Advertisement intelligent hosting method based on big data

Abstract

The invention discloses an intelligent advertisement hosting method based on big data, belongs to the field of advertisement hosting, and solves the problem of how to effectively supervise an advertisement putting process to realize man-machine cooperation so as to avoid limitation of an algorithm and neglect user experience during intelligent advertisement hosting. According to the method, the putting effect coefficient is calculated through the putting conversion rate and the user dissatisfaction degree, whether the advertisement putting effect reaches the expectation or not can be judged preliminarily, and resources are saved; the putting effect coefficient, the putting accuracy and the putting density are combined to calculate a putting early warning value, and the advertisement putting effect can be comprehensively evaluated from multiple dimensions; according to the method and the system, the influence of the three is integrated into one index according to the advertisement putting early warning value, the problem in the advertisement effect can be revealed, and an accurate early warning signal is provided, so that the bad putting strategy is improved in time in the advertisement putting supervision process, and the intelligent hosting of the target advertisement is more efficient.

Description

Technical Field

[0001] The present invention belongs to the field of advertising hosting, and specifically is an advertising intelligent hosting method based on big data. Background Art

[0002] Ad Smart Management is an advertising management method based on artificial intelligence and big data analysis technology, which aims to automate and optimize the advertising delivery process. Through Ad Smart Management, advertisers can manage advertising delivery more efficiently and accurately, achieving better delivery results and higher return on investment.

[0003] In the prior art, when advertising is intelligently hosted, the advertising delivery strategy basically relies entirely on intelligent analysis software. Although the advertising delivery strategy can be flexibly changed during intelligent hosting, there are also some potential problems and challenges, and it is not perfect. For example, some delivery strategy algorithms may have limitations and ignore user experience. It is still necessary to supervise the advertising delivery process to achieve human-machine collaboration. Summary of the invention

[0004] The present invention aims to solve at least one of the technical problems existing in the prior art. To this end, the present invention proposes an advertising intelligent hosting method based on big data, which solves how to effectively supervise the advertising delivery process to achieve human-machine collaboration, thereby avoiding the limitations of the algorithm during advertising intelligent hosting and the problem of ignoring user experience.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] An advertising intelligent hosting method based on big data, comprising:

[0007] Step S1: obtaining user behavior data of target advertisements delivered in the past preset time period, obtaining the delivery conversion rate and user dissatisfaction of the target advertisements in the past preset time period based on the user behavior data analysis; and evaluating the delivery effect of the target advertisements in the past preset time period based on the delivery conversion rate and user dissatisfaction;

[0008] Step S2: obtaining user characteristic data and delivery frequency data of the target advertisement in the past preset time period, obtaining the delivery accuracy of the target advertisement in the past preset time period based on the user characteristic data analysis, and obtaining the delivery density of the target advertisement in the past preset time period based on the delivery frequency data analysis;

[0009] Step S3: When the delivery effect of the target advertisement in the past preset time period is poor, a comprehensive analysis is performed based on the delivery effect, delivery accuracy and delivery density of the target advertisement in the past preset time period to issue an early warning for the current intelligent hosting delivery strategy.

[0010] Furthermore, obtaining the delivery conversion rate and user dissatisfaction of the target advertisement in the past preset time period based on the user behavior data analysis includes:

[0011] Traverse the user behavior data of the target advertisements delivered in the past preset time period, and obtain the browsing behavior data, conversion behavior data and user feedback data of the target advertisements in the past preset time period; wherein the browsing behavior data includes the total number of users who have browsed the target advertisements in the past preset time period and the total time each user browses the target advertisements, which are marked as I and Ti respectively; i represents the user number, i=1, 2...I; the conversion behavior data includes the number of users participating in different conversion event types Ej in the past preset time period, wherein j represents the number of conversion event types, j=1, 2...J; J represents the total number of conversion event types of the target advertisements; the user feedback data includes the total number of users who have negative feedback on the target advertisements in the past preset time period, the number of negative feedbacks by each user who has negative feedback, and the total time each user who has negative feedback browses the target advertisements, and are marked as G, Ug and Tg respectively; g=1, 2...G; g represents the number of users who have negative feedback on the target advertisements in the past preset time period;

[0012] Based on the browsing behavior data and conversion behavior data analysis, the target advertisement delivery conversion rate in the past preset time period is obtained. The calculation formula is:

[0013] In the formula, CR represents the delivery conversion rate; kj represents the preset proportional coefficient of the target advertisement conversion event type; and kj is greater than 0; T0 represents the complete playback time of the target advertisement;

[0014] Based on the browsing behavior data and user feedback data analysis, the user dissatisfaction of the target advertisement in the past preset time period is obtained. The calculation formula is:

[0015] In the formula, DR represents user dissatisfaction.

[0016] Furthermore, the evaluation of the delivery effect of the target advertisement in the past preset time period based on the delivery conversion rate and user dissatisfaction includes:

[0017] Obtain the delivery conversion rate CR and user dissatisfaction DR of the target advertisement in the past preset time period;

[0018] According to the calculation formula Calculate and obtain the delivery effect coefficient EC of the target advertisement in the past preset time period; where α and β represent the preset weight factors of delivery conversion rate and user dissatisfaction respectively; and α and β are both greater than 0;

[0019] The calculated delivery effect coefficient of the target advertisement in the past preset time period is compared with the preset delivery effect coefficient threshold to determine the delivery effect of the target advertisement in the past preset time period;

[0020] If the calculated delivery effect coefficient of the target advertisement in the past preset time period is greater than or equal to the preset delivery effect coefficient threshold, a signal is generated that the delivery effect of the target advertisement in the past preset time period is good;

[0021] If the calculated delivery effect coefficient of the target advertisement in the past preset time period is less than the preset delivery effect coefficient threshold, a signal indicating that the delivery effect of the target advertisement in the past preset time period is poor is generated.

[0022] Furthermore, the user characteristic data of the target advertisement includes target audience data and actual audience observation data; wherein the target audience data refers to the user characteristics that match the target audience; the actual audience observation data refers to the user characteristics of each actual user who has viewed the target advertisement; the target advertisement delivery frequency data includes the frequency of displaying the target advertisement to the user group in the past preset time period;

[0023] Furthermore, the method of obtaining the delivery accuracy of the target advertisement in the past preset time period based on the user characteristic data analysis includes:

[0024] Compare the actual audience observation data obtained with the target audience data, and count the number of users who fully meet the target audience data in the past preset time period, marking them as NP;

[0025] According to the calculation formula Get the delivery accuracy of the target advertisement in the past preset time period; where I represents the total number of users who have viewed the target advertisement in the past preset time period.

[0026] Furthermore, the obtaining of the delivery density of the target advertisement in the past preset time period based on delivery frequency data analysis includes:

[0027] Divide the past preset time period into several time intervals according to different time period types in a day; divide the time intervals belonging to the same time period type into one group;

[0028] Then, the number of times the target advertisement is delivered in different time intervals in the past preset time period is counted and marked as Cf,h; wherein f represents the time period type, f=1,2...F; F is the total number of time period types; h represents the number of the time interval in the same group, h=1,2...H; H represents the total number of time intervals in the same group;

[0029] According to the calculation formula Get the delivery density of the target advertisement in the past preset time period; where I represents the total number of users who have viewed the target advertisement in the past preset time period; λf represents the preset proportion coefficient of different time periods; and λf is greater than 0.

[0030] Further, when the delivery effect of the target advertisement in the past preset time period is poor, the delivery effect coefficient EC, delivery accuracy QR and delivery density MR of the target advertisement in the past preset time period are obtained;

[0031] According to the calculation formula Obtain the delivery warning value WV of the target advertisement in the past preset time period; where EC0, QR0 and MR0 represent the delivery effect value threshold, delivery accuracy standard value and delivery density standard value respectively;

[0032] Compare the calculated delivery warning value with the preset comprehensive delivery efficiency threshold to determine whether the target advertisement is normal in the current delivery strategy;

[0033] If the calculated delivery warning value is greater than or equal to the preset comprehensive delivery efficiency threshold, a normal signal is generated and sent to the hosting backend; after the hosting backend receives the normal signal, the relevant personnel do not need to pay attention to it;

[0034] If the calculated delivery warning value is less than the preset comprehensive delivery efficiency threshold, a warning signal is generated and sent to the hosting backend; after the hosting backend receives the warning signal, relevant personnel need to handle it in a timely manner.

[0035] Compared with the prior art, the present invention has the following beneficial effects:

[0036] 1. In the present invention, the advertising delivery effect is analyzed in combination with the advertising delivery conversion rate and user dissatisfaction. Such a dual analysis focuses on both improving the conversion effect and improving the user experience, thereby achieving a higher return on investment and brand value. Specifically, a high conversion rate indicates that the advertisement effectively attracts users and promotes a variety of conversion events, but if it is accompanied by high user dissatisfaction, it may cause damage to the brand image. By combining these two indicators, advertisers can reduce user dissatisfaction while improving the conversion rate, thereby avoiding the advertising effect being offset by negative feedback. Therefore, combining the delivery conversion rate and user dissatisfaction can more accurately reflect the delivery effect of the advertisement, and provide the primary judgment condition for whether the delivery strategy needs to be optimized in the future.

[0037] 2. In the present invention, delivery accuracy measures whether the advertisement is accurately delivered to users who meet the characteristics of the target audience. If the delivery accuracy is low, it means that the advertisement may not effectively reach the users who are truly interested. However, too high delivery accuracy may also cause user fatigue, thereby affecting the delivery effect of the advertisement. Delivery density reflects the coverage frequency of the advertisement in the actual user group. High density may cause visual fatigue of the user, while low density may fail to fully cover the target users. By analyzing these two indicators, it can be determined whether the effect problem is caused by improper advertising accuracy or improper delivery density, or both. This provides a data basis for subsequent advertising supervision and early warning.

[0038] 3. In the present invention, the delivery effect coefficient is first calculated by the delivery conversion rate and user dissatisfaction, which can help to preliminarily judge whether the effect of advertising delivery meets the expectations; if the delivery effect coefficient indicates that the advertising delivery effect is poor, the delivery accuracy and delivery density of the advertisement will be further analyzed, which can avoid unnecessary in-depth analysis when the advertising effect itself is very good, thereby saving resources; the delivery effect coefficient, delivery accuracy and delivery density are combined to calculate the delivery warning value, which can comprehensively evaluate the effect of advertising delivery from multiple dimensions; the delivery effect coefficient provides an overall effect evaluation, while delivery accuracy and delivery density are supplemented by analysis from the perspectives of audience matching and display frequency; the delivery warning value integrates the influence of these three into one indicator, which helps to reveal problems in advertising effectiveness and provide accurate warning signals, so that bad delivery strategies can be improved in a timely manner during the process of advertising delivery supervision, making the intelligent hosting of target advertisements more efficient. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 The present invention is a flowchart of an advertising intelligent hosting method based on big data. DETAILED DESCRIPTION

[0040] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than 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.

[0041] like Figure 1 As shown, an advertising intelligent hosting method based on big data includes:

[0042] Step S1: obtaining user behavior data of target advertisements delivered in the past preset time period, obtaining the delivery conversion rate and user dissatisfaction of the target advertisements in the past preset time period based on the user behavior data analysis; and evaluating the delivery effect of the target advertisements in the past preset time period based on the delivery conversion rate and user dissatisfaction;

[0043] Step S2: obtaining user characteristic data and delivery frequency data of the target advertisement in the past preset time period, obtaining the delivery accuracy of the target advertisement in the past preset time period based on the user characteristic data analysis, and obtaining the delivery density of the target advertisement in the past preset time period based on the delivery frequency data analysis;

[0044] Step S3: When the delivery effect of the target advertisement in the past preset time period is poor, a comprehensive analysis is performed based on the delivery effect, delivery accuracy and delivery density of the target advertisement in the past preset time period to issue an early warning for the current intelligent hosting delivery strategy.

[0045] The present invention adopts the method of evaluating the delivery effect of the target advertisement in the past preset time period based on the delivery conversion rate and user dissatisfaction; if the delivery effect of the target advertisement in the past preset time period is good, no warning is required; if the delivery effect of the target advertisement in the past preset time period is poor, it means that there is a problem with both or either of the delivery conversion rate and user dissatisfaction, which may be that the advertisement content itself is not attractive enough, or that the delivery density is too high, causing users to feel tired and not having a good user experience, or that the existing delivery strategy cannot achieve a good delivery effect, and a comprehensive analysis and warning are required to remind manual intervention;

[0046] In step S1, the user behavior data of the target advertisement delivered in the past preset time period includes browsing behavior data, conversion behavior data, user feedback data, etc.; the preset time period may be one day, one week, etc., depending on the advertiser's revenue target and budget;

[0047] The method of obtaining the delivery conversion rate and user dissatisfaction of the target advertisement in the past preset time period based on the user behavior data analysis includes:

[0048] The user behavior data of the acquired target advertisement delivered in the past preset time period is traversed to obtain the browsing behavior data, conversion behavior data and user feedback data of the target advertisement in the past preset time period; wherein the browsing behavior data includes the total number of users who have browsed the target advertisement in the past preset time period and the total time each user browses the target advertisement, which are marked as I and Ti respectively; i represents the user number, i=1, 2...I; it should be noted that in the present invention, when the target advertisement is delivered to a large number of users, if the corresponding user has browsed the target advertisement, then the total time of browsing the target advertisement is greater than 0, if the corresponding user directly ignores the target advertisement every time he sees it in the past preset time period, then the total time of browsing the target advertisement is equal to 0; the conversion behavior data includes the number of users participating in different conversion event types Ej in the past preset time period, wherein j represents the number of the conversion event type, j=1,2...J; J represents the total number of conversion event types of the target advertisement; the conversion event type is determined by the advertisement attribute, for example, the conversion event types that the product advertisement requires the user to generate may include click, add to cart, purchase, etc., and the conversion event types that the software advertisement requires the user to generate may include click, download, register, log in, and membership recharge, etc.; the user feedback data includes the total number of users who have negative feedback on the target advertisement in the past preset time period, the number of negative feedbacks of each user with negative feedback, and the total time each user with negative feedback browses the target advertisement, and are marked as G, Ug and Tg respectively; g=1,2...G; g represents the number of users who have negative feedback on the target advertisement in the past preset time period; it is understandable that the same user may upload multiple negative feedbacks based on the same advertisement;

[0049] Based on the browsing behavior data and conversion behavior data analysis, the target advertisement delivery conversion rate in the past preset time period is obtained. The calculation formula is:

[0050] In the formula, CR represents the delivery conversion rate; kj represents the preset ratio coefficient of the target advertisement conversion event type; and kj is greater than 0; T0 represents the complete broadcasting time of the target advertisement; it should be noted that the value of kj is obtained by calculating the delivery conversion rate based on a large amount of browsing behavior data and conversion behavior data of the target advertisement in the early stage, and analyzing it in combination with the conversion rate effect expected by the advertiser;

[0051] When calculating the delivery conversion rate within a preset time period in the past, by measuring the product of the user participation rate of all types of conversion events and the total browsing completion ratio, it can more accurately reflect the actual degree of users' viewing of the advertising content; this method is different from the traditional conversion rate calculation method, which usually ignores the user's stay time on the advertising page or the complete viewing of the advertisement. The total browsing completion ratio, as a reflection of the user's interest level, can more accurately evaluate the actual impact of the advertisement; generally, the longer the user's browsing time, the stronger their interest in the advertising content, and thus the more likely they are to complete the conversion behavior; this method considers the depth of interaction between users and advertisements, not only focusing on the user's conversion behavior, but also evaluating the user's participation in the advertisement, making the conversion rate evaluation more comprehensive and representative;

[0052] In addition, by setting the preset proportional coefficient of the conversion event type, the relative value of different conversion behaviors can be more accurately reflected when calculating the conversion rate, so that the calculation results are more in line with the actual needs of advertisers; because different conversion behaviors may involve different degrees of user operation difficulty or effort, the preset proportional coefficient of the conversion event type can fairly reflect these differences in the calculation, avoiding the excessive impact of certain simple conversion behaviors on the overall conversion rate, thereby improving the effectiveness of advertising;

[0053] At the same time, based on the browsing behavior data and user feedback data analysis, the user dissatisfaction of the target advertisement in the past preset time period is obtained. The calculation formula is:

[0054] In the formula, DR represents user dissatisfaction;

[0055] Understandably, It indicates the proportion of users who provided negative feedback among all users who viewed the target advertisement in the past preset time period, reflecting the breadth of users' overall satisfaction with the advertisement; it reflects how many users, among all users who were exposed to the advertisement, were dissatisfied with or had negative emotions towards the advertisement; the higher the proportion, the more common the negative reaction caused by the advertisement among the target audience, which may mean that there are problems with the content, format or positioning of the advertisement, thus leading to greater user dissatisfaction; The ratio of the total number of negative feedbacks in the past preset time period to the browsing completion ratio of users with negative feedback reflects the intensity of negative feedback generated by users during the advertising participation process; specifically, this ratio reflects the frequency and intensity of negative feedback in the process of users browsing the advertisement completely; if the ratio is high, it means that users frequently generate negative feedback in a long advertising browsing process, indicating that the advertising content may have caused great negative emotions or dissatisfaction to users; this measurement method can reveal the overall response quality of users to advertisements, especially when they have sufficient exposure to advertising content; the user dissatisfaction in the present invention can comprehensively consider the prevalence of negative feedback and the intensity of feedback. This method not only reflects how many users provide negative feedback, but also considers the actual participation level and feedback frequency of these users, thereby providing a more comprehensive and accurate user dissatisfaction evaluation, effectively reducing the deviation that may be caused by a single indicator, presenting the dissatisfaction of advertisements more realistically, and providing a more valuable reference basis for advertising optimization;

[0056] The evaluation of the delivery effect of the target advertisement in the past preset time period based on the delivery conversion rate and user dissatisfaction includes:

[0057] Obtain the delivery conversion rate CR and user dissatisfaction DR of the target advertisement in the past preset time period;

[0058] According to the calculation formula Calculate and obtain the delivery effect coefficient EC of the target advertisement in the past preset time period;

[0059] In the formula, α and β represent the preset weight factors of delivery conversion rate and user dissatisfaction respectively; and α and β are both greater than 0; CRα can amplify or weaken the impact of delivery conversion rate on delivery effect coefficient through the preset weight factor α, and β determines the decay speed of user dissatisfaction on delivery effect coefficient, ensuring that even if the delivery conversion rate CR value is very high, but the user dissatisfaction DR is also very high, the delivery effect coefficient EC will not be exaggerated; Among them, the values ​​of α and β are obtained in the early stage of the target advertising delivery test based on a large number of delivery effect coefficient calculations and the expected analysis of the advertiser; Among them, if the delivery conversion rate CR value is larger and the user dissatisfaction DR value is smaller, the delivery effect coefficient EC value is larger, indicating that the delivery effect of the target advertisement in the past preset time period is better;

[0060] The calculated delivery effect coefficient of the target advertisement in the past preset time period is compared with the preset delivery effect coefficient threshold to determine the delivery effect of the target advertisement in the past preset time period; wherein the preset delivery effect coefficient threshold is obtained through a large number of advertisement delivery effect calculations and advertiser expectation analysis in the early stage;

[0061] If the calculated delivery effect coefficient of the target advertisement in the past preset time period is greater than or equal to the preset delivery effect coefficient threshold, a signal is generated that the delivery effect of the target advertisement in the past preset time period is good;

[0062] If the calculated delivery effect coefficient of the target advertisement in the past preset time period is less than the preset delivery effect coefficient threshold, a signal that the delivery effect of the target advertisement in the past preset time period is poor is generated;

[0063] It should be noted that, in this embodiment, when calculating user dissatisfaction and delivery effect coefficient, the purpose of adding 1 to the denominator is to prevent the denominator from being zero, thereby ensuring the validity and stability of the calculation formula;

[0064] It is understandable that combining the advertising conversion rate and user dissatisfaction to analyze the advertising effect, such a dual analysis focuses on both improving the conversion effect and improving the user experience, thereby achieving a higher return on investment and brand value; specifically, a high conversion rate indicates that the advertisement effectively attracts users and promotes a variety of conversion events, but if it is accompanied by high user dissatisfaction, it may cause damage to the brand image; by combining these two indicators, advertisers can reduce user dissatisfaction while increasing the conversion rate, thereby avoiding the advertising effect being offset by negative feedback; therefore, combining the advertising conversion rate and user dissatisfaction can more accurately reflect the advertising effect, and provide the primary judgment condition for whether the subsequent delivery strategy needs to be optimized;

[0065] In step S2, the user characteristic data of the target advertisement includes target audience data and actual audience observation data; wherein the target audience data refers to user characteristics that meet the target audience, such as the user's age, gender, interests, geographic location, etc.; the actual audience observation data refers to the user characteristics of each actual user who has browsed the target advertisement;

[0066] The method of obtaining the delivery accuracy of the target advertisement in the past preset time period based on the analysis of user characteristic data includes:

[0067] Compare the actual audience observation data obtained with the target audience data, and count the number of users who fully meet the target audience data in the past preset time period, marking them as NP;

[0068] According to the calculation formula Obtain the delivery accuracy of the target advertisement in the past preset time period; where I represents the total number of users who have viewed the target advertisement in the past preset time period;

[0069] It is understandable that delivery accuracy measures whether the advertisement is delivered to users who meet the data characteristics of the target audience. If the delivery accuracy is high, it means that the target advertisement is displayed to the user group that is more in line with the target audience data. These users are more likely to be interested in the advertisement content, thereby improving the conversion rate; if the delivery accuracy is low, it means that the advertisement fails to reach the user group that is interested in it, resulting in inefficient delivery, which in turn affects the overall effect of the advertisement. Delivery accuracy affects the audience matching of the advertisement. If the advertisement is not delivered to the correct user, the conversion rate is usually low, which may lead to a decrease in the effect coefficient; but it should also be noted that too high delivery accuracy can also cause user fatigue, thereby affecting the delivery effect of the advertisement;

[0070] The target advertisement delivery frequency data includes the frequency of displaying the target advertisement to the user group in the past preset time period; the delivery density of the target advertisement in the past preset time period obtained based on delivery frequency data analysis includes:

[0071] Divide the past preset time period into several time intervals according to different time period types in a day; divide the time intervals belonging to the same time period type into one group;

[0072] Then, the number of times the target advertisement is delivered in different time intervals in the past preset time period is counted and marked as Cf,h; wherein f represents the time period type, f=1,2...F; F is the total number of time period types; h represents the number of the time interval in the same group, h=1,2...H; H represents the total number of time intervals in the same group;

[0073] According to the calculation formula Obtain the delivery density of the target advertisement in the past preset time period; where I represents the total number of users who have viewed the target advertisement in the past preset time period; λf represents the preset ratio coefficient of different time periods; and λf is greater than 0; the value of λf is obtained based on the analysis of user activity in different time periods. Generally, during the peak hours in the morning, afternoon, and evening, there are more users with free time, and a higher preset ratio coefficient can be set for these time periods; at other times, there are fewer users, and a lower preset ratio coefficient can be set for these time periods; the specific numerical setting is set by the relevant technical personnel of the advertising hosting platform;

[0074] It is understandable that high delivery density means that advertisements are displayed more frequently during the time period when users may be active, thereby increasing the exposure of advertisements. The calculation method of delivery density in the present invention is based on the user activity in different time periods and introduces a preset proportional coefficient. This method not only takes into account the number of advertisement displays, but also combines the user's active time period, allowing advertisers to increase advertisement delivery during high-activity periods and reduce delivery during low-activity periods, thereby achieving optimal resource allocation; however, it should be noted that if the delivery density is too high, it may cause user visual fatigue and reduce the advertising effect; if the delivery density is too low, it may not be able to fully cover the target users, affecting the conversion effect; appropriate delivery density can increase the exposure opportunities of advertisements while avoiding the negative effects of excessive display;

[0075] In the present invention, delivery accuracy measures whether the advertisement is accurately delivered to users who meet the characteristics of the target audience; if the delivery accuracy is low, it means that the advertisement may not effectively reach the users who are truly interested; but too high delivery accuracy may also cause user fatigue, thereby affecting the delivery effect of the advertisement; delivery density reflects the coverage frequency of the advertisement in the actual user group. High density may cause user visual fatigue, while low density may fail to fully cover the target users; by analyzing these two indicators, it can be determined whether the effect problem is caused by improper advertising accuracy or improper delivery density, or both; providing a data basis for subsequent advertising supervision and early warning.

[0076] In step S3, when the delivery effect of the target advertisement in the past preset time period is poor, the delivery effect coefficient EC, delivery accuracy QR and delivery density MR of the target advertisement in the past preset time period are obtained;

[0077] According to the calculation formula Obtain the target advertisement delivery warning value WV in the past preset time period; where EC0, QR0 and MR0 represent delivery effect value threshold, delivery accuracy standard value and delivery density standard value respectively; the values ​​of QR0 and MR0 are obtained by relevant technical personnel based on a large number of previous advertisement delivery analyses, and are generally specific values, which will not be elaborated on here;

[0078] It can be understood that the larger the delivery effect coefficient, the smaller the delivery precision ratio, and the smaller the delivery density, the larger the delivery warning value, indicating that the overall delivery effect is better and no warning is needed; otherwise, a warning is needed;

[0079] When the delivery effect is poor, it is necessary to further analyze the external performance of the delivery strategy. This can be done by combining delivery accuracy and delivery density. If the result of the comprehensive analysis still requires an early warning, it means that the current advertising delivery strategy requires manual intervention.

[0080] Compare the calculated delivery warning value with the preset comprehensive delivery efficiency threshold to determine whether the target advertisement is normal in the current delivery strategy;

[0081] If the calculated delivery warning value is greater than or equal to the preset comprehensive delivery efficiency threshold, a normal signal is generated and sent to the hosting backend; after the hosting backend receives the normal signal, the relevant personnel do not need to pay attention to it;

[0082] If the calculated delivery warning value is less than the preset comprehensive delivery efficiency threshold, a warning signal is generated and sent to the hosting backend. After the hosting backend receives the warning signal, relevant personnel need to handle it in a timely manner, analyze the problems with the current delivery strategy, and handle it accordingly.

[0083] In the present invention, the delivery effect coefficient is first calculated by the delivery conversion rate and user dissatisfaction, which can help to preliminarily judge whether the effect of advertising delivery meets the expectations; if the delivery effect coefficient indicates that the advertising delivery effect is poor, the delivery accuracy and delivery density of the advertisement will be further analyzed, which can avoid unnecessary in-depth analysis when the advertising effect itself is very good, thereby saving resources; the delivery effect coefficient, delivery accuracy and delivery density are combined to calculate the delivery warning value, which can comprehensively evaluate the effect of advertising delivery from multiple dimensions; the delivery effect coefficient provides an overall effect evaluation, while delivery accuracy and delivery density are supplemented by analysis from the perspective of audience matching and display frequency; the delivery warning value integrates the influence of these three into one indicator, which helps to reveal problems in advertising effectiveness and provide accurate warning signals, so that bad delivery strategies can be improved in a timely manner during the process of advertising delivery supervision, making the intelligent hosting of target advertisements more efficient.

[0084] The above formulas are all calculated by removing dimensions and taking numerical values. The formula is a formula that is closest to the actual situation obtained by collecting a large amount of data and performing software simulation. The preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained by simulating a large amount of data.

[0085] In the embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the modules is only a logical function division, and there may be other division methods in actual implementation; the modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the method of this embodiment.

[0086] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. An advertising intelligent hosting method based on big data, characterized by: include: Step S1: obtaining user behavior data of target advertisements delivered in the past preset time period, obtaining the delivery conversion rate and user dissatisfaction of the target advertisements in the past preset time period based on the user behavior data analysis; and evaluating the delivery effect of the target advertisements in the past preset time period based on the delivery conversion rate and user dissatisfaction; Step S2: obtaining user characteristic data and delivery frequency data of the target advertisement in the past preset time period, obtaining the delivery accuracy of the target advertisement in the past preset time period based on the user characteristic data analysis, and obtaining the delivery density of the target advertisement in the past preset time period based on the delivery frequency data analysis; Step S3: When the delivery effect of the target advertisement in the past preset time period is poor, a comprehensive analysis is performed based on the delivery effect, delivery accuracy and delivery density of the target advertisement in the past preset time period to issue an early warning for the current intelligent hosting delivery strategy.

2. The method for intelligent advertising hosting based on big data according to claim 1, characterized in that: The method of obtaining the delivery conversion rate and user dissatisfaction of the target advertisement in the past preset time period based on the user behavior data analysis includes: Traverse the user behavior data of the target advertisements delivered in the past preset time period, and obtain the browsing behavior data, conversion behavior data and user feedback data of the target advertisements in the past preset time period; wherein the browsing behavior data includes the total number of users who have browsed the target advertisements in the past preset time period and the total time each user browses the target advertisements, which are marked as I and Ti respectively; i represents the user number, i=1, 2...I; the conversion behavior data includes the number of users participating in different conversion event types Ej in the past preset time period, wherein j represents the number of conversion event types, j=1, 2...J; J represents the total number of conversion event types of the target advertisements; the user feedback data includes the total number of users who have negative feedback on the target advertisements in the past preset time period, the number of negative feedbacks by each user who has negative feedback, and the total time each user who has negative feedback browses the target advertisements, and are marked as G, Ug and Tg respectively; g=1, 2...G; g represents the number of users who have negative feedback on the target advertisements in the past preset time period; Based on the browsing behavior data and conversion behavior data analysis, the target advertisement delivery conversion rate in the past preset time period is obtained. The calculation formula is: In the formula, CR represents the delivery conversion rate; kj represents the preset proportional coefficient of the target advertisement conversion event type; and kj is greater than 0; T0 represents the complete playback time of the target advertisement; Based on the browsing behavior data and user feedback data analysis, the user dissatisfaction of the target advertisement in the past preset time period is obtained. The calculation formula is: In the formula, DR represents user dissatisfaction.

3. The method for intelligent advertising hosting based on big data according to claim 2, characterized in that: The evaluation of the delivery effect of the target advertisement in the past preset time period based on the delivery conversion rate and user dissatisfaction includes: Obtain the delivery conversion rate CR and user dissatisfaction DR of the target advertisement in the past preset time period; According to the calculation formula Calculate and obtain the delivery effect coefficient EC of the target advertisement in the past preset time period; where α and β represent the preset weight factors of delivery conversion rate and user dissatisfaction respectively; and α and β are both greater than 0; The calculated delivery effect coefficient of the target advertisement in the past preset time period is compared with the preset delivery effect coefficient threshold to determine the delivery effect of the target advertisement in the past preset time period; If the calculated delivery effect coefficient of the target advertisement in the past preset time period is greater than or equal to the preset delivery effect coefficient threshold, a signal is generated that the delivery effect of the target advertisement in the past preset time period is good; If the calculated delivery effect coefficient of the target advertisement in the past preset time period is less than the preset delivery effect coefficient threshold, a signal indicating that the delivery effect of the target advertisement in the past preset time period is poor is generated.

4. The method for intelligent advertising hosting based on big data according to claim 1, characterized in that: The user characteristic data of the target advertisement includes the target audience data and the actual audience observation data; among which, the target audience data refers to the user characteristics that match the target audience; the actual audience observation data refers to the user characteristics of each actual user who has viewed the target advertisement; the target advertisement delivery frequency data includes the frequency of displaying the target advertisement to the user group in the past preset time period.

5. The method for intelligent advertising hosting based on big data according to claim 4, characterized in that: The method of obtaining the delivery accuracy of the target advertisement in the past preset time period based on the analysis of user characteristic data includes: Compare the actual audience observation data obtained with the target audience data, and count the number of users who fully meet the target audience data in the past preset time period, marking them as NP; According to the calculation formula Get the delivery accuracy of the target advertisement in the past preset time period; where I represents the total number of users who have viewed the target advertisement in the past preset time period.

6. The method for intelligent advertising hosting based on big data according to claim 4, characterized in that: The method of obtaining the delivery density of the target advertisement in the past preset time period based on delivery frequency data analysis includes: Divide the past preset time period into several time intervals according to different time period types within a day; divide the time intervals belonging to the same time period type into one group; Then, the number of times the target advertisement is delivered in different time intervals in the past preset time period is counted and marked as Cf,h; wherein f represents the time period type, f=1,2...F; F is the total number of time period types; h represents the number of the time interval in the same group, h=1,2...H; H represents the total number of time intervals in the same group; According to the calculation formula Get the delivery density of the target advertisement in the past preset time period; where I represents the total number of users who have viewed the target advertisement in the past preset time period; λf represents the preset proportion coefficient of different time periods; and λf is greater than 0.

7. The method for intelligent advertising hosting based on big data according to claim 1, characterized in that: When the delivery effect of the target advertisement in the past preset time period is poor, the delivery effect coefficient EC, delivery accuracy QR and delivery density MR of the target advertisement in the past preset time period are obtained; According to the calculation formula Obtain the delivery warning value WV of the target advertisement in the past preset time period; where EC0, QR0 and MR0 represent the delivery effect value threshold, delivery accuracy standard value and delivery density standard value respectively; Compare the calculated delivery warning value with the preset comprehensive delivery efficiency threshold to determine whether the target advertisement is normal in the current delivery strategy; If the calculated delivery warning value is greater than or equal to the preset comprehensive delivery efficiency threshold, a normal signal is generated and sent to the hosting backend; After the hosting backend receives a normal signal, the relevant personnel do not need to pay attention to it; If the calculated delivery warning value is less than the preset comprehensive delivery efficiency threshold, a warning signal is generated and sent to the hosting backend; After the hosting backend receives the early warning signal, relevant personnel need to handle it in a timely manner.

Images