Specific embodiment
To keep the purposes, technical schemes and advantages of the application clearer, below in conjunction with the application specific embodiment and
Technical scheme is clearly and completely described in corresponding attached drawing.Obviously, described embodiment is only the application one
Section Example, instead of all the embodiments.Based on the embodiment in the application, those of ordinary skill in the art are not doing
Every other embodiment obtained under the premise of creative work out, shall fall in the protection scope of this application.
In order to improve recommendation content changing effect, this specification embodiment provides a kind of content recommendation method, device
And equipment.In order to be clearly understood from the application scenarios for each technical solution that this specification embodiment provides, Fig. 1 is first combined below
A kind of running environment of this specification embodiment is illustrated.
As shown in Figure 1, one or more user terminals 101 (one is only shown in Fig. 1) can by network 102 and one or
Multiple servers 103 (one is only shown in Fig. 1) are connected, to carry out data communication or interaction.Wherein, user terminal 101 can be with
It is PC (personal computer, PC), laptop, tablet computer, smart phone, electronic reader, network
TV, wearable device etc. being capable of smart machines connected to the network.
In this specification embodiment, client can be installed in user terminal 101, which can be third party
Application software is also possible to browser, which can provide the user with service together with server end, such as provide payment
Service, for pay to transaction, etc..
It should be noted that content recommendation method and device that this specification embodiment provides, can be applied to user's end
In the client installed in end 101.
A kind of content recommendation method provided below this specification embodiment is illustrated.
As shown in Fig. 2, this specification embodiment provides a kind of content recommendation method, may include:
Step 202 determines the characteristic point for needing the first content recommended on target exhibition position, and the characteristic point is for inhaling
Quote the characteristic information at family.
Exhibition position, can be implement this specification embodiment provide content recommendation method client display interface in,
For showing or exposing the display area of content (Item).As shown in figure 3, being pushed away with implementing the content of this specification embodiment offer
For the client for recommending method is the client being mounted in smart phone, exhibition position be can be after user opens the client,
First area 31, second area 32, third region 33 and the fourth region 34 shown in one display interface of the client,
That is first area 31, second area 32 shown in Fig. 3, third region 33 and the fourth region 34 can be counted as
Four exhibition positions illustrate different coupon informations (content) on this four exhibition positions.Certainly, in a display interface
The quantity of exhibition position can arbitrarily be set, for example, it is also possible to be one, two, three or four or more, without limiting to institute in Fig. 3
Four shown.
First content is a content for needing to show on target exhibition position.One content can be an information, and such as one
Coupon information, an advertising information etc.;Wherein, it can specifically include the title, excellent of discount coupon in a coupon information
The information such as the brief introduction of favour certificate, discount coupon be discussed in detail, the validity period of the usage scenario of discount coupon and discount coupon.
The characteristic point of one content, can refer to can attract in the extension information of this content and this content user or
It is able to use the interested characteristic information in family.For the supplier of client behind, since these characteristic points be used to inhale
The attention for quoting family, brings interests with the supplier for client and the businessman cooperated with the supplier, therefore, in one
The characteristic point of appearance can also be visually known as profit point.For example, for a coupon information, due to its brand,
Special discount has led the information such as number, affiliated classification, affiliated commercial circle, label that can preferably cause user's note that therefore, this
A little information can be used as the characteristic point of a discount coupon.
As an example, step 201 may include: the extension letter for obtaining the first content and the first content
Breath, from the extension information of the first content and the first content, extracts the characteristic point of the first content.Specifically
, can after obtaining corresponding with this content extension information of a content, by way of regular expression matching, from
In extract the characteristic point of this content, the mode for hereinafter extracting the characteristic point of the second content is similar, hereinafter will no longer weigh
Multiple description.
The extension information of one content, refers to and is not included in this content, but has with this content and be associated with
The other information of relationship.For example, the extension information of this content can wrap if a content is a coupon information
It includes and one of information such as the brand of the discount coupon, businessman and the Xian Xia shops that the discount coupon can be used or a variety of is provided.
It due to needing the content recommended on an exhibition position is under normal conditions screened from many alternating contents,
Also, it when a commending contents are carried out on multiple exhibition positions, needs to select from alternating content consistent with booth quantity
A plurality of content is shown with corresponding on the multiple exhibition position, for example, as shown in figure 3, a commending contents are needed at four
Four contents are shown on exhibition position, this just needs to extract the characteristic point of a plurality of content respectively every time, leads to commending contents efficiency not
It is high.
Therefore, in order to improve commending contents efficiency, optionally, as another example, step 201 may include: from
Read the characteristic point of the first content in one database, the characteristic point of the first content is from the first content and described
It is extracted in the extension information of first content.Namely it from each alternating content (including first content) and can also be somebody's turn to do in advance
In the extension information of alternating content, the characteristic point of this alternating content and preservation are extracted, for example be stored in a database
In, in this way when needing the characteristic point using any alternating content therein, read from first database, therefore can be with
Improve the efficiency of commending contents.Specific extracting mode can be identical as a upper example.
The official documents and correspondence column information of step 204, the characteristic point classification based on the first content and the target exhibition position generates
Multiple official documents and correspondence templates.
After step 202 determines the characteristic point of first content, that is, it can determine that the classification of the characteristic point of first content,
Under normal conditions, for a content, a characteristic point corresponds to a kind of characteristic point classification.For example, with first content for one
For coupon information, if it is determined that the characteristic point gone out includes that entitled xx supermarket, discount are full 50 yuan and subtract 10 yuan and effectively
Phase is three characteristic points of day xxxx xx month xx, then the feature vertex type of this content includes title, discount and three kinds of validity period
Classification.
As an example, the characteristic point classification for determining first content may include: in the feature for determining first content
After point, it is directly based upon the characteristic point of first content, determines the characteristic point classification of first content.
Alternatively, as another example, determining that the characteristic point classification of first content may include: to read from the second database
The characteristic point classification of the first content is taken, the characteristic point classification of the first content is the characteristic point based on the first content
Determining.Namely each alternating content can also be determined in advance after the characteristic point for determining each alternating content
Characteristic point classification and be stored in the second database, needing the characteristic point classification using any alternating content therein in this way
When, it is read from the second database.
After determining the characteristic point classification of first content, it can be designed with the official documents and correspondence column information of combining target exhibition position
Multiple official documents and correspondence templates out, wherein the official documents and correspondence column information of exhibition position can be in the exhibition position for showing the official documents and correspondence letter of official documents and correspondence information
Cease position and the quantity on column.
For example, if first content includes 3 kinds of classifications of A, B and C, target exhibition position include 2 official documents and correspondence information bars (such as
Shown in Fig. 3, comprising the first official documents and correspondence column 311 and the second official documents and correspondence column 312 in the corresponding exhibition position in first area 31), then can be with
Design following several official documents and correspondence templates: the first, the first official documents and correspondence column 311 is used to show the characteristic point of A and B classification, the second text
Case column 312 is used to show the characteristic point of C classification, for example A is title, and B is discount, and C is to have led number, then the first official documents and correspondence column
311 for showing title and discount, and the second official documents and correspondence column 312 has led number for showing;Second, the first official documents and correspondence column 311
For showing the characteristic point of A classification, the second official documents and correspondence column 312 is used to show the characteristic point of B and C classification;The third, the first official documents and correspondence
Column 311 is used to show the characteristic point of C classification, and the second official documents and correspondence column 312 is used to show the characteristic point, etc. of A and B classification.
It is appreciated that when first content characteristic point classification be it is multiple, can when the official documents and correspondence column of target exhibition position is also multiple
To design multiple official documents and correspondence templates by different permutation and combination methods.
The characteristic point of the first content is inserted the multiple official documents and correspondence template respectively by step 206, is obtained in described first
The multiple recommendation official documents and correspondences held.
For example, if the quantity for the official documents and correspondence template designed in step 204 is 4, then by the characteristic point of first content
Specifying information is inserted after this 4 official documents and correspondence templates, the available 4 recommendation official documents and correspondences about first content.
Step 208, by testing the multiple recommendation official documents and correspondence, determine target from the multiple official documents and correspondence template
Official documents and correspondence template, the target official documents and correspondence template is used as recommends the official documents and correspondence template of the second content on the target exhibition position, and described second
The characteristic point classification of content and the characteristic point classification of the first content are identical.
As an example, in a step 208, institute can be obtained by carrying out AB test to the multiple recommendation official documents and correspondence
State multiple test results for recommending official documents and correspondence;Determine the target of test result the second preset condition of satisfaction in the multiple recommendation official documents and correspondence
Official documents and correspondence, specifically, the test result can be conversion ratio, it is most high that the second preset condition can be conversion ratio;By the target
Official documents and correspondence template used in official documents and correspondence is determined as the target official documents and correspondence template.
AB tests (A/B test), is exactly that (for example this specification is real for two schemes of the same target making in simple terms
Apply two official documents and correspondence templates in example), allow a part of user using A scheme, another part user uses B scheme, records user
Service condition, see that the effect of which scheme is more preferable.AB test is tested between separate type group, it is therefore intended that passes through section
Experimental design, sample representativeness, flow, which are divided with modes such as small flow rate tests, obtains representative experiment
Conclusion, and firmly believe that the conclusion is believable in being generalized to whole flows.
Certainly, in addition to AB is tested, target text can also be selected from this multiple official documents and correspondence template by other test modes
The modes such as case template, such as sampling assessment.
It is appreciated that for target exhibition position after determining target official documents and correspondence template based on first content, optionally, Fig. 2
Shown in method can also include:
Based on the characteristic point of the target official documents and correspondence template and the first content, the recommendation text of the first content is generated
Case;Specifically, the characteristic point of the first content is inserted the target official documents and correspondence template, the recommendation text of the first content is obtained
Case;
The recommendation official documents and correspondence that the first content is shown on the target exhibition position, realizes the recommendation of the first content.
Alternatively, optionally, method shown in Fig. 2 is also based on the target official documents and correspondence template recommended characteristics point classification and first
Identical second content of content, is detailed in embodiment shown in Fig. 4, wouldn't repeat herein.
A kind of content recommendation method that this specification embodiment provides, due to can be based on the characteristic point classification of first content
With the official documents and correspondence column information of target exhibition position, multiple official documents and correspondence templates are generated, and select one from this multiple official documents and correspondence template by test
A template, as the target official documents and correspondence template of recommended characteristics point classification the second content identical with the characteristic point classification of first content,
Rather than all contents are recommended using identical official documents and correspondence template, therefore can more attract the attention of user after recommending,
To improve the first content of recommendation and the changing effect of the second content.
Optionally, as shown in figure 4, after above-mentioned steps 208, a kind of commending contents side of this specification embodiment offer
Method can also include:
Step 210 determines second content for needing to recommend on the target exhibition position.
In one example, similar with the above determining mode of first content, it can also be from multiple candidates to be recommended
Second content is determined in content.Specifically, second content is determined from multiple alternating contents to be recommended, it can
To include:
Sub-step 1 obtains multiple alternating contents to be recommended.
For example, can be based on preset strategy (as random), from 10000 contents, roughing goes out 100 as to be recommended
Multiple alternating contents.
Sub-step 2, the ranking score based on default estimation the multiple alternating content of model pre-estimating.
For example, 100 gone out using Wide&Deep model (other Ordering and marking models can also be used certainly) to roughing
The click-through-rate (Click Through Rate, CTR) of alternating content is estimated, using the row as 100 alternating contents
Sequence score.
CTR is the common term of Internet advertising, refers to the web advertisement (display advertising, copy, AdWords, row
Name advertisement, video ads etc.) click arrival rate, i.e. the actual click number of the advertisement (for stringent, can be arrival mesh
Mark the quantity of the page) divided by the amount of showing (Show content) of advertisement.CTR is that one of measurement Internet advertising effect is important
Index.CTR, which estimates needs and provides, accurately clicks probability, and A clicking rate 0.3%, B clicking rate 0.13% etc., there are many proposed algorithm
When only need to obtain an optimal order A > B > C, be exposed based on ranking results combination scene.
Sub-step 3, the ranking score based on the multiple alternating content are ranked up the multiple alternating content.
For example, can be ranked up according to the sequence of click-through-rate from high to low to 100 alternating contents.
The alternating content that sequence meets the first preset condition is determined as second content by sub-step 4.
As an example, specifically using which content in a plurality of alternating content after sequence as the second content, with
The position of target exhibition position is related.If target exhibition position is the corresponding exhibition position in first area 31 in display interface shown in Fig. 3, then
Will sequence first (the first preset condition) alternating content as the first content;If target exhibition position is shown in Fig. 3
Display interface in the corresponding exhibition position of second area 32, then will sequence second (the first preset condition) alternating content make
For the first content, and so on, this specification does not repeat one by one.
Step 212, the characteristic point for determining second content.
As an example, step 210 may include: the extension letter for obtaining second content and second content
Breath extracts the characteristic point of second content from the extension information of second content and second content.
As another example, step 210 may include: that the feature of second content is read from first database
Point, the characteristic point of second content are extracted from the extension information of second content and second content.
This can also be extracted in advance from each alternating content (including the second content) and the extension information of this alternating content
The characteristic point of alternating content and preservation, for example be stored in a database, it is needing in this way using any candidate therein
When the characteristic point of content, read from first database, commending contents efficiency can be improved in this.
Step 214, the characteristic point based on the target official documents and correspondence template and second content generate second content
Recommend official documents and correspondence.
The characteristic point of second content is inserted into the target official documents and correspondence template, the recommendation of second content can be obtained
Official documents and correspondence.
Step 216, the recommendation official documents and correspondence that second content is shown on the target exhibition position realize second content
Recommend.
A kind of content recommendation method that this specification embodiment provides, the content provided compared to embodiment shown in Fig. 2
Recommended method recommends the second content due to can use the preferable target official documents and correspondence template determined based on first content,
The changing effect of the second content can be promoted.
Optionally, as shown in figure 5, in another embodiment, before above-mentioned steps 202, Fig. 2 or one kind shown in Fig. 4
Content recommendation method can also include:
Step 218 determines the first content from multiple alternating contents to be recommended.
As an example, the first content is determined from multiple alternating contents to be recommended, may include:
Sub-step 1 obtains multiple alternating contents to be recommended.
For example, can be based on preset strategy (as random), from 10000 contents, roughing goes out 100 as to be recommended
Multiple alternating contents.
Sub-step 2, the ranking score based on default estimation the multiple alternating content of model pre-estimating.
For example, click-through-rate (the Click of 100 alternating contents gone out using Wide&Deep model to roughing
Through Rate, CTR) it is estimated, using the ranking score as 100 alternating contents.
Sub-step 3, the ranking score based on the multiple alternating content are ranked up the multiple alternating content.
For example, can be ranked up according to the sequence of click-through-rate from high to low to 100 alternating contents.
The alternating content that sequence meets the first preset condition is determined as the first content by sub-step 4.
As an example, specifically using which content in a plurality of alternating content after sequence as first content, with
The position of target exhibition position is related.If target exhibition position is the corresponding exhibition position in first area 31 in display interface shown in Fig. 3, then
It will sort in primary alternating content as the first content;If target exhibition position is in display interface shown in Fig. 3
The corresponding exhibition position of second area 32, then will sequence in deputy alternating content as the first content, and so on, this theory
It is bright not repeat one by one.
A kind of content recommendation method that this specification embodiment provides is provided compared to Fig. 2 or embodiment shown in Fig. 4
Content recommendation method can not only determine the preferable target official documents and correspondence template for recommending first content and the second content, and
And the first content recommended is that the ranking score determined from multiple alternating contents to be recommended meets the first preset condition
Alternating content, therefore can preferably cause user's note that so as to more after recommendation first content on target exhibition position
Further promote the changing effect of first content.
As shown in fig. 6, this specification embodiment also provides a kind of content recommendation method, this method may include:
Step 602 determines the second content for needing to recommend on target exhibition position.
The second content can be filtered out from multiple alternating contents to be recommended, specifically, available to be recommended more
A alternating content;Ranking score based on default estimation the multiple alternating content of model pre-estimating;Based in the multiple candidate
The ranking score of appearance is ranked up the multiple alternating content;The alternating content that sequence meets the first preset condition is determined as
Second content.
It is appreciated that this scheme for selecting the second content from multiple alternating contents according to ranking score, because select
Two contents are the alternating contents met certain condition, therefore, can be further after the second content is recommended on target exhibition position
Ground causes user's note that further promote the changing effect of the second content.
Step 604, the characteristic point for determining second content, the characteristic point are the characteristic informations for attracting user.
As an example, step 604 may include: the extension letter for obtaining second content and second content
Breath extracts the characteristic point of second content from the extension information of second content and second content.
As another example, step 604 may include: that the feature of second content is read from first database
Point, the characteristic point of second content are extracted from the extension information of second content and second content.
This can also be extracted in advance from each alternating content (including the second content) and the extension information of this alternating content
The characteristic point of alternating content and preservation, for example be stored in a database, it is needing in this way using any candidate therein
When the characteristic point of content, read from first database, commending contents efficiency can be improved in this.
Step 606, the characteristic point based on target official documents and correspondence template and second content generate the recommendation of second content
Official documents and correspondence, wherein the target official documents and correspondence template is to select to obtain by testing from multiple official documents and correspondence templates, the multiple official documents and correspondence mould
The official documents and correspondence column information that version is characteristic point classification based on first content and the target exhibition position includes generates, in described first
The characteristic point classification of appearance is identical as the characteristic point classification of second content.
Selection target official documents and correspondence template please refers to the description to embodiment shown in Fig. 2 above from multiple official documents and correspondence templates, this
Place repeats no more.
Characteristic point based on target official documents and correspondence template and second content generates the recommendation official documents and correspondence of second content, can
To include: that the characteristic point of second content is inserted the target official documents and correspondence template, the recommendation official documents and correspondence of the second content can be obtained.
Step 608, the recommendation official documents and correspondence that second content is shown on the target exhibition position realize second content
Recommend.
A kind of content recommendation method that this specification embodiment provides, due to that can be based in target official documents and correspondence template and second
The characteristic point of appearance generates the recommendation official documents and correspondence of the second content and recommends on target exhibition position, and target official documents and correspondence template therein is logical
Test is crossed to select from multiple official documents and correspondence templates, and the multiple official documents and correspondence template be characteristic point classification based on first content and
In the characteristic point classification and described second of the first content that the official documents and correspondence column information that the target exhibition position includes generates and described
The characteristic point classification of appearance is identical, rather than is recommended the second content using changeless official documents and correspondence template, therefore recommends the
The attention of user can preferably be attracted after two contents, so as to improve recommendation the second content changing effect.
Optionally, before step 606, a kind of content recommendation method shown in fig. 6 can also include:
Determine the characteristic point of the first content;
The official documents and correspondence column information of characteristic point classification and the target exhibition position based on the first content generates the multiple
Official documents and correspondence template;
The characteristic point of the first content is inserted into the multiple official documents and correspondence template respectively, obtains the multiple of the first content
Recommend official documents and correspondence;
By testing the multiple recommendation official documents and correspondence, the target official documents and correspondence is determined from the multiple official documents and correspondence template
Template.
Wherein, the characteristic point of the determination first content, comprising:
The extension information for obtaining the first content and the first content, from the first content and the first content
Extension information in, extract the characteristic point of the first content;Alternatively,
The characteristic point of the first content is read from first database, the characteristic point of the first content is from described
It is extracted in one content and the extension information of the first content.
Optionally, before step 606, a kind of content recommendation method shown in fig. 6 can also include:
Based on the characteristic point of the target official documents and correspondence template and the first content, the recommendation text of the first content is generated
Case;Specifically, the characteristic point of the first content is inserted the target official documents and correspondence template, the recommendation text of the first content is obtained
Case;
The recommendation official documents and correspondence that the first content is shown on the target exhibition position, realizes the recommendation of the first content.
Optionally, before the characteristic point for determining the first content, a kind of content recommendation method shown in fig. 6 may be used also
To include: to determine the first content from multiple alternating contents to be recommended.
It may include: that acquisition is to be recommended specifically, determining the first content from multiple alternating contents to be recommended
Multiple alternating contents;Ranking score based on default estimation the multiple alternating content of model pre-estimating;Based on the multiple time
The ranking score of content is selected to be ranked up the multiple alternating content;The alternating content that sequence is met the first preset condition is true
It is set to the first content.
It is also understood that this scheme for selecting first content from multiple alternating contents according to ranking score, because selecting
First content be the alternating content for meeting certain predetermined condition, therefore, when first content by target exhibition position recommend after, energy
Further cause user's note that further promote the changing effect of first content.
Wherein, described by testing the multiple recommendation official documents and correspondence, institute is determined from the multiple official documents and correspondence template
State target official documents and correspondence template, comprising: AB test is carried out to the multiple recommendation official documents and correspondence, obtains the multiple test knot for recommending official documents and correspondence
Fruit;Determine the target official documents and correspondence of test result the second preset condition of satisfaction in the multiple recommendation official documents and correspondence;By the target official documents and correspondence institute
The official documents and correspondence template used is determined as the target official documents and correspondence template.
Wherein, the characteristic point of the determination second content, comprising: obtain second content and second content
Extension information extract the characteristic point of second content from the extension information of second content and second content;
Alternatively, reading the characteristic point of second content from first database, the characteristic point of second content is from described second
It is extracted in the extension information of content and second content.
Wherein, the characteristic point of the determination first content, comprising: obtain the first content and the first content
Extension information extract the characteristic point of the first content from the extension information of the first content and the first content;
Alternatively, reading the characteristic point of the first content from first database, the characteristic point of the first content is from described first
It is extracted in content and the extension information of the first content.
It is appreciated that when the characteristic point for the characteristic point for reading the first content (or second content) from first database
When, it can further promote the efficiency for recommending the first content (or second content).
It is that a kind of explanation of content recommendation method is provided this specification above, the electronics provided below this specification is set
It is standby to be introduced.
Fig. 7 is the structural schematic diagram for the electronic equipment that one embodiment of this specification provides.Referring to FIG. 7, in hardware
Level, the electronic equipment include processor, optionally further comprising internal bus, network interface, memory.Wherein, memory can
It can include memory, such as high-speed random access memory (Random-Access Memory, RAM), it is also possible to further include non-easy
The property lost memory (non-volatile memory), for example, at least 1 magnetic disk storage etc..Certainly, which is also possible to
Including hardware required for other business.
Processor, network interface and memory can be connected with each other by internal bus, which can be ISA
(Industry Standard Architecture, industry standard architecture) bus, PCI (Peripheral
Component Interconnect, Peripheral Component Interconnect standard) bus or EISA (Extended Industry Standard
Architecture, expanding the industrial standard structure) bus etc..The bus can be divided into address bus, data/address bus, control always
Line etc..Only to be indicated with a four-headed arrow in Fig. 8, it is not intended that an only bus or a type of convenient for indicating
Bus.
Memory, for storing program.Specifically, program may include program code, and said program code includes calculating
Machine operational order.Memory may include memory and nonvolatile memory, and provide instruction and data to processor.
Processor is from the then operation into memory of corresponding computer program is read in nonvolatile memory, in logical layer
Content recommendation device is formed on face.Processor executes the program that memory is stored, and is specifically used for executing following operation:
Determine the characteristic point for needing the first content recommended on target exhibition position, the characteristic point is for attracting user
Characteristic information;
The official documents and correspondence column information of characteristic point classification and the target exhibition position based on the first content, generates multiple official documents and correspondences
Template;
The characteristic point of the first content is inserted into the multiple official documents and correspondence template respectively, obtains the multiple of the first content
Recommend official documents and correspondence;
By testing the multiple recommendation official documents and correspondence, target official documents and correspondence mould is determined from the multiple official documents and correspondence template
Version, the target official documents and correspondence template are used as the official documents and correspondence template for recommending the second content on the target exhibition position, second content
Characteristic point classification is identical as the characteristic point classification of the first content.
Alternatively, processor, executes the program that memory is stored, and it is specifically used for executing following operation:
Determine the second content for needing to recommend on target exhibition position;
Determine that the characteristic point of second content, the characteristic point are the characteristic informations for attracting user;
Characteristic point based on target official documents and correspondence template and second content generates the recommendation official documents and correspondence of second content,
In, the target official documents and correspondence template is to select to obtain by testing from multiple official documents and correspondence templates, and the multiple official documents and correspondence template is base
It is generated in the official documents and correspondence column information that the characteristic point classification of first content and the target exhibition position include, the spy of the first content
Sign point classification is identical as the characteristic point classification of second content;
The recommendation official documents and correspondence of second content is shown on the target exhibition position, realizes the recommendation of second content.
Content recommendation method disclosed in the above-mentioned any embodiment as shown in this specification Fig. 2, Fig. 4-Fig. 6 can be applied
It is realized in processor, or by processor.Processor may be a kind of IC chip, the processing capacity with signal.
During realization, each step of the above method can pass through the integrated logic circuit or software form of the hardware in processor
Instruction complete.Above-mentioned processor can be general processor, including central processing unit (Central Processing
Unit, CPU), network processing unit (Network Processor, NP) etc.;It can also be digital signal processor (Digital
Signal Processor, DSP), specific integrated circuit (Application Specific Integrated Circuit,
ASIC), field programmable gate array (Field-Programmable Gate Array, FPGA) or other programmable logic
Device, discrete gate or transistor logic, discrete hardware components.It may be implemented or execute this specification one or more
Disclosed each method, step and logic diagram in embodiment.General processor can be microprocessor or the processor
It can be any conventional processor etc..The step of method in conjunction with disclosed in this specification one or more embodiment, can be straight
Connect and be presented as that hardware decoding processor executes completion, or in decoding processor hardware and software module combination executed
At.Software module can be located at random access memory, and flash memory, read-only memory, programmable read only memory or electrically-erasable can
In the storage medium of this fields such as programmable memory, register maturation.The storage medium is located at memory, and processor reads storage
Information in device, in conjunction with the step of its hardware completion above method.
The electronic equipment can also carry out content recommendation method shown in any embodiment, this specification in 2, Fig. 4-Fig. 6 and exist
This is repeated no more.
Certainly, other than software realization mode, other implementations are not precluded in the electronic equipment of this specification, such as
Logical device or the mode of software and hardware combining etc., that is to say, that the executing subject of following process flow is not limited to each
Logic unit is also possible to hardware or logical device.
This specification embodiment also proposed a kind of computer readable storage medium, the computer-readable recording medium storage
One or more programs, the one or more program include instruction, and the instruction is when by the portable electric including multiple application programs
When sub- equipment executes, the method that the portable electronic device can be made to execute Fig. 2, Fig. 4 or embodiment illustrated in fig. 5, and specifically use
The operation below executing:
Determine the characteristic point for needing the first content recommended on target exhibition position, the characteristic point is for attracting user
Characteristic information;
The official documents and correspondence column information of characteristic point classification and the target exhibition position based on the first content, generates multiple official documents and correspondences
Template;
The characteristic point of the first content is inserted into the multiple official documents and correspondence template respectively, obtains the multiple of the first content
Recommend official documents and correspondence;
By testing the multiple recommendation official documents and correspondence, target official documents and correspondence mould is determined from the multiple official documents and correspondence template
Version, the target official documents and correspondence template are used as the official documents and correspondence template for recommending the second content on the target exhibition position, second content
Characteristic point classification is identical as the characteristic point classification of the first content.
This specification embodiment also proposed a kind of computer readable storage medium, the computer-readable recording medium storage
One or more programs, the one or more program include instruction, and the instruction is when by the portable electric including multiple application programs
When sub- equipment executes, the method that the portable electronic device can be made to execute embodiment illustrated in fig. 6, and be specifically used for executing following
Operation:
Determine the second content for needing to recommend on target exhibition position;
Determine that the characteristic point of second content, the characteristic point are the characteristic informations for attracting user;
Characteristic point based on target official documents and correspondence template and second content generates the recommendation official documents and correspondence of second content,
In, the target official documents and correspondence template is to select to obtain by testing from multiple official documents and correspondence templates, and the multiple official documents and correspondence template is base
It is generated in the official documents and correspondence column information that the characteristic point classification of first content and the target exhibition position include, the spy of the first content
Sign point classification is identical as the characteristic point classification of second content;
The recommendation official documents and correspondence of second content is shown on the target exhibition position, realizes the recommendation of second content.
The content recommendation device provided below this specification is illustrated.
As shown in figure 8, one embodiment of this specification provides a kind of content recommendation device 800, it is real in a kind of software
It applies in mode, the content recommendation device 800 can include: fisrt feature point determining module 801, official documents and correspondence template generation module 802,
One official documents and correspondence generation module 803 and template selecting module 804.
Fisrt feature point determining module 801, for determining the characteristic point for needing the first content recommended on target exhibition position,
The characteristic point is the characteristic information for attracting user.
As an example, fisrt feature point determining module 801 can be used for obtaining in the first content and described first
The extension information of appearance extracts the feature of the first content from the extension information of the first content and the first content
Point.
As another example, fisrt feature point determining module 801 can be used for reading described first from first database
The characteristic point of content, the characteristic point of the first content are mentioned from the extension information of the first content and the first content
It takes out.
Official documents and correspondence template generation module 802, for based on the first content characteristic point classification and the target exhibition position
Official documents and correspondence column information generates multiple official documents and correspondence templates.
As an example, the characteristic point classification for determining first content may include: in the feature for determining first content
After point, it is directly based upon the characteristic point of first content, determines the characteristic point classification of first content.
Alternatively, as another example, determining that the characteristic point classification of first content may include: to read from the second database
The characteristic point classification of the first content is taken, the characteristic point classification of the first content is the characteristic point based on the first content
Determining.
After determining the characteristic point classification of first content, it can be designed with the official documents and correspondence column information of combining target exhibition position
Multiple official documents and correspondence templates out, wherein the official documents and correspondence column information of exhibition position can be in the exhibition position for showing the official documents and correspondence letter of official documents and correspondence information
Cease position and the quantity on column.
It is appreciated that when first content characteristic point classification be it is multiple, can when the official documents and correspondence column of target exhibition position is also multiple
To design multiple official documents and correspondence templates by different permutation and combination methods.
First official documents and correspondence generation module 803, for the characteristic point of the first content to be inserted the multiple official documents and correspondence mould respectively
Version, obtains multiple recommendation official documents and correspondences of the first content.
Template selecting module 804, for by testing the multiple recommendation official documents and correspondence, from the multiple official documents and correspondence template
In determine target official documents and correspondence template, the target official documents and correspondence template is used as the official documents and correspondence mould for recommending the second content on the target exhibition position
Version, the characteristic point classification of second content and the characteristic point classification of the first content are identical.
As an example, template selecting module 804 can be used for carrying out AB test to the multiple recommendation official documents and correspondence, obtain
The multiple test result for recommending official documents and correspondence;Determine the mesh of test result the second preset condition of satisfaction in the multiple recommendation official documents and correspondence
Mark official documents and correspondence;Official documents and correspondence template used in the target official documents and correspondence is determined as the target official documents and correspondence template.
Certainly, template selecting module 804 can be also used for carrying out testing by other to the multiple recommendation official documents and correspondence, such as
The modes such as sampling assessment obtain the multiple test result for recommending official documents and correspondence;Determine test result in the multiple recommendation official documents and correspondence
Meet the target official documents and correspondence of the second preset condition;Official documents and correspondence template used in the target official documents and correspondence is determined as the target official documents and correspondence mould
Version.
Optionally, device 800 shown in Fig. 8 can also include: the first display module, for being based on the target official documents and correspondence mould
The characteristic point of version and the first content, generates the recommendation official documents and correspondence of the first content;Described in being shown on the target exhibition position
The recommendation official documents and correspondence of first content, realizes the recommendation of the first content.
A kind of content recommendation device 800 that this specification embodiment provides, due to can be based on the characteristic point of first content
The official documents and correspondence column information of classification and target exhibition position generates multiple official documents and correspondence templates, and is selected from this multiple official documents and correspondence template by test
A template is selected, the target official documents and correspondence mould as recommended characteristics point classification the second content identical with the characteristic point classification of first content
Version, rather than all contents are recommended using identical official documents and correspondence template, therefore can more attract the attention of user after recommending
Power, to improve the first content of recommendation and the changing effect of the second content.
Optionally, as shown in figure 9, the device 800 that this specification embodiment provides, can also include: the first determining module
805, second feature point determining module 806, the second official documents and correspondence generation module 807 and the second display module 808.
First determining module 805, second content recommended on the target exhibition position for determining needs.
In one example, with hereinafter determine the mode of first content it is similar, the first determining module 805 can also to
Second content is determined in the multiple alternating contents recommended.
Specifically, the first determining module 805 can be used for:
Obtain multiple alternating contents to be recommended;
Ranking score based on default estimation the multiple alternating content of model pre-estimating;
Ranking score based on the multiple alternating content is ranked up the multiple alternating content;
The alternating content that sequence meets the first preset condition is determined as second content.
Second feature point determining module 806, for determining the characteristic point of second content.
As an example, second feature point determining module 806 can be used for obtaining in second content and described second
The extension information of appearance extracts the feature of second content from the extension information of second content and second content
Point.
As another example, second feature point determining module 806 can be used for reading described second from first database
The characteristic point of content, the characteristic point of second content are mentioned from the extension information of second content and second content
It takes out.
Second official documents and correspondence generation module 807, it is raw for the characteristic point based on the target official documents and correspondence template and second content
At the recommendation official documents and correspondence of second content.
Second display module 808 realizes institute for showing the recommendation official documents and correspondence of second content on the target exhibition position
State the recommendation of the second content.
A kind of content recommendation device 800 that embodiment shown in Fig. 9 provides is provided compared to embodiment shown in Fig. 8
Content recommendation method recommends the second content due to can use the preferable target official documents and correspondence template determined based on first content,
Therefore the changing effect of the second content can be promoted.
Optionally, as shown in Figure 10, this specification embodiment provide device 800, can also include:
Second determining module 809, for before determining the characteristic point of first content that needs are recommended on target exhibition position,
The first content is determined from multiple alternating contents to be recommended.
As an example, second determining module 809 can include:
Acquisition submodule, for obtaining multiple alternating contents to be recommended.
Score estimates submodule, for the ranking score based on default estimation the multiple alternating content of model pre-estimating.
Sorting sub-module arranges the multiple alternating content for the ranking score based on the multiple alternating content
Sequence.
Determine submodule, the alternating content for sequence to be met the first preset condition is determined as the first content.
A kind of content recommendation device 800 that embodiment shown in Fig. 10 provides, compared to Fig. 8 or embodiment shown in Fig. 9
The content recommendation device 800 of offer can not only determine the preferable target text for recommending first content and the second content
Case template, and the first content recommended is that the ranking score determined from multiple alternating contents to be recommended meets certain item
The alternating content of part, thus on target exhibition position recommend first content after, can preferably cause user's note that so as to
Further to promote the changing effect of first content.
It should be noted that content recommendation device 800 can be realized Fig. 2, Fig. 4 or embodiment of the method shown in fig. 5 provides
Method, specifically refer to the content recommendation method of Fig. 2, Fig. 4 or embodiment illustrated in fig. 5, repeat no more.
As shown in figure 11, this specification embodiment also provides a kind of content recommendation device 1100, and this method may include:
One determining module 1101, second feature point determining module 1102, the second official documents and correspondence generation module 1103 and the second display module 1104.
First determining module 1101, the second content recommended on target exhibition position for determining needs.
Specifically, the first determining module 1101 can filter out the second content from multiple alternating contents to be recommended, tool
The screening mode of body refers to the description above to embodiment shown in Fig. 9, and details are not described herein again.
Second feature point determining module 1102, for determining that the characteristic point of second content, the characteristic point are to be used for
Attract the characteristic information of user.
As an example, second feature point determining module 1102 can be used for obtaining second content and described second
The extension information of content extracts the spy of second content from the extension information of second content and second content
Sign point;Alternatively,
As another example, second feature point determining module 1102 can be used for reading described from first database
The characteristic point of two contents, the characteristic point of second content are from the extension information of second content and second content
It extracts.
Second official documents and correspondence generation module 1103 is generated for the characteristic point based on target official documents and correspondence template and second content
The recommendation official documents and correspondence of second content, wherein the target official documents and correspondence template is selected by testing from multiple official documents and correspondence templates
It arrives, the official documents and correspondence column letter that the multiple official documents and correspondence template is characteristic point classification based on first content and the target exhibition position includes
What breath generated, the characteristic point classification of the first content is identical as the characteristic point classification of second content.
Specifically, the second official documents and correspondence generation module 1103, for the characteristic point of second content to be inserted the target text
Case template, to obtain the recommendation official documents and correspondence of the second content.
Second display module 1104 is realized for showing the recommendation official documents and correspondence of second content on the target exhibition position
The recommendation of second content.
A kind of content recommendation device 1100 that this specification embodiment provides, since target official documents and correspondence template and the can be based on
The characteristic point of two contents generates the recommendation official documents and correspondence of the second content and recommends on target exhibition position, and target official documents and correspondence template therein
It is to be selected by testing from multiple official documents and correspondence templates, and the multiple official documents and correspondence template is the characteristic point class based on first content
Not and the target exhibition position official documents and correspondence column information that includes generate and the first content characteristic point classification and described the
The characteristic point classification of two contents is identical, rather than is recommended using changeless official documents and correspondence template the second content, therefore push away
Recommend the attention that can preferably attract user after the second content, so as to improve recommendation the second content changing effect.
Optionally, content recommendation device 1100 can also include: fisrt feature point determining module, official documents and correspondence template generation mould
Block, the first official documents and correspondence generation module and template selecting module.
Fisrt feature point determining module, for generating the recommendation text of second content in the second official documents and correspondence generation module
Before case, the characteristic point of the first content is determined.
As an example, the fisrt feature point determining module, can be used for obtaining the first content and described first
The extension information of content extracts the spy of the first content from the extension information of the first content and the first content
Sign point.
As another example, the fisrt feature point determining module can be used for reading described from first database
The characteristic point of one content, the characteristic point of the first content are from the extension information of the first content and the first content
It extracts.
Official documents and correspondence template generation module, the official documents and correspondence for characteristic point classification and the target exhibition position based on the first content
Column information generates the multiple official documents and correspondence template.
First official documents and correspondence generation module, for the characteristic point of the first content to be inserted the multiple official documents and correspondence template respectively,
Obtain multiple recommendation official documents and correspondences of the first content.
Template selecting module, for by testing the multiple recommendation official documents and correspondence, from the multiple official documents and correspondence template
Determine the target official documents and correspondence template.
In one example, the template selecting module can be used for carrying out AB test to the multiple recommendation official documents and correspondence, obtain
The multiple test result for recommending official documents and correspondence;Determine the mesh of test result the second preset condition of satisfaction in the multiple recommendation official documents and correspondence
Mark official documents and correspondence;Official documents and correspondence template used in the target official documents and correspondence is determined as the target official documents and correspondence template.
Optionally, content recommendation device 1100 can also include:
First display module, for the characteristic point based on the target official documents and correspondence template and the first content, described in generation
The recommendation official documents and correspondence of first content;The recommendation official documents and correspondence of the first content is shown on the target exhibition position, is realized in described first
The recommendation of appearance.
Optionally, content recommendation device 1100 can also include:
Second determining module, for before the characteristic point for determining the first content, out of, multiple candidates to be recommended
The first content is determined in appearance.
Specifically, second determining module can include:
Acquisition submodule, for obtaining multiple alternating contents to be recommended;
Score estimates submodule, for the ranking score based on default estimation the multiple alternating content of model pre-estimating;
Sorting sub-module arranges the multiple alternating content for the ranking score based on the multiple alternating content
Sequence;
Determine submodule, the alternating content for sequence to be met the first preset condition is determined as the first content.
Sorting sub-module arranges the multiple alternating content for the ranking score based on the multiple alternating content
Sequence;
Determine submodule, the alternating content for sequence to be met the first preset condition is determined as the first content.
It is appreciated that this scheme for selecting first content from multiple alternating contents according to ranking score, because select
One content is the alternating content for meeting certain predetermined condition, therefore, when first content by target exhibition position recommend after, can more into
Cause to one step user's note that further promote the changing effect of first content.
It should be noted that the method that content recommendation device 1100 can be realized embodiment of the method shown in fig. 6, specifically may be used
With reference to the content recommendation method of embodiment illustrated in fig. 6, repeat no more.
Above-mentioned that this specification specific embodiment is described, other embodiments are in the scope of the appended claims
It is interior.In some cases, the movement recorded in detail in the claims or step can be come according to the sequence being different from embodiment
It executes and desired result still may be implemented.In addition, process depicted in the drawing not necessarily require show it is specific suitable
Sequence or consecutive order are just able to achieve desired result.In some embodiments, multitasking and parallel processing be also can
With or may be advantageous.
All the embodiments in this specification are described in a progressive manner, same and similar portion between each embodiment
Dividing may refer to each other, and each embodiment focuses on the differences from other embodiments.Especially for device reality
For applying example, since it is substantially similar to the method embodiment, so being described relatively simple, related place is referring to embodiment of the method
Part explanation.
In short, being not intended to limit the protection of this specification the foregoing is merely the preferred embodiment of this specification
Range.With within principle, made any modification, changes equivalent replacement all spirit in this specification one or more embodiment
Into etc., it should be included within the protection scope of this specification one or more embodiment.
System, device, module or the unit that above-described embodiment illustrates can specifically realize by computer chip or entity,
Or it is realized by the product with certain function.It is a kind of typically to realize that equipment is computer.Specifically, computer for example may be used
Think personal computer, laptop computer, cellular phone, camera phone, smart phone, personal digital assistant, media play
It is any in device, navigation equipment, electronic mail equipment, game console, tablet computer, wearable device or these equipment
The combination of equipment.
Computer-readable medium includes permanent and non-permanent, removable and non-removable media can be by any method
Or technology come realize information store.Information can be computer readable instructions, data structure, the module of program or other data.
The example of the storage medium of computer includes, but are not limited to phase change memory (PRAM), static random access memory (SRAM), moves
State random access memory (DRAM), other kinds of random access memory (RAM), read-only memory (ROM), electric erasable
Programmable read only memory (EEPROM), flash memory or other memory techniques, read-only disc read only memory (CD-ROM) (CD-ROM),
Digital versatile disc (DVD) or other optical storage, magnetic cassettes, tape magnetic disk storage or other magnetic storage devices
Or any other non-transmission medium, can be used for storage can be accessed by a computing device information.As defined in this article, it calculates
Machine readable medium does not include temporary computer readable media (transitory media), such as the data-signal and carrier wave of modulation.
It should also be noted that, the terms "include", "comprise" or its any other variant are intended to nonexcludability
It include so that the process, method, commodity or the equipment that include a series of elements not only include those elements, but also to wrap
Include other elements that are not explicitly listed, or further include for this process, method, commodity or equipment intrinsic want
Element.When not limiting more, the element that is limited by sentence "including a ...", it is not excluded that in the mistake including the element
There is also other identical elements in journey, method, commodity or equipment.
All the embodiments in this specification are described in a progressive manner, same and similar portion between each embodiment
Dividing may refer to each other, and each embodiment focuses on the differences from other embodiments.Especially for system reality
For applying example, since it is substantially similar to the method embodiment, so being described relatively simple, related place is referring to embodiment of the method
Part explanation.