An advertisement effect evaluation management method and system based on fused data

Abstract

The application provides an advertisement effect evaluation management method and system based on fusion data, belongs to the technical field of data processing, and specifically comprises the following steps: based on the delivery data of the target strategy delivered advertisement and the coincidence degree of the user portraits between different target strategy delivered advertisements, determining an analysis management method of the advertisement effect of the target strategy delivered advertisement between different platforms, determining the recognition result of the associated user identity between different platforms by using the analysis management method, determining the available analysis groups between the platforms based on the recognition result, and determining the evaluation management method of the to-be-delivered advertisement in different groups based on the platform association degree between the available analysis groups and the platform data corresponding to the available analysis groups, so that the comprehensiveness of the evaluation analysis processing of the advertisement promotion effect is improved.

Description

Technical Field

[0001] This invention belongs to the field of data processing technology, and in particular relates to a method and system for evaluating and managing advertising effectiveness based on fused data. Background Technology

[0002] Traditional advertising effectiveness evaluation often uses "click attribution," attributing conversion success to the advertising channels users encounter. For example, in invention patent application CN202511204676.7, "An Artificial Intelligence-Based Advertising Effectiveness Evaluation Method and System," multi-source data, including user behavior data streams and advertising performance metrics streams, are acquired simultaneously to generate a spatiotemporally synchronized multidimensional data cube. After feature decoupling of the multidimensional data cube, a spatiotemporal distribution map of advertising conversion probability is finally output. Based on this map, a set of adversarial optimization advertising strategies that satisfy Pareto optimality is output, which can improve the accuracy and real-time performance of advertising effectiveness evaluation. However, the above technical solution has the following drawbacks: When performing multi-source analysis, the reliability of the correlation analysis of user identities between different platforms is crucial. Therefore, determining the placement management strategies for different advertisements to be placed, and then conducting correlation analysis of user identities between different platforms, is a technical problem that urgently needs to be solved in order to ensure the reliability of the correlation analysis of user identities for different advertisements and improve the reliability of the advertising analysis effect.

[0003] Therefore, there is an urgent need for an advertising effectiveness evaluation and management method and system based on integrated data. Summary of the Invention

[0004] To achieve the objectives of this invention, the following technical solution is adopted: Specifically, this application provides a method for evaluating and managing advertising effectiveness based on fused data, which includes: S1 uses the user profile of the ads to be delivered to divide the ads to be delivered into different groups, and determines the ads to be delivered using the target strategy based on the ads to be delivered in different groups, and delivers the ads using the target strategy. S2 delivers ads across all platforms based on the target strategy. Based on the ad delivery data and the degree of overlap in user profiles between different target strategy ads, a method for analyzing and managing the ad performance of target strategy ads across different platforms is determined. The analysis and management method is used to determine the identification results of associated user identities between different platforms. Based on the identification results, available analysis groups between the platforms are determined. S3 determines the evaluation and management methods for advertisements to be placed in different groups based on the degree of platform correlation between the available analysis groups and the platform data corresponding to the available analysis groups.

[0005] Furthermore, the advertisements to be delivered are divided into different groups, specifically including: Ads with consistent user profiles are grouped into the same group.

[0006] Furthermore, the method for determining the target strategy for ad delivery is as follows: Based on the group data, determine the number of groups; By utilizing the ads to be delivered in different groups, the number of ads to be delivered in each group is determined; By using the number of groups and the number of ads to be delivered in different groups, the target strategy ads are determined among the ads to be delivered.

[0007] Furthermore, the method for determining the analysis and management method of the advertising effect of the target strategy across different platforms is as follows: Based on the ad delivery data of the target strategy, determine the number of ads delivered under the target strategy; Based on the degree of overlap in user profiles between ads targeting different target strategies, determine the amount of overlap between user profiles of the ads targeting a specific target strategy and ads targeting other target strategies. A method for analyzing and managing the advertising effectiveness of targeted strategy ads across different platforms is used to determine the number of ads delivered using the target strategy and the amount of overlap in user profiles between the targeted strategy ads and other targeted strategy ads.

[0008] Furthermore, the identification result of the associated user identity is determined based on the identification results of the same user belonging to the same user in the same target strategy advertising on two platforms.

[0009] In a second aspect, the present invention provides a computer system comprising: a memory and a processor connected in communication, and a computer program stored in the memory and capable of running on the processor, wherein the processor executes the aforementioned advertising effectiveness evaluation and management method based on fused data when running the computer program.

[0010] The beneficial effects of this invention are as follows: Based on the ads to be delivered in different groups, the ads to be delivered using the target strategy are identified. Specifically, the number of groups and the number of ads to be delivered in different groups are used to determine which ads to be delivered using the target strategy. The more groups there are and the more ads to be delivered in different groups, the greater the need for user correlation analysis between different platforms. That is, to determine whether users on different platforms belong to the same user, thus increasing the need for evaluating and analyzing the effectiveness of advertising by analyzing the browsing time of a single ad. By utilizing the need for user correlation analysis between different platforms, the target strategy ads are determined, thus laying the foundation for analyzing and processing user correlation between different platforms.

[0011] Based on the platform correlation between available analytics groups and the platform data corresponding to those groups, we determine the evaluation and management methods for ads to be placed in different groups. Using the platform data corresponding to the available analytics groups, we determine the reliability of user identity analysis when conducting a full-platform ad performance evaluation and analysis based on the current available analytics groups across different platforms. Furthermore, considering the platform correlation between available analytics groups, we determine whether the number of platforms in different groups suitable for correlation analysis is sufficient in the current state. Based on the reliability of user identity analysis and the sufficiency of the number of platforms suitable for correlation analysis in different groups when conducting a full-platform ad performance evaluation and analysis, we determine the evaluation and management methods for ads to be placed in the groups. This determines which groups are used for full-platform correlation analysis, thus ensuring both the reliability of ad performance evaluation and analysis, and the reliability of user correlation analysis across different platforms.

[0012] Other features and advantages will be set forth in the following description, and the objects and other advantages of the invention are realized and obtained through the structures particularly pointed out in the description and the drawings.

[0013] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0014] The above and other features and advantages of the present invention will become more apparent from a detailed description of exemplary embodiments thereof with reference to the accompanying drawings.

[0015] Figure 1 This is a flowchart of an advertising effectiveness evaluation and management method based on fused data; Figure 2 This is a flowchart illustrating the method for determining the target strategy for ad delivery; Figure 3 This is a flowchart illustrating the methods for determining the analysis and management of advertising effectiveness across different platforms using a targeted strategy. Detailed Implementation

[0016] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this specification, and not all embodiments. Based on the embodiments of this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this specification.

[0017] Example 1 like Figure 1 As shown, this application provides a method for evaluating and managing advertising effectiveness based on fused data, specifically including: S1 uses the user profile of the ads to be delivered to divide the ads to be delivered into different groups, and determines the ads to be delivered using the target strategy based on the ads to be delivered in different groups, and delivers the ads using the target strategy. The user profile refers to the user characteristic description corresponding to the target audience of the advertisement to be placed, including comprehensive profile data of users' age group, consumption preferences, interest tags, geographical location, etc.; the advertisement to be placed refers to the advertising materials that the enterprise plans to place on various platforms and their corresponding targeting parameters; the group refers to the set formed by merging and classifying the advertisements to be placed according to the principle of user profile consistency, that is, advertisements to be placed with consistent user profiles are divided into the same group; the targeted strategy advertisement refers to the advertisement selected from the advertisements to be placed in each group to undertake the task of inter-platform user correlation analysis based on the comprehensive evaluation result of the number of groups and the number of advertisements.

[0018] Suppose a company plans to launch a batch of targeted ads, each corresponding to a different user profile combination. Ads with completely identical user profiles are grouped into the same group, forming multiple groups. The number of ads in each group is counted. Based on the number of groups and the number of ads in different groups, the need for cross-platform user association analysis is assessed, and the scope of the target strategy for ad placement is determined.

[0019] This step establishes a technical path from user profile classification to determining the target strategy for ad delivery. Its significance lies in the centralized processing of ads to be delivered to users with consistent user profiles through group management, enabling subsequent inter-platform user association analysis to be carried out within user groups with consistent profiles, improving the accuracy of user association identification, and laying the foundation for reliable evaluation of advertising effectiveness.

[0020] Furthermore, the advertisements to be delivered are divided into different groups, specifically including: Ads with consistent user profiles are grouped into the same group.

[0021] Furthermore, such as Figure 2 As shown, the method for determining the target strategy for ad delivery is as follows: In this embodiment, the number of groups and the number of ads to be delivered in different groups are used to determine which ads to be delivered as the target strategy. The more groups there are and the more ads to be delivered in different groups, the greater the need for user correlation analysis between different platforms. That is, to determine whether users on different platforms are the same user. This increases the need for evaluating and analyzing the effectiveness of ads by analyzing the browsing time of a user on a single ad. By utilizing the need for user correlation analysis between different platforms, the target strategy ads are determined among the ads to be delivered, thus laying the foundation for analyzing and processing user correlation between different platforms.

[0022] S11 determines the number of groups based on the group data.

[0023] The group data refers to the composition information of each group after the advertisements are classified according to the principle of user profile consistency, including the list of advertisements to be delivered and the corresponding user profile information contained in each group; the number of groups refers to the total number of advertisement groups with different user profiles after the division, which is used to measure the degree of diversity of user profiles covered by the current advertisements to be delivered.

[0024] Suppose that a batch of ads to be placed are classified according to the principle of user profile consistency, forming multiple groups. The total number of all groups is counted and compared with a preset group number threshold to determine the degree of diversity of user profiles and the overall demand intensity for conducting inter-platform user association analysis.

[0025] This step, by counting the number of groups, is significant in that it quantifies the diversity of user profiles covered by the ads to be placed into a comparable value, providing a basis for branch judgments in the subsequent target strategy ad placement logic, and ensuring that a wider range of analysis and coverage strategies can be adopted when there are many types of user profiles.

[0026] S12 uses the ads to be delivered in different groups to determine the number of ads to be delivered in each group; The number of ads to be placed in the group refers to the number of ads to be placed that have been merged into each group, which is used to reflect the concentration of ad placement under each user profile; the preset ad placement number threshold is a critical value used to determine whether the number of ads in a certain group is sufficient to support user association analysis between platforms; the placement analysis demand weight value is a weight index determined based on the number of ads to be placed in each group. The more ads to be placed in a group, the greater the placement analysis demand weight value of that group.

[0027] Assuming the composition information of each group is determined, the number of ads to be placed in each group is counted to form a dataset of ad quantity for each group. This dataset is used to calculate the demand weight value for ad placement analysis and to determine whether each group meets the preset conditions.

[0028] This step, by statistically analyzing the number of ads in each group, aims to quantify the concentration of ad placements under different user profiles. This provides a data foundation for subsequent analysis of ad placement needs based on the number of ads in each group, as well as for determining the selection criteria for ad placements under the target strategy. It ensures that groups with a larger number of ads receive priority consideration in the determination of ad placements under the target strategy.

[0029] S13 uses the number of groups and the number of ads to be delivered in different groups to determine the target strategy ads among the ads to be delivered.

[0030] The preset group number threshold refers to a critical value used to determine whether the number of groups is too large and whether the demand for user association analysis between platforms is high; the preset condition refers to a group whose number of ads to be placed is greater than the preset ad placement number threshold; the second preset condition refers to a group whose number of ads to be placed is greater than the preset ad placement number value, wherein the preset ad placement number value is greater than the preset ad placement number threshold, that is, the second preset condition is more stringent than the preset condition; the weight sum refers to the sum of the weight values ​​of the ad placement analysis demand of different groups.

[0031] Assume the number of groups and the number of ads in each group have been counted. Set a preset threshold for the number of groups. Compare the total number of groups with this threshold. Based on the comparison results, proceed to different decision branches. In each branch, further combine preset conditions or a second preset condition to determine which ad to randomly select from each group that meets the conditions as the target ad for the strategy.

[0032] This step, through a comprehensive judgment of the number of groups and the number of advertisements, is significant in that it dynamically determines the screening scope of the target strategy advertisements based on the current distribution of advertisements to be placed. When the demand for inter-platform user correlation analysis is high, a wider range of screening conditions is adopted, while when the demand is relatively low, the screening criteria are flexibly adjusted according to the weight distribution. This ensures that the target strategy advertisements cover the main user profiles sufficiently, while avoiding unnecessary consumption of analysis resources.

[0033] Understandably, if the number of groups exceeds a preset group number threshold, then in order to evaluate and analyze the advertising effect, it is necessary to determine the correlation between the click data of the same user on different platforms for the same advertisement. Therefore, the target strategy advertisement in the advertisement to be delivered is to randomly select one advertisement from the groups that meet the preset conditions as the target strategy advertisement. This ensures that on which platforms user identity analysis can be reliably performed, and also ensures that the user profiles of the groups with a large number of advertisements to be delivered can be reliably analyzed on which platforms.

[0034] The groups that meet the preset conditions refer to groups where the number of ads to be placed is greater than the preset threshold for the number of ads to be placed. When the number of groups exceeds the preset threshold for the number of groups, it indicates that the user profiles covered by the ads to be placed are more diverse and the overall demand for user association analysis between platforms is higher. In this case, the target strategy is selected to place ads only from the groups with a large number of ads to ensure sufficient coverage of user profiles.

[0035] Assuming the number of groups is greater than a preset group number threshold, select groups from all groups whose number of ads to be delivered is greater than the preset ad delivery threshold as groups that meet the preset conditions. Randomly select one ad to be delivered from each group that meets the preset conditions as the target strategy ad delivery ad, forming the target strategy ad delivery set.

[0036] This approach, through a screening mechanism based on preset conditions, is significant because when user profiles are highly diverse, it ensures that the targeted advertising strategy covers the main user profile groups with a large number of ads, providing representative analytical samples for subsequent inter-platform user association analysis and improving the reliability of association analysis results under different user profiles.

[0037] Additionally, it is understood that if the number of groups is not greater than a preset group number threshold, the following is included: The weight value of the advertising demand for each group is determined by the number of ads to be delivered in each group, i.e., by multiplying the user's preset scaling factor by the number of ads to be delivered. It should be noted that the more ads to be delivered in a group, the greater the weight value of the advertising demand for that group.

[0038] The weight value of the ad placement analysis demand is determined based on the number of ads to be placed in the group. The more ads there are, the greater the weight value, which is used to quantify the priority of the group in the process of selecting ads for the target strategy. The preset weight threshold is a critical value used to judge whether the average level of the ad placement analysis demand weight value is high. The preset threshold is a critical value used to judge whether the weight has reached the threshold for using a wider range of screening conditions.

[0039] Assuming the number of groups is no greater than the preset group number threshold, calculate the ad placement analysis demand weight value for each group, and calculate the average and sum of the ad placement analysis demand weight values ​​for all groups. Based on the comparison results of the average value and the preset weight threshold, as well as the comparison results of the weight sum and the preset threshold, determine whether to use the preset condition or the second preset condition to screen the target strategy ad placement.

[0040] This situation involves a dual judgment based on the average value and the sum of weights of the demand analysis values. Its significance lies in that when the number of groups does not exceed the preset group number threshold, the strength of the overall analysis demand is further evaluated through the weight distribution characteristics. When the overall demand is strong, a preset condition with a wider coverage is adopted, and when the overall demand is relatively weak, a more stringent second preset condition is adopted. This achieves dynamic adjustment of the target strategy's ad selection criteria and fine-tunes the analysis demand of the current ad placement layout.

[0041] Case 1: If the average weight value of the demand analysis of different groups is greater than the preset weight threshold, then the target strategy advertisement in the advertisement to be delivered will be a randomly selected advertisement from the groups that meet the preset conditions as the target strategy advertisement.

[0042] The average value of the demand weight value for ad placement analysis refers to the arithmetic mean of the demand weight values ​​for ad placement analysis of each group. It is used to measure the overall level of the number of ads and the intensity of the demand analysis for all groups. When the average value of the demand weight value for ad placement analysis is greater than the preset weight threshold, it indicates that the overall demand analysis is strong and the preset conditions should be used to select ads for the target strategy.

[0043] Assuming the average weight value of the ad placement analysis demand for all groups is calculated, and a preset weight threshold is set to a certain value, if the average value is greater than the preset weight threshold, it indicates that the number of ads to be placed in the current group is relatively large, and there is a high demand for inter-platform user correlation analysis. Randomly select one ad to be placed from each of the groups that meet the preset conditions (groups with more ads than the preset ad placement threshold) as the target strategy ad placement ad.

[0044] This situation, determined by the average value of the demand weight in the ad delivery analysis, is significant because when the overall number of ads in each group is high, a wider range of preset conditions are used to select ads for the target strategy, ensuring that the ads for the target strategy cover the main user profile sufficiently. The same filtering mechanism is used when the number of groups exceeds the preset group number threshold, achieving consistency of the strategy in both scenarios.

[0045] Scenario 2: If the average weight value of the ad delivery analysis demand of different groups is not greater than the preset weight threshold, the weight sum is determined by the sum of the weight values ​​of the ad delivery analysis demand of different groups. It is then determined whether the weight sum is greater than the preset threshold. If it is, the target strategy ad in the ad to be delivered is a randomly selected ad in the group that meets the preset conditions. If not, the target strategy ad in the ad to be delivered is determined to be a randomly selected ad in the group that meets the second preset condition.

[0046] The weighted sum refers to the sum of the weight values ​​of the ad placement analysis needs of all groups. It is used to assess the overall analysis needs from the total volume dimension when the average weight value of each group is not high. The second preset condition is a group whose number of ads to be placed is greater than the preset value of the number of ads to be placed. The preset value of the number of ads to be placed is greater than the preset threshold of the number of ads to be placed. That is, the second preset condition is more stringent than the preset condition. When the weighted sum is not greater than the preset threshold, it indicates that the overall analysis needs are weak. The more stringent second preset condition should be adopted, and the target strategy should be selected to place ads only from the group with the largest number of ads.

[0047] Assuming that the average value of the demand weights for each group is not greater than the preset weight threshold, the weight sum is further calculated. Let the preset threshold be a certain value. If the weight sum is greater than the preset threshold, the target strategy will be randomly selected from the groups that meet the conditions and the ads will be delivered. If the weight sum is not greater than the preset threshold, the second preset condition (groups with more ads than the preset value of the number of ads to be delivered) will be used, and only one ad to be delivered will be randomly selected from each group with more ads as the target strategy ad.

[0048] This situation, supplemented by weighted sums, is significant because when the average and total weight values ​​of each group are low, more stringent screening conditions are adopted to focus on the core group with the most ads. This reduces unnecessary consumption of analysis resources while ensuring that the targeted strategy's ad placements are concentrated on the most representative user profile group, providing higher-quality analytical samples for subsequent inter-platform user association analysis.

[0049] This embodiment achieves intelligent determination of target strategy advertising through comprehensive judgment in steps S11 to S13 and various other scenarios. Its core value is reflected in three aspects: First, by dividing users into groups based on the principle of user profile consistency, it centrally manages ads to be delivered that have consistent user profiles, providing a unified analysis unit for subsequent correlation analysis. Second, through multi-dimensional evaluation of the number of groups and the weight value of the demand for delivery analysis, it dynamically selects the screening criteria for target strategy advertising based on the distribution characteristics of the current advertising delivery layout. Third, by flexibly switching between preset conditions and second preset conditions, it expands the coverage of target strategy advertising when the demand for analysis is strong, and focuses on core groups when the demand is weak, achieving a dynamic balance between the allocation of analysis resources and the sufficiency of analysis coverage.

[0050] S2 delivers ads across all platforms based on the target strategy. Based on the ad delivery data and the degree of overlap in user profiles between different target strategy ads, a method for analyzing and managing the ad performance of target strategy ads across different platforms is determined. The analysis and management method is used to determine the identification results of associated user identities between different platforms. Based on the identification results, available analysis groups between the platforms are determined. The delivery data refers to the data generated after the targeted strategy ads are actually delivered on various platforms, including ad impressions, clicks, and conversions; the overlap of user profiles refers to the degree of cross-over between user profiles targeted by different targeted strategy ads, with higher overlap indicating more similar user profiles targeted by the two targeted strategy ads; the analysis and management method refers to the specific method for conducting user identity association analysis between different platforms under what conditions; the identification result of the associated user identity is determined based on the identification results of belonging to the same user in the same targeted strategy ad on two platforms; the available analysis group refers to the set of groups that can support reliable user identity association analysis processing among a specific combination of platforms.

[0051] This step establishes a complete technical path from the data of targeted advertising to the identification of available analytical groups. Its significance lies in determining the appropriate inter-platform user association analysis strategies for each targeted advertising strategy by assessing the difficulty of analysis and the degree of overlap in user profiles. Then, by assessing the association identification results and the sufficiency of the number of users identified, it determines which groups can conduct reliable user association analysis between specific platform combinations, providing reliable group foundation data for the final determination of advertising effectiveness evaluation and management methods.

[0052] Furthermore, such as Figure 3 As shown, the method for determining the analysis and management method of the advertising effect of the target strategy across different platforms is as follows: In this embodiment, the difficulty of analysis and processing under the current number of targeted strategy advertisements is determined by utilizing the advertising data of the targeted strategy. Furthermore, the degree of overlap in user profiles between different targeted strategy advertisements is combined to determine the reliability of user association analysis and processing on multiple platforms under different user profiles. Based on the reliability of association analysis and processing under different user profiles and the difficulty of analysis and processing, a method for analyzing and managing the advertising effect of different targeted strategy advertisements on different platforms is determined. This not only reduces the difficulty of data analysis and processing but also ensures the reliability of user association analysis and processing on multiple platforms for different user profiles.

[0053] S21 determines the number of ads to be placed under the target strategy based on the ad placement data of the target strategy.

[0054] The number of ads targeted by the strategy refers to the total number of ads selected and placed on various platforms under the target strategy, used to assess the scale and complexity of the current analysis task; the preset ad placement threshold is a critical value used to determine whether the number of ads targeted by the strategy is too small and whether the analysis and processing difficulty is too low.

[0055] Assuming that all targeted strategy ads have been deployed on all platforms, count the total number of targeted strategy ads and compare this number with a preset ad deployment threshold. If the number is small, the analysis task is small, and a full analysis of all targeted strategy ads can be performed directly. If the number is large, the degree of overlap in user profiles needs to be further considered to determine a differentiated analysis and management method.

[0056] This step involves counting the number of ads placed under the target strategy. Its significance lies in quantifying the scale of the analysis task into a measurable numerical indicator. When the number is small, the most comprehensive analysis strategy is adopted directly to avoid omissions. When the number is large, a differentiated strategy is determined through subsequent user profile overlap analysis, thus ensuring the quality of the analysis while reasonably controlling the complexity of the analysis process.

[0057] In the above steps, the number of ads placed under the target strategy is obtained. If the number of ads placed under the target strategy is less than the preset ad placement threshold, then the method for analyzing and managing the advertising effect of all ads placed under the target strategy across different platforms is to determine whether the same user exists on different platforms as long as click data exists on any platform.

[0058] The analysis and management method is a strategy of "conducting correlation analysis whenever click data exists on any platform". This is the most comprehensive analysis strategy, which means that the same user identity is identified across platforms for every user with click records. When the number of ads placed by the target strategy is small, the analysis task is small in scale and has the conditions to perform correlation analysis on all scenarios with click records. Therefore, this full-scale analysis strategy is directly adopted.

[0059] Assuming the total number of ads delivered under the target strategy is less than the preset ad delivery threshold, it indicates that the scale of this analysis task is small. The unified analysis and management method for all ads delivered under the target strategy is as follows: as long as click data of the ad delivered under the target strategy is detected on any platform, the user identity association identification process between that platform and other platforms is initiated to determine whether the clicking user is the same user.

[0060] The significance of directly adopting a full-scale analysis strategy in this case is that when the analysis task is small in scale, it ensures the comprehensiveness of user association identification by maximizing the analysis coverage, avoiding the omission of effective user association data due to strategy limitations, and accumulating sufficient association identification sample data for the subsequent determination of usable analysis groups.

[0061] Additionally, it can be understood that if the number of ads delivered by the target strategy is not less than the preset ad delivery threshold, then proceed to step S22.

[0062] Assuming the total number of ads delivered under the target strategy is not less than the preset ad delivery threshold, it indicates that the analysis task is large in scale. It is necessary to further determine differentiated analysis and management methods for ads delivered under different target strategies based on the degree of overlap in user profiles between ads delivered under different target strategies, so as to ensure the quality of analysis while avoiding excessive data processing pressure caused by full-scale analysis.

[0063] This situation is addressed by performing differentiated analysis in S22. The significance of this approach is that when the analysis task is large-scale, by assessing the degree of overlap in user profiles, it can identify target strategy ads that have reliable analysis conditions under different profile dimensions. This avoids the risk of forcing full correlation analysis on ads with insufficient analysis conditions, which could lead to unreliable analysis results.

[0064] S22 determines the degree of overlap between the user profiles of the targeted strategy ads and other targeted strategy ads based on the degree of overlap between the user profiles of the ads delivered under different target strategies.

[0065] The overlap of user profiles refers to the number of profile tags that overlap between ads targeting a certain strategy and ads targeting other strategies in terms of user profile dimension. The greater the overlap, the higher the similarity between the ads targeting a certain strategy and other ads in terms of user profile. The number of ads targeting different user profiles refers to the number of ads targeting different strategies that have an overlapping relationship under each user profile (i.e., the user profile corresponding to each group). The preset ad quantity threshold is a critical value used to determine whether the number of ads targeting different strategies that have an overlapping relationship under a certain user profile is sufficient. The reliable analysis strategy refers to a full-scale analysis strategy that "conducts correlation analysis as long as click data exists on any platform".

[0066] Assuming the user profile information for each target strategy's ads is already determined, the user profiles of each target strategy's ads are compared with those of other target strategy's ads one by one. The number of overlaps between user profiles of each target strategy's ads and other ads is counted. Furthermore, the number of target strategy ads with overlapping relationships under each user profile is counted and compared with a preset ad quantity threshold to determine whether the number of overlaps under each user profile is sufficient to support reliable correlation analysis.

[0067] This step, through the statistical analysis of the number of overlapping user profiles and the analysis of the number of advertisements under different user profiles, is significant in that it assesses the supporting conditions for correlation analysis of the user profile environment in which advertisements for each target strategy are placed. The greater the number of overlaps, the richer the advertisement samples participating in the correlation analysis under that profile, and the higher the reliability of the correlation analysis, thus providing a quantitative basis for determining the subsequent differentiated analysis management methods.

[0068] Furthermore, the above steps include the following: S221 determines the number of targeted strategy ads under different user profiles by the number of overlaps between the target strategy ads and other targeted strategy ads. It then determines whether the average number of targeted strategy ads under different user profiles is less than a preset ad quantity threshold. If so, it determines that the ad performance analysis and management method for all targeted strategy ads across different platforms is that as long as click data exists on any platform, it determines whether the same clicking user exists on different platforms. If not, it proceeds to step S222.

[0069] The average number of targeted strategy ads placed under different user profiles refers to the arithmetic mean of the number of targeted strategy ads with overlapping relationships under each user profile. It is used to measure the overall sufficiency of the analysis support conditions under each user profile. When the average number is less than the preset ad quantity threshold, it indicates that there are relatively few ad samples with overlapping relationships under each user profile, and the sample sufficiency of the correlation analysis is insufficient. It is necessary to uniformly adopt the most comprehensive click data trigger analysis strategy for all targeted strategy ads.

[0070] Suppose we calculate the number of ads placed under the target strategy that have overlapping relationships under each user profile and calculate their average value. Let the preset ad quantity threshold be a certain value. If the average value is less than the preset ad quantity threshold, it indicates that the degree of overlap between ads placed under the target strategy is generally low under each user profile dimension. The unified analysis and management method for all ads placed under the target strategy is to start the correlation analysis as long as there is click data on any platform. If the average value is not less than the preset ad quantity threshold, then proceed to S222 for more refined overlap quantity judgment.

[0071] This step, by judging the average number of ads under different user profiles, is significant because when the overall correlation analysis support conditions of each user profile dimension are insufficient, a unified full-volume analysis strategy is adopted to maximize analysis coverage. This ensures that sufficient correlation identification data can still be accumulated even with a small sample size, and avoids a serious lack of correlation analysis data for some user profiles due to differentiated strategies.

[0072] S222 Determine whether the maximum number of overlaps between the user profiles of the target strategy ad and other target strategy ads is greater than a preset overlap threshold. If so, determine that the method for analyzing and managing the ad performance of the target strategy ad across different platforms is a reliable analysis strategy. That is, as long as there is click data on any platform, it is determined whether the same click user exists on different platforms. If not, proceed to step S23. The maximum number of overlapping user profiles refers to the largest number of overlapping user profiles between a target strategy ad and all other target strategy ads; the preset overlap threshold is a critical value used to determine whether the maximum number of overlapping user profiles between a target strategy ad and other ads reaches the condition for using a reliable analysis strategy; the reliable analysis strategy refers to a strategy that uses all click data to trigger correlation analysis for target strategy ads with the highest number of overlapping user profiles, ensuring the reliability of the correlation analysis.

[0073] Assuming the number of user profile overlaps between a target strategy ad and all other target strategy ads has been counted, the maximum value is taken. A preset overlap threshold is set. If the maximum value is greater than the preset overlap threshold, it indicates that the target strategy ad has a high degree of similarity to at least one other ad in terms of user profile, and the correlation analysis sample at that profile level is sufficient, thus confirming the analysis management method as a reliable analysis strategy. If the maximum value is not greater than the preset overlap threshold, proceed to S23 for comprehensive judgment.

[0074] This step, by determining the maximum number of overlaps, is significant because when the target strategy's advertising has a sufficiently high degree of similarity to the maximum user profile of other advertisements, a reliable analysis strategy can be directly determined, avoiding unnecessary comprehensive calculations and improving the efficiency of determining analysis and management methods. It also ensures that a full analysis strategy can be adopted in a timely manner when the supporting conditions for correlation analysis are sufficient.

[0075] S23 uses the number of ads delivered by the target strategy and the number of overlaps in user profiles between the ads delivered by the target strategy and other ads delivered by the target strategy to determine the analysis and management method of the advertising effect of the ads delivered by the target strategy across different platforms.

[0076] The "analysis deviation user profile" refers to the user profile corresponding to the target strategy ads that were not identified as reliable analysis strategies in the target strategy ad placements under different user profiles; the "preset deviation user profile proportion threshold" refers to the critical value used to judge whether the number of analysis deviation user profiles accounts for a high proportion of the total number of user profiles; the "other analysis strategies" refer to the strategy of "determining whether the converted user exists in different platforms as long as conversion data exists in any platform". Compared with the reliable analysis strategy, the condition for triggering correlation analysis has been upgraded from click data to conversion data, and the analysis triggering conditions are more stringent.

[0077] Assuming that, based on the judgment in S222, the maximum number of overlapping ads for some target strategies does not exceed the preset overlap threshold, further statistics are compiled on the target strategy ads that have been identified as reliable analysis strategies under different user profiles. The user profiles corresponding to the remaining ads that have not been identified as reliable analysis strategies are marked as analysis deviation user profiles. The proportion of analysis deviation user profiles to all user profiles is calculated and compared with the preset deviation user profile proportion threshold. Based on the comparison results, the final analysis and management method for the target strategy ads is determined.

[0078] This step, by analyzing the proportion of user profiles with discrepancies, is significant in that it allows for targeted advertising based on a strategy where the maximum number of overlaps does not exceed a preset threshold. By analyzing the overall degree of discrepancy, it determines whether a reliable analysis strategy should still be used. When the proportion of user profiles with discrepancies is high, it indicates a significant gap in the overall analysis conditions, and a reliable analysis strategy should be uniformly adopted to avoid excessive analysis omissions. When the proportion is low, other analysis strategies should be used to appropriately lower the trigger threshold, ensuring the quality of core analysis data while reducing unnecessary analysis processing.

[0079] It should be noted that in the above steps, based on the target strategy advertising with reliable analysis strategies under different user profiles, the user profiles of target strategy advertising without reliable analysis strategies are used as analysis deviation user profiles. It is determined whether the proportion of analysis deviation user profiles in the user profiles of the target strategy advertising is greater than a preset deviation user profile proportion threshold. If so, the method for analyzing and managing the advertising effect of the target strategy advertising across different platforms is determined to be a reliable analysis strategy. That is, as long as click data exists on any platform, it is determined whether the same clicking user exists on different platforms. If not, the method for analyzing and managing the advertising effect of the target strategy advertising across different platforms is determined to be another analysis strategy. That is, as long as conversion data exists on any platform, it is determined whether the same converting user exists on different platforms.

[0080] The percentage of user profiles with analytical bias refers to the proportion of user profiles with analytical bias to the total number of user profiles involved in the target strategy's advertising. A higher percentage indicates that more user profiles cannot be analyzed using a reliable analysis strategy. When this percentage is greater than a preset threshold for the percentage of user profiles with analytical bias, the overall degree of analytical bias is high, and a reliable analysis strategy is uniformly adopted to ensure comprehensive analysis coverage. When this percentage is not greater than the preset threshold for the percentage of user profiles with analytical bias, the degree of analytical bias is low, and other analysis strategies are adopted for target strategy advertising for which an analysis strategy has not yet been determined. Conversion data is used as a trigger condition to reduce the pressure of full analysis of click data while ensuring the quality of core conversion data analysis.

[0081] Assuming that, based on the judgment in S222, some target strategy ads have been identified as reliable analysis strategies, the remaining target strategy ads proceed to S23. The proportion of deviation user profiles to the total number of user profiles is statistically analyzed. Let a preset threshold for the proportion of deviation user profiles be a certain value. If this proportion is greater than the preset threshold, then reliable analysis strategies are uniformly applied to target strategy ads for which analysis strategies have not yet been determined; if this proportion is not greater than the preset threshold, then other analysis strategies are applied to these ads, with the appearance of conversion data serving as the condition for triggering correlation analysis.

[0082] This situation, judged by analyzing the proportion of user profile deviations, is significant because when the problem of deviation is relatively common, a reliable analysis strategy is uniformly adopted to ensure the comprehensiveness of user association identification as a whole; when the problem of deviation is limited, other analysis strategies are adopted to lower the trigger threshold, so as to reasonably control the overall complexity of data analysis and processing while ensuring the quality of association analysis in core scenarios.

[0083] Furthermore, the identification result of the associated user identity is determined based on the identification results of the same user belonging to the same user in the same target strategy advertising on two platforms.

[0084] Furthermore, the method for determining the available analysis groups among the platforms is as follows: In this embodiment, the identification results of associated user identities between different platforms in the platform combination are used to determine the sufficiency of the number of associated users identified between platforms under different user profiles. The sufficiency of the number of associated users is used to determine the reliability of the association analysis and processing of user identities between different platforms in the platform combination. The reliability of the association analysis and processing of user identities between different platforms is used to determine the available analysis groups in the group, thereby ensuring the reliability of the evaluation and analysis processing of the advertising effect when the above group is processed on multiple platforms in the above platform combination.

[0085] S31 combines platforms in pairs to form a platform combination. Using the identification results of the associated user identities between different platforms in the platform combination, it determines the users who belong to the same user across different platforms in the target strategy's advertising, and uses them as associated identification users. The platform pairing refers to pairing all advertising platforms in pairs to form a platform combination, with each platform combination containing two different advertising platforms; the identification result of the associated user identity is determined based on the identification result of being identified as belonging to the same user in the same target strategy advertising on both platforms; the associated identified user refers to a user who, after association analysis, is confirmed to have behavioral records on both platforms and belongs to the same natural person.

[0086] Assuming there are multiple advertising platforms, these platforms are paired to form multiple platform combinations. For each platform combination, using established analysis and management methods (reliable analysis strategies or other analysis strategies), related users belonging to the same user on both platforms are identified one by one during the advertising campaigns targeting the same objective strategy. The number of related users in each platform combination is then counted.

[0087] This step, through pairwise platform combinations and the identification of associated users, is significant in that it breaks down the analysis of user associations between platforms from a platform-wide perspective to a platform combination perspective, and evaluates the identification of user associations between each pair of platforms separately, providing a detailed data foundation for subsequent assessment of the sufficiency of the number of associated users based on user profile dimensions.

[0088] S32 uses the associated users identified in the ads delivered by different target strategies and the user profiles of the ads delivered by the target strategies to determine the number of associated users under different user profiles. The number of associated users under different user profiles refers to the total number of users identified as associated under each user profile (i.e., the user profile corresponding to each group) dimension, which is used to measure the sufficiency of the sample for user identity association analysis between platforms under a specific user profile dimension; the user profile of the target strategy ad is consistent with the user profile of the group to which it belongs, so the user profile can be traced through the group to which the target strategy ad belongs.

[0089] Assuming that the associated users under each platform combination have been counted, we will associate each associated user with the user profile of the group to which the target strategy advertising belongs, and count the total number of associated users under each user profile dimension to form the distribution data of the number of associated users under different user profiles for each platform combination.

[0090] This step involves statistically analyzing the number of users identified under different user profiles. Its significance lies in refining the number of identified users by user profile dimension, assessing the sufficiency of the inter-platform correlation analysis samples under a specific user profile, and providing quantitative data basis for subsequent screening of available analysis groups based on user profile dimension. This ensures that the determination of available analysis groups can accurately reflect the reliability of the analysis of user profiles of each group under a specific platform combination.

[0091] S33, based on the number of users associated with different user profiles, determine the available analysis groups among the platforms corresponding to the platform combination.

[0092] The available analysis group refers to a group that, after evaluation, is confirmed to have reliable user identity association analysis conditions under a specific platform combination; the preset associated user number threshold refers to a critical value used to determine whether the total number of associated users identified under all user profiles for a certain platform combination is sufficient; the reliable analysis profile refers to a user profile under a specific platform combination whose number of associated identified users is greater than the preset number of identified users, indicating that the sample sufficiency of inter-platform association analysis under that user profile has reached a reliable level; the preset identified user number refers to a critical value used to determine whether the number of associated identified users under a certain user profile is sufficient; the preset reliable analysis profile refers to a critical number used to determine whether the number of reliable analysis profiles fully covers each user profile; the preset reliable profile proportion threshold refers to a critical value used to determine whether the proportion of reliable analysis profiles in a certain group of user profiles meets the conditions for including the group in the available analysis group.

[0093] Assuming that the number of users associated with each platform combination under different user profiles has been counted, and setting a certain threshold for the number of associated users, a certain number of identified users, a certain number of reliable analysis profiles, and a certain threshold for the proportion of reliable profiles, the available analysis groups corresponding to each platform combination are determined according to the judgment logic of Case 1, Case 2, and Case 3 in sequence.

[0094] This step involves a hierarchical assessment of three scenarios. Its significance lies in gradually filtering out usable analysis groups that meet the conditions for reliable correlation analysis under a specific platform combination, based on the sufficiency of the number of users identified through association, from three dimensions: overall sufficiency, number of reliable analysis profiles, and sufficiency of group profile coverage. This ensures that the determination of subsequent advertising effectiveness evaluation and management methods can be based on reliable correlation analysis.

[0095] It should be noted that, based on the number of users identified through association under different user profiles, the available analysis groups among the platforms corresponding to the platform combination are determined, specifically including: Scenario 1: If the sum of the number of associated users under different user profiles among the platforms corresponding to the platform combination is greater than the preset threshold for the number of associated users, it indicates that identity association processing can be effectively performed between the platforms. Therefore, the available analysis group among the platforms corresponding to the platform combination is determined to be all groups. That is, the advertisements to be placed in all groups can utilize user identity association analysis among the above platforms to achieve reliable analysis and processing of advertising effects. This not only considers the conversion of the final advertising link, but also the user's educational experience with the advertisement in the early stage.

[0096] The sum of the number of associated identified users refers to the total number of associated identified users under all user profiles of a certain platform combination. When the total number is greater than the preset threshold for the number of associated users, it indicates that the platform combination as a whole has accumulated sufficient user association identification data, and the user identity association analysis between platforms has a high degree of reliability. At this time, all groups are included in the available analysis groups of the platform combination, making full use of the association analysis capabilities between platforms to achieve a comprehensive evaluation of the advertising effect, taking into account both the final conversion of the advertising link and the multiple touches of users during the advertising display stage.

[0097] If the total number of associated users identified by a platform combination across all user profiles is greater than a preset threshold for the number of associated users, it indicates that the platform combination has accumulated sufficient associated identification data under each user profile. The available analysis group for the platform combination is determined to be all groups. The advertising effectiveness evaluation based on user association analysis can be carried out for the advertisements to be placed in all groups across the platforms covered by the platform combination.

[0098] This situation, through the overall judgment of the total number of users identified by association, is significant because when the overall association identification data of the platform combination is sufficient, it maximizes the coverage of the available analysis groups, ensures that the advertising effect of all groups can be fully and reliably evaluated in the platform combination, and gives full play to the analytical value of multi-platform integrated data.

[0099] Scenario 2: If the sum of the number of associated users under different user profiles among the platforms corresponding to the platform combination is not greater than the preset threshold for the number of associated users, the user profile with the number of associated users greater than the preset threshold for the number of associated users is taken as a reliable analysis profile. If the number of reliable analysis profiles among the platforms is greater than the preset threshold for reliable analysis profiles, then the available analysis groups among the platforms corresponding to the platform combination are determined to be all groups.

[0100] The reliable analysis profile refers to the user profile that has a number of associated and identified users that is greater than the preset number of identified users under a specific platform combination. When the number of reliable analysis profiles is greater than the preset number of reliable analysis profiles, it indicates that although the total number of associated and identified users does not exceed the preset threshold for the number of associated users, the number of reliable associated and identified users has been reached in a sufficient number of user profile dimensions. The association analysis conditions of each group are generally sufficient, and all groups can still be included in the available analysis groups.

[0101] Assuming the total number of associated users identified by a certain platform combination does not exceed a preset threshold for the number of associated users, user profiles with an associated user count exceeding the preset threshold are further counted and marked as reliable analysis profiles. The number of reliable analysis profiles is then calculated. Let the preset threshold for the number of reliable analysis profiles be a certain value. If the number of reliable analysis profiles exceeds this value, it indicates that the association analysis conditions are sufficient for most user profiles, and all groups are still considered usable for analysis.

[0102] This situation is supplemented by a reliable number of user profiles. The significance of this is that when the total number of users identified is insufficient but the analysis conditions for each user profile dimension are generally sufficient, the full coverage of the available analysis groups is still maintained, avoiding excessive restrictions on the scope of available analysis groups due to deviations in the total number of indicators.

[0103] Scenario 3: If the number of reliable analysis profiles between the platforms is not above the preset reliable analysis profile, based on the user profile of the advertisement to be delivered in the group, determine whether the proportion of reliable analysis profiles in the user profile of the advertisement to be delivered in the group is above the preset reliable profile proportion threshold. If yes, then determine that the group belongs to the available analysis group between the platforms corresponding to the platform combination. If no, then determine that the group does not belong to the available analysis group between the platforms corresponding to the platform combination.

[0104] The reliable profile ratio refers to the proportion of reliable profiles among the user profiles involved in the ads to be delivered to a certain group, out of the total number of user profiles in that group. The preset reliable profile ratio threshold is a critical value used to determine whether the reliable profile ratio of user profiles in a certain group reaches the condition for including the group in the available analysis group. When the reliable profile ratio is above the preset reliable profile ratio threshold, it indicates that the user profiles corresponding to the group have sufficient correlation analysis support under the platform combination, and the group is included in the available analysis group; otherwise, the correlation analysis support of the user profiles in the group is insufficient, and it is not included in the available analysis group to ensure the reliability of the correlation analysis of the available analysis group.

[0105] Assuming the number of reliable analysis profiles is not above the preset number of reliable analysis profiles, the proportion of reliable analysis profiles in the user profiles of each group is determined one by one. Let the preset threshold for the proportion of reliable profiles be a certain value. If the proportion of reliable analysis profiles in a group is above this threshold, then that group is included in the available analysis groups of the platform combination; if the proportion of reliable analysis profiles in a group is below this threshold, then that group is not included in the available analysis groups of the platform combination.

[0106] This situation, through the refined judgment of the proportion of reliable analytical profiles in the user profile of a group, is significant because when the overall number of reliable analytical profiles is insufficient, differentiated evaluations are conducted for each group. Only groups with a high coverage ratio of reliable analytical profiles in the user profile are included in the available analytical groups, ensuring that the correlation analysis conditions in the available analytical groups all meet the reliability requirements. This provides a high-quality available analytical group data foundation for determining subsequent advertising effectiveness evaluation and management methods.

[0107] This embodiment, through S21 to S33 and comprehensive judgment of various situations, realizes the determination of the target strategy advertising analysis and management method and the intelligent screening of available analysis groups between platforms. Its core value is reflected in four aspects: First, through multi-dimensional evaluation of the overlap between the number of ads placed under the target strategy and user profiles, the analysis and management method of each ad is dynamically determined, ensuring the comprehensiveness of the analysis while reasonably controlling the complexity of the analysis; Second, through flexible switching between reliable analysis strategies and other analysis strategies, click data is used to trigger correlation analysis when the correlation analysis conditions are sufficient, and conversion data is used to trigger it when the conditions are limited, balancing analysis quality and processing efficiency; Third, through the identification of related users and the statistics of the number under different user profiles, the correlation analysis capability between platforms is refined to the user profile dimension, providing a data foundation for the accurate screening of available analysis groups; Fourth, through three-level judgment of the total number of users identified by correlation, the number of reliable analysis profiles, and the sufficiency of group profile coverage, available analysis groups with reliable correlation analysis conditions are gradually screened out, ensuring the reliability of subsequent advertising effect evaluation.

[0108] S3 determines the evaluation and management methods for advertisements to be placed in different groups based on the degree of platform correlation between the available analysis groups and the platform data corresponding to the available analysis groups.

[0109] The degree of platform correlation is determined based on the overlap of available analysis groups between different platform combinations. A greater overlap indicates more consistent available analysis groups between different platform combinations, and a higher degree of correlation between platforms. The platform data corresponding to the available analysis group includes the number of platform combinations to which the group belongs. A higher number indicates that the group can perform correlation analysis processing across more platform combinations. The evaluation management method refers to a specific method for specifying which groups' advertisements need to be processed for placement across all platforms to conduct cross-platform advertising performance correlation analysis. The first strategy refers to placing all advertisements to be placed in groups that belong to available analysis groups across all platform combinations across all platforms. The second strategy refers to placing all advertisements to be placed in a group across all platform combinations if the proportion of platform combinations belonging to available analysis groups in all platform combinations exceeds a preset platform combination proportion threshold. The third strategy, based on the second strategy, adds a condition that the placement analysis demand weight value is greater than a preset weight threshold to further filter groups with higher placement demand.

[0110] Assuming the available analysis groups for each platform combination have been determined, calculate the analysis matching ratio for each platform combination. Set a preset analysis matching ratio threshold, a preset platform combination percentage threshold, a preset overlap threshold, a preset demand threshold, and a preset weight threshold. Based on the average analysis matching ratio, the average overlap, and the correlation analysis demand value, determine in turn whether to adopt the first, second, or third strategy to determine the evaluation and management method for each group.

[0111] This step establishes a complete decision-making path from available analytics group data to the determination of evaluation and management methods. Its significance lies in the fact that by comprehensively evaluating the coverage of available analytics groups across the entire platform and the degree of correlation between platforms, the advertising effectiveness evaluation and management methods for each group are dynamically determined. This ensures both the reliability of advertising effectiveness evaluation and the sufficiency of user correlation analysis and processing between different platforms.

[0112] Furthermore, the degree of platform correlation between the available analytics groups is determined based on the amount of overlap in the available analytics groups between different platform combinations.

[0113] Furthermore, the platform data corresponding to the available analytics group is determined by including the number of platform combinations to which the group belongs.

[0114] Furthermore, the method for determining the evaluation and management method for advertisements to be delivered in the group is as follows: In this embodiment, based on the platform data corresponding to the available analysis groups, the reliability of user identity analysis processing is determined when evaluating and analyzing the advertising effect across all platforms, based on the available analysis groups across different platforms. That is, the fewer the number of available analysis groups across different platforms, the lower the reliability of user analysis processing. Combining the platform correlation between available analysis groups, it is determined whether the number of platforms in different groups that can undergo correlation analysis processing is sufficient in the current state. That is, the more platform combinations a group belongs to, the more sufficient the number of platforms in different groups that can undergo correlation analysis processing. Based on the reliability of user identity analysis processing when evaluating and analyzing the advertising effect across all platforms and whether the number of platforms in different groups that can undergo correlation analysis processing is sufficient, the method for evaluating and managing advertisements to be placed in a group is determined. This determines which groups are used for correlation analysis processing across all platforms, thus ensuring both the reliability of advertising effect evaluation and analysis processing and the reliability of user correlation analysis processing across different platforms.

[0115] S41 uses the platform data corresponding to the available analysis groups to determine the available analysis groups in different platform combinations, and uses the proportion of the available analysis groups in the platform combination to all groups to determine the analysis matching ratio of the platform combination. The analysis matching ratio refers to the proportion of the number of available analysis groups of a certain platform combination to the total number of all groups, which is used to measure the degree of group coverage that the platform combination can support for association analysis; the analysis deviation platform combination refers to the platform combination whose analysis matching ratio is less than the preset analysis matching ratio threshold, indicating that the association analysis coverage capability of the platform combination is insufficient.

[0116] Assuming that the available analysis groups for each platform combination have been determined, calculate the analysis matching ratio for each platform combination (i.e., the number of available analysis groups ÷ the total number of all groups) to form an analysis matching ratio dataset for each platform combination. Calculate the average analysis matching ratio for different platform combinations and compare it with the preset analysis matching ratio threshold.

[0117] This step, by analyzing the calculation of the matching ratio, is significant in that it quantifies the ability of each platform combination to cover the entire group in association analysis. This provides a basis for subsequent assessment of the reliability of overall association analysis based on the average value of the matching ratio, identifies platform combinations with insufficient association analysis coverage, and provides key reference information for the selection of evaluation management methods.

[0118] The above steps include the following: Scenario 1: If the average of the analysis matching ratios of different platform combinations is less than the preset analysis matching ratio threshold, then the first strategy is adopted to determine the evaluation and management method of the advertisements to be placed in the group.

[0119] The average of the analysis matching ratio refers to the arithmetic mean of the analysis matching ratios of each platform combination. When the average value is less than the preset analysis matching ratio threshold, it indicates that the overall correlation analysis coverage of each platform combination is weak, and the reliability of user identity analysis is low when evaluating the advertising effect across all platforms. In this case, the first strategy is adopted, that is, only the advertisements to be placed in the groups that belong to the available analysis groups in all platform combinations are placed on all platforms to maximize the reliability of analysis under limited correlation analysis conditions.

[0120] Assuming the analysis matching ratio of each platform combination has been calculated, calculate its average value. Set the preset analysis matching ratio threshold to a certain value. If the average value is less than the preset analysis matching ratio threshold, it indicates that the overall correlation analysis coverage is insufficient. In this case, the first strategy is adopted, which involves placing all ads to be delivered in groups that belong to the available analysis groups in all platform combinations on all platforms.

[0121] The significance of adopting the first strategy in this situation lies in using the most conservative evaluation and management method when the overall reliability of the correlation analysis is low. This involves conducting cross-platform correlation analysis only on the groups with the most sufficient correlation analysis conditions (groups that are available for analysis in all platform combinations). This ensures that the groups included in the overall platform evaluation have the highest reliability guarantee for correlation analysis. It is better to narrow the scope of the evaluation than to fail to ensure the quality of the evaluation, thereby further determining the reliability of user correlation analysis processing in different platform combinations.

[0122] Case 2: If the average analysis matching ratio of different platform combinations is not less than the preset analysis matching ratio threshold, the platform combinations with analysis matching ratios less than the preset analysis matching ratio threshold are regarded as analysis deviation platform combinations. It is determined whether there are analysis deviation platform combinations. If not, proceed to step S42. If yes, the first strategy is adopted to determine the evaluation and management method of the advertisements to be placed in the group.

[0123] The platform combination with analysis deviation refers to the platform combination whose analysis matching ratio is less than the preset analysis matching ratio threshold. When the average analysis matching ratio is not less than the preset analysis matching ratio threshold but there is a platform combination with analysis deviation, it indicates that although the average level is acceptable, there is a specific platform combination with significantly insufficient correlation analysis capability. In order to ensure the overall reliability of the correlation analysis, the first strategy is still adopted.

[0124] Assuming the average analysis matching ratio is not less than the preset analysis matching ratio threshold, further check whether there are platform combinations with an analysis matching ratio lower than the preset analysis matching ratio threshold. If there are platform combinations with analysis deviation, it indicates that the correlation analysis coverage of some platform combinations is insufficient. The first strategy is still adopted, and the ads to be delivered in the groups that belong to the available analysis groups in all platform combinations are delivered to all platforms. If there are no platform combinations with analysis deviation, then the analysis matching ratio of all platform combinations is above the threshold, and proceed to S42 for more refined platform correlation analysis.

[0125] The significance of this analysis of platform combinations with biases is that even if the overall average matching ratio meets the standard, if there are obviously weak platform combinations, a conservative strategy should be adopted to ensure that the groups included in the overall platform evaluation are not affected by individual weak platform combinations, thus maintaining the robustness of the advertising effectiveness evaluation results.

[0126] S42 determines the number of overlapping available analysis groups between different platform combinations based on the degree of platform correlation among the available analysis groups; The degree of platform correlation between available analysis groups is determined based on the number of overlaps in available analysis groups between different platform combinations. The number of overlaps refers to the number of groups that two different platform combinations belong to in common available analysis groups. The greater the overlap, the higher the degree of overlap between the two platform combinations in available analysis groups, and the stronger the correlation between the platforms. The average number of overlaps refers to the arithmetic mean of the number of overlaps in available analysis groups between all platform combinations, which is used to measure the overall degree of correlation between platforms.

[0127] Assuming that the available analytical groups for each platform combination have been determined, the number of overlapping available analytical groups between each pair of platform combinations is calculated one by one, forming a data matrix of the number of overlapping groups between each pair of platform combinations. The average number of overlapping groups is calculated and compared with a preset overlap threshold. Based on the comparison results, the subsequent evaluation and management methods are selected to determine the strategy.

[0128] This step, by calculating the number of overlapping available analysis groups, is significant in assessing the consistency of different platform combinations across available analysis groups. A higher overlap indicates that the correlation analysis coverage of each platform combination is more consistent, and that different groups have good correlation analysis conditions in multiple platform combinations, providing a quantitative basis for the selection of subsequent assessment and management methods based on the degree of platform correlation.

[0129] S43 determines the evaluation and management method for ads to be placed in the group based on the analysis matching ratio in different platform combinations and the number of overlaps in available analysis groups between different platform combinations.

[0130] The correlation analysis demand value is determined based on the difference between 1 and the average percentage of platform combinations belonging to the available analysis groups for different groups. The smaller the average percentage of platform combinations belonging to the available analysis groups for different groups, the larger the correlation analysis demand value, indicating that the overall coverage of platform combinations that can be used for correlation analysis for each group is relatively weak, and the need for expanded correlation analysis is more urgent. The preset demand threshold is a critical value used to determine whether the correlation analysis demand value is high enough to adopt the second strategy. The preset overlap threshold is a critical value used to determine whether the average number of overlapping available analysis groups among platform combinations is small. The third strategy adds the condition that the placement analysis demand weight value is greater than the preset weight threshold on the basis of the second strategy, that is, only groups that meet both the platform combination percentage condition and have a high placement analysis demand weight value are processed for full-platform placement.

[0131] Assuming the average number of overlapping available analytical groups across different platform combinations has been calculated, and a preset threshold for the number of overlapping groups is set, if the average is less than the threshold, then the degree of overlap of available analytical groups in different platform combinations for each group is low, and the second strategy is adopted; if the average is not less than the threshold, then the correlation analysis requirement value is further calculated, and the evaluation and management method for each group is determined by the second strategy or the third strategy based on the comparison result between the correlation analysis requirement value and the preset requirement threshold.

[0132] This step involves a multi-layered comprehensive judgment based on the average value of overlapping quantities, the correlation analysis demand value, and the weight value of the ad placement analysis demand value. Its significance lies in dynamically selecting the most suitable evaluation and management method for the current ad placement layout based on the degree of correlation between platforms and the intensity of correlation analysis demand for each group. When the degree of correlation between platforms is low, a second strategy with a wider coverage is adopted. When the degree of correlation between platforms is high but the correlation analysis demand is high, the second strategy is still adopted. When the correlation analysis demand is low, a more precise third strategy is adopted. This ensures the reliability of ad performance evaluation while also ensuring the sufficiency of user correlation analysis processing between different platforms.

[0133] In the above steps, S431 determines whether the average number of overlaps in the available analysis groups between different platform combinations is less than a preset overlap threshold. If so, the second strategy is used to determine the evaluation and management method of the advertisements to be placed in the group. If not, the process proceeds to step S432. The second strategy is as follows: if the group belongs to a platform combination of available analysis groups and its proportion in all platform combinations is above a preset platform combination proportion threshold, then all ads to be delivered in the group will be delivered on all platforms to more comprehensively determine the reliability of user association analysis processing between different platform combinations. When the average overlap is small, it indicates that the available analysis groups of each platform combination are different and the association analysis conditions of different groups in each platform combination are uneven. The second strategy is adopted to include groups with a high platform combination coverage ratio into the full platform delivery processing to maximize the expansion of association analysis processing opportunities in these groups.

[0134] Assuming the average number of overlaps is less than the preset overlap threshold, the second strategy is adopted. For each group, it is determined whether the proportion of the platform combination that it belongs to in the available analysis group is above the preset platform combination proportion threshold. If so, all the ads to be placed in the group will be placed on all platforms. If not, the group will not be included in the full platform placement in this round of evaluation and management.

[0135] The significance of adopting the second strategy in this situation is that when the consistency of available analytical groups among platform combinations is low, the broad platform coverage of the group itself is used as the screening criterion to ensure that the groups included in the full platform campaign have the conditions for correlation analysis in as many platform combinations as possible, thereby improving the overall reliability of multi-platform advertising performance evaluation.

[0136] S432 determines the correlation analysis demand value based on the average percentage of platform combinations belonging to available analysis groups in different groups, and determines whether the correlation analysis demand value is greater than a preset demand threshold. If so, the second strategy is adopted to determine the evaluation and management method of the advertisements to be placed in the group; otherwise, the third strategy is adopted to determine the evaluation and management method of the advertisements to be placed in the group.

[0137] The correlation analysis demand value is determined based on the average percentage of platform combinations belonging to available analysis groups for different groups. The smaller this average percentage, the greater the correlation analysis demand value. The preset demand threshold is a critical value used to determine whether the correlation analysis demand is high and whether the second strategy needs to be adopted. When the correlation analysis demand value is greater than the preset demand threshold, it indicates that there are relatively few platform combinations that can participate in the correlation analysis for each group as a whole, and the demand for expanding the correlation analysis is more urgent. The second strategy is adopted to lower the screening criteria to expand the analysis coverage. When the correlation analysis demand value is not greater than the preset demand threshold, it indicates that the overall conditions for correlation analysis are relatively sufficient. The third strategy is adopted to further require the group's campaign analysis demand weight value to meet the conditions based on the second strategy, focusing on the most valuable group to carry out full-platform campaign processing.

[0138] Assuming the average number of overlaps is not less than a preset overlap threshold, the correlation analysis demand value is further calculated. Let the preset demand threshold be a certain value. If the correlation analysis demand value is greater than the preset demand threshold, the second strategy is adopted; if the correlation analysis demand value is not greater than the preset demand threshold, the third strategy is adopted. In addition to meeting the platform combination ratio condition, the group's placement analysis demand weight value also needs to be greater than a preset weight threshold before all ads in that group can be placed on all platforms.

[0139] This step, through the determination of the required value of correlation analysis, is significant because, when the consistency of available analysis groups among platform combinations is high, the second or third strategy can be dynamically selected based on the overall level of the number of platform combinations that can participate in the correlation analysis for each group. When the overall conditions for correlation analysis are sufficient, the more precise third strategy is adopted, focusing on high-value groups with a large number of ads to carry out cross-platform correlation analysis across all platforms, thereby further improving the comprehensive value of ad performance evaluation.

[0140] Furthermore, the third strategy is as follows: if the group belongs to the platform combination of the available analysis group, its proportion in all platform combinations is above the preset platform combination proportion threshold, and the weight value of the ad placement analysis demand in the group is greater than the preset weight threshold, then all the ads to be placed in the group will be placed on all platforms. This will more comprehensively determine the reliability of the correlation analysis processing between users of different platform combinations, and will also utilize groups with larger ad placement analysis demand weight values, i.e., groups with a larger number of ads to be placed, to achieve reliable correlation analysis processing between users of different platform combinations.

[0141] The third strategy adds a dual screening condition based on the second strategy: the weight value of the ad placement analysis demand is greater than the preset weight threshold. By combining the sufficiency of platform coverage (the proportion of platform combinations belonging to the available analysis group) and the representativeness of the number of ads (the weight value of the ad placement analysis demand), it ensures that the groups included in the full platform placement have both sufficient correlation analysis coverage and represent a high ad placement demand, thereby achieving accurate allocation of correlation analysis resources.

[0142] Assuming the third strategy is adopted, two conditions are determined for each group: first, whether the proportion of the platform combination belonging to the available analysis group in this group is above a preset platform combination proportion threshold; second, whether the weight value of the group's ad placement analysis demand is greater than a preset weight threshold. Groups that meet both conditions will have all their ads placed on all platforms processed for placement; groups that meet only one condition or neither condition will not be included in the full-platform placement process in this round.

[0143] This step, through the dual screening of the third strategy, is significant because when the overall conditions for correlation analysis are sufficient, it further focuses on core groups with a large number of ads to carry out full-platform placement and correlation analysis. This allows limited analytical resources to be concentrated on the groups with the highest demand for ad placement, while ensuring the reliability of correlation analysis through the conditions of sufficient platform coverage, thus achieving accurate and efficient allocation of resources for ad performance evaluation.

[0144] Example 2 In a second aspect, the present invention provides a computer system comprising: a memory and a processor connected in communication, and a computer program stored in the memory and capable of running on the processor, wherein the processor executes the aforementioned advertising effectiveness evaluation and management method based on fused data when running the computer program.

[0145] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the embodiments of apparatus, devices, and non-volatile computer storage media are basically similar to the method embodiments, so the descriptions are relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0146] The foregoing has described specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than that shown in the embodiments and may still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require the specific or sequential order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0147] The above description is merely one or more embodiments of this specification and is not intended to limit this specification. Various modifications and variations can be made to the one or more embodiments of this specification by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of one or more embodiments of this specification should be included within the scope of the claims of this specification.

Claims

1. A method for evaluating and managing advertising effectiveness based on fused data, characterized in that, Specifically, it includes: Based on the user profile of the ads to be delivered, the ads to be delivered are divided into different groups. Based on the ads to be delivered in different groups, the ads to be delivered using the target strategy are determined and the ads are delivered using the target strategy. Based on the target strategy, the advertisements are delivered to all platforms. Based on the delivery data of the target strategy advertisements and the degree of overlap of user profiles between different target strategy advertisements, the method for analyzing and managing the advertising effect of the target strategy advertisements on different platforms is determined. The analysis and management method is used to determine the identification results of related user identities between different platforms. Based on the identification results, the available analysis groups between the platforms are determined. Based on the degree of platform correlation between the available analytics groups and the platform data corresponding to the available analytics groups, the evaluation and management methods for advertisements to be placed in different groups are determined.

2. The advertising effectiveness evaluation and management method based on fused data as described in claim 1, characterized in that, The advertisements to be delivered are divided into different groups, specifically including: Ads with consistent user profiles are grouped into the same group.

3. The advertising effectiveness evaluation and management method based on fused data as described in claim 1, characterized in that, The method for determining the target strategy for ad delivery is as follows: Based on the group data, determine the number of groups; By utilizing the ads to be delivered in different groups, the number of ads to be delivered in each group is determined; By using the number of groups and the number of ads to be delivered in different groups, the target strategy ads are determined among the ads to be delivered.

4. The advertising effectiveness evaluation and management method based on fused data as described in claim 3, characterized in that, If the number of groups is greater than the preset group number threshold, then the target strategy advertisement in the advertisement to be delivered will be randomly selected from the groups that meet the preset conditions as the target strategy advertisement.

5. The advertising effectiveness evaluation and management method based on fused data as described in claim 4, characterized in that, The preset condition is a group whose number of advertisements to be placed is greater than a preset threshold for the number of advertisements to be placed.

6. The advertising effectiveness evaluation and management method based on fused data as described in claim 1, characterized in that, The method for determining the analysis and management method of the advertising effect of the target strategy across different platforms is as follows: Based on the ad delivery data of the target strategy, determine the number of ads delivered under the target strategy; Based on the degree of overlap in user profiles between ads targeting different target strategies, determine the amount of overlap between user profiles of the ads targeting a specific target strategy and ads targeting other target strategies. A method for analyzing and managing the advertising effectiveness of targeted strategy ads across different platforms is used to determine the number of ads delivered using the target strategy and the amount of overlap in user profiles between the targeted strategy ads and other targeted strategy ads.

7. The advertising effectiveness evaluation and management method based on fused data as described in claim 6, characterized in that, The number of ads placed under the target strategy is obtained. If the number of ads placed under the target strategy is less than the preset ad placement threshold, then the analysis and management method for the advertising effect of all ads placed under the target strategy across different platforms is determined as follows: if click data exists on any platform, it is determined whether the same user clicks on different platforms.

8. The advertising effectiveness evaluation and management method based on fused data as described in claim 6, characterized in that, Based on the target strategy advertising based on reliable analysis strategies under different user profiles, user profiles of target strategy advertising without reliable analysis strategies are used as analysis deviation user profiles. It is determined whether the proportion of analysis deviation user profiles in the user profiles of the target strategy advertising is greater than a preset deviation user profile proportion threshold. If so, the method for analyzing and managing the advertising effect of the target strategy advertising across different platforms is determined to be a reliable analysis strategy. That is, as long as click data exists on any platform, it is determined whether the same clicking user exists on different platforms. If not, the method for analyzing and managing the advertising effect of the target strategy advertising across different platforms is determined to be another analysis strategy. That is, as long as conversion data exists on any platform, it is determined whether the same converting user exists on different platforms.

9. The advertising effectiveness evaluation and management method based on fused data as described in claim 1, characterized in that, The method for determining the evaluation and management method for advertisements to be placed in the group is as follows: Based on the platform data corresponding to the available analysis groups, determine the available analysis groups in different platform combinations, and determine the analysis matching ratio of the platform combination based on the proportion of the available analysis groups in the platform combination among all groups; Based on the degree of platform correlation among the available analysis groups, determine the amount of overlap in available analysis groups among different platform combinations; The evaluation and management method for ads to be placed in the groups is determined based on the analysis matching ratio of different platform combinations and the amount of overlap of available analysis groups between different platform combinations.

10. A computer system, comprising: A memory and processor connected by communication, and a computer program stored in the memory and capable of running on the processor, characterized in that, when the processor runs the computer program, it executes an advertising effectiveness evaluation and management method based on fused data as described in any one of claims 1-9.

Images