CN110189173A - Advertisement generation method and device - Google Patents
Advertisement generation method and device Download PDFInfo
- Publication number
- CN110189173A CN110189173A CN201910451519.4A CN201910451519A CN110189173A CN 110189173 A CN110189173 A CN 110189173A CN 201910451519 A CN201910451519 A CN 201910451519A CN 110189173 A CN110189173 A CN 110189173A
- Authority
- CN
- China
- Prior art keywords
- material segment
- primary
- advertisement
- segment
- pointer network
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0276—Advertisement creation
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Accounting & Taxation (AREA)
- Development Economics (AREA)
- Strategic Management (AREA)
- Finance (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Economics (AREA)
- Marketing (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention proposes a kind of advertisement generation method and device, wherein this method comprises: obtaining multiple ad material segments；A plurality of candidate locations are generated according to the multiple ad material segment and preset pointer network；Targeted advertisements are determined from a plurality of candidate locations according to default rule.Thus, permutation and combination is carried out to multiple ad material segments based on preset pointer network and automatically generates the good a plurality of candidate locations of smoothness, and it introduces risk-control mechanism and a plurality of candidate locations is filtered to generate targeted advertisements, realize the ad pool for expanding advertiser, the advertisement that advertiser can be allowed to launch can be there is a greater chance that being able to expose or click.
Description
Technical field
The present invention relates to technical field of advertisement more particularly to a kind of advertisement generation method and devices.
Background technique
Currently, information flow advertiser will write advertising creative by advertiser, advertisement accurately is then carried out by algorithm and is pushed away
It send.However, advertiser is due to itself writing official documents and correspondence scarce capacity or not knowing about user for the hobby of WORDS AND PHRASES IN ADVERTISEMENT, so that writing
Advertisement lack attraction, the advertisement form write is not abundant enough.As advertisement recommendation side, analysis mass data point can be passed through
Analysis and conclusion, common some simple methods, such as the replacement of synonym, the setting including dynamic creative word packet can be with one
Determine degree to expand single advertisement, be pushed away with being easier to hit the keyword that advertisement is recommended to obtain relatively high advertisement
Ranking is recommended to help advertiser's optimization and enrich its advertising creative, as advertiser write primary advertisement " purchase 0 down payment of vehicle, is purchased by stages,
Light buying car " helps it to expand advertisement " purchase 0 down payment of vehicle, purchase by stages, light buying car, you do not look at also fastly " by algorithm.So
And its expansion capacity of such methods is limited, the rewriting amplitude for the primary advertisement write to advertiser is smaller.
Summary of the invention
The present invention is directed to solve at least some of the technical problems in related technologies.
For this purpose, the first purpose of this invention is to propose a kind of advertisement generation method.
Second object of the present invention is to propose a kind of advertisement generating means.
Third object of the present invention is to propose a kind of computer equipment.
Fourth object of the present invention is to propose a kind of computer readable storage medium.
In order to achieve the above object, first aspect present invention embodiment proposes a kind of advertisement generation method, comprising:
Obtain multiple ad material segments；
A plurality of candidate locations are generated according to the multiple ad material segment and preset pointer network；
Targeted advertisements are determined from a plurality of candidate locations according to default rule.
In one possible implementation, the method also includes:
Sample set is obtained, the sample set includes the corresponding material segment collection of multiple primary advertising copies, each primary wide
Accuse at least one material segment that the corresponding material segment collection of sample includes each primary advertising copy and it is corresponding at least one
Expand material segment；
More wheel training are carried out to initial pointer network according to the corresponding material segment collection of each primary advertising copy, are obtained
The preset pointer network.
In one possible implementation, it is described according to the corresponding material segment collection of each primary advertising copy to initial
Pointer network carry out more wheel training, obtain the preset pointer network, comprising:
After currently wheel training, according to when the primary advertising copy of front-wheel and when the calculating of front-wheel reality output result is current
Take turns the corresponding loss function of primary advertising copy, wherein when front-wheel reality output the result is that described the primary advertisement sample of front-wheel will be worked as
This corresponding material segment collection is input to what the initial pointer network obtained；
The initial pointer network is updated according to the loss function and stochastic gradient descent method, is obtained more
Pointer network after new carries out next round training using the updated pointer network, according to the primary advertising copy of next round
Loss function corresponding with the next round reality output result calculating primary advertising copy of next round, repeats this step, until
All training finishes primary advertising copy in the sample set, wherein next round reality output is the result is that by the next round
The corresponding material segment collection of primary advertising copy is input to what the updated pointer network obtained；
Using updated pointer network as the preset pointer network.
In one possible implementation, the multiple ad material segments of acquisition include:
It obtains at least one material segment of the primary advertisement and is obtained at least by expanding the primary advertisement
One expansion material segment；
At least one material segment of the primary advertisement and at least one described expansion material segment are broken up, obtained
To multiple ad material segments.
In one possible implementation, described to obtain at least one expansion material piece by expanding the primary advertisement
The mode of section includes one of following manner or a variety of:
At least one the material piece to match with each material segment of the primary advertisement is obtained from ad material library
At least one material segment that each material segment with the primary advertisement matches is determined as by expanding the original by section
At least one expansion material segment that raw advertisement obtains；
Or,
History material segment corresponding with the advertiser for writing the primary advertisement is obtained, from the history material segment
At least one the history material segment to match with each material segment of the primary advertisement is obtained, it will be with the primary advertisement
At least one history material segment for matching of each material segment be determined as by expand the primary advertisement obtain to
A few expansion material segment；
Or,
The corresponding general segment of the primary advertisement generic is obtained, the general segment is determined as by expanding institute
It states at least one that primary advertisement obtains and expands material segment.
In one possible implementation, described to determine target from a plurality of candidate locations according to default rule
Advertisement includes:
Show a plurality of candidate locations；
According to the selection information to a plurality of candidate locations, determine that the target is wide from a plurality of candidate locations
It accuses；
Or,
The a plurality of candidate locations are screened according to preset screening model, the candidate locations filtered out are determined as
The targeted advertisements.
Advertisement generation method provided in an embodiment of the present invention, by obtaining multiple ad material segments；According to the multiple
Ad material segment and preset pointer network generate a plurality of candidate locations；According to default rule from a plurality of candidate locations
Middle determining targeted advertisements.Permutation and combination is carried out to multiple ad material segments based on preset pointer network as a result, to automatically generate
The good a plurality of candidate locations of smoothness, and introduce risk-control mechanism a plurality of candidate locations are filtered it is wide to generate target
It accuses, realizes the ad pool for expanding advertiser, the advertisement that advertiser can be allowed to launch can be there is a greater chance that be able to expose or point
It hits.
In order to achieve the above object, second aspect of the present invention embodiment proposes a kind of advertisement generating means, comprising:
Module is obtained, for obtaining multiple ad material segments；
Generation module, it is a plurality of candidate wide for being generated according to the multiple ad material segment and preset pointer network
It accuses；
Determining module, for determining targeted advertisements from a plurality of candidate locations according to default rule.
In one possible implementation, described device further include: training module；
The acquisition module, is also used to obtain sample set, and the sample set includes the corresponding element of multiple primary advertising copies
Material segment collection, the corresponding material segment collection of each primary advertising copy include at least one material piece of each primary advertising copy
Section and at least one corresponding expansion material segment；
The training module, for according to the corresponding material segment collection of each primary advertising copy to initial pointer network
More wheel training are carried out, the preset pointer network is obtained.
In one possible implementation, the training module is specifically used for:
After currently wheel training, according to when the primary advertising copy of front-wheel and when the calculating of front-wheel reality output result is current
Take turns the corresponding loss function of primary advertising copy, wherein when front-wheel reality output the result is that described the primary advertisement sample of front-wheel will be worked as
This corresponding material segment collection is input to what the initial pointer network obtained；
The initial pointer network is updated according to the loss function and stochastic gradient descent method, is obtained more
Pointer network after new carries out next round training using the updated pointer network, according to the primary advertising copy of next round
Loss function corresponding with the next round reality output result calculating primary advertising copy of next round, repeats this step, until
All training finishes primary advertising copy in the sample set, wherein next round reality output is the result is that by the next round
The corresponding material segment collection of primary advertising copy is input to what the updated pointer network obtained；
Using updated pointer network as the preset pointer network.
In one possible implementation, the acquisition module includes acquiring unit and breaks up unit:
The acquiring unit, for obtaining at least one material segment of the primary advertisement and by expanding the original
At least one expansion material segment that raw advertisement obtains；
It is described to break up unit, at least one material segment of the primary advertisement and at least one described expansion is plain
Material segment is broken up, and multiple ad material segments are obtained.
In one possible implementation, the acquiring unit is specifically used for:
At least one the material piece to match with each material segment of the primary advertisement is obtained from ad material library
At least one material segment that each material segment with the primary advertisement matches is determined as by expanding the original by section
At least one expansion material segment that raw advertisement obtains；
Or,
History material segment corresponding with the advertiser for writing the primary advertisement is obtained, from the history material segment
At least one the history material segment to match with each material segment of the primary advertisement is obtained, it will be with the primary advertisement
At least one history material segment for matching of each material segment be determined as by expand the primary advertisement obtain to
A few expansion material segment；
Or,
The corresponding general segment of the primary advertisement generic is obtained, the general segment is determined as by expanding institute
It states at least one that primary advertisement obtains and expands material segment.
In one possible implementation, the determining module is specifically used for:
Show a plurality of candidate locations；
According to the selection information to a plurality of candidate locations, determine that the target is wide from a plurality of candidate locations
It accuses；
Or,
The a plurality of candidate locations are screened according to preset screening model, the candidate locations filtered out are determined as
The targeted advertisements.
Advertisement generating means provided in an embodiment of the present invention, by obtaining multiple ad material segments；According to the multiple
Ad material segment and preset pointer network generate a plurality of candidate locations；According to default rule from a plurality of candidate locations
Middle determining targeted advertisements.Permutation and combination is carried out to multiple ad material segments based on preset pointer network as a result, to automatically generate
The good a plurality of candidate locations of smoothness, and introduce risk-control mechanism a plurality of candidate locations are filtered it is wide to generate target
It accuses, realizes the ad pool for expanding advertiser, the advertisement that advertiser can be allowed to launch can be there is a greater chance that be able to expose or point
It hits.
In order to achieve the above object, third aspect present invention embodiment proposes a kind of computer equipment, including memory, processing
On a memory and the computer program that can run on a processor, when processor execution described program, is realized for device and storage
Advertisement generation method as described above.
To achieve the goals above, fourth aspect present invention embodiment proposes a kind of computer readable storage medium, when
When instruction in the storage medium is executed by processor, advertisement generation method as described above is realized.
The additional aspect of the present invention and advantage will be set forth in part in the description, and will partially become from the following description
Obviously, or practice through the invention is recognized.
Detailed description of the invention
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, in which:
Fig. 1 is a kind of flow diagram of advertisement generation method provided in an embodiment of the present invention；
Fig. 2 is the flow diagram of another advertisement generation method provided in an embodiment of the present invention；
Fig. 3 is a kind of structural schematic diagram of advertisement generating means provided in an embodiment of the present invention；
Fig. 4 is the structural schematic diagram of another advertisement generating means provided in an embodiment of the present invention；
Fig. 5 is a kind of structural schematic diagram of computer equipment provided in an embodiment of the present invention.
Specific embodiment
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, it is intended to is used to explain the present invention, and is not considered as limiting the invention.
Below with reference to the accompanying drawings the advertisement generation method and device of the embodiment of the present invention are described.
Fig. 1 is a kind of flow diagram of advertisement generation method provided in an embodiment of the present invention.Present embodiments provide one
Kind advertisement generation method, executing subject are advertisement generating means, which is made of hardware and/or software.Advertisement is raw
At device be specifically as follows the software installed on hardware device, such as terminal device, background server etc. or hardware device or
Application program etc..
As shown in Figure 1, the advertisement generation method, comprising the following steps:
S101, multiple ad material segments are obtained.
In the present embodiment, in order to greatly expand the richness of targeted advertisements, ad material segment is in addition to from by wide
Main write primary advertising is accused, also expands material segment from by expanding at least one that primary advertisement obtains.
Specifically, after getting the primary advertisement write by advertiser, using segment as granularity to primary advertisement into
Row cutting process obtains at least one material segment of primary advertisement.
For example, primary advertisement is " " continent of wind " Mmorpg, AR catches pet, and cute pet is taken a risk hand in hand ", in cutting process,
Using separator of the punctuation mark in primary advertisement between segment, 3 material segments cutting be respectively " here you
" continent of wind " Mmorpg ", " AR catches pet ", " cute pet venture hand in hand ".
Specifically, one of following manner or various ways can expand at least one expansion that primary advertisement obtains
Material segment, but it is not limited to this.
The first example:
" by way of expanding primary advertisement and obtaining at least one expansion material segment " are as follows: obtained from ad material library
At least one material segment to match with each material segment of the primary advertisement, by each element with the primary advertisement
At least one material segment that material segment matches is determined as expanding element by expanding at least one that the primary advertisement obtains
Material segment.
By taking above-mentioned primary advertisement " " continent of wind " Mmorpg, AR catches pet, and cute pet is taken a risk hand in hand " as an example, from ad material
At least one extracted in library expands material segment and just send for Japanese evil spirit hand trip, the imperial valued knife click of Tu.
In the present embodiment, the ad material of magnanimity is stored in ad material library.It is available rich by ad material library
Rich expansion material segment realizes the synonymous expansion to primary advertisement, greatly expands the richness of targeted advertisements, as much as possible
The problems such as business of avoiding the occurrence of distorts, mismatch.
Specifically, according to the difference of application scenarios, each material with the primary advertisement is obtained from ad material library
The mode at least one material segment that segment matches is different.
For example, after cutting to obtain each material segment to primary advertisement, it can be to each material piece of primary advertisement
Duan Jinhang semantic analysis is extracted according to the semantic analysis result of each material segment from ad material library with identical semanteme
One or more material segments.
In another example after cutting to obtain each material segment to primary advertisement, it can be to each material of primary advertisement
Each word of segment carries out semantic analysis, according to the semantic analysis result of each word of each material segment from ad material library
Extract one or more material segments with synonym.
Second of example:
" by way of expanding primary advertisement and obtaining at least one expansion material segment " are as follows: obtain and write and is described primary
The corresponding history material segment of the advertiser of advertisement obtains each element with the primary advertisement from the history material segment
At least one history material segment that material segment matches, each material segment with the primary advertisement is matched at least
One history material segment is determined as expanding material segment by expanding at least one that the primary advertisement obtains.
In the present embodiment, is obtained from the history material segment that same advertiser is write and expand material segment, Neng Goujin
Possibly guarantee under identical services scene, targeted advertisements have the material segment abundant that can describe business scenario, improve
The practicability of targeted advertisements reduces the risk of targeted advertisements.It should be pointed out that advertisement generating means are in advance by each advertisement
The history material segment for the primary advertisement that main history is write is stored.
For example, the manufacture of advertiser 1 is product 1, what advertiser 2 manufactured is product 2, from the history material piece of advertiser 1
The expansion material segment extracted in section is related to product 1, the expansion material piece extracted from the history material segment of advertiser 2
Duan Junyu product 2 is related, it is clear that expansion material piece is extracted from the history material segment of same advertiser can guarantee as far as possible
There is not the material segment for not meeting the business scenario for being associated with advertiser in ground, has greatly carried out risk control, and practical performance is good.
Specifically, according to the difference of application scenarios, each element with the primary advertisement is obtained from history material segment
The mode at least one material segment that material segment matches is different.
For example, after cutting to obtain each material segment to primary advertisement, it can be to each material piece of primary advertisement
Duan Jinhang semantic analysis, being extracted from history material segment according to the semantic analysis result of each material segment has identical semanteme
One or more material segments.
In another example after cutting to obtain each material segment to primary advertisement, it can be to each material of primary advertisement
Each word of segment carries out semantic analysis, according to the semantic analysis result of each word of each material segment from history material segment
It is middle to extract one or more material segments with synonym.
The third example:
" by way of expanding primary advertisement and obtaining at least one expansion material segment " are as follows: obtain the primary advertisement institute
Belong to the corresponding general segment of classification, the general segment is determined as at least one expansion obtained by expanding the primary advertisement
Fill material segment.
In the present embodiment, generic general segment can provide the same category advertisement with primary advertisement for targeted advertisements
The applicable official documents and correspondence material of some high frequencies.It should be pointed out that have collected each commercial paper in advance other general for advertisement generating means
Segment, and storage is classified to general segment according to advertisement classification.Advertisement classification is, for example, game class, electric business class, news
Class etc..
S102, a plurality of candidate locations are generated according to the multiple ad material segment and preset pointer network.
In the present embodiment, since pointer network (Pointer Networks) can effectively simply solve Combinatorial Optimization
Problem, for this purpose, the preset pointer network obtained by the sample training of magnanimity, and based on the preset pointer network to input
Multiple ad material segments carry out permutation and combination to form the good multiple candidate locations of smoothness, realize and expand the wide of advertiser
Pond is accused, the advertisement that advertiser can be allowed to launch can be there is a greater chance that being able to expose or click.
It should be pointed out that when multiple ad material segments at least derive from primary advertisement, ad material library, same advertisement
Main write history material segment, with primary advertisement generic general segment when, combined by preset pointer network lineup
A plurality of candidate locations generated have certain guarantee to the consistency etc. of smoothness, holding and service product, as far as possible
Ground avoids the problem that business is distorted and mismatch.
S103, targeted advertisements are determined from a plurality of candidate locations according to default rule.
In the present embodiment, although by the good multiple candidate locations of the available smoothness degree of preset pointer network,
The introducing that portion material segment may also occur brings some unnecessary risks, some unnecessary in order to be avoided as much as
Risk introduces risk-control mechanism, filters out the candidate locations there are risk, to guarantee the quality of the targeted advertisements of output.
According to the difference of application scenarios, the mode of targeted advertisements is determined not from a plurality of candidate locations according to default rule
Together.
For example, the implementation of " determining targeted advertisements from a plurality of candidate locations according to default rule " are as follows: aobvious
Show a plurality of candidate locations；According to the selection information to a plurality of candidate locations, determined from a plurality of candidate locations
The targeted advertisements.
In the present embodiment, by way of manual intervention, it is reduced as far as the generation of unnecessary risk.Specifically,
A plurality of candidate locations are shown to advertiser directly, voluntarily select target to be put wide from a plurality of candidate locations by advertiser
It accuses.
By taking advertisement generating means is the advertisement launching platforms of content supplier as an example, a choosing is shown in advertisement launching platform
Interface is selected, which includes a plurality of candidate locations；User is received to the selection information of a plurality of candidate locations, from a plurality of candidate
Targeted advertisements are determined in advertisement.
In another example the implementation of " determining targeted advertisements from a plurality of candidate locations according to default rule " are as follows:
The a plurality of candidate locations are screened according to preset screening model, the candidate locations filtered out are determined as the target
Advertisement.
In the present embodiment, targeted advertisements are filtered out from a plurality of candidate locations automatically by preset screening model.Wherein,
Preset screening model is set according to actual needs.For example, preset screening model is passes through the risk advertisement institute of magnanimity
The ad classification device of building can be realized and judge whether advertisement to be identified is risk advertisement.Wherein, risk advertisement is for example, normal
See advertisement, the advertisement with sensitive word etc. of mistake collocation, but it is not limited to this.
Table 1 show the targeted advertisements expanded based on primary advertisement.
Table 1
Advertisement generation method provided in an embodiment of the present invention obtains multiple ad material segments；According to the multiple advertisement
Material segment and preset pointer network generate a plurality of candidate locations；It is true from a plurality of candidate locations according to default rule
Set the goal advertisement.Permutation and combination is carried out to multiple ad material segments based on preset pointer network as a result, and automatically generates smoothness
The good a plurality of candidate locations of property, and introduce risk-control mechanism and a plurality of candidate locations are filtered to generate targeted advertisements,
Realize the ad pool for expanding advertiser, the advertisement that advertiser can be allowed to launch can be there is a greater chance that being able to expose or click
Deng.
Fig. 2 is the flow diagram of another advertisement generation method provided in an embodiment of the present invention.The present embodiment is to default
Training stage of pointer network be illustrated.In conjunction with reference Fig. 2, on the basis of embodiment shown in Fig. 1, step S101 it
Before, the advertisement generation method is further comprising the steps of:
S104, sample set is obtained.
Specifically, sample set includes the corresponding material segment collection of multiple primary advertising copies, each primary advertising copy pair
The material segment collection answered include each primary advertising copy at least one material segment and it is corresponding at least one expand material
Segment.
Wherein, each primary advertising copy it is corresponding at least one expand material segment, can from ad material library,
The history material and the generic general segment of primary advertisement of the advertiser of primary advertising copy are write, but it is not limited to this.
It should be pointed out that it is granularity to primary wide that at least one material segment of each primary advertising copy, which is using segment,
It accuses sample and carries out what cutting process obtained.
It should be pointed out that the implementation that at least one obtained by primary advertising copy expands material segment can be with
It is no longer superfluous herein referring to the mode of " expand at least one that primary advertisement obtains and expand material segment " of above-described embodiment introduction
It states.
For example, sample set includes the material segment collection etc. of the material segment collection of primary advertising copy A, primary advertising copy B；
Wherein, the material segment collection of primary advertising copy A includes obtaining at least one material segment to primary advertising copy A cutting process
And expand material segment by expanding at least one that primary advertising copy A is obtained；Wherein, the material of primary advertising copy B
Segment collection includes at least one material segment being obtained to primary advertising copy B cutting process and by expanding primary advertising copy
At least one expansion material segment that B is obtained.
S105, more trainings in rotation are carried out to initial pointer network according to the corresponding material segment collection of each primary advertising copy
Practice, obtains the preset pointer network.
Specifically, pointer network is once trained using a primary advertising copy corresponding material segment collection, often
Secondary training, which is finished, is updated optimization to the parameter of pointer network, and trained desired output result is the primary advertisement of input every time
Sample.
Since each primary advertising copy not only includes cutting primary advertising copy to obtain one or more in the present embodiment
Material segment further includes expanding primary advertising copy to obtain one or more expansion material segments, and training finishes default as a result,
Pointer network have permutation and combination carried out to multiple ad material segments of input to form good multiple candidate wide of smoothness
The ability of announcement realizes the ad pool for expanding advertiser, and the advertisement that advertiser can be allowed to launch can be there is a greater chance that be exposed
Light or click etc..
In the present embodiment, the specific implementation process of S105 the following steps are included:
S1, after current wheel training, according to when the primary advertising copy of front-wheel and when front-wheel reality output result calculates
When the corresponding loss function of the primary advertising copy of front-wheel, wherein the result is that will be described when front-wheel is primary wide when front-wheel reality output
It accuses the corresponding material segment collection of sample and is input to what the initial pointer network obtained.
Specifically, the gap degree of loss function characterization prediction and real data is the finger for measuring the quality of model prediction
Mark, more introductions about loss function are detailed in the relevant technologies.
S2, the initial pointer network is updated according to the loss function and stochastic gradient descent method, is obtained
To updated pointer network, next round training is carried out using the updated pointer network, according to the primary advertisement of next round
Sample and next round reality output result calculate the corresponding loss function of the primary advertising copy of next round, repeat this step,
Until all training finishes the primary advertising copy in the sample set, wherein next round reality output is the result is that under will be described
The corresponding material segment collection of the one primary advertising copy of wheel is input to what the updated pointer network obtained.
Specifically, stochastic gradient descent method first provides a class value of parameter at random, and then undated parameter, makes to update every time
Structure afterwards can allow loss function to become smaller, and be finally reached minimum.More introductions about stochastic gradient descent method
The relevant technologies are detailed in, are introduced about according to the more of parameter of loss function and stochastic gradient descent method more new model referring to phase
Pass technology.
In the present embodiment, after first round training, according to current loss function and stochastic gradient descent method to initial
The parameter of pointer network be adjusted, obtain updated pointer network；It, all can be according to current after subsequent each wheel training
Loss function is updated last round of obtained updated pointer network.
S3, using updated pointer network as the preset pointer network.
In the present embodiment, by obtaining sample set, sample set is obtained, the sample set includes multiple primary advertising copies pair
The material segment collection answered, the corresponding material segment collection of each primary advertising copy include at least one of each primary advertising copy
Material segment and at least one corresponding expansion material segment；According to the corresponding material segment collection of each primary advertising copy to first
The pointer network of beginning carries out more wheel training, obtains the preset pointer network.The preset pointer network that training finishes as a result,
Have and permutation and combination is carried out to form the ability of the good multiple candidate locations of smoothness, in fact to multiple ad material segments of input
Now expand the ad pool of advertiser, the advertisement that advertiser can be allowed to launch can be there is a greater chance that being able to expose or click.
Fig. 3 is a kind of structural schematic diagram of advertisement generating means provided in an embodiment of the present invention.Present embodiments provide one
Kind advertisement generating means, which is the executing subject of advertisement generation method, which is made of hardware and/or software.
As shown in figure 3, the advertisement generating means include: to obtain module 11, generation module 12, determining module 13.
Module 11 is obtained, for obtaining multiple ad material segments；
Generation module 12, it is a plurality of candidate wide for being generated according to the multiple ad material segment and preset pointer network
It accuses；
Determining module 13, for determining targeted advertisements from a plurality of candidate locations according to default rule.
In one possible implementation, the acquisition module 11 includes acquiring unit and breaks up unit:
The acquiring unit, for obtaining at least one material segment of the primary advertisement and by expanding the original
At least one expansion material segment that raw advertisement obtains；
It is described to break up unit, at least one material segment of the primary advertisement and at least one described expansion is plain
Material segment is broken up, and multiple ad material segments are obtained.
In one possible implementation, the acquiring unit 11 is specifically used for:
At least one the material piece to match with each material segment of the primary advertisement is obtained from ad material library
At least one material segment that each material segment with the primary advertisement matches is determined as by expanding the original by section
At least one expansion material segment that raw advertisement obtains；
Or,
History material segment corresponding with the advertiser for writing the primary advertisement is obtained, from the history material segment
At least one the history material segment to match with each material segment of the primary advertisement is obtained, it will be with the primary advertisement
At least one history material segment for matching of each material segment be determined as by expand the primary advertisement obtain to
A few expansion material segment；
Or,
The corresponding general segment of the primary advertisement generic is obtained, the general segment is determined as by expanding institute
It states at least one that primary advertisement obtains and expands material segment.
In one possible implementation, the determining module 13 is specifically used for:
Show a plurality of candidate locations；
According to the selection information to a plurality of candidate locations, determine that the target is wide from a plurality of candidate locations
It accuses；
Or,
The a plurality of candidate locations are screened according to preset screening model, the candidate locations filtered out are determined as
The targeted advertisements.
It should be noted that the aforementioned advertisement for being also applied for the embodiment to the explanation of advertisement generation method embodiment
Generating means, details are not described herein again.
Advertisement generating means provided in an embodiment of the present invention, by obtaining multiple ad material segments；According to the multiple
Ad material segment and preset pointer network generate a plurality of candidate locations；According to default rule from a plurality of candidate locations
Middle determining targeted advertisements.Permutation and combination is carried out to multiple ad material segments based on preset pointer network as a result, to automatically generate
The good a plurality of candidate locations of smoothness, and introduce risk-control mechanism a plurality of candidate locations are filtered it is wide to generate target
It accuses, realizes the ad pool for expanding advertiser, the advertisement that advertiser can be allowed to launch can be there is a greater chance that be able to expose or point
It hits.
Fig. 4 is the structural schematic diagram of another advertisement generating means provided in an embodiment of the present invention.In conjunction with reference Fig. 4,
On the basis of embodiment illustrated in fig. 3, described device further include: training module 14；
Wherein, the acquisition module 11, is also used to obtain sample set, and the sample set includes multiple primary advertising copies pair
The material segment collection answered, the corresponding material segment collection of each primary advertising copy include at least one of each primary advertising copy
Material segment and at least one corresponding expansion material segment；
The training module 14, for according to the corresponding material segment collection of each primary advertising copy to initial pointer net
Network carries out more wheel training, obtains the preset pointer network.
In one possible implementation, the training module 14 is specifically used for:
After currently wheel training, according to when the primary advertising copy of front-wheel and when the calculating of front-wheel reality output result is current
Take turns the corresponding loss function of primary advertising copy, wherein when front-wheel reality output the result is that described the primary advertisement sample of front-wheel will be worked as
This corresponding material segment collection is input to what the initial pointer network obtained；
The initial pointer network is updated according to the loss function and stochastic gradient descent method, is obtained more
Pointer network after new carries out next round training using the updated pointer network, according to the primary advertising copy of next round
Loss function corresponding with the next round reality output result calculating primary advertising copy of next round, repeats this step, until
All training finishes primary advertising copy in the sample set, wherein next round reality output is the result is that by the next round
The corresponding material segment collection of primary advertising copy is input to what the updated pointer network obtained；
Using updated pointer network as the preset pointer network.
It should be noted that the aforementioned advertisement for being also applied for the embodiment to the explanation of advertisement generation method embodiment
Generating means, details are not described herein again.
Advertisement generating means provided in an embodiment of the present invention obtain sample set, the sample set packet by obtaining sample set
Include the corresponding material segment collection of multiple primary advertising copies, the corresponding material segment collection of each primary advertising copy includes each original
At least one material segment of raw advertising copy and at least one corresponding expansion material segment；According to each primary advertising copy
Corresponding material segment collection carries out more wheel training to initial pointer network, obtains the preset pointer network.It trains as a result,
It is good to form smoothness that the preset pointer network finished has multiple ad material segments progress permutation and combination to input
The ability of multiple candidate locations realizes the ad pool for expanding advertiser, and the advertisement that advertiser can be allowed to launch can have bigger
Chance is able to expose or click.
Fig. 5 is a kind of structural schematic diagram of computer equipment provided in an embodiment of the present invention.The computer equipment includes:
Memory 1001, processor 1002 and it is stored in the calculating that can be run on memory 1001 and on processor 1002
Machine program.
Processor 1002 realizes the advertisement generation method provided in above-described embodiment when executing described program.
Further, computer equipment further include:
Communication interface 1003, for the communication between memory 1001 and processor 1002.
Memory 1001, for storing the computer program that can be run on processor 1002.
Memory 1001 may include high speed RAM memory, it is also possible to further include nonvolatile memory (non-
Volatile memory), a for example, at least magnetic disk storage.
Processor 1002 realizes advertisement generation method described in above-described embodiment when for executing described program.
If memory 1001, processor 1002 and the independent realization of communication interface 1003, communication interface 1003, memory
1001 and processor 1002 can be connected with each other by bus and complete mutual communication.The bus can be industrial standard
Architecture (Industry Standard Architecture, referred to as ISA) bus, external equipment interconnection
(Peripheral Component, referred to as PCI) bus or extended industry-standard architecture (Extended Industry
Standard Architecture, referred to as EISA) bus etc..The bus can be divided into address bus, data/address bus, control
Bus processed etc..Only to be indicated with a thick line in Fig. 5, it is not intended that an only bus or a type of convenient for indicating
Bus.
Optionally, in specific implementation, if memory 1001, processor 1002 and communication interface 1003, are integrated in one
It is realized on block chip, then memory 1001, processor 1002 and communication interface 1003 can be completed mutual by internal interface
Communication.
Processor 1002 may be a central processing unit (Central Processing Unit, referred to as CPU), or
Person is specific integrated circuit (Application Specific Integrated Circuit, referred to as ASIC) or quilt
It is configured to implement one or more integrated circuits of the embodiment of the present invention.
The present embodiment also provides a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that
The program realizes advertisement generation method as described above when being executed by processor.
In the description of this specification, reference term " one embodiment ", " some embodiments ", " example ", " specifically show
The description of example " or " some examples " etc. means specific features, structure, material or spy described in conjunction with this embodiment or example
Point is included at least one embodiment or example of the invention.In the present specification, schematic expression of the above terms are not
It must be directed to identical embodiment or example.Moreover, particular features, structures, materials, or characteristics described can be in office
It can be combined in any suitable manner in one or more embodiment or examples.In addition, without conflicting with each other, the skill of this field
Art personnel can tie the feature of different embodiments or examples described in this specification and different embodiments or examples
It closes and combines.
In addition, term " first ", " second " are used for descriptive purposes only and cannot be understood as indicating or suggesting relative importance
Or implicitly indicate the quantity of indicated technical characteristic.Define " first " as a result, the feature of " second " can be expressed or
Implicitly include at least one this feature.In the description of the present invention, the meaning of " plurality " is at least two, such as two, three
It is a etc., unless otherwise specifically defined.
Any process described otherwise above or method description are construed as in flow chart or herein, and expression includes
It is one or more for realizing custom logic function or process the step of executable instruction code module, segment or portion
Point, and the range of the preferred embodiment of the present invention includes other realization, wherein can not press shown or discussed suitable
Sequence, including according to related function by it is basic simultaneously in the way of or in the opposite order, Lai Zhihang function, this should be of the invention
Embodiment person of ordinary skill in the field understood.
Expression or logic and/or step described otherwise above herein in flow charts, for example, being considered use
In the order list for the executable instruction for realizing logic function, may be embodied in any computer-readable medium, for
Instruction execution system, device or equipment (such as computer based system, including the system of processor or other can be held from instruction
The instruction fetch of row system, device or equipment and the system executed instruction) it uses, or combine these instruction execution systems, device or set
It is standby and use.For the purpose of this specification, " computer-readable medium ", which can be, any may include, stores, communicates, propagates or pass
Defeated program is for instruction execution system, device or equipment or the dress used in conjunction with these instruction execution systems, device or equipment
It sets.The more specific example (non-exhaustive list) of computer-readable medium include the following: there is the electricity of one or more wirings
Interconnecting piece (electronic device), portable computer diskette box (magnetic device), random access memory (RAM), read-only memory
(ROM), erasable edit read-only storage (EPROM or flash memory), fiber device and portable optic disk is read-only deposits
Reservoir (CDROM).In addition, computer-readable medium can even is that the paper that can print described program on it or other are suitable
Medium, because can then be edited, be interpreted or when necessary with it for example by carrying out optical scanner to paper or other media
His suitable method is handled electronically to obtain described program, is then stored in computer storage.
It should be appreciated that each section of the invention can be realized with hardware, software, firmware or their combination.Above-mentioned
In embodiment, software that multiple steps or method can be executed in memory and by suitable instruction execution system with storage
Or firmware is realized.Such as, if realized with hardware in another embodiment, following skill well known in the art can be used
Any one of art or their combination are realized: have for data-signal is realized the logic gates of logic function from
Logic circuit is dissipated, the specific integrated circuit with suitable combinational logic gate circuit, programmable gate array (PGA), scene can compile
Journey gate array (FPGA) etc..
Those skilled in the art are understood that realize all or part of step that above-described embodiment method carries
It suddenly is that relevant hardware can be instructed to complete by program, the program can store in a kind of computer-readable storage medium
In matter, which when being executed, includes the steps that one or a combination set of embodiment of the method.
It, can also be in addition, each functional unit in each embodiment of the present invention can integrate in a processing module
It is that each unit physically exists alone, can also be integrated in two or more units in a module.Above-mentioned integrated mould
Block both can take the form of hardware realization, can also be realized in the form of software function module.The integrated module is such as
Fruit is realized and when sold or used as an independent product in the form of software function module, also can store in a computer
In read/write memory medium.
Storage medium mentioned above can be read-only memory, disk or CD etc..Although having been shown and retouching above
The embodiment of the present invention is stated, it is to be understood that above-described embodiment is exemplary, and should not be understood as to limit of the invention
System, those skilled in the art can be changed above-described embodiment, modify, replace and become within the scope of the invention
Type.
Claims (14)
1. a kind of advertisement generation method characterized by comprising
Obtain multiple ad material segments；
A plurality of candidate locations are generated according to the multiple ad material segment and preset pointer network；
Targeted advertisements are determined from a plurality of candidate locations according to default rule.
2. the method according to claim 1, wherein further include:
Sample set is obtained, the sample set includes the corresponding material segment collection of multiple primary advertising copies, each primary advertisement sample
This corresponding material segment collection includes at least one material segment and at least one corresponding expansion of each primary advertising copy
Material segment；
More wheel training are carried out to initial pointer network according to the corresponding material segment collection of each primary advertising copy, are obtained described
Preset pointer network.
3. according to the method described in claim 2, it is characterized in that, described according to the corresponding material piece of each primary advertising copy
Section collection carries out more wheel training to initial pointer network, obtains the preset pointer network, comprising:
After current wheel training, according to when the primary advertising copy of front-wheel and when front-wheel reality output result is calculated when front-wheel is former
The corresponding loss function of life advertising copy, wherein when front-wheel reality output the result is that described the primary advertising copy pair of front-wheel will be worked as
The material segment collection answered is input to what the initial pointer network obtained；
The initial pointer network is updated according to the loss function and stochastic gradient descent method, after obtaining update
Pointer network, next round training is carried out using the updated pointer network, according to the primary advertising copy of next round and under
One wheel reality output result calculates the corresponding loss function of the primary advertising copy of next round, repeats this step, until described
All training finishes primary advertising copy in sample set, wherein next round reality output is the result is that primary by the next round
The corresponding material segment collection of advertising copy is input to what the updated pointer network obtained；
Using updated pointer network as the preset pointer network.
4. the method according to claim 1, wherein the multiple ad material segments of acquisition include:
At least one for obtaining at least one material segment of the primary advertisement and being obtained by the expansion primary advertisement
Expand material segment；
At least one material segment of the primary advertisement and at least one described expansion material segment are broken up, obtained more
A ad material segment.
5. according to the method described in claim 4, it is characterized in that, described obtain at least one by expanding the primary advertisement
It includes one of following manner or a variety of for expanding the mode of material segment:
At least one the material segment to match with each material segment of the primary advertisement is obtained from ad material library, it will
It is determined as at least one material segment that each material segment of the primary advertisement matches described primary wide by expanding
It accuses at least one obtained and expands material segment；
Or,
History material segment corresponding with the advertiser for writing the primary advertisement is obtained, is obtained from the history material segment
At least one the history material segment to match with each material segment of the primary advertisement, will be every with the primary advertisement
At least one history material segment that a material segment matches is determined as obtain by expanding the primary advertisement at least one
A expansion material segment；
Or,
The corresponding general segment of the primary advertisement generic is obtained, the general segment is determined as by expanding the original
At least one expansion material segment that raw advertisement obtains.
6. the method according to claim 1, wherein it is described according to default rule from a plurality of candidate locations
Middle determining targeted advertisements include:
Show a plurality of candidate locations；
According to the selection information to a plurality of candidate locations, the targeted advertisements are determined from a plurality of candidate locations；
Or,
The a plurality of candidate locations are screened according to preset screening model, the candidate locations filtered out are determined as described
Targeted advertisements.
7. a kind of advertisement generating means characterized by comprising
Module is obtained, for obtaining multiple ad material segments；
Generation module, for generating a plurality of candidate locations according to the multiple ad material segment and preset pointer network；
Determining module, for determining targeted advertisements from a plurality of candidate locations according to default rule.
8. device according to claim 7, which is characterized in that further include: training module；
The acquisition module, is also used to obtain sample set, and the sample set includes the corresponding material piece of multiple primary advertising copies
Section collection, the corresponding material segment collection of each primary advertising copy include each primary advertising copy at least one material segment and
At least one corresponding expansion material segment；
The training module, for being carried out according to the corresponding material segment collection of each primary advertising copy to initial pointer network
More wheel training, obtain the preset pointer network.
9. device according to claim 8, which is characterized in that the training module is specifically used for:
After current wheel training, according to when the primary advertising copy of front-wheel and when front-wheel reality output result is calculated when front-wheel is former
The corresponding loss function of life advertising copy, wherein when front-wheel reality output the result is that described the primary advertising copy pair of front-wheel will be worked as
The material segment collection answered is input to what the initial pointer network obtained；
The initial pointer network is updated according to the loss function and stochastic gradient descent method, after obtaining update
Pointer network, next round training is carried out using the updated pointer network, according to the primary advertising copy of next round and under
One wheel reality output result calculates the corresponding loss function of the primary advertising copy of next round, repeats this step, until described
All training finishes primary advertising copy in sample set, wherein next round reality output is the result is that primary by the next round
The corresponding material segment collection of advertising copy is input to what the updated pointer network obtained；
Using updated pointer network as the preset pointer network.
10. device according to claim 7, which is characterized in that the acquisition module includes acquiring unit and breaks up unit:
The acquiring unit is described primary wide for obtaining at least one material segment of the primary advertisement and passing through expansion
It accuses at least one obtained and expands material segment；
It is described to break up unit, for by least one material segment of the primary advertisement and it is described at least one expand material piece
Duan Jinhang breaks up, and obtains multiple ad material segments.
11. device according to claim 10, which is characterized in that the acquiring unit is specifically used for:
At least one the material segment to match with each material segment of the primary advertisement is obtained from ad material library, it will
It is determined as at least one material segment that each material segment of the primary advertisement matches described primary wide by expanding
It accuses at least one obtained and expands material segment；
Or,
History material segment corresponding with the advertiser for writing the primary advertisement is obtained, is obtained from the history material segment
At least one the history material segment to match with each material segment of the primary advertisement, will be every with the primary advertisement
At least one history material segment that a material segment matches is determined as obtain by expanding the primary advertisement at least one
A expansion material segment；
Or,
The corresponding general segment of the primary advertisement generic is obtained, the general segment is determined as by expanding the original
At least one expansion material segment that raw advertisement obtains.
12. device according to claim 7, which is characterized in that the determining module is specifically used for:
Show a plurality of candidate locations；
According to the selection information to a plurality of candidate locations, the targeted advertisements are determined from a plurality of candidate locations；
Or,
The a plurality of candidate locations are screened according to preset screening model, the candidate locations filtered out are determined as described
Targeted advertisements.
13. a kind of computer equipment characterized by comprising
Memory, processor and storage are on a memory and the computer program that can run on a processor, which is characterized in that institute
It states when processor executes described program and realizes such as advertisement generation method as claimed in any one of claims 1 to 6.
14. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the program is by processor
Such as advertisement generation method as claimed in any one of claims 1 to 6 is realized when execution.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910451519.4A CN110189173B (en) | 2019-05-28 | 2019-05-28 | Advertisement generation method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910451519.4A CN110189173B (en) | 2019-05-28 | 2019-05-28 | Advertisement generation method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110189173A true CN110189173A (en) | 2019-08-30 |
| CN110189173B CN110189173B (en) | 2021-11-30 |
Family
ID=67718286
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910451519.4A Active CN110189173B (en) | 2019-05-28 | 2019-05-28 | Advertisement generation method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110189173B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112767024A (en) * | 2021-01-15 | 2021-05-07 | 北京搜狗科技发展有限公司 | Method and related device for generating advertisement creativity |
| CN112862516A (en) * | 2021-01-14 | 2021-05-28 | 北京达佳互联信息技术有限公司 | Resource delivery method and device, electronic equipment and storage medium |
| CN113099267A (en) * | 2021-06-04 | 2021-07-09 | 武汉卓尔数字传媒科技有限公司 | Video generation method and device, electronic equipment and storage medium |
| CN115471263A (en) * | 2022-09-16 | 2022-12-13 | 深圳掌屿软件有限公司 | Advertisement generation method, device, equipment and storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1321278A (en) * | 1998-10-01 | 2001-11-07 | 国际商业机器公司 | Systems, methods and computer program products for assigning, generating and delivering content to intranet users |
| US7920689B2 (en) * | 2005-05-25 | 2011-04-05 | Radziewicz Clifford J | Ringback replacement insertion system |
| CN108280112A (en) * | 2017-06-22 | 2018-07-13 | 腾讯科技（深圳）有限公司 | Abstraction generating method, device and computer equipment |
| CN109241536A (en) * | 2018-09-21 | 2019-01-18 | 浙江大学 | It is a kind of based on deep learning from the sentence sort method of attention mechanism |
| CN109493116A (en) * | 2018-10-15 | 2019-03-19 | 上海基分文化传播有限公司 | A kind of method and system that advertisement automatically generates |
| CN109544201A (en) * | 2018-10-11 | 2019-03-29 | 北京奇虎科技有限公司 | A kind of generation method of Advertising Copy, device and electronic equipment |
-
2019
- 2019-05-28 CN CN201910451519.4A patent/CN110189173B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1321278A (en) * | 1998-10-01 | 2001-11-07 | 国际商业机器公司 | Systems, methods and computer program products for assigning, generating and delivering content to intranet users |
| US7920689B2 (en) * | 2005-05-25 | 2011-04-05 | Radziewicz Clifford J | Ringback replacement insertion system |
| CN108280112A (en) * | 2017-06-22 | 2018-07-13 | 腾讯科技（深圳）有限公司 | Abstraction generating method, device and computer equipment |
| CN109241536A (en) * | 2018-09-21 | 2019-01-18 | 浙江大学 | It is a kind of based on deep learning from the sentence sort method of attention mechanism |
| CN109544201A (en) * | 2018-10-11 | 2019-03-29 | 北京奇虎科技有限公司 | A kind of generation method of Advertising Copy, device and electronic equipment |
| CN109493116A (en) * | 2018-10-15 | 2019-03-19 | 上海基分文化传播有限公司 | A kind of method and system that advertisement automatically generates |
Non-Patent Citations (2)
| Title |
|---|
| JINGJING GONG: "End-to-End Neural Sentence Ordering Using Pointer Network", 《HTTPS://ARXIV.ORG/ABS/1611.04953》 * |
| 栾克鑫: "基于注意力机制的句子排序方法", 《中文信息学报》 * |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112862516A (en) * | 2021-01-14 | 2021-05-28 | 北京达佳互联信息技术有限公司 | Resource delivery method and device, electronic equipment and storage medium |
| CN112862516B (en) * | 2021-01-14 | 2024-03-12 | 北京达佳互联信息技术有限公司 | Resource release method and device, electronic equipment and storage medium |
| CN112767024A (en) * | 2021-01-15 | 2021-05-07 | 北京搜狗科技发展有限公司 | Method and related device for generating advertisement creativity |
| CN112767024B (en) * | 2021-01-15 | 2024-02-20 | 北京搜狗科技发展有限公司 | Advertisement creative generation method and related device |
| CN113099267A (en) * | 2021-06-04 | 2021-07-09 | 武汉卓尔数字传媒科技有限公司 | Video generation method and device, electronic equipment and storage medium |
| CN113099267B (en) * | 2021-06-04 | 2021-11-12 | 武汉卓尔数字传媒科技有限公司 | Video generation method and device, electronic equipment and storage medium |
| CN115471263A (en) * | 2022-09-16 | 2022-12-13 | 深圳掌屿软件有限公司 | Advertisement generation method, device, equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110189173B (en) | 2021-11-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110189173A (en) | Advertisement generation method and device | |
| US20220391773A1 (en) | Method and system for artificial intelligence learning using messaging service and method and system for relaying answer using artificial intelligence | |
| CN104317959B (en) | Data digging method based on social platform and device | |
| KR101600640B1 (en) | Automatic customization and rendering of ads based on detected features in a web page | |
| US8423349B1 (en) | Filtering phrases for an identifier | |
| CN103207904B (en) | The offer method of Search Results and search engine | |
| CN109657213A (en) | Text similarity detection method and device and electronic equipment | |
| CN108446813A (en) | A kind of method of electric business service quality overall merit | |
| US20090049020A1 (en) | System and method for providing personalized recommended word and computer readable recording medium recording program for implementing the method | |
| CN109543022A (en) | Text error correction method and device | |
| CN110795627B (en) | Information recommendation method and device and electronic equipment | |
| CN108304426B (en) | Identification obtaining method and device | |
| CN106844341A (en) | News in brief extracting method and device based on artificial intelligence | |
| CN111767393A (en) | Text core content extraction method and device | |
| CN110334356A (en) | Article matter method for determination of amount, article screening technique and corresponding device | |
| CN108681541A (en) | Image searching method, device and computer equipment | |
| CN110399505A (en) | Semantic label generation method and equipment, computer storage medium | |
| JP2013073626A (en) | Search method and device using concept keyword extension dataset and computer readable recording medium | |
| CN111737961B (en) | Method and device for generating story, computer equipment and medium | |
| CN113901263B (en) | Label generation method and device for video material | |
| CN113886568A (en) | Text abstract generation method and device | |
| CN108829882A (en) | Formation gathering method, device, terminal and medium | |
| CN110968687A (en) | Method and device for classifying texts | |
| CN107748801A (en) | News recommends method, apparatus, terminal device and computer-readable recording medium | |
| CN109033241A (en) | News recommended method, device and electronic equipment |
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 |