Advertisement pushing method and system based on Internet platform

Abstract

The invention relates to the technical field of advertisement pushing, and particularly discloses an advertisement pushing method and system based on an internet platform, and the method comprises the following steps: S1, obtaining the browsing data of an undetermined advertisement, calculating a standard time length based on the browsing data, and determining a target advertisement; obtaining a first keyword, and obtaining a first word vector based on the first keyword; s2, acquiring a second keyword, acquiring a second word vector based on the second keyword, and determining an effective advertisement; drawing a first curve, determining a second curve, and determining a target curve on the second curve; and S3, calculating a selection value based on the target curve, determining a target platform, and putting an advertisement on the target platform. According to the invention, the maximization of the advertisement putting effect can be realized at the optimal cost.

Description

Technical Field

[0001] The present invention relates to the technical field of advertisement pushing, and particularly relates to an advertisement pushing method and system based on an Internet platform. Background Art

[0002] The Internet platform is an important carrier in the digital age. By aggregating a large number of users and diversified content, it realizes the efficient dissemination of information and value interaction. In the competitive digital age, making full use of the advertisement resources of the Internet platform can not only quickly expand the audience coverage, but also optimize the marketing cost, helping products gain a head start in the market.

[0003] Relying on the efficient dissemination attribute of the Internet platform and rich advertisement resources, brand owners can accurately lock in target users and achieve deep interaction with potential consumers. In the diversified digital environment, through big data mining and real-time delivery strategies, advertisement information can quickly reach a wide range of people, opening up a broader market for brands. However, to achieve this effect, brand owners often need a lot of capital investment. Therefore, how to balance the promotion effect and cost has become an urgent problem to be solved. Summary of the Invention

[0004] The purpose of the present invention is to provide an advertisement pushing method and system based on an Internet platform, and solve the following technical problems:

[0005] In the diversified digital environment, through big data mining and real-time delivery strategies, advertisement information can quickly reach a wide range of people, opening up a broader market for brands. However, to achieve this effect, brand owners often need a lot of capital investment.

[0006] The purpose of the present invention can be achieved through the following technical solutions:

[0007] An advertisement pushing method based on an Internet platform includes the following steps:

[0008] S1: Mark the advertisements browsed by users as pending advertisements, obtain the browsing data of the pending advertisements, where the browsing data includes the total return times F1, the average time interval F2 between two adjacent returns, and the total browsing duration F3, calculate the standard duration FBZ = SF * F2 / (1 + η * F1), η represents a preset first correction coefficient, and mark the pending advertisements with a standard duration greater than the preset market threshold as target advertisements;

[0009] S2: Obtain the keywords in the target advertisements, use them as the first keywords, and obtain the first word vectors based on the first keywords;

[0010] Obtain the keywords in the advertisements placed by the brand side, use them as the second keywords, obtain the second word vectors based on the second keywords, determine the cosine value of the angle between the first word vector and the second word vector, and mark the target advertisements with the corresponding cosine value less than the preset cosine value threshold of the angle as effective advertisements;

[0011] Periodically obtain the number of effective advertisements within a preset monitoring period, draw a curve of the number of effective advertisements changing with time, denoted as the first curve, obtain the second curve based on the first curves corresponding to m monitoring periods, m is a preset number, use the straight line passing through the point (0, Dys) on the y-axis and parallel to the x-axis as the reference line, Dys represents the preset threshold value of the number of effective advertisements, and use the part of the second curve above the reference line as the target curve;

[0012] S3: Calculate the selection value based on the target curve, and its calculation formula is:

[0013]

[0014] where, β represents the preset second correction coefficient, n represents the total number of the target curves corresponding to a single Internet platform, Ti, sta and Ti, end respectively represent the time points corresponding to the start point and the end point of the i-th target curve, and the i-th pending placement period Ti = Ti, end - Ti, sta;

[0015] Determine the target platform based on the selection value, and place advertisements on the target platform.

[0016] As a further solution of the present invention: In the step S2, the following steps are further included:

[0017] Input the first keyword into a preset word vector model to obtain a first sub-vector, and add the values of each dimension of the first sub-vector to obtain a first word vector;

[0018] Input the second keyword into the word vector model to obtain a second sub-vector, and add the values of each dimension of the second sub-vector to obtain a second word vector.

[0019] As a further solution of the present invention: In the step S3, the process of determining the target platform specifically includes:

[0020] Calculate the estimated price of placing advertisements on the k-th Internet platform QBJk represents the price of pushing advertisements for a preset duration t on the k-th Internet platform, and Nk represents the total number of the target curves corresponding to the k-th Internet platform;

[0021] Sort the Internet platforms in descending order according to the selection values to obtain the target sorting. Starting from the first Internet platform in the target sorting, determine the target sorting position C in the target sorting, where the estimated prices corresponding to the first C Internet platforms in the target sorting are less than the preset target price threshold and the estimated prices corresponding to the first C + 1 Internet platforms in the target sorting are greater than or equal to the total price threshold;

[0022] Take the first C Internet platforms in the target sorting as the target platforms.

[0023] As a further solution of the present invention: During the process of obtaining the target sorting, when two or more selection values are the same, the smaller the corresponding estimated price, the more forward the position of the Internet platform in the target sorting.

[0024] As a further solution of the present invention: In step S3, the process of placing an advertisement on the target platform specifically includes:

[0025] Place an advertisement during the to-be-placed time period corresponding to the target platform.

[0026] As a further solution of the present invention: In step S2, the process of obtaining the second curve specifically includes:

[0027] Select a number of reference points on the first curve at a preset time interval, calculate the mean value of the effective advertisement quantities corresponding to the same reference points, generate coordinate points (a, Da), where Da represents the mean value of the effective advertisement quantities corresponding to reference point a, and fit the coordinate points to obtain the second curve.

[0028] As a further solution of the present invention: During the process of calculating the mean value of the effective advertisement quantities corresponding to the same reference points, when the difference between the effective advertisement quantity corresponding to reference point a within the monitoring period x and the mean value is greater than the preset threshold, remove the monitoring period x and set a new monitoring period again to replace the monitoring period x and calculate the mean value of the effective advertisement quantities corresponding to reference point a within m monitoring periods again.

[0029] An advertisement push system based on an Internet platform, including:

[0030] A collection module: Mark the advertisements browsed by users as to-be-determined advertisements, obtain the browsing data of the to-be-determined advertisements, where the browsing data includes the total return times F1, the average time interval F2 between two adjacent returns, and the total browsing duration F3, calculate the standard duration FBZ = SF * F2 / (1 + η * F1), where η represents a preset first correction coefficient, and mark the to-be-determined advertisements with a standard duration greater than the preset market threshold as target advertisements;

[0031] Analysis module: Obtain the keywords in the target advertisement, use them as the first keywords, and obtain the first word vectors based on the first keywords;

[0032] Obtain the keywords in the advertisements placed by the brand party, use them as the second keywords, obtain the second word vectors based on the second keywords, determine the cosine value of the included angle between the first word vectors and the second word vectors, and mark the target advertisements with the corresponding cosine value of the included angle less than the preset threshold of the cosine value of the included angle as effective advertisements;

[0033] Periodically obtain the number of effective advertisements within a preset monitoring period, draw a curve of the number of effective advertisements changing with time, denoted as the first curve, obtain the second curve based on the first curves corresponding to m monitoring periods, m is a preset number, use the straight line passing through the point (0, Dys) on the y-axis and parallel to the x-axis as the reference line, Dys represents the preset threshold of the number of effective advertisements, and use the part of the second curve above the reference line as the target curve;

[0034] Delivery module: Calculate the selection value based on the target curve, and its calculation formula is:

[0035]

[0036] where, β represents the preset second correction coefficient, n represents the total number of the target curves corresponding to a single Internet platform, Ti, sta and Ti, end respectively represent the time points corresponding to the start and end of the i-th target curve, and the i-th pending delivery period Ti = Ti, end - Ti, sta;

[0037] Determine the target platform based on the selection value, and place advertisements on the target platform.

[0038] Advantages of the present invention: In this solution, first, by analyzing the user browsing behavior data, the advertisements attractive to users are screened out. Based on the actual browsing behavior of the user on the current Internet platform (such as the number of returns, etc.), the types of advertisements that the user is interested in on this platform can be accurately found, so as to accurately understand the user's needs and preferences, and thus clarify which types of advertisements are attractive on different Internet platforms, providing a reference standard for the subsequent advertisement placement of the brand party; then, by comparing the target advertisement with the advertisement content of the brand party, the advertisement with a higher similarity is selected as the effective advertisement. By matching the keywords and content of the target advertisement with the brand party's advertisement, the advertisement that not only meets the interests of the platform users but also fits the brand promotion content (i.e., the effective advertisement) is found, which is used to calculate the subsequent selection value. The effective advertisement refers to the target advertisement that is similar to both the product advertisement of the brand party and the attractive advertisement type on the Internet platform. Therefore, the number of effective advertisements can approximately reflect the effect of the brand party's advertisement placement on the Internet platform, that is, the advertisement revenue; finally, by quantifying the change trend of the advertisement benefit, the placement effect of each Internet platform is scientifically evaluated. By dynamically understanding the dissemination effect of the effective advertisement at different time periods and on different platforms, data support is provided for determining the best placement time and platform combination; through the scientific analysis and comparison of the target curve, the platform with the best placement effect can be accurately found, avoiding the dispersion of resources to the platforms with poor effects, and realizing cost-effective advertisement placement. Brief Description of the Drawings

[0039] The present invention will be further described below with reference to the accompanying drawings.

[0040] Figure 1 It is a flowchart of an advertisement push method based on an Internet platform according to the present invention. Detailed Embodiments

[0041] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without making creative efforts belong to the scope of protection of the present invention.

[0042] Please refer to Figure 1 As shown, the present invention is an advertisement push method based on an Internet platform, including the following steps:

[0043] S1: Mark the advertisements browsed by the user as pending advertisements, obtain the browsing data of the pending advertisements, where the browsing data includes the total number of returns F1, the average time interval F2 between two adjacent returns, and the total browsing duration F3, calculate the standard duration FBZ = SF * F2 / (1 + η * F1), η represents a preset first correction coefficient, and mark the pending advertisements with a standard duration greater than the preset market threshold as target advertisements;

[0044] S2: Obtain the keywords in the target advertisements as the first keywords, and obtain the first word vectors based on the first keywords;

[0045] Obtain the keywords in the advertisements placed by the brand side as the second keywords, obtain the second word vectors based on the second keywords, determine the cosine value of the angle between the first word vector and the second word vector, and mark the target advertisements with a corresponding cosine value of the angle less than the preset cosine value threshold as effective advertisements;

[0046] Periodically obtain the number of effective advertisements within a preset monitoring period, draw a curve of the number of effective advertisements changing with time, denoted as the first curve, obtain the second curve based on the first curves corresponding to m monitoring periods, m is a preset number, use the straight line passing through the point (0, Dys) on the y-axis and parallel to the x-axis as the reference line, Dys represents the preset effective advertisement number threshold, and use the part of the second curve above the reference line as the target curve;

[0047] S3: Calculate the selection value based on the target curve, and its calculation formula is:

[0048]

[0049] where β represents a preset second correction coefficient, n represents the total number of the target curves corresponding to a single Internet platform, Ti, sta and Ti, end respectively represent the start and end time points of the i-th target curve, and the i-th pending placement period Ti = Ti, end - Ti, sta;

[0050] Determine the target platform based on the selection value, and place advertisements on the target platform.

[0051] It should be noted that by analyzing the user browsing behavior data, attractive ads for users can be screened out. Based on the actual browsing behavior of users on the current Internet platform (such as the number of returns, etc.), the types of ads that users are interested in on this platform can be accurately found, so as to accurately understand the user needs and preferences, and thus clarify which ad types are attractive on different Internet platforms, providing a reference standard for the subsequent ad placement of brand owners; then, by comparing the target ads with the ad content of brand owners, ads with a relatively high similarity are selected as effective ads. By matching the keywords and content of the target ads with those of brand owners, ads that not only meet the interests of platform users but also fit the brand promotion content (i.e., effective ads) are found for calculating the subsequent selection value. An effective ad refers to a target ad that is similar to both the product ad of the brand owner and the attractive ad types on the Internet platform. Therefore, the number of effective ads can approximately reflect the effect of the ad placement of brand owners on the Internet platform, that is, the ad revenue; finally, by quantifying the change trend of ad benefits, the placement effects of each Internet platform can be scientifically evaluated. By dynamically understanding the communication effects of effective ads at different time periods and on different platforms, data support is provided for determining the best placement time and platform combination; through the scientific analysis and comparison of the target curve, the platform with the best placement effect can be accurately found, avoiding the dispersion of resources to platforms with poor effects and achieving cost-effective ad placement.

[0052] It is worth noting that the browsing data is collected under the premise of obtaining prior user authorization.

[0053] In another preferred embodiment of the present invention, in step S2, the following steps are further included:

[0054] Input the first keyword into a preset word vector model to obtain a first sub-vector, and add the values on each dimension of the first sub-vector to obtain a first word vector;

[0055] Input the second keyword into the word vector model to obtain a second sub-vector, and add the values on each dimension of the second sub-vector to obtain a second word vector.

[0056] It is worth noting that through the word vector model, keywords can be transformed into vector representations with numerical characteristics, and this representation can retain the semantic information and potential associations of keywords. Further, by adding the values on each dimension of the sub-vectors, the complex vector data is simplified into a single numerical form of word vector, which is convenient for subsequent keyword correlation calculation; by quantifying the semantic features of the first keyword and the second keyword through the word vector model, the semantic similarity or association degree between the two can be more intuitively evaluated.

[0057] In another preferred embodiment of the present invention, in step S3, the process of determining the target platform specifically includes:

[0058] Calculate the estimated price of advertising on the k-th Internet platform. QBJk represents the price of advertising pushed by the k-th Internet platform for a preset duration t, and Nk represents the total number of target curves corresponding to the k-th Internet platform;

[0059] Sort the Internet platforms in descending order according to the size of the selection value to obtain a target sorting. Starting from the Internet platform at the top of the target sorting, determine the target sorting position C in the target sorting. The estimated prices corresponding to the first C Internet platforms in the target sorting are less than the preset target price threshold, and the estimated prices corresponding to the first C + 1 Internet platforms in the target sorting are greater than or equal to the total price threshold;

[0060] Take the first C Internet platforms in the target sorting as the target platforms.

[0061] In another preferred embodiment of the present invention, during the process of obtaining the target sorting, when two or more selection values are the same, the Internet platform with the smaller corresponding estimated price is ranked higher in the target sorting.

[0062] In another preferred embodiment of the present invention, in step S3, the process of advertising on the target platform specifically includes:

[0063] Place advertisements during the time period to be placed corresponding to the target platform.

[0064] In another preferred embodiment of the present invention, in step S2, the process of obtaining the second curve specifically includes:

[0065] Select a number of reference points on the first curve at a preset time interval, calculate the average value of the effective advertisement quantities corresponding to the same reference points, generate coordinate points (a, Da), where Da represents the average value of the effective advertisement quantities corresponding to reference point a, and fit the coordinate points to obtain the second curve.

[0066] In another preferred embodiment of the present invention, during the process of calculating the average value of the effective advertisement quantities corresponding to the same reference points, when the difference between the effective advertisement quantity corresponding to reference point a within the monitoring time period x and the average value is greater than the preset threshold, remove the monitoring time period x and set a new monitoring time period again to replace the monitoring time period x and calculate the average value of the effective advertisement quantities corresponding to reference point a within m monitoring time periods again.

[0067] An advertisement push system based on an Internet platform, comprising:

[0068] Collection module: Mark the advertisements browsed by users as pending advertisements, obtain the browsing data of the pending advertisements, where the browsing data includes the total number of returns F1, the average time interval F2 between two adjacent returns, and the total browsing duration F3, calculate the standard duration FBZ = SF * F2 / (1 + η * F1), where η represents a preset first correction coefficient, and mark the pending advertisements with a standard duration greater than the preset market threshold as target advertisements;

[0069] Analysis module: Obtain the keywords in the target advertisements as the first keywords, and obtain the first word vectors based on the first keywords;

[0070] Obtain the keywords in the advertisements put by the brand side as the second keywords, obtain the second word vectors based on the second keywords, determine the cosine value of the angle between the first word vector and the second word vector, and mark the target advertisements with a corresponding cosine value of the angle less than the preset cosine value threshold as effective advertisements;

[0071] Periodically obtain the number of effective advertisements within a preset monitoring period, draw a curve of the number of effective advertisements changing with time, denoted as the first curve, obtain the second curve based on the first curves corresponding to m monitoring periods, where m is a preset number, use the straight line passing through the point (0, Dys) on the y-axis and parallel to the x-axis as the reference line, where Dys represents the preset threshold of the number of effective advertisements, and use the part of the second curve above the reference line as the target curve;

[0072] Delivery module: Calculate the selection value based on the target curve, and its calculation formula is:

[0073]

[0074] where β represents a preset second correction coefficient, n represents the total number of the target curves corresponding to a single Internet platform, Ti, sta and Ti, end respectively represent the start and end time points of the i-th target curve, and the i-th pending delivery period Ti = Ti, end - Ti, sta;

[0075] Determine the target platform based on the selection value, and deliver advertisements on the target platform.

[0076] The above has described an embodiment of the present invention in detail, but the above content is only a preferred embodiment of the present invention and cannot be considered as limiting the scope of implementation of the present invention. All equivalent changes and improvements made according to the scope of the application of the present invention should still fall within the scope covered by the patent of the present invention.

Claims

1. An advertisement push method based on an Internet platform, characterized in that, It includes the following steps: S1: Mark the advertisements browsed by users as pending advertisements, obtain the browsing data of the pending advertisements, where the browsing data includes the total number of returns F1, the average time interval F2 between two adjacent returns, and the total browsing duration F3. Calculate the standard duration FBZ = SF * F2 / (1 + η * F1), where η represents a preset first correction coefficient. Mark the pending advertisements with a standard duration greater than the preset market threshold as target advertisements; S2: Obtain the keywords in the target advertisements as the first keywords, and obtain the first word vectors based on the first keywords; Obtain the keywords in the advertisements put by the brand side as the second keywords, obtain the second word vectors based on the second keywords, determine the cosine value of the angle between the first word vector and the second word vector, and mark the target advertisements with the corresponding cosine value of the angle less than the preset cosine value threshold as effective advertisements; Periodically obtain the number of effective advertisements within a preset monitoring period, draw a curve of the number of effective advertisements changing with time, denoted as the first curve. Obtain the second curve based on the first curves corresponding to m monitoring periods, where m is a preset number. Use the straight line passing through the point (0, Dys) on the y-axis and parallel to the x-axis as the reference line, where Dys represents the preset threshold of the number of effective advertisements. Take the part of the second curve above the reference line as the target curve; S3: Calculate the selection value based on the target curve, and its calculation formula is: where β represents a preset second correction coefficient, n represents the total number of the target curves corresponding to a single Internet platform, Ti, sta and Ti, end respectively represent the start and end time points of the i-th target curve, and the i-th pending placement period Ti = Ti, end - Ti, sta; Determine the target platform based on the selection value, and place advertisements on the target platform.

2. The method for pushing advertisements based on an Internet platform according to claim 1, characterized in that In the step S2, the following steps are further included: Input the first keywords into a preset word vector model to obtain the first sub-vector, and add the values of each dimension of the first sub-vector to obtain the first word vector; Input the second keywords into the word vector model to obtain the second sub-vector, and add the values of each dimension of the second sub-vector to obtain the second word vector.

3. The method for advertising push based on an Internet platform according to claim 2, characterized in that, In the step S3, the process of determining the target platform specifically includes: Calculate the estimated price of advertising on the k-th Internet platform QBJk represents the price of advertising with a preset duration t pushed by the k-th Internet platform, and Nk represents the total number of target curves corresponding to the k-th Internet platform; Sort the Internet platforms in descending order according to the selection value to obtain the target sorting. Starting from the Internet platform at the top of the target sorting, determine the target sorting position C in the target sorting, where the estimated prices corresponding to the first C Internet platforms in the target sorting are less than the preset target price threshold and the estimated prices corresponding to the first C + 1 Internet platforms in the target sorting are greater than or equal to the total price threshold; Take the first C Internet platforms in the target sorting as the target platform.

4. The advertising push method based on an Internet platform according to claim 3, wherein During the process of obtaining the target sorting, when two or more selection values are the same, the Internet platform with a smaller corresponding estimated price is ranked higher in the target sorting.

5. A method for advertising push based on an Internet platform according to claim 1, characterized in that, In the step S3, the process of placing advertisements on the target platform specifically includes: Advertise during the to-be-delivered period corresponding to the target platform.

6. A method for advertising push based on an Internet platform according to claim 1, characterized in that In step S2, the process of obtaining the second curve specifically includes: Select a number of reference points on the first curve at a preset time interval, calculate the average value of the effective advertisement quantities corresponding to the same reference point, generate coordinate points (a, Da), where Da represents the average value of the effective advertisement quantities corresponding to reference point a, and fit the coordinate points to obtain the second curve.

7. A method for advertising push based on an Internet platform according to claim 6, characterized in that, During the process of calculating the average value of the effective advertisement quantities corresponding to the same reference point, when the difference between the effective advertisement quantity corresponding to reference point a within the monitoring period x and the average value is greater than the preset threshold, remove the monitoring period x and set a new monitoring period again, replace the monitoring period x and calculate the average value of the effective advertisement quantities corresponding to reference point a within m monitoring periods again.

8. An advertisement push system based on an Internet platform, characterized in that, It includes: Collection module: Mark the advertisements browsed by users as pending advertisements, obtain the browsing data of the pending advertisements, where the browsing data includes the total number of returns F1, the average time interval F2 between two adjacent returns, and the total browsing duration F3, calculate the standard duration FBZ = SF * F2 / (1 + η * F1), η represents the preset first correction coefficient, and mark the pending advertisements with a standard duration greater than the preset market threshold as target advertisements; Analysis module: Obtain the keywords in the target advertisements, use them as the first keywords, and obtain the first word vectors based on the first keywords; Obtain the keywords in the advertisements placed by the brand party, use them as the second keywords, obtain the second word vectors based on the second keywords, determine the cosine value of the angle between the first word vector and the second word vector, and mark the target advertisements with the corresponding cosine value of the angle less than the preset cosine value threshold of the angle as effective advertisements; Periodically obtain the effective advertisement quantity within the preset monitoring period, draw a curve of the effective advertisement quantity changing with time, denoted as the first curve, obtain the second curve based on the first curves corresponding to m monitoring periods, m is the preset quantity, use the straight line passing through the point (0, Dys) on the y-axis and parallel to the x-axis as the reference line, Dys represents the preset effective advertisement quantity threshold, and use the part of the second curve above the reference line as the target curve; Delivery module: Calculate the selection value based on the target curve, and its calculation formula is: where β represents the preset second correction coefficient, n represents the total number of the target curves corresponding to a single Internet platform, Ti, sta and Ti, end respectively represent the start time point and the end time point corresponding to the i-th target curve, and the i-th to-be-delivered period Ti = Ti, end - Ti, sta; Determine the target platform based on the selection value and advertise on the target platform.

Images