An intelligent evaluation method for advertising effect

Abstract

The present invention relates to the technical field of advertising effect evaluation, and specifically to an intelligent evaluation method for advertising delivery effect. It includes determining the advertising delivery time period, obtaining the click evaluation value, conversion evaluation value, playback evaluation value and heat evaluation value of the advertising delivery plan during the advertising delivery time period, and obtaining the actual effect value during the advertising delivery time period based on the obtained click evaluation value, conversion evaluation value, playback evaluation value and heat evaluation value. By determining the verification operation and verifying the user account through the false touch time threshold or the playback time threshold, the risk account is determined, and the predicted login IP address is determined with the help of the generated risk account, so as to avoid the risk account with unreasonable swiping and influence of the advertising delivery data, so as to make the collection and calculation of the click evaluation value, conversion evaluation value, playback evaluation value and heat evaluation value more accurate, thereby improving the accuracy of the evaluation of the advertising delivery effect.

Description

Technical Field

[0001] The present invention relates to the technical field of advertising effect evaluation, and in particular to an intelligent evaluation method for advertising delivery effect. Background Art

[0002] With the popularity of the Internet and mobile devices, advertising has become an important means for companies to promote their products and services. However, how to evaluate advertising effectiveness and improve the ROI (return on investment) of advertising has become the focus of attention for companies. Traditional advertising effectiveness evaluation methods mainly rely on questionnaire surveys, telephone surveys, etc., but these methods have problems such as small sample size, long survey cycle, and inaccurate data. Therefore, advertising effectiveness evaluation with the help of big data and artificial intelligence technology has become a hot topic in current research.

[0003] Among them, when evaluating the effectiveness of advertising, when using big data and artificial intelligence technology to analyze or predict the effectiveness of advertising, it is usually necessary to use the viewing, click-through rate and other characteristics of the advertising page for analysis. However, after the advertisement is released, in the actual application process, if there are some unreasonable or malicious operations, such as the viewing rate, click-through rate and other characteristics of the swiping advertisements, this will lead to the subsequent analysis and evaluation of the advertising effectiveness through the collected viewing rate, click-through rate and other characteristics. The evaluation of the advertising data will produce huge errors, affecting the analysis, decision-making and optimization of the advertising plan. Summary of the invention

[0004] In view of the above-mentioned shortcomings of the prior art, the present invention provides an intelligent evaluation method for advertising delivery effects, which can effectively solve the problem in the prior art that after the advertisement is delivered, if there are some unreasonable or malicious operations, which leads to huge errors in data evaluation when subsequently analyzing and evaluating the advertising delivery effect through the collected viewing rate, click rate and other characteristics.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0006] The present invention provides an intelligent evaluation method for advertising delivery effect, comprising the following steps:

[0007] Determine the advertising delivery time period and obtain the click evaluation value of the advertising delivery plan during the advertising delivery time period , conversion evaluation value , playback evaluation value And the heat evaluation value ,in:

[0008] Set the false touch time threshold separately and playback time threshold , with false touch time threshold or playback time threshold Confirm the verification operation and verify the user account, and obtain the registration time included in the verification user account , browsing data information duration And login ip address , thereby determining whether the verified user account is a risky account;

[0009] If there are multiple risky accounts, obtain the corresponding login IP address , determine multiple login IP addresses Whether it is an associated boundary area to determine whether to generate a predicted login IP address , if the predicted login IP address is generated Based on the risk registration time , browsing data information duration , login ip address Or predict login IP address Predict risky accounts and avoid collecting advertising data to obtain click evaluation values , conversion evaluation value , playback evaluation value And the heat evaluation value ;

[0010] Based on the click evaluation value obtained , conversion evaluation value , playback evaluation value And the heat evaluation value Get the actual effect value during the advertising period ;

[0011] Click on the evaluation value Calculated according to the following relationship:

[0012]

[0013] Where: The total number of times users click on your ads. is the total number of ad impressions. Determined by the sum of web page ad impressions, video ad impressions, mobile app ad impressions, and hierarchical ad impressions. is the weight coefficient.

[0014] Conversion evaluation value Calculated according to the following relationship:

[0015]

[0016] Where: The total number of conversions completed by users in the ad. A conversion is when a user completes a step by clicking on registration, logging in, or purchasing based on the content in the ad. is the weight coefficient.

[0017] Playback evaluation value Calculated according to the following relationship:

[0018]

[0019] Where: is the total duration of the ad being played. is the total number of times the ad was played. is the weight coefficient, where:

[0020] When determining the total playback time and total playback count, set the playback time threshold , to meet the playback time threshold The playing time is determined as the sampleable playing time and the number of playing times;

[0021] When it is less than the playback time threshold, the playback time and playback times are not sampled.

[0022] Heat evaluation value Calculated according to the following relationship:

[0023]

[0024] Where: is the total amount of advertising delivered, is the total number of times the ad was shared. is the total sales volume of the product corresponding to the advertisement, is the weight coefficient.

[0025] False touch time threshold and playback time threshold When determining, the following steps are also included:

[0026] S100: When the advertisement exists, monitoring the false touch time threshold or playback time threshold End and immediately close the user operation of the advertisement, and obtain whether the user operation is multiple. If multiple, mark them all as verification operations;

[0027] S200, the advertising delivery terminal generates information acquisition rights corresponding to the user information to the website or APP according to the verification operation, and the advertising delivery terminal determines the user account corresponding to the verification operation and marks it as the verification user account;

[0028] S300. Clearly verify the registration time of the user account , browsing data information duration And login ip address ,in:

[0029] If there are multiple verification user accounts corresponding to the registration time , browsing data information duration And login ip address The preset risk registration time , browsing data information duration And login ip address If it corresponds, the verified user account is marked as a risk account, a collection avoidance instruction is generated not to obtain the advertising data of the risk account, and a collection avoidance specified execution time is generated;

[0030] Among them, the advertising data of risky accounts that are not obtained is the total number of clicks. , total impressions , Total playback time , Total number of plays and total number of shares .

[0031] Login IP address Register multiple user accounts at risk , browsing data information duration Under the condition of being in the same login ip address .

[0032] After the collection avoidance instruction is generated, the following steps are also included:

[0033] S310. Obtain the corresponding login IP address from multiple risk accounts Are they consistent?

[0034] If some of the risk accounts are the same among multiple risk accounts, they are divided into multiple accounts with the same login IP address. Risk account number;

[0035] S320. Obtain the login IP address corresponding to the risk account , analyze multiple login IP addresses Whether it is an associated boundary area. The associated boundary area refers to multiple login IP addresses. The corresponding city is at the border, get the login IP address corresponding to the risk account , to log in with the IP address Get the predicted login IP address associated with it Therefore, if there is a user account that meets the preset risk registration time , browsing data information duration , login ip address Or predict login IP address When a user receives a risky account, it is marked as a risky account and the step S300 is executed for it.

[0036] Actual effect value Calculated according to the following relationship:

[0037]

[0038] Where: by are the corresponding weight coefficients, for as well as the sum of is a constant correction factor.

[0039] Actual effect value When confirmed, perform the following judgment steps:

[0040] The actual effect value The preset scale effect value for the advertising delivery period Compare, if present:

[0041] , the advertising effect during the advertising delivery period reaches the expected result;

[0042] On the contrary, the advertising effect during the advertising delivery period does not meet expectations, and information for improvement is generated.

[0043] Compared with the known prior art, the technical solution provided by the present invention has the following beneficial effects:

[0044] By setting the false touch time threshold and the playback time threshold, the false touch time threshold or the playback time threshold is used to determine the verification operation and the verification user account, and the registration time, browsing data information duration and login IP address of the verified user account are obtained, thereby determining the risk account, and using the generated risk account to determine the predicted login IP address, thereby achieving the avoidance of risk accounts with unreasonable swiping and influence on advertising data, so as to make the collection and calculation of click evaluation values, conversion evaluation values, playback evaluation values ​​and popularity evaluation values ​​more accurate, thereby improving the calculation and evaluation accuracy of the actual effect value during the advertising delivery time period, improving the accuracy of the advertising delivery effect evaluation, and facilitating the analysis of advertising delivery plans or optimizing plan configuration. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0046] Figure 1 It is a schematic diagram of the overall algorithm of the present invention;

[0047] Figure 2 It is a schematic diagram of the flow of the advertising delivery effect method of the present invention. DETAILED DESCRIPTION

[0048] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0049] The present invention will be further described below in conjunction with the embodiments.

[0050] Example 1 (see Figure 1 and Figure 2 ): A method for intelligently evaluating the effect of advertising, comprising the following steps:

[0051] Determine the advertising delivery time period (the time between the initial time of advertising delivery and the current time that needs to be evaluated), and obtain the click evaluation value of the advertising delivery plan during the advertising delivery time period , conversion evaluation value , playback evaluation value And the heat evaluation value , among which:

[0052] Click on the evaluation value Calculated according to the following relationship:

[0053]

[0054] Where: The total number of times users click on your ads. It is the total number of ad impressions (the number of times an ad is displayed to users within a certain period of time, including web ads, each time an ad is displayed when the page is loaded and until it is closed, it is counted as an impression; video ads, each time a video starts playing, it is counted as an impression; mobile app ads, each time an app is opened and displays an ad until it is closed, it is counted as an impression; layered ads, which exist on a page or video for a long time and are counted as an impression until they are closed (such as web ads), are counted as an impression again when the user clicks to open the ad for browsing. Among them, web ads, video ads, layered ads and mobile app ads all have preset durations. The total number of ad impressions is determined by the sum of web ad impressions, video ad impressions, layered ad impressions and mobile app ad impressions). is the weight coefficient, which is pre-set for the personnel.

[0055] In the above, the false touch time threshold is pre-set , false touch time threshold The default setting is 2s (based on actual conditions). This is to take into account the possibility that users may accidentally click on an ad and then close it when browsing Internet pages or video information, and the ad playback time is within the accidental click time threshold. , and will not browse the ad page. This situation applies to hierarchical ads. Therefore, under hierarchical ads, if the false touch time threshold is triggered, , the layered ads will not be displayed once, but only recorded as one display, and the click count will be ignored. and This is to determine the total effective clicks and total impressions of the advertisement, so as to accurately analyze the advertising effectiveness data.

[0056] Conversion evaluation value Calculated according to the following relationship:

[0057]

[0058] Where: The total number of conversions completed by users in the ad. A conversion is a complete step in which a user clicks to register, log in, or purchases something based on the content in the ad. This is recorded as a conversion. is the weight coefficient, which is pre-set for the personnel.

[0059] Playback evaluation value Calculated according to the following relationship:

[0060]

[0061] Where: is the total playing time of the ad (the sum of the playing time corresponding to multiple play times). is the total number of times the ad was played. is the weight coefficient. It should be noted that when determining the total playback time and total playback times, the playback time threshold is set. , to meet the current playback time threshold The playback time is determined as the sampleable playback time and the number of playbacks. It can be pre-set to 0-2S. Therefore, when it is less than the playback time threshold, it means that the user quickly closed the playing advertisement page and did not watch the advertisement. In this case, the playback time and the number of playbacks will not be sampled and will not be included in the evaluation of the delivery effect, so as to ensure the accuracy of the advertising effect evaluation.

[0062] In the above, it should be noted that when the false touch time threshold is set and playback time threshold When the advertisement is displayed, the effective data generated by the advertisement can be collected and extracted to determine the effect of the advertisement. However, it should be considered that if there is a time threshold for users to avoid accidental touches, and playback time threshold If the effective advertising data is refreshed and changed in this way, the accuracy of the subsequent evaluation of the advertising effect will still be affected. Therefore, this solution sets a false touch time threshold. and playback time threshold On this basis, the following steps are also included:

[0063] S100: When the advertisement exists, monitoring the false touch time threshold or playback time threshold End and immediately close the user operation of the advertisement, and obtain whether the user operation is multiple. If multiple, mark them all as verification operations;

[0064] S200, the advertising delivery terminal generates information acquisition rights corresponding to the user information to the website or APP based on the verification operation. It should be noted that when the user initially registers an account on the website or APP, the website or APP clearly displays a pop-up window to collect the user's registration information, and explains that it is only used to evaluate the advertising delivery effect data and does not involve other operations. Therefore, the advertising delivery terminal here determines the user account corresponding to the verification operation and marks it as the verification user account;

[0065] S300. Clearly verify the registration time of the user account , browsing data information duration And login ip address ,in:

[0066] If there are multiple verification user accounts corresponding to the registration time , browsing data information duration And login ip address The preset risk registration time , browsing data information duration And login ip address If the user account is marked as a risk account, it should be noted that the preset risk registration time 0-7 days, browsing data information duration It means that after the user account is registered, the user account logs in to the website or APP only within the time threshold of accidental touch or playback time threshold End, that is, the time to exit the website or APP, that is, the false touch time threshold or playback time threshold The end time is the standard, login IP address This means that multiple user accounts meet the above risk registration time , browsing data information duration Under the condition of being in the same login ip address (Login IP addresses of multiple user accounts The same), then the user account is determined to be registered recently and is only used for advertising data ( ) is maliciously swiped or changed, thereby determining that the verified user account is a risky account, generating a collection avoidance instruction not to obtain the advertising data of the risky account, and generating a collection avoidance specified execution time, where the execution time can be preset to 7, 10 days, etc. without specific limitation, thereby making the click evaluation value determined above , conversion evaluation value And playback evaluation value It is more accurate and eliminates the need to collect and analyze advertising data in risky accounts, thus improving data processing and analysis efficiency.

[0067] Furthermore, in the above scheme, a collection avoidance instruction is generated to avoid the collection and analysis of the advertising data generated by the risky account, which further includes the following steps:

[0068] S310. Obtain the corresponding login IP address from multiple risk accounts Are they consistent?

[0069] If some of the risk accounts are the same among multiple risk accounts, they are divided into multiple accounts with the same login IP address. Risk accounts, for example: 9 risk accounts, with 3 different login IP addresses , but there are at least 2 risk accounts with the same login IP address ;

[0070] S320. Obtain the login IP address corresponding to the risk account , analyze multiple login IP addresses Whether it is an associated boundary area. The associated boundary area refers to multiple login IP addresses. The corresponding cities are at the border, and there is no gap between adjacent cities. Login IP addresses of other cities Therefore, if this happens, consider that there is a malicious switch login IP address , because of the login ip address Switch to the login IP address of the junction The distance is relatively close, so it is relatively easier to switch, so that the advertising data can be refreshed and changed. Therefore, the login IP address corresponding to the risk account is obtained. , to log in with the IP address Get the predicted login IP address associated with it , that is, predict the login IP address Is the login ip address The associated boundary area is determined, thereby generating multiple predicted login IP addresses Therefore, if there is a user account that meets the preset risk registration time , browsing data information duration , login ip address Or predict login IP address When the account is marked as a risk account, the step S300 is executed for it. Thus, even if there is no login IP address corresponding to the risk account determined in S300, When using the predicted login IP address Realize the prediction and positioning of risky accounts so that the click evaluation value determined above , conversion evaluation value And playback evaluation value It is more accurate and improves the efficiency of analyzing and evaluating the effectiveness of advertising.

[0071] It should be noted that multiple login IP addresses are determined in the above Or predict login IP address When you log in to the IP address, you can log in to the IP address later. Or predict login IP address To circumvent this, in order to determine a valid click evaluation value , conversion evaluation value And playback evaluation value , which is conducive to the analysis and evaluation of advertising effectiveness.

[0072] Heat evaluation value Calculated according to the following relationship:

[0073]

[0074] Where: is the total amount of advertising delivered, The total number of times the ad was shared (it should be noted that Also determined by means of steps S100-S320 above), is the total sales volume of the product corresponding to the advertisement, is the weight coefficient, which is pre-set for the personnel;

[0075] In summary, then for the above Normalization is performed to obtain the actual effect value during the advertising delivery period. , the actual effect value Calculated according to the following relationship:

[0076]

[0077] Where: as well as are the corresponding weight coefficients, which can be preset by personnel. for as well as the sum of is a constant correction coefficient, and then the actual effect value after the current advertisement is released is calculated. , by taking the actual effect value The preset scale effect value for the advertising delivery period Compare, if present:

[0078] , it means that the advertising effect in the advertising delivery time period has reached the expectation; otherwise, it means that the advertising effect in the advertising delivery time period has not reached the expectation. Therefore, information for improvement is generated to indicate that there is room for improvement in the advertising delivery plan in the advertising delivery time period, so that advertising terminal personnel can optimize the advertising delivery plan.

[0079] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements will not cause the essence of the corresponding technical solutions to deviate from the protection scope of the technical solutions of the embodiments of the present invention.

Claims

1. An intelligent evaluation method for advertising delivery effect, characterized in that: The steps include: Determine the advertising delivery time period and obtain the click evaluation value of the advertising delivery plan during the advertising delivery time period , conversion evaluation value , playback evaluation value And the heat evaluation value ,in: Set the false touch time threshold separately and playback time threshold , with false touch time threshold or playback time threshold Confirm the verification operation and verify the user account, and obtain the registration time included in the verification user account , browsing data information duration And login ip address , thereby determining whether the verified user account is a risky account; If there are multiple risky accounts, obtain the corresponding login IP address , determine multiple login IP addresses Whether it is an associated boundary area to determine whether to generate a predicted login IP address , if the predicted login IP address is generated Based on the risk registration time , browsing data information duration , login ip address Or predict login IP address Predict risky accounts and avoid collecting advertising data to obtain click evaluation values , conversion evaluation value , playback evaluation value And the heat evaluation value ; Based on the click evaluation value obtained , conversion evaluation value , playback evaluation value And the heat evaluation value Get the actual effect value during the advertising period ; False touch time threshold and playback time threshold When determining, the following steps are also included: S100: When the advertisement exists, monitoring the false touch time threshold or playback time threshold End and immediately close the user operation of the advertisement, and obtain whether the user operation is multiple. If multiple, mark them all as verification operations; S200, the advertising delivery terminal generates information acquisition rights corresponding to the user information to the website or APP according to the verification operation, and the advertising delivery terminal determines the user account corresponding to the verification operation and marks it as the verification user account; S300. Clearly verify the registration time of the user account , browsing data information duration And login ip address ,in: If there are multiple verification user accounts corresponding to the registration time , browsing data information duration And login ip address The preset risk registration time , browsing data information duration And login ip address If the verification is successful, the verified user account is marked as a risk account, a collection avoidance instruction is generated to prevent the acquisition of the advertising data of the risk account, and an execution time of the collection avoidance instruction is generated; Among them, the advertising data of risky accounts that are not obtained is the total number of clicks. , total impressions , Total playback time , Total number of plays and total number of shares ; After the collection avoidance instruction is generated, the following steps are also included: S310. Obtain the corresponding login IP address from multiple risk accounts Are they consistent? If some of the risk accounts are the same among multiple risk accounts, they are divided into multiple accounts with the same login IP address. Risk account number; S320. Obtain the login IP address corresponding to the risk account , analyze multiple login IP addresses Whether it is an associated boundary area. The associated boundary area refers to multiple login IP addresses. The corresponding city is at the border, get the login IP address corresponding to the risk account , to log in with the IP address Get the predicted login IP address associated with it Therefore, if there is a user account that meets the preset risk registration time , browsing data information duration , login ip address Or predict login IP address When a user receives a message from a third party, the user is marked as a risky account and the step S300 is executed for the account. Risk Registration Time 0-7 days; The execution time is preset to 7 or 10 days.

2. The method for intelligently evaluating the effect of advertising according to claim 1, characterized in that: The click evaluation value Calculated according to the following relationship: Where: The total number of times users click on your ads. is the total number of ad impressions. Determined by the sum of web page ad impressions, video ad impressions, mobile app ad impressions, and hierarchical ad impressions. is the weight coefficient.

3. The method for intelligently evaluating the effect of advertising according to claim 1, characterized in that: The conversion evaluation value Calculated according to the following relationship: Where: The total number of conversions completed by users in the ad. A conversion is when a user completes a step by clicking on registration, logging in, or purchasing based on the content in the ad. is the weight coefficient.

4. The method for intelligently evaluating the effect of advertising according to claim 1, characterized in that: The playback evaluation value Calculated according to the following relationship: Where: is the total duration of the ad being played. is the total number of times the ad was played. is the weight coefficient, where: When determining the total playback time and total playback count, set the playback time threshold , to meet the playback time threshold The playback time determines the sampling playback time and the number of playbacks; When it is less than the playback time threshold, the playback time and playback times are not sampled.

5. The method for intelligently evaluating the effect of advertising according to claim 1, characterized in that: The heat evaluation value Calculated according to the following relationship: Where: is the total amount of advertising delivered, is the total number of times the ad was shared. is the total sales volume of the product corresponding to the advertisement, is the weight coefficient.

6. The method for intelligently evaluating the effect of advertising according to claim 1, characterized in that: The login ip address Register multiple user accounts at risk , browsing data information duration Under the condition of being in the same login ip address .

7. The method for intelligently evaluating the effect of advertising according to claim 1, characterized in that: The actual effect value Calculated according to the following relationship: Where: as well as are the corresponding weight coefficients, for as well as the sum of is a constant correction factor.

8. The method for intelligently evaluating the effect of advertising according to claim 1, characterized in that: The actual effect value When confirmed, perform the following judgment steps: The actual effect value The preset scale effect value for the advertising delivery period Compare, if present: , the advertising effect during the advertising delivery period reaches the expected result; On the contrary, the advertising effect during the advertising delivery period does not meet expectations, and information for improvement is generated.

Images