the background with Keywords matching.Server is to being broadcast live product
The normalized mode of the volumes of searches of class can be:Live streaming category (including this hair of preset quantity is obtained through the above way
The bright live streaming category) in each category volumes of searches, then calculate being averaged for volumes of searches in the live streaming category of the preset quantity
Value, according to the average value of volumes of searches, maximum value and minimum value carry out normalizing to the volumes of searches of live streaming category of the present invention
Change is handled, and obtains the numerical value after the volumes of searches normalized of live streaming category of the present invention.
Server obtains the creation time of live streaming category, calculates time difference of the creation time apart from current time, then right
The time difference is normalized.Server can be to the mode of the time difference normalized:Pass through above-mentioned side
Each category creation time is apart from current time in the category (including live streaming category of the present invention) of formula acquisition preset quantity
Time difference, further according to category creation time in the live streaming category of preset quantity in the time difference of current time most
The creation time of live streaming category of the present invention is normalized in big value and minimum value, obtains of the present invention
The numerical value after the Temporal-distance value normalized of category creation time and current time is broadcast live.
Server obtains live streaming category of the present invention in the precedence data using store, and using other in store
The precedence data of category is broadcast live, and place is normalized in the precedence data to live streaming category of the present invention in application store
Reason.The precedence data mode that is normalized of the server to live streaming category of the present invention in application store can be with
It is:The precedence data using the live streaming category (including live streaming category of the present invention) of preset quantity in store is obtained, is utilized
The precedence data of maximum value and minimum value to live streaming category of the present invention in application store in the precedence data is returned
One change is handled, and obtains numerical value of the live streaming category of the present invention after the precedence data normalized in application store.Specifically
Normalized is represented by:X=(x-min)/(max-min).Wherein X indicates live streaming category of the present invention using quotient
The precedence data in city be normalized after corresponding numerical value, x indicates live streaming category of the present invention using store
Precedence data be normalized before rank value, min indicates minimum value described above, and max indicates above-mentioned institute
The maximum value stated.
Server obtains the main broadcaster's number that starts broadcasting of the live streaming category in preset time period.Server is to live streaming of the present invention
The normalized mode of the main broadcaster's number that starts broadcasting of category can be:Live streaming category (the packet of preset quantity is obtained through the above way
Include live streaming category of the present invention) in each category the main broadcaster's number that starts broadcasting, then calculate and opened in the live streaming category of the preset quantity
The average value for broadcasting main broadcaster's number, according to the average value for the main broadcaster's number that starts broadcasting, maximum value and minimum value are to live streaming product of the present invention
The main broadcaster's number that starts broadcasting of class is normalized, and obtains the numerical value after the main broadcaster's number normalized of the present invention that starts broadcasting.
It should be noted that in the precedence data to live streaming category in application store, creation time and current time when
Between be previously mentioned in distance value and volumes of searches, the main broadcaster's number that starts broadcasting the step of being normalized【The live streaming product of preset quantity
Class (including live streaming category of the present invention))】Refer to the same meaning.It that is to say, server is using same category packet to straight
The precedence data in application store of category is broadcast, the Temporal-distance value and volumes of searches of creation time and current time, start broadcasting master
Number is broadcast to be normalized.Therefore, when just can ensure that each relevant parameter to live streaming category is handled, using same
Reference standard.It that is to say during normalized, other of reference are broadcast live what categories were to determine, to ensure to this hair
The accuracy for the numerical value that the relevant parameter of the bright live streaming category obtains after being normalized.
Server is respectively handled volumes of searches and the main broadcaster's number that starts broadcasting that category is broadcast live according to least square method, is obtained
It is broadcast live the variation tendency value of category volumes of searches and live streaming category starts broadcasting the variation tendency value of main broadcaster's number.Can be specifically with minimum
Square law calculates the volumes of searches of live streaming category and the variation tendency value for the main broadcaster's number that starts broadcasting.
In the present embodiment, the ranking according to the live streaming category obtained by processing mode described above in application store
Data be normalized after numerical value, place is normalized in the Temporal-distance value that category creation time and current time is broadcast live
The numerical value that the volumes of searches of category is normalized is broadcast live in numerical value after reason, and main broadcaster's number progress of starting broadcasting of category is broadcast live
The numerical value that normalized obtains, and the variation tendency value of live streaming category volumes of searches that is obtained according to least square method and start broadcasting
The variation tendency value of main broadcaster's number, obtains corresponding multidimensional characteristic vectors.Each element corresponds to said one numerical value in feature vector.
S400 pushes category according to the score value of the category to be pushed.
Be calculated according to above-mentioned steps after the score value of category to be pushed, server according to the score value of category to be pushed to
APP client push categories.Preferably, when the score value when push category reaches the default score value of server, to APP client
Push the category to be pushed.When this is when the score value for pushing category is not up to server and presets score value, server obtains other points
Value reaches the category of the default score value.
In one embodiment, step S400 may include:Obtain multiple categories to be pushed in the first preset time period
Score value;The multiple category to be pushed is ranked up from big to small according to score value, the preceding preset quantity that sorts is obtained and waits pushing away
Category is sent, preset quantity category to be pushed is pushed.It that is to say, the first preset time period is obtained according to step S100-S300
The score value of interior multiple categories to be pushed from big to small arranges multiple categories to be pushed according to the score value of each category to be pushed
Sequence, the category to be pushed for obtaining the preceding preset quantity that sorts are pushed.First preset time period can be in one week.It is default
Quantity can be determined according to the quantity for the category for being pushed to APP clients from the background.
In one embodiment, as shown in figure 4, step S400 may comprise steps of:
S410 obtains the score value of multiple categories to be pushed in the first preset time period.
S430, push score value reach the category to be pushed in default score range.
In the present embodiment, server obtains multiple in the first preset time period wait for respectively according to step S100- steps 300
Push the score value of category.Statistics score value reaches the category to be pushed in default score range, and this waits pushing away to APP client push
Send category.
In one embodiment, as shown in figure 5, step S430 includes the following steps:
S431 obtains the quantity that score value reaches the category to be pushed in default score range.
S433, judges whether the quantity is less than preset quantity.
S435, if so, the difference of preset quantity and the quantity is calculated, using difference as quantity to be obtained;It is preset from first
The category of quantity to be obtained is obtained in category list;Push the score value reach category to be pushed in default score range and from
The category of the quantity to be obtained obtained in the first preset category list.
S437, if it is not, pushing the score value reaches category to be pushed in default score range.
In the present embodiment, server determines in multiple categories to be pushed that score value reaches and waits pushing away in default score range
The quantity for sending category, judges whether the quantity is less than preset quantity.Wherein, preset quantity is pushed away to APP clients according to backstage
The quantity of category is sent to determine.
If the quantity is less than preset quantity, i.e., algorithm data-base calculates the push category of output by algorithm in server
Quantity is less than preset quantity, then further calculates preset quantity and reach the number of the category to be pushed in default score range with score value
The difference of amount obtains the category of the quantity to be obtained using the difference as quantity to be obtained from the first preset category list.First is pre-
It is that the preset category of server is guaranteed the minimum list to set category list, to reach default score range to APP client push score values
The category of interior category to be pushed and the quantity to be obtained obtained from the first preset category list.It that is to say, when algorithm calculates
When the quantity of the push category of output is less than the quantity of the default category for being pushed to APP clients of server, from the guarantor of server
Score value is obtained in the category list of bottom and reaches the category in default score range, this two-part category is pushed to APP clients.
If the quantity is not less than preset quantity, the category to be pushed that algorithm is directly calculated to output is pushed to APP client
End.
In one embodiment, as shown in fig. 6, step S430 may comprise steps of:
S432 obtains score value in the second preset category list and reaches the category in default score range.
S434, push score value reach in category to be pushed and the second preset category list in default score range
Score value reaches the category in default score range.
In the present embodiment, server waits pushing from getting score value in algorithm data-base and reach in default score range
After category, while obtaining score value from the second preset category list and reaching the category in default score range.Wherein, second is pre-
It is category list of the backstage according to user demand to APP client push to set category list, or is recommended from the background to APP clients
Category list.Such as, the category list that newest category is formed, the category list etc. that the high category of nearest temperature is formed.Therefore,
The category of APP clients can be pushed to by backstage operation regulation and control.
In a specific embodiment, category to be pushed is live streaming category.Server is obtained by step S100- steps 300
Category is broadcast live to APP client push according to the score value of live streaming category in the score value for getting multiple live streaming categories.Can be specifically,
Multiple live streaming categories are ranked up according to the score value of live streaming category, are sorted preceding preset quantity to APP client push
Category is broadcast live.It can also be to obtain score value to reach the live streaming category in default score range, reached to APP client push score values
To the live streaming category in default score range.
In one embodiment, the category to be pushed is live streaming category.As shown in fig. 7, step S430 may include following
Step:
S436 obtains score value and reaches minimum in the second preset time period of the live streaming category in default score range and starts broadcasting
Main broadcaster's number and/or minimum popularity number.
S438, the minimum main broadcaster's number that starts broadcasting of push are more than the second preset value more than the first preset value and/or minimum popularity number
Category is broadcast live.
In the present embodiment, server gets the score value of multiple categories to be pushed from algorithm data-base and determines point
After value reaches the live streaming category in default score range, further confirm that score value reaches the live streaming category in default score range
Minimum main broadcaster's number and/or the minimum popularity number of starting broadcasting in the second preset time period.Confirm live streaming category start broadcasting main broadcaster's number or
Minimum popularity number can be executed in backstage operation, can also be directly to be executed in algorithm data-base.
Further, server obtains the minimum main broadcaster's number that starts broadcasting and is more than the more than the first preset value and/or minimum popularity number
The live streaming category of the part is pushed to live streaming client by the live streaming category of two preset values.Therefore, it can further improve and be issued to
The user satisfaction of the live streaming category of user terminal improves the live streaming competitiveness of product in market.
In one embodiment, the category to be pushed is live streaming category.As shown in figure 8, step S400 includes following step
Suddenly:
S450 obtains the score value of multiple live streaming categories.
S460 determines the live streaming category issued according to the score value of the multiple live streaming category.
S470, the viewing number of users of the live streaming category issued described in acquisition.
S480 pushes live streaming category according to the viewing number of users of the live streaming category issued.
In the present embodiment, server obtains the score value of multiple live streaming categories, and the live streaming product issued are determined according to the score value
Class.Meanwhile the viewing number of users of the live streaming category issued is further obtained, according to the viewing user of the live streaming category issued
Number push live streaming category.In a particular embodiment, can be obtain preset quantity live streaming category it is (including of the present invention
Live streaming category) in the viewing number of users of category is each broadcast live, then calculate the viewing number of users of the live streaming category of the preset quantity
Average value, according to viewing number of users average value, maximum value and minimum value return live streaming category of the present invention
One change is handled, and obtains the numerical value after normalized.In addition, also can be by least square method to live streaming category of the present invention
Viewing number of users handled, obtain viewing number of users variation tendency value.Server according to viewing number of users normalization at
Numerical value and its variation tendency value after reason adjust and push the live streaming category issued.Wherein, the live streaming category that server obtains
Viewing number of users with push live streaming category can be synchronous progress, can also be asynchronous progress.
In other embodiments, the live streaming category issued first can be pushed to live streaming client by server, in advance
If being obtained after the period in live streaming client and watching number of users, the live streaming category issued is adjusted according to viewing number of users, will be watched
The live streaming category that number of users is more than preset value is pushed to live streaming client.Can also be, server issue live streaming category before,
The viewing number of users that category is broadcast live in the live streaming category issued in historical time is first obtained, number of users will be watched and be more than preset value
Live streaming category is pushed to live streaming client.
In one embodiment, step S400 includes:Category is obtained according to the score value of the category to be pushed;Described in acquisition
Position, title and the icon that the category of acquisition is shown in APP client homepages;To APP clients send the position, title and
Icon.
In the present embodiment, after server is pushed to the category of APP clients according to the acquisition of the score value of category to be pushed,
Further obtain position, title and icon that the category is shown in APP client homepages, to APP clients issue the position,
Title and icon.APP clients are after receiving the position that push category is shown in homepage, displaying push on the position
The title and icon of category.Wherein, the category is in the homepage that the position that APP client homepages are shown can be APP clients
The position in newest heat trip prefecture.
A kind of category method for pushing provided by the invention, obtains the characteristic information data of category to be pushed, and will wait pushing
The characteristic information data of category is converted into feature vector, and point of the category to be pushed can be obtained according to default score value operational model
Value, to push category to subscription client according to the score value of category to be pushed.Wherein, default score value operational model, which characterizes, waits for
Push the incidence relation of the feature vector of category and the score value of final output.It that is to say, believed by the feature of each category to be pushed
The feature vector of breath data conversion can obtain one and the category characteristic information pair to be pushed according to default score value operational model
The score value answered.Therefore, corresponding score value can be matched according to the hobby of user, pushing matching score value to user corresponds to
Category to be pushed, to improve the competitiveness of product in market.
Further, the characteristic information data of category to be pushed is multiple, by the feature feature that obtains later of conversion to
Amount is multidimensional characteristic vectors, and the score value for the category to be pushed that final operation obtains is joined with reference to multiple correlations of category to be pushed
Number.This improves the reference values of the score value of category to be pushed so that the category of push is more accurate, more meets user's
Demand.
Preferably, category to be pushed is live streaming category, and characteristic information data is included in the precedence data using store, creates
The Temporal-distance value and volumes of searches of time and current time, start broadcasting main broadcaster's number.The present invention will be respectively by aforementioned four reference number
According to being normalized, the data after normalized are obtained.Meanwhile according to least square method respectively to above-mentioned volumes of searches and
The main broadcaster's number that starts broadcasting is handled, and obtains corresponding variation tendency value.Further according to after normalized data and variation become
Gesture value obtains the corresponding multidimensional characteristic vectors.It that is to say, be pushed to the live streaming category of live streaming client with reference to live streaming
Precedence data, the main broadcaster's number that starts broadcasting, volumes of searches and the creation time that category is broadcast live, to ensure that the live streaming category of push more accords with
Be fated the demand of lower user, to improve the competitiveness of product in market.
For a kind of category method for pushing provided by the invention, one is provided specifically below in conjunction with common application environment
Embodiment, the implementation process of the category method for pushing is further described:
In the present embodiment, the category of push is live streaming category.APP clients are live streaming client.Client is broadcast live
The live streaming client used that starts broadcasting including main broadcaster and user watch the live streaming client used.Server is specially that fortune is broadcast live
The server of battalion.Server background gives live streaming client push newest heat trip live streaming category.
As shown in figure 9, server 110 includes algorithm data-base 111 and operation module 113.Algorithm data-base 111 will lead to
It crosses default score value model and calculates the score value of live streaming category, and export the recommendation list that score value reaches the most hot category of default score value
And the weight (corresponding with the score value described in above-described embodiment) of category is each broadcast live in recommendation list.Run 113 basis of module
The recommendation list that algorithm data-base 111 exports carries out operation intervention, increases new recommendation live streaming category in recommendation list, or
Shield certain live streaming categories in recommendation list.Final server 110 passes through interface integration algorithm database 111 and operation module
The table data of 113 outputs, and live streaming client 310 is issued to by network 210.Server 110 is issued to live streaming client
310 data include:Block icon, block name and the column list of category is broadcast live.Column list includes【gameid】、
【Broadcasting number】.The live streaming category list that live streaming client 310 will be issued in the page newest heat trip position displaying server 110.
Specifically, algorithm data-base 111 provides " new hot column " sorted lists according to certain rule-based algorithm, including gameid (category ID)
With weight (weight height represents more forward).Server 110 issues live streaming category list using Summer interfaces to client.One
In embodiment, the sorted lists that Summer interfaces are provided according to algorithm data-base, in conjunction with the operation intervention (modification of backstage configuration
Weight, strategy of guaranteeing the minimum etc.), data are issued to client.When client loads homepage, according to specific data are issued, show
【Newest heat trip】Block.
Wherein, algorithm data-base 111 is as follows to the processing procedure that category is broadcast live:
First, the data that algorithm operation is used include:The ranking of category is broadcast live in app store, it is straight in a week
The main broadcaster's number that starts broadcasting under category is broadcast, the creation time of category is broadcast live, the live streaming category searching times in a week.Algorithm operation
Process includes:The Feature Conversion that above-mentioned data are done with different linear and nonlinears obtains the feature vector of one 8 dimension, then right
This feature vector carries out nonlinear operation and obtains a score, and score higher position illustrates it is new heat.The game live streaming of new heat is aobvious
Show in newest heat trip region.Specifically calculating process is:
0, the category searching times in a week:According to the search term of user, corresponding category is obtained with Keywords matching
Under volumes of searches, then calculate average value+minimax normalized.
1, the live streaming category searching times in a week:And obtain the variation of category volumes of searches with least-squares algorithm
Trend value.
2, the main broadcaster's number that starts broadcasting under the live streaming category in a week:It averages+minimax normalized.
3, the main broadcaster's number that starts broadcasting under the live streaming category in a week:Obtain starting broadcasting the change of main broadcaster's number with least-squares algorithm
Change Trend value.
4, the creation time of category is broadcast live:From the number of days that the same day calculates time phase difference, then at minimax normalization
Reason.
5, the ranking of category is broadcast live in app store:It is normalized, is arranged using minimax (x-min)/(max-min)
Name data are reptile obtains from the website of appstore daily data, and with integrated in server broadcasting live streaming category
Processing.
6, the viewing number of users in the category next week:It averages+minimax normalized.
7, the viewing number of users in the category next week:The variation that viewing number of users is obtained with least-squares algorithm becomes
Gesture value.
The corresponding score value of live streaming category is obtained finally by following formula:
Score=(1.0*x [4]+(x [1]+0.8*x [1] * x [0]+x [0]+0.5*x [2] * x [3]+x [2]+x [3]+x
[4]*x[3]+1.0*x[4]*x[0])/4+1.0*x[5])。
Wherein, x [0] indicates that the data obtained in step 0, x [1] indicate that the data obtained in step 1, x [2] indicate step
The data obtained in 2, x [3] indicate that the data obtained in step 3, x [4] indicate that the data obtained in step 4, x [5] indicate step
The data obtained in rapid 5.
For the processing of the viewing number of users of live streaming category in step 6 and step 7, the adjustment for subsequently pushing category,
Only do data preparation processing herein.
Module 113 is runed for the intervention processing procedure of push live streaming category referring to Fig.1 shown in 0.Described in Figure 10
Shown in processing procedure is explained as follows:
In server 110, algorithm data-base 111 and operation module 113 are as follows for the integrated disposal processing that category is broadcast live:
The present invention also provides a kind of category pushers, and as shown in figure 11, which includes the first acquisition mould
Block 101, the second acquisition module 103, computing module 105 and pushing module 107.
First acquisition module 101 is used to obtain the characteristic information data of category to be pushed.In the present embodiment, server obtains
Take the characteristic information data of category to be pushed.Category to be pushed can be live streaming category, short-sighted frequency category, APP categories etc..It waits pushing away
Sending the characteristic information data of category can be:The creation time of category to be pushed, ranking of the category to be pushed on operation platform
Data, the popularity of category to be pushed, the volumes of searches etc. of category to be pushed.Wherein, the characteristic information data of category to be pushed
Can be one or more.In order to improve the accuracy of push category, multiple characteristic information numbers of category to be pushed can be acquired
According to.Multiple characteristic information datas herein can be with the characteristic of category to be pushed and with the relevant feature of the interest-degree of user
Information data so that the score value that characteristic information data subsequently based on category to be pushed obtains can more embody the category for
The push of user group is worth.
Second acquisition module 103 is used to the characteristic information data carrying out Feature Conversion, obtains feature vector.Further
The characteristic information data of category to be pushed is carried out Feature Conversion by ground, server, to obtain the feature of the category to be pushed to
Amount.The technical implementation way of Feature Conversion can be, to characteristic information data carry out linear process either Nonlinear Processing or
Linear process is carried out to partial data in characteristic information data, other part data carry out Nonlinear Processing.Specific feature turns
The mode of changing can be characteristic information data is normalized, and/or, according to least square method to characteristic information data
It is handled.
In the present embodiment, the characteristic information data of category to be pushed is multiple.Multiple characteristic information datas are carried out special
After sign conversion, multidimensional characteristic vectors are obtained.Specifically, server carries out Feature Conversion to each characteristic information data respectively, obtains
To corresponding multidimensional characteristic vectors.Wherein, in multidimensional characteristic vectors, each element corresponds to a characteristic information data and carries out feature
Numerical value after conversion.
Computing module 105 is used to, according to score value operational model is preset, calculate the corresponding score value of described eigenvector, will be described
The score value of score value category to be pushed as described in;Wherein, the default score value operational model for characterize described eigenvector with
The incidence relation of the category score value to be pushed.In the present embodiment, server, which inputs described eigenvector, presets score value fortune
It calculates in model, to obtain the score value of category to be pushed.Wherein, score value operational model characteristic feature vector and category to be pushed are preset
The incidence relation of score value.Default score value operational model can be the score value operational model trained according to sample data, also may be used
To be existing score value operational model.Such as the score value operational model trained according to neural network model or multiple ginsengs
The score value operational model that the sum function value of number (can be the parameter in characteristic information data) product is formed.
The score value that pushing module 107 is used for the category to be pushed according to pushes category.It is calculated according to above-mentioned module
After the score value of category to be pushed, server is according to the score value of category to be pushed to APP client push categories.Preferably, when
This is when the score value for pushing category reaches the default score value of server, to the APP pushes customers category to be pushed.When this waits pushing
The score value of category be not up to server preset score value when, server obtains the category that other score values reach the default score value.
In other embodiments, the modules in category pusher provided by the invention are additionally operable to execute institute of the present invention
In the category method for pushing stated, the operation that corresponding each step executes no longer is described in detail herein.
The present invention also provides a kind of storage devices.A plurality of instruction is stored in the storage device;Described instruction be suitable for by
Reason device loads and executes the category method for pushing described in any of the above-described embodiment.
Storage device described in the present embodiment can be memory.Memory can be built-in storage or external memory, or
Person includes both built-in storage and external memory.Built-in storage may include read-only memory (ROM), programming ROM (PROM),
Electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), flash memory or random access memory.External storage
Device may include hard disk, floppy disk, ZIP disks, USB flash disk, tape etc..Memory disclosed in this invention includes but not limited to these types
Memory.Memory disclosed in this invention is only used as example rather than as restriction.
The present invention also provides a kind of computer equipments.A kind of computer equipment includes:One or more processors;Storage
Device;One or more application program.Wherein one or more of application programs are stored in the memory and are configured
To be executed by one or more of processors, one or more of application programs are configured to carry out any of the above-described embodiment
The category method for pushing.
Figure 12 is the structural schematic diagram of the computer equipment in one embodiment of the invention.Equipment described in the present embodiment can be with
It is computer equipment.Such as server, personal computer and the network equipment.As shown in figure 12, equipment include processor 1203,
The devices such as memory 1205, input unit 1207 and display unit 1209.It will be understood by those skilled in the art that Figure 12 is shown
Device structure device do not constitute the restriction to all devices, may include than illustrating more or fewer components or group
Close certain components.Memory 1205 can be used for storing application program 1201 and each function module, the operation storage of processor 1203
In the application program 1201 of memory 1205, to execute various function application and the data processing of equipment.Memory can be with
It is built-in storage or external memory, or including both built-in storage and external memory.Built-in storage may include read-only storage
Device (ROM), programming ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), flash
Device or random access memory.External memory may include hard disk, floppy disk, ZIP disks, USB flash disk, tape etc..It is disclosed in this invention to deposit
Reservoir includes but not limited to the memory of these types.Memory disclosed in this invention is only used as example rather than as restriction.
Input unit 1207 is used to receive the input of signal, and receives keyword input by user.Input unit 1207
It may include touch panel and other input equipments.Touch panel collect user on it or neighbouring touch operation (such as
User uses the operations of any suitable object or attachment on touch panel or near touch panel such as finger, stylus), and
Corresponding attachment device is driven according to a pre-set procedure;Other input equipments can include but is not limited to physical keyboard, work(
It is one or more in energy key (such as broadcasting control button, switch key etc.), trace ball, mouse, operating lever etc..Display unit
1209 can be used for showing information input by user or be supplied to the information of user and the various menus of computer equipment.Display is single
The forms such as liquid crystal display, Organic Light Emitting Diode can be used in member 1209.Processor 1203 is the control centre of computer equipment,
Using the various pieces of various interfaces and the entire computer of connection, by run or execute be stored in it is soft in memory 1203
Part program and/or module, and the data being stored in memory are called, perform various functions and handle data.
In one embodiment, equipment includes one or more processors 1203, and one or more memories 1205,
One or more application program 1201.Wherein one or more of application programs 1201 are stored in memory 1205 and quilt
Be configured to be executed by one or more of processors 1203, one or more of application programs 1201 be configured to carry out with
Category method for pushing method described in upper embodiment.
In addition, each functional unit in each embodiment of the present invention can be integrated in a processing module, it can also
That each unit physically exists alone, can also two or more units be integrated in a module.Above-mentioned integrated mould
The form that hardware had both may be used in block is realized, can also be realized in the form of software function module.The integrated module is such as
Fruit is realized in the form of software function module and when sold or used as an independent product, can also be stored in a computer
In read/write memory medium.
One of ordinary skill in the art will appreciate that realizing that all or part of step of above-described embodiment can pass through hardware
It completes, relevant hardware can also be instructed to complete by program, which can be stored in a computer-readable storage medium
In matter, storage medium may include memory, disk or CD etc..
The above is only some embodiments of the present invention, it is noted that for the ordinary skill people of the art
For member, various improvements and modifications may be made without departing from the principle of the present invention, these improvements and modifications are also answered
It is considered as protection scope of the present invention.
Claims (12)
1. a kind of category method for pushing, which is characterized in that include the following steps:
Obtain the characteristic information data of category to be pushed;
The characteristic information data is subjected to Feature Conversion, obtains feature vector;
According to default score value operational model, the corresponding score value of described eigenvector is calculated, the score value is waited pushing as described in
The score value of category;Wherein, the default score value operational model is for characterizing described eigenvector and the category score value to be pushed
Incidence relation;
Category is pushed according to the score value of the category to be pushed.
2. according to the method described in claim 1, it is characterized in that, the characteristic information data is included in the ranking using store
The Temporal-distance value and volumes of searches of data, creation time and current time;It is described respectively by multiple characteristic information datas
Feature Conversion is carried out, corresponding multidimensional characteristic vectors are obtained, including:
Respectively by the precedence data in application store, the Temporal-distance value and volumes of searches of creation time and current time
It is normalized, obtains the data after normalized;
Described search amount is handled according to least square method, obtains the variation tendency value of described search amount;
According to after the normalized data and described search amount change Trend value, obtain the corresponding multidimensional characteristic to
Amount.
3. according to the method described in claim 1, it is characterized in that, the category to be pushed is live streaming category;The feature letter
Breath data are included in the precedence data using store, and the Temporal-distance value and volumes of searches of creation time and current time start broadcasting
Main broadcaster's number;It is described that multiple characteristic information datas are subjected to Feature Conversion respectively, corresponding multidimensional characteristic vectors are obtained, are wrapped
It includes:
Respectively by it is described application store precedence data, the Temporal-distance value and volumes of searches of creation time and current time,
The main broadcaster's number that starts broadcasting is normalized, and obtains the data after normalized;
Described search amount and the main broadcaster's number that starts broadcasting are handled respectively according to least square method, obtain described search amount
The variation tendency value of variation tendency value and the main broadcaster's number that starts broadcasting;
Become according to the variation tendency value of data, described search amount after the normalized and the variation for starting broadcasting main broadcaster's number
Gesture value obtains the corresponding multidimensional characteristic vectors.
4. according to the method described in claim 1, it is characterized in that, the score value of category to be pushed described in the basis pushes product
Class, including:
It obtains in the first preset time period, the score value of multiple categories to be pushed;
Push score value reaches the category to be pushed in default score range;Or
The multiple category to be pushed is ranked up from big to small according to score value, the preceding preset quantity that sorts is obtained and waits pushing
Category pushes preset quantity category to be pushed.
5. according to the method described in claim 4, it is characterized in that, the push score value reaches waits pushing away in default score range
Category is sent, including:
Obtain the quantity that score value reaches the category to be pushed in default score range;
Judge whether the quantity is less than preset quantity;
If so, the difference of preset quantity and the quantity is calculated, using difference as quantity to be obtained;From the first preset category list
Obtain the category of quantity to be obtained;It pushes the score value and reaches category to be pushed in default score range and pre- from described first
Set the category of the quantity to be obtained obtained in category list;
If it is not, pushing the score value reaches category to be pushed in default score range.
6. according to the method described in claim 4, it is characterized in that, the push score value reaches waits pushing away in default score range
Category is sent, including:
It obtains score value in the second preset category list and reaches category in default score range;
Push score value reaches score value in category to be pushed and the second preset category list in default score range and reaches
Category in default score range.
7. according to the method described in claim 4, it is characterized in that, the category to be pushed is live streaming category;The push point
Value reaches the category to be pushed in default score range, including:
Obtain score value reach minimum in the second preset time period of the live streaming category in default score range start broadcasting main broadcaster's number with/
Or minimum popularity number;
Push the live streaming category that the minimum main broadcaster's number that starts broadcasting is more than the second preset value more than the first preset value and/or minimum popularity number.
8. according to the method described in claim 1, it is characterized in that, the category to be pushed is live streaming category;It is described according to institute
The score value push category of category to be pushed is stated, including:
Obtain the score value of multiple live streaming categories;
The live streaming category issued is determined according to the score value of the multiple live streaming category;
The viewing number of users of the live streaming category issued described in acquisition;
Live streaming category is pushed according to the viewing number of users of the live streaming category issued.
9. according to the method described in claim 1, it is characterized in that, the score value of category to be pushed described in the basis pushes product
Class, including:
Category is obtained according to the score value of the category to be pushed;
Obtain position, title and icon that the category of the acquisition is shown in APP client homepages;
The position, title and icon are sent to APP clients.
10. a kind of category pusher, which is characterized in that including:
First acquisition module, the characteristic information data for obtaining category to be pushed;
Second acquisition module obtains feature vector for the characteristic information data to be carried out Feature Conversion;
Computing module, for according to score value operational model is preset, calculating the corresponding score value of described eigenvector, the score value being made
For the score value of the category to be pushed;Wherein, the default score value operational model is waited for for characterizing described eigenvector with described
Push the incidence relation of category score value;
Pushing module, the score value for the category to be pushed according to push category.
11. a kind of storage device, which is characterized in that be stored thereon with a plurality of instruction;Described instruction is suitable for by processor load simultaneously
Execute the category method for pushing described in any one of the claims 1-9.
12. a kind of computer equipment, which is characterized in that it includes:
One or more processors;
Memory;
One or more application program, wherein one or more of application programs are stored in the memory and are configured
To be executed by one or more of processors, one or more of application programs are configured to carry out according to claim 1
To 9 any one of them category method for pushing.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810460322.2A CN108683734B (en) | 2018-05-15 | 2018-05-15 | Method and device for pushing classes, storage equipment and computer equipment |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810460322.2A CN108683734B (en) | 2018-05-15 | 2018-05-15 | Method and device for pushing classes, storage equipment and computer equipment |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108683734A true CN108683734A (en) | 2018-10-19 |
CN108683734B CN108683734B (en) | 2021-04-09 |
Family
ID=63806135
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810460322.2A Active CN108683734B (en) | 2018-05-15 | 2018-05-15 | Method and device for pushing classes, storage equipment and computer equipment |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108683734B (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110390056A (en) * | 2019-06-18 | 2019-10-29 | 平安科技(深圳)有限公司 | Big data processing method, device, equipment and readable storage medium storing program for executing |
CN111753158A (en) * | 2020-01-20 | 2020-10-09 | 全息空间(深圳)智能科技有限公司 | Live broadcast platform commodity searching method and device, computer equipment and storage medium |
CN112579905A (en) * | 2020-12-25 | 2021-03-30 | 百果园技术(新加坡)有限公司 | Message push object determining method, device, equipment and storage medium |
CN113487361A (en) * | 2021-07-13 | 2021-10-08 | 广州虎牙科技有限公司 | Method, device, equipment and storage medium for predicting platform user value |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102982079A (en) * | 2012-10-30 | 2013-03-20 | 北京奇虎科技有限公司 | Method and device for personalized website navigation |
CN103559206A (en) * | 2013-10-10 | 2014-02-05 | Tcl集团股份有限公司 | Information recommending method and system |
CN103971170A (en) * | 2014-04-17 | 2014-08-06 | 北京百度网讯科技有限公司 | Method and device for forecasting changes of feature information |
CN104601438A (en) * | 2014-04-28 | 2015-05-06 | 腾讯科技(深圳)有限公司 | Friend recommendation method and device |
US20150213492A1 (en) * | 2014-01-24 | 2015-07-30 | Flipboard, Inc. | Recommending Magazines to Users of a Digital Magazine Server |
CN104881798A (en) * | 2015-06-05 | 2015-09-02 | 北京京东尚科信息技术有限公司 | Device and method for personalized search based on commodity image features |
CN105787061A (en) * | 2016-02-29 | 2016-07-20 | 广东顺德中山大学卡内基梅隆大学国际联合研究院 | Information pushing method |
CN105843953A (en) * | 2016-04-12 | 2016-08-10 | 乐视控股(北京)有限公司 | Multimedia recommendation method and device |
-
2018
- 2018-05-15 CN CN201810460322.2A patent/CN108683734B/en active Active
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102982079A (en) * | 2012-10-30 | 2013-03-20 | 北京奇虎科技有限公司 | Method and device for personalized website navigation |
CN103559206A (en) * | 2013-10-10 | 2014-02-05 | Tcl集团股份有限公司 | Information recommending method and system |
US20150213492A1 (en) * | 2014-01-24 | 2015-07-30 | Flipboard, Inc. | Recommending Magazines to Users of a Digital Magazine Server |
CN103971170A (en) * | 2014-04-17 | 2014-08-06 | 北京百度网讯科技有限公司 | Method and device for forecasting changes of feature information |
CN104601438A (en) * | 2014-04-28 | 2015-05-06 | 腾讯科技(深圳)有限公司 | Friend recommendation method and device |
CN104881798A (en) * | 2015-06-05 | 2015-09-02 | 北京京东尚科信息技术有限公司 | Device and method for personalized search based on commodity image features |
CN105787061A (en) * | 2016-02-29 | 2016-07-20 | 广东顺德中山大学卡内基梅隆大学国际联合研究院 | Information pushing method |
CN105843953A (en) * | 2016-04-12 | 2016-08-10 | 乐视控股(北京)有限公司 | Multimedia recommendation method and device |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110390056A (en) * | 2019-06-18 | 2019-10-29 | 平安科技(深圳)有限公司 | Big data processing method, device, equipment and readable storage medium storing program for executing |
CN111753158A (en) * | 2020-01-20 | 2020-10-09 | 全息空间(深圳)智能科技有限公司 | Live broadcast platform commodity searching method and device, computer equipment and storage medium |
CN112579905A (en) * | 2020-12-25 | 2021-03-30 | 百果园技术(新加坡)有限公司 | Message push object determining method, device, equipment and storage medium |
CN113487361A (en) * | 2021-07-13 | 2021-10-08 | 广州虎牙科技有限公司 | Method, device, equipment and storage medium for predicting platform user value |
Also Published As
Publication number | Publication date |
---|---|
CN108683734B (en) | 2021-04-09 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108683734A (en) | Category method for pushing, device and storage device, computer equipment | |
US11250072B2 (en) | Apparatus, system, and method for natural language processing | |
JP6262764B2 (en) | Method and system for pushing mobile applications | |
US20230185857A1 (en) | Method and system for providing context based query suggestions | |
CN102073699B (en) | For improving the method for Search Results, device and equipment based on user behavior | |
CN109325132A (en) | Expertise recommended method, device, computer equipment and storage medium | |
US20080201287A1 (en) | Dissimilar item recommendation method, device, and program thereof | |
CN104428769B (en) | The information of text file reader is provided | |
CN109144997A (en) | Data correlation method, device and storage medium | |
US20190087884A1 (en) | Theme recommendation method and apparatus | |
US20080059416A1 (en) | Software system for rules-based searching of data | |
KR100782047B1 (en) | Automatic Generation Method of the Initial Query Set for Image Search on the Mobile Web | |
CN108667877B (en) | Method and device for determining recommendation information, computer equipment and storage medium | |
KR20100059516A (en) | Method, system and computer-readable recording medium for selecting and providing related search keyword based on plural search keywords received in sequence from user terminal | |
WO2021129259A1 (en) | Method for dynamically and quickly loading module according to usage habits of user | |
US20150120705A1 (en) | Cuisine search device, cuisine search method, program, and computer-readable storage medium | |
JP5296745B2 (en) | Query suggestion providing apparatus and method | |
CN105956168B (en) | A kind of information retrieval method and information indexing device | |
CN110134773A (en) | A kind of search recommended method and system | |
CN114117025B (en) | Information query method, device, storage medium and system | |
CN113094602B (en) | Hotel recommendation method, system, equipment and medium | |
CN114154014A (en) | Video cold start recommendation method and device | |
CN113268692A (en) | Method and system for automatically collecting customer options, electronic equipment and storage medium | |
JP3395232B2 (en) | Search item determination method | |
JP7078702B1 (en) | Question management system, question management method and program |
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 |