CN114331543A - Advertisement propagation method for large-scale crowd orientation and dynamic scene matching - Google Patents
Advertisement propagation method for large-scale crowd orientation and dynamic scene matching Download PDFInfo
- Publication number
- CN114331543A CN114331543A CN202111650242.1A CN202111650242A CN114331543A CN 114331543 A CN114331543 A CN 114331543A CN 202111650242 A CN202111650242 A CN 202111650242A CN 114331543 A CN114331543 A CN 114331543A
- Authority
- CN
- China
- Prior art keywords
- user
- advertisement
- portrait
- bidding
- accurate
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses an advertisement propagation method for large-scale crowd orientation and dynamic scene matching, which comprises the following steps: the dynamic user accurate portrait sketching is realized in different outdoor scenes by utilizing the movement detection technology of a camera or a probe; calculating in real time, and finding the optimal and most matched advertisement aiming at a target audience in a dynamic scene and time period through a bidding algorithm; the user portrait is merged by analyzing the vehicle information of the user, the user portrait and the big data of the user behavior, and the user portrait becomes more accurate through a machine learning algorithm model; through user feedback when the advertisement is watched, the user portrait is corrected, and the user portrait is more accurate through a correction algorithm, so that more advertisements are accurately thrown. The advertisement propagation model can comprehensively improve the real-time relevance between advertisement delivery and target audiences and scenes, realize the accurate delivery of advertisements, accurately reach the audiences, improve the propagation effect, and realize more and more economical advertisement delivery and more accurate advertisement delivery.
Description
Technical Field
The invention relates to the technical field of advertisement putting, in particular to an advertisement propagation method for large-scale crowd orientation and dynamic scene matching.
Background
With the development of technology and the transition of times, the development of outdoor advertisements is driven by programming-digitization-datamation, and the 'precision' and 'data support' are essential characteristics of the continuous improvement of the advertising effect in the intelligent advertisement operation mode, and the business mode change of the advertising industry is necessarily evolved around the basic logic of the improvement of the technology-datamation driven advertising effect, and even the business mode change is defined as the 'original year' of the programming intelligence of the outdoor advertisements in 2021.
Tracing the development of the advertising media industry, it can be found that the traditional advertising transaction and propagation mode is a pure linear mode, and advertisers directly purchase media advertising spots through agents, so that the system has absolute control right, the reach of users and audiences, and the effect is not regarded. With the popularization of the internet, a large amount of user behaviors are digitalized. The big data analysis processing technology can directly help the advertiser to more accurately locate and find the target crowd. Therefore, a data management platform of an integrated user, an advertisement demand side platform for helping an advertiser analyze data and perform advertisement putting, and a media supplier platform for gathering more media long-tail resources are gradually formed, and an advertisement trading platform for connecting all supply and demand and data resources and providing trading places is generated, so that a systematic and automatic programmed purchase and information propagation mode is formed. As shown in fig. 1, this phase is referred to as "automated information pushing phase based on programmed digitization", and in this phase, programmed purchasing of advertisements to advertisement propagation develops from a programmed mode, and it realizes programmed docking between advertisers, agents and network media in a digitized, automated and systematic manner, helps advertisers to accurately find target audiences and users associated and matched with advertisement information, and makes programmed delivery implement automation of the whole outdoor advertisement industry chain through the whole process from advertisers to media to exposure. The method has the greatest characteristic of realizing automation of a series of processes such as programmed medium purchasing, advertisement putting, put report feedback and the like in an advertisement propagation mode of programmed putting.
Disclosure of Invention
Aiming at the defects of the prior art, the invention provides an advertisement propagation method for large-scale crowd orientation and dynamic scene matching, which segments and clusters a target audience through the driving of data, and accurately outlines a target consumer portrait; the accuracy of advertisement putting and audience reaching is improved, different advertisements represent different commodities, and the advertisement player is suitable for different crowds.
Aiming at the problems in the prior art, the invention provides an advertisement propagation method for large-scale crowd orientation and dynamic scene matching, which comprises the following steps:
(1) the dynamic user accurate portrait sketching is realized in different outdoor scenes by utilizing the movement detection technology of a camera or a probe;
(2) calculating in real time, and finding the optimal and most matched advertisement aiming at a target audience in a dynamic scene and time period through a bidding algorithm;
(3) the user portrait is merged by analyzing the vehicle information of the user, the user portrait and the big data of the user behavior, and the user portrait becomes more accurate through a machine learning algorithm model;
(4) through user feedback when the advertisement is watched, the user portrait is corrected, and the user portrait is more accurate through a correction algorithm, so that more advertisements are accurately thrown.
In a further limited technical solution of the present application, in the advertisement propagation method for large-scale crowd targeting and dynamic scene matching, the bidding algorithm adopts a nonlinear comprehensive bidding algorithm, and a mathematical formula is used to describe the bidding scheme as follows:
subject to NT∫θb(θ)θw(b(θ))pθ(θ)dθ≤B
in the formula: theta is the number of predicted advertisement releases, p (theta) is a probability density function of the number of releases, w (B (theta)) is a winning rate function, N is the number of advertisement bids, and B is a fixed budget;
the bidding function becomes an optimization problem with constraint conditions, and the Lagrange multiplier method is used for solving the extremum and simplifying the method:
because the bidding function and the winning rate function contain the distribution characteristic of advertisement putting frequency, p (theta) is eliminated in the process, and a specific relationship exists between the bidding function and the winning rate function; the goal is then to statistically derive a win rate function win (b) based on historical data.
Derivation of winning bid rate function: the historical data which can be observed by the platform side of the demand side is only the bid price of the platform of the demand side and the actual bid price after winning the bid price; when the maximum price of the platform of the demand party is larger than the price of the soft floor, a second high price bidding mechanism is adopted, so that the platform of the demand party can observe the real winning price of the bidding for the advertisement activity successful in bidding, namely the second high price. For a record of a failure in a bidding activity, only the price at which the bid failed is available, which is the lower bound of the winning price in this bidding activity, since the winning price is unlikely to be lower than the price at which the bid failed, and therefore the second highest price cannot be observed in this case. A brief overview of a most basic win-win rate statistical scheme follows.
Calculating the expected value, namely the average value, of the actual payment price of the advertising campaign with the winning bid, and using the mu expression; and the variance of all actual payment prices of the demand side platform are expressed by sigma, parameters gamma and sigmoid functions are introduced, the formula is as follows,
and substituting the winning bid rate function into the simplified relational expression to obtain the bidding function.
In the advertisement propagation method for large-scale crowd targeting and dynamic scene matching, the expression of the machine learning algorithm model is as follows:
viis the hidden vector of the ith characteristic, the length of the hidden vector is k (k < n), and k factors describing the characteristic are included;
all binomial parameters wijA symmetric matrix W can be formed, which is decomposed into W = VTV, wherein V = (V)1，V2，…，Vn0t，Vi=(vi1，vi2，…，vik) (ii) a Matrix VTEach row of matrix V represents the relevance of a certain user to different characteristics, and each row of matrix V represents the relevance of a certain characteristic to different advertisements;
for a machine learning algorithm model algorithm, a middle k-dimensional vector can be trained by analyzing the vehicle information of a user, the track and age of the user and the matching degree relation between the gender and the advertisement, and the k-dimensional vectors of the user and the commodity can be trained, so that the portrait of the user can be better corrected, and the portrait of the user is more accurate.
Further, in the advertisement propagation method for large-scale crowd targeting and dynamic scene matching, the correction algorithm adopts the following method for correction:
1) correcting the user image by guiding the user to perform interaction before the screen and taking the feedback of the interaction as an input item;
2) the method comprises the following steps of (1) guiding a user to actively register and select a personal favorite label;
3) through the modes of facial expression recognition and eyeball attention recognition, the user portrait is corrected through the feedback of the facial expression and the eyeball attention during advertisement watching;
4) by interfacing with third-party internet desensitization data, the user representation is rectified.
The invention has the beneficial effects that: the intelligent advertisement is scene intelligent marketing based on big data and artificial intelligence technology. The operation mechanism is that according to a specific user and a specific situation, an advertisement which is most matched with the specific user is determined through an efficient algorithm, and accurate originality, production, putting, propagation and interaction are carried out, so that the matching problem of advertisement information, the user and a scene is solved. The personalized accurate recommendation method has the greatest characteristic that the personalized accurate recommendation method obtains, stores and analyzes user behavior characteristic data through big data analysis and artificial intelligence algorithm and optimization of a screen-front recognition CV technology, sends appropriate content to a user in scenes such as time and place suitable for the user, and meets dynamic information requirements of the user in different scenes. The intelligent advertisement under the advertisement mode can instantly discover the needs of the user under a specific scene, dynamically predict the needs of the user and bind the user to create services. The invention not only improves the accuracy of advertisement propagation, but also needs to meet the requirements of consumers in specific scenes, the consumption behavior and media contact in the mobile internet era are extremely fragmented, the outdoor scene is used as the core entrance of advertisement information propagation, and the advertisement propagation mode is developed to the advertisement propagation mode with crowd oriented and dynamic scene matching as the core under the energization of big data and artificial intelligence.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the description of the embodiments are briefly introduced below, it is obvious that the drawings in the following description are only some embodiments of the present invention, the embodiments in the drawings do not constitute any limitation to the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
Fig. 1 is a block diagram of a conventional advertisement propagation pattern system.
Fig. 2 is a flowchart of advertisement propagation according to the present embodiment.
Fig. 3 is a block diagram of the outdoor advertisement propagation model according to the embodiment.
Fig. 4 is a diagram of the core algorithm structure of the present invention.
Detailed Description
The technical solution of the present invention will be further described in detail with reference to the accompanying drawings and embodiments, which are preferred embodiments of the present invention. It is to be understood that the described embodiments are merely a subset of the embodiments of the invention, and not all embodiments; it should be noted that the embodiments and features of the embodiments may be combined with each other without conflict. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In order to better understand the technical solutions, the technical solutions will be described in detail below with reference to the drawings and specific embodiments.
The present embodiment provides an advertisement dissemination process for large-scale crowd targeting and dynamic scene matching, as shown in fig. 2,
1) the detection equipment for the vehicle characteristics and the person characteristics transmits the identified characteristic tag values to Ad Exchanges;
2) ad Exchanges sends bidding requests to a plurality of DSPs connected with the Ad Exchanges;
3) the DSP deduces the user attribute through a DMP data management platform or a user database of the DSP, judges whether to participate in bidding through a bidding engine of the DSP, and gives a bidding price if the participation is judged;
4) the DSP transmits the bidding response to Ad Exchanges;
5) ad Exchanges receives the responses of all the DSPs or reaches the deadline, and determines the winning DSP;
6) ad Exchanges synthesizes the characteristic tag information + bidding information of vehicles and people, matches the advertisement content most suitable for users in front of the screen, and carries out the advertisement putting and displaying.
The noun of the present embodiment explains:
and (4) DSP: the Demand Side Platform provides a comprehensive management Platform for advertisers and agent companies, and by means of the DSP, the advertisers can bid RTB Real-Time Bidding on online advertisements in an advertisement trading Platform AD Exchange in Real Time so as to efficiently manage advertisement pricing;
SSP: the Sell-Side Platform is a media service Platform, intelligently manages the inventory of media advertisement spots through a crowd targeting technology, optimizes the delivery of advertisements, helps network media to realize the optimization of advertisement resources, improves the value of the advertisement resources and achieves the purpose of helping the media to improve the income.
DMP: Data-Management Platform, Data Management Platform. The data management platform can help all parties involved in purchasing and selling the advertising inventory to manage data, more conveniently use third-party data, enhance understanding of all data, return data or transfer customized data to a platform for better positioning.
Ad Exchange: the advertisement transaction platform is connected with a DSP (buyer platform) and an SSP (seller platform), collects and processes data belonging to advertisement target customers by collecting a large amount of media flow through accessing the SSP, and Ad Exchange is a transaction place for realizing accurate marketing.
The bidding algorithm of the present embodiment adopts a nonlinear comprehensive bidding algorithm, and describes the bidding scheme using a mathematical formula as follows:
subject to NT∫θb(θ)θw(b(θ))pθ(θ)dθ≤B
in the formula: theta is the number of predicted advertisement releases, p (theta) is a probability density function of the number of releases, w (B (theta)) is a winning rate function, N is the number of advertisement bids, and B is a fixed budget;
the bidding function becomes an optimization problem with constraint conditions, and the Lagrange multiplier method is used for solving the extremum and simplifying the method:
because the bidding function and the winning rate function contain the distribution characteristic of advertisement putting frequency, p (theta) is eliminated in the process, and a specific relationship exists between the bidding function and the winning rate function; the goal is then to statistically derive a win rate function win (b) based on historical data.
Derivation of winning bid rate function: the historical data which can be observed by the platform side of the demand side is only the bid price of the platform of the demand side and the actual bid price after winning the bid price; when the maximum price of the platform of the demand party is larger than the price of the soft floor, a second high price bidding mechanism is adopted, so that the platform of the demand party can observe the real winning price of the bidding for the advertisement activity successful in bidding, namely the second high price. For a record of a failure in a bidding activity, only the price at which the bid failed is available, which is the lower bound of the winning price in this bidding activity, since the winning price is unlikely to be lower than the price at which the bid failed, and therefore the second highest price cannot be observed in this case. A brief overview of a most basic win-win rate statistical scheme follows.
Calculating the expected value, namely the average value, of the actual payment price of the advertising campaign with the winning bid, and using the mu expression; and the variance of all actual payment prices of the demand side platform are expressed by sigma, parameters gamma and sigmoid functions are introduced, the formula is as follows,
and substituting the winning bid rate function into the simplified relational expression to obtain the bidding function.
Expression of the machine learning algorithm model:
viis the hidden vector of the ith characteristic, the length of the hidden vector is k (k < n), and k factors describing the characteristic are included;
all binomial parameters wijA symmetric matrix W can be formed, which is decomposed into W = VTV, wherein V = (V)1，V2，…，Vn)T，Vi=(vi1，vi2，…，vik) (ii) a Matrix VTEach row of (a) represents the facies of a user with different featuresRelevance, each row of the matrix V representing the relevance of a certain feature to a different advertisement;
for a machine learning algorithm model algorithm, a middle k-dimensional vector can be trained by analyzing the vehicle information of a user, the track and age of the user and the matching degree relation between the gender and the advertisement, and the k-dimensional vectors of the user and the commodity can be trained, so that the portrait of the user can be better corrected, and the portrait of the user is more accurate.
The correction algorithm of the embodiment performs correction in the following manner:
1) correcting the user image by guiding the user to perform interaction before the screen and taking the feedback of the interaction as an input item;
2) the method comprises the following steps of (1) guiding a user to actively register and select a personal favorite label;
3) through the modes of facial expression recognition and eyeball attention recognition, the user portrait is corrected through the feedback of the facial expression and the eyeball attention during advertisement watching;
4) by interfacing with third-party internet desensitization data, the user representation is rectified.
According to the advertisement propagation method in the scheme, programmed advertisement purchasing is adopted, compared with conventional manual advertisement purchasing, the advertisement efficiency, the scale and the advertisement putting strategy are greatly improved, the accurate realization of programmed advertisement is that proper brand and product information is transmitted aiming at definite target users or groups, influence factors from the purchase intention of the users to the final consumption are very many, and the pursuit of an advertiser on the effect cannot be met only by a digital, automatic and systematic advertisement putting model.
The outdoor advertisement propagation model enters a large-scale crowd orientation and dynamic scene matching stage based on big data and artificial intelligence technologies, as shown in fig. 3, the popularization of the internet, the big data and the artificial intelligence technologies and the development of the pre-screen recognition technology comprise 'thousand-person thousand-face' -large-scale face recognition technology and 'thousand-car thousand-face' -large-scale vehicle information recognition technology; the method can directly help the advertiser to more accurately position and search the target crowd and accurately reach the target crowd in a dynamic scene so as to achieve the optimal propagation effect.
At this stage, smart advertising is a scenic smart marketing based on big data and artificial intelligence technology. The operation mechanism is that according to a specific user and a specific situation, an advertisement which is most matched with the specific user is determined through an efficient algorithm, and accurate originality, production, putting, propagation and interaction are carried out, so that the matching problem of advertisement information, the user and a scene is solved. The personalized accurate recommendation is the biggest characteristic of the method, and the method adopts big data analysis and artificial intelligence algorithm and adopts a screen-front CV recognition technology, and comprises the following steps: the 'thousand-person thousand-face' -large-scale face recognition technology 'thousand-car thousand-face' -large-scale vehicle information recognition technology optimizes the acquisition, storage and analysis of user behavior characteristic data, and sends appropriate content to a user in scenes such as time and place suitable for the user, so that the dynamic information requirements of the user in different scenes are met. The intelligent advertisement under the advertisement mode can instantly discover the needs of the user under a specific scene, dynamically predict the needs of the user and bind the user to create services.
In the advertisement propagation model based on the large-scale crowd targeting and dynamic scene matching of the big data + artificial intelligence technology, a core algorithm structure diagram is shown in fig. 4 and includes the following four parts:
1) data acquisition: the dynamic user accurate portrait sketching is realized in different outdoor scenes by utilizing the mobile detection technologies such as a camera and a probe;
2) and (3) calculating in real time: calculating in real time, and finding the optimal and most matched advertisement aiming at a target audience in a dynamic scene and time period through a bidding algorithm;
3) and (3) off-line calculation: by analyzing the user vehicle information, the user portrait, the user behavior and other user big data, the user portrait is merged, and the user portrait becomes more accurate through a machine learning algorithm model.
4) And (3) user feedback: through user feedback when the advertisement is watched, the user portrait is corrected, and the user portrait is more accurate through an algorithm, so that more accurate advertisement throwing is realized.
It should be noted that, for simplicity of description, the above-mentioned method embodiments are described as a series of acts or combination of acts, but those skilled in the art will recognize that the present invention is not limited by the order of acts, as some steps may occur in other orders or concurrently in accordance with the invention. Further, those skilled in the art should also appreciate that the embodiments described in the specification are exemplary embodiments and that the acts and modules illustrated are not necessarily required to practice the invention.
The above embodiments may be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented in software, the above-described embodiments may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. The procedures or functions described in accordance with the embodiments of the present application are generated in whole or in part when the computer program instructions are loaded or executed on a computer. The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable device. The computer instructions may be stored on a computer readable storage medium or transmitted from one computer readable storage medium to another computer readable storage medium, for example, the computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wire (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server, data center, etc. that contains one or more collections of available media. The available media may be magnetic media (e.g., floppy disks, hard disks, tapes), optical media (e.g., DVDs), or semiconductor media. The semiconductor medium may be a solid state disk.
Those of ordinary skill in the art will appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.
In the foregoing embodiments, the descriptions of the respective embodiments have respective emphasis, and for parts that are not described in detail in a certain embodiment, reference may be made to related descriptions of other embodiments.
It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described systems, apparatuses and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.
In the embodiments provided in the present invention, it should be understood that the disclosed system, apparatus and method may be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, the above-described division of units is only one type of division of logical functions, and other divisions may be realized in practice, for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may be in an electrical, mechanical or other form.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.
The integrated unit may be stored in a computer-readable storage medium if it is implemented in the form of a software functional unit and sold or used as a separate product. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the above method according to the embodiments of the present invention. And the aforementioned storage medium includes: various media capable of storing program codes, such as a usb disk, a removable hard disk, a read-only memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.
The present invention has been described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (devices) and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
It is noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the same element.
Finally, it should be noted that: the above embodiments are only used to illustrate the technical solution of the present invention, and not to limit the same; while the invention has been described in detail and with reference to the foregoing embodiments, it will be understood by those skilled in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; and the modifications or the substitutions do not make the essence of the corresponding technical solutions depart from the scope of the technical solutions of the embodiments of the present invention.
Claims (4)
1. An advertisement propagation method for large-scale crowd orientation and dynamic scene matching is characterized by comprising the following steps:
(1) the dynamic user accurate portrait sketching is realized in different outdoor scenes by utilizing the movement detection technology of a camera or a probe;
(2) calculating in real time, and finding the optimal and most matched advertisement aiming at a target audience in a dynamic scene and time period through a bidding algorithm;
(3) the user portrait is merged by analyzing the vehicle information of the user, the user portrait and the big data of the user behavior, and the user portrait becomes more accurate through a machine learning algorithm model;
(4) through user feedback when the advertisement is watched, the user portrait is corrected, and the user portrait is more accurate through a correction algorithm, so that more advertisements are accurately thrown.
2. The large-scale crowd targeting and dynamic scene matching advertisement dissemination method according to claim 1, characterized by: the bidding algorithm adopts a nonlinear comprehensive bidding algorithm, and uses a mathematical formula to describe a bidding scheme as follows:
subject to NT∫θb(θ)θw(b(θ))pθ(θ)dθ≤B
in the formula: theta is the number of predicted advertisement releases, p (theta) is a probability density function of the number of releases, w (B (theta)) is a winning rate function, N is the number of advertisement bids, and B is a fixed budget;
the bidding function becomes an optimization problem with constraint conditions, and the Lagrange multiplier method is used for solving the extremum and simplifying the method:
because the bidding function and the winning rate function contain the distribution characteristic of advertisement putting frequency, p (theta) is eliminated in the process, and a specific relationship exists between the bidding function and the winning rate function;
calculating the expected value, namely the average value, of the actual payment price of the advertising campaign with the winning bid, and using the mu expression; and the variance of all actual payment prices of the demand side platform are expressed by sigma, parameters gamma and sigmoid functions are introduced, the formula is as follows,
and substituting the winning bid rate function into the simplified relational expression to obtain the bidding function.
3. The large-scale crowd targeting and dynamic scene matching advertisement dissemination method according to claim 1, characterized by: an expression of the machine learning algorithm model:
viis the hidden vector of the ith characteristic, the length of the hidden vector is k (k < n), and k factors describing the characteristic are included;
all binomial parameters wijA symmetric matrix W may be formed, which is decomposed into W ═ VTV, wherein V ═ V1，V2，…，Vn)T，Vi＝(vi1，vi2，…，vik) (ii) a Matrix VTEach row of matrix V represents the relevance of a certain user to different characteristics, and each row of matrix V represents the relevance of a certain characteristic to different advertisements;
for a machine learning algorithm model algorithm, a middle k-dimensional vector can be trained by analyzing the vehicle information of a user, the track and age of the user and the matching degree relation between the gender and the advertisement, and the k-dimensional vectors of the user and the commodity can be trained, so that the portrait of the user can be better corrected, and the portrait of the user is more accurate.
4. The large-scale crowd targeting and dynamic scene matching advertisement dissemination method according to claim 1, characterized by: the correction algorithm performs correction in the following manner:
1) correcting the user image by guiding the user to perform interaction before the screen and taking the feedback of the interaction as an input item;
2) the method comprises the following steps of (1) guiding a user to actively register and select a personal favorite label;
3) through the modes of facial expression recognition and eyeball attention recognition, the user portrait is corrected through the feedback of the facial expression and the eyeball attention during advertisement watching;
4) by interfacing with third-party internet desensitization data, the user representation is rectified.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111650242.1A CN114331543A (en) | 2021-12-29 | 2021-12-29 | Advertisement propagation method for large-scale crowd orientation and dynamic scene matching |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111650242.1A CN114331543A (en) | 2021-12-29 | 2021-12-29 | Advertisement propagation method for large-scale crowd orientation and dynamic scene matching |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114331543A true CN114331543A (en) | 2022-04-12 |
Family
ID=81018105
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111650242.1A Pending CN114331543A (en) | 2021-12-29 | 2021-12-29 | Advertisement propagation method for large-scale crowd orientation and dynamic scene matching |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114331543A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115423537A (en) * | 2022-11-02 | 2022-12-02 | 北京车讯互联网股份有限公司 | Internet automobile industry accurate advertisement putting method based on artificial intelligence |
| CN115423510A (en) * | 2022-08-30 | 2022-12-02 | 成都智元汇信息技术股份有限公司 | Media service processing method based on subway associated data |
| CN116452272A (en) * | 2023-06-09 | 2023-07-18 | 北京万物镜像数据服务有限公司 | Advertisement information processing method, device and equipment in virtual space |
-
2021
- 2021-12-29 CN CN202111650242.1A patent/CN114331543A/en active Pending
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115423510A (en) * | 2022-08-30 | 2022-12-02 | 成都智元汇信息技术股份有限公司 | Media service processing method based on subway associated data |
| CN115423537A (en) * | 2022-11-02 | 2022-12-02 | 北京车讯互联网股份有限公司 | Internet automobile industry accurate advertisement putting method based on artificial intelligence |
| CN116452272A (en) * | 2023-06-09 | 2023-07-18 | 北京万物镜像数据服务有限公司 | Advertisement information processing method, device and equipment in virtual space |
| CN116452272B (en) * | 2023-06-09 | 2023-09-05 | 北京万物镜像数据服务有限公司 | Advertisement information processing method, device and equipment in virtual space |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108476334B (en) | Cross-screen optimization of advertisement placement | |
| CN108352025B (en) | Television advertisement slot targeting based on consumer online behavior | |
| Bose et al. | Quantitative models for direct marketing: A review from systems perspective | |
| CN106570718B (en) | Information delivery method and delivery system | |
| CN114331543A (en) | Advertisement propagation method for large-scale crowd orientation and dynamic scene matching | |
| US20150095166A1 (en) | System, method and computer program for providing qualitative ad bidding | |
| US20160132935A1 (en) | Systems, methods, and apparatus for flexible extension of an audience segment | |
| US11216850B2 (en) | Predictive platform for determining incremental lift | |
| US20140278973A1 (en) | System and method for audience targeting | |
| Paulson et al. | Efficient large-scale internet media selection optimization for online display advertising | |
| CN104657879A (en) | User engagement-based contextually-dependent automated pricing for non-guaranteed delivery | |
| McStay | 3.1 Micro-moments, liquidity, intimacy and automation: Developments in programmatic Ad-tech | |
| US20120022920A1 (en) | Eliciting customer preference from purchasing behavior surveys | |
| CN115860870A (en) | Commodity recommendation method, system and device and readable medium | |
| US20200202403A1 (en) | Distribution service providing method connecting initial provider and buyer via mid-distribution man | |
| CN116132749B (en) | Advertisement marketing method combined with live broadcast and carried goods | |
| US20230316106A1 (en) | Method and apparatus for training content recommendation model, device, and storage medium | |
| CN112348590A (en) | Method and device for determining value of article, electronic equipment and storage medium | |
| JP2022531410A (en) | Digital anthropology and ethnographic system | |
| Ahmed et al. | Big data analytics in the entertainment Industry: audience behavior analysis, content recommendation, and Revenue maximization | |
| US20210090098A1 (en) | Computational Methods and Systems for Marketing Analysis | |
| CN113052653A (en) | Financial product content recommendation method and system and computer readable storage medium | |
| CN113516496A (en) | Advertisement conversion rate pre-estimation model construction method, device, equipment and medium thereof | |
| Li | Evaluation of Artificial Intelligence models and wireless network applications for enterprise sales management innovation under the new retail format | |
| KR102422409B1 (en) | System and method of auction linked to Social Network Service |
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 |