Digital integrated marketing service platform

Abstract

The invention discloses a digital integrated marketing service platform, and relates to the technical field of digital marketing resource allocation, and the platform comprises the steps: collecting the installation height and horizontal position of an offline contact, and in-screen layout parameters of an online contact, and constructing a contact record vector comprising a vertical height difference, a horizontal distance, a normalized vertical direction position and a nominal exposure weight; calculating a vertical sight deflection angle, a comfortable sight center angle and steady dispersion of the vertical sight deflection angle based on the contact record vector to obtain a sight reachable contribution coefficient of each contact; and in combination with the total exposure quota and the nominal exposure weight of the current marketing period, determining the exposure quota of each contact, and generating a cross-channel putting instruction set. According to the invention, the putting effect and the resource utilization efficiency of integrated marketing can be improved.

Description

Technical Field

[0001] This invention relates to the field of digital marketing resource allocation technology, and in particular to a digital integrated marketing service platform. Background Technology

[0002] In today's digital marketing landscape, brands typically leverage multiple touchpoints simultaneously, including offline store screens, mall wayfinding displays, outdoor media, and online in-app ads and news feed recommendations, for integrated campaigns. Marketing platforms need to plan and dynamically allocate exposure resources across these different touchpoints within a defined budget. Because the target audience's gaze is influenced by factors such as posture, walking path, screen height, and page layout, the level of attention given to the same content varies significantly depending on its vertical position. Without a quantifiable description of the relative comfort zones of each touchpoint, it becomes difficult to maximize overall exposure while simultaneously improving the efficiency of limited exposure resources across multiple touchpoints.

[0003] In existing technologies, digital marketing platforms primarily allocate resources based on channel type, traffic volume, or historical click-through rate data. They lack unified modeling for factors such as the relationship between the installation height of offline touchpoints and the eye level of the target audience, the spatial distance between offline touchpoints and the audience's activity path, and the vertical position of online touchpoints on different terminal screens. Often, they rely on experience to roughly adjust by dividing areas into upper, middle, and lower sections or fixing ad placement levels, failing to reflect the actual eye distribution of different audiences in real-world scenarios. Furthermore, online and offline touchpoints use independent coordinate systems and statistical methods, lacking a standardized data structure that can simultaneously describe spatial location and exposure intensity parameters within a single platform. This results in coarse-grained cross-channel exposure resource allocation, with some locations having concentrated exposure resources but insufficient actual visibility, while other locations have long been underestimated, leading to overall campaign effectiveness that needs further improvement. Summary of the Invention

[0004] The purpose of this invention is to solve the problem of mismatch between the allocation of multi-touchpoint exposure resources and the actual visual attention of the target audience in the existing technology, and to propose a digital integrated marketing service platform.

[0005] To address the problems existing in the prior art, the present invention adopts the following technical solution:

[0006] A digital integrated marketing service platform, comprising:

[0007] The touch point recording vector module is used to calculate the touch point recording vector of each touch point in the touch point set. The touch point recording vector includes the vertical height difference, horizontal distance, normalized vertical position, normalized horizontal position, and nominal exposure weight.

[0008] The line-of-sight statistics module is used to calculate the vertical line-of-sight deflection angle based on the touch point recording vector, and to calculate the robustness of the comfortable line-of-sight center angle and the vertical line-of-sight deflection angle based on the vertical line-of-sight deflection angle and the nominal exposure weight.

[0009] The reachability contribution module is used to calculate the reachability contribution coefficient of the line of sight based on the vertical line of sight deflection angle, the comfortable line of sight center angle, and the robust dispersion of the vertical line of sight deflection angle.

[0010] The exposure quota module is used to obtain the total exposure quota for the current marketing cycle and determine the exposure quota for each touchpoint based on the total exposure quota, nominal exposure weight, and visibility reach contribution coefficient.

[0011] The instruction generation module is used to generate cross-channel delivery instruction sets based on exposure quotas.

[0012] Preferably, calculating the contact record vector for each contact in the contact set includes:

[0013] Obtain a set of touchpoints, which includes offline touchpoints and online touchpoints;

[0014] Acquire data on the installation height and horizontal position of offline touchpoints, as well as the average eye height of the target population;

[0015] The vertical height difference of the offline contact point is calculated based on the installation height and the average eye height data of the target population.

[0016] The horizontal distance to the offline contact point is calculated based on the horizontal position and the reference line-of-sight position of the target group;

[0017] Obtain the in-screen layout parameters of the online touchpoints, and normalize the in-screen layout parameters based on the terminal screen size and pixel density to obtain the normalized horizontal and vertical positions of the online touchpoints.

[0018] Nominal exposure weights are generated based on campaign logs and reach statistics;

[0019] The vertical height difference of the lower touch point, the horizontal distance of the lower touch point, the normalized horizontal position of the upper touch point, the normalized vertical position of the upper touch point, and the nominal exposure weight are used to form the touch point recording vector.

[0020] Preferably, calculating the vertical line-of-sight angle based on the contact point recording vector includes:

[0021] Obtain the touch point recording vector for each touch point. The touch point recording vector includes the vertical height difference and horizontal distance of the offline touch points, the normalized horizontal position and normalized vertical position of the online touch points, and the nominal exposure weight.

[0022] The angle between the vertical height difference and the horizontal distance is calculated to obtain the vertical line-of-sight angle of the offline contact point;

[0023] The screen size and pixel density of the terminal are converted to obtain the in-screen normalized scale;

[0024] The angle between the normalized vertical position and the screen normalized scale is calculated to obtain the vertical line-of-sight angle of the online touch point.

[0025] Preferably, calculating the comfortable line-of-sight center angle includes:

[0026] The comfortable viewing center angle is obtained by weighting the vertical viewing angle of each contact point with the nominal exposure weight.

[0027] Preferably, calculating the robust dispersion of the vertical line-of-sight deflection angle includes:

[0028] The median absolute deviation of the vertical line-of-sight angle at each contact point is calculated to obtain the robust dispersion of the vertical line-of-sight angle at each contact point.

[0029] Preferably, the achievable contribution coefficient of the line of sight is calculated based on the vertical line of sight deflection angle, the comfortable line of sight center angle, and the robust dispersion of the vertical line of sight deflection angle, including:

[0030] Calculate the absolute difference between the vertical line-of-sight angle and the center angle of the comfortable line-of-sight at each contact point, and perform a weighted average calculation on the absolute difference based on the nominal exposure weight to obtain the line-of-sight gravity mismatch.

[0031] The mismatch modulation scale is obtained by adding the robust dispersion of the vertical line of sight deflection angle to the line of sight gravity skew mismatch.

[0032] Based on the mismatch modulation scale, the absolute difference between the vertical line-of-sight deflection angle of each contact point and the center angle of the comfortable line-of-sight is subjected to exponential decay mapping to obtain the line-of-sight reachable contribution coefficient.

[0033] Preferably, the exposure quota for each touchpoint is determined based on the total exposure quota, nominal exposure weight, and line-of-sight reach contribution coefficient, wherein the formula for calculating the exposure quota is:

[0034] ;

[0035] In the formula, For the first Exposure quota per touchpoint Total exposure quota, For the first Nominal exposure weight per touchpoint For the first The line of sight at each contact point can contribute to the coefficient. For the set of contact points, For index variables, This is the result of summing the products of the nominal exposure weights and the reachable contribution coefficients of all touchpoints in the touchpoint set.

[0036] Preferably, a cross-channel delivery instruction set is generated based on the exposure quota, including:

[0037] Obtain the unit resource cost of each touchpoint, and based on the ratio of the exposure quota of each touchpoint to the unit resource cost, obtain the instruction quantity of each touchpoint;

[0038] A cross-channel delivery instruction set is constructed based on the instruction quantity of each touchpoint. The cross-channel delivery instruction set includes touchpoint identifiers and instruction quantities.

[0039] Compared with the prior art, the beneficial effects of the present invention are:

[0040] 1. This invention collects the vertical height difference and horizontal distance of offline touchpoints, as well as the normalized vertical and horizontal positions and nominal exposure weights of online touchpoints in the touchpoint recording vector module. In the gaze statistics module, it calculates the vertical gaze deviation angle, the comfortable gaze center angle, and the robust dispersion of the vertical gaze deviation angle. This allows touchpoints of different channels and forms to be quantitatively described under the same data structure and the same gaze coordinate system. As a result, the positional relationships that were originally scattered in physical space and screen space are transformed into calculable gaze offset and concentration indicators, thus realizing a fine modeling of the natural gaze distribution of the target audience.

[0041] 2. This invention further calculates the reachability contribution coefficient based on the robust dispersion of the vertical viewing angle, the comfortable viewing center angle, and the vertical viewing angle through the reachability contribution module. In the exposure quota module, the total exposure quota of the current marketing cycle is combined with the nominal exposure weight and the reachability contribution coefficient to calculate the exposure quota of each touchpoint. This makes the allocation of exposure resources not only consider the importance of historical or planned deployments, but also explicitly introduce the distance relationship between the touchpoint and the comfortable viewing area and the overall offset degree. It realizes an allocation mechanism that comprehensively weights visibility and importance under the condition of limited total amount, thereby improving the utilization efficiency of limited exposure resources among multiple touchpoints and reducing the waste of resources in high-cost but low-visibility positions.

[0042] 3. This invention also obtains the unit resource cost of each touchpoint, converts the exposure quota into the instruction quantity of each touchpoint, and constructs a cross-channel delivery instruction set containing touchpoint identifiers and instruction quantities. This enables the platform to directly drive the delivery execution of different offline media and online applications under unified budget and exposure quota constraints. It realizes the automatic distribution of quotas optimized according to line-of-sight related parameters and cross-channel linkage control, thereby reducing manual splitting and communication configuration work, improving the consistency and response speed of delivery execution, and ensuring that the optimized resource allocation results can be stably implemented in the actual delivery process, further improving the overall marketing campaign delivery effect and management efficiency. Attached Figure Description

[0043] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0044] Figure 1 This is a functional module diagram of a digital integrated marketing service platform provided in an embodiment of the present invention. Detailed Implementation

[0045] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0046] Example: This example provides a digital integrated marketing service platform. See [link / reference] Figure 1 Specifically, including:

[0047] The touch point recording vector module is used to calculate the touch point recording vector of each touch point in the touch point set. The touch point recording vector includes the vertical height difference, horizontal distance, normalized vertical position, normalized horizontal position, and nominal exposure weight.

[0048] Specifically, the touchpoint recording vector for each touchpoint in the touchpoint set is calculated, and the vertical height difference, horizontal distance, normalized vertical position, normalized horizontal position, and nominal exposure weight are simultaneously introduced into the touchpoint recording vector. This is to comprehensively characterize the key information of offline and online touchpoints in both spatial geometric position and exposure intensity within the same data structure, enabling the platform to compare and perform calculations on multiple touchpoints in a unified format. The vertical height difference and horizontal distance reflect the actual spatial offset of offline touchpoints relative to the natural line of sight of the target audience, while the normalized vertical and horizontal positions are used to measure the relative positions of different terminals. Under the conditions of screen size and resolution, the in-screen position measurement of online touchpoints is unified. The nominal exposure weight is used to characterize the exposure distribution of each touchpoint in the existing campaign. Through this recording method, the subsequent gaze statistics module can directly calculate the vertical gaze angle, comfortable gaze center angle and robust dispersion of the vertical gaze angle based on the touchpoint recording vector without adding extra data conversion steps. These results are then used continuously in the reachable contribution module and the exposure quota module to generate the gaze reachable contribution coefficient and exposure quota, thereby ensuring that the data transmission chain of the entire digital integrated marketing service platform is clear and consistent in multi-touchpoint and multi-channel scenarios.

[0049] In an embodiment of the present invention, calculating the contact record vector of each contact in the contact set includes:

[0050] Obtain a set of touchpoints, which includes offline touchpoints and online touchpoints;

[0051] Acquire data on the installation height and horizontal position of offline touchpoints, as well as the average eye height of the target population;

[0052] The vertical height difference of the offline contact point is calculated based on the installation height and the average eye height data of the target population.

[0053] The horizontal distance to the offline contact point is calculated based on the horizontal position and the reference line-of-sight position of the target group;

[0054] Specifically, a touchpoint set refers to a collection of multiple touchpoints selected by the platform in the current marketing campaign to carry out content display or advertising. This includes offline touchpoints set in physical spaces and online touchpoints set in application pages, web pages, or information flow interfaces. Offline touchpoints refer to display carriers installed in actual scenarios, such as wall screens, window screens, or store signs. Their location can be characterized by installation height and horizontal position. The vertical height difference of offline touchpoints refers to the difference between the installation height of offline touchpoints and the average eye height of the target audience, used to describe the degree to which the touchpoint requires the user to look up or down relative to natural eye level.

[0055] Specifically, when acquiring the touchpoint set, the platform collects all marketing touchpoint information that has been integrated into the current marketing system through the preset touchpoint access registration module. The touchpoints are then categorized to form a touchpoint set that includes both offline and online touchpoints. Offline touchpoints include physical display touchpoints such as store digital signage, location wayfinding screens, and offline event display screens, while online touchpoints include virtual display touchpoints such as in-app advertising slots, information feed recommendation slots, and live e-commerce advertising slots.

[0056] Specifically, when acquiring data on the installation height and horizontal position of offline touchpoints, as well as the average eye height of the target audience, the installation height of offline touchpoints is obtained by offline surveyors using a laser rangefinder to measure the vertical height of the touchpoint screen's center point from the ground. The horizontal position of offline touchpoints is obtained by GPS positioning or site plan coordinate calibration to acquire the horizontal coordinates of the touchpoint screen's center point relative to a site reference point (such as the center point of a store entrance or the center point of a shopping mall atrium). This measurement data is then entered into the platform by the surveyors. The platform calls the historical user data storage module to filter out the average height data of the target audience for the current marketing campaign. Then, it calculates the average eye height data of the target audience based on the general ergonomic conversion standard (average eye height = average height × 0.92). Before collecting and using user data, the platform clearly informs users of the purpose and method of processing, and collects and processes the information only after obtaining their consent.

[0057] Specifically, when calculating the vertical height difference of the offline touchpoint based on the installation height and the average eye height of the target population, the calculation method is to subtract the average eye height of the target population from the installation height of the offline touchpoint. When the calculation result is positive, it indicates that the offline touchpoint is higher than the average eye height of the target population, and when it is negative, it indicates that the offline touchpoint is lower than the average eye height of the target population. This operation can quantify the vertical orientation deviation of the offline touchpoint relative to the line of sight of the target population, providing key vertical dimension data for subsequently mapping the spatial position of the offline touchpoint to the gravity-constrained natural line of sight coordinates.

[0058] Specifically, the platform first combines the heat map data of the crowd at the offline venue to set the center point of the main activity area of ​​the target crowd as the reference line of sight position of the target crowd and determine its horizontal coordinate. Then, the straight-line distance between the horizontal coordinate of the offline touch point and the horizontal coordinate of the reference line of sight position is calculated by the distance calculation formula between the two points, so as to obtain the horizontal distance of the offline touch point. This operation can quantify the horizontal orientation deviation of the offline touch point relative to the line of sight of the target crowd. Together with the vertical height difference, it forms complete spatial position deviation data of the offline touch point, providing complete spatial dimension basic data for subsequent unified calculation of the vertical line of sight deflection angle of the touch point.

[0059] Obtain the in-screen layout parameters of the online touchpoints, and normalize the in-screen layout parameters based on the terminal screen size and pixel density to obtain the normalized horizontal and vertical positions of the online touchpoints.

[0060] Nominal exposure weights are generated based on campaign logs and reach statistics;

[0061] The vertical height difference of the lower touch point, the horizontal distance of the lower touch point, the normalized horizontal position of the upper touch point, the normalized vertical position of the upper touch point, and the nominal exposure weight are used to form the touch point recording vector.

[0062] Specifically, the platform establishes data interface connections with the terminal applications (such as apps, information flow platforms, and live e-commerce platforms) to which each online touchpoint belongs. It then calls the interface layout management module of each terminal application to obtain the in-screen layout parameters of the online touchpoints. These in-screen layout parameters are specifically the pixel coordinate information of the online touchpoints within the terminal screen, including the horizontal and vertical coordinates of the top-left and bottom-right pixels of the touchpoint's area. By calculating the average of these four pixel coordinates, the pixel center coordinates of the online touchpoint (i.e., the core data of the in-screen layout parameters) can be obtained. Simultaneously, the platform obtains the screen size parameters (specifically including the horizontal length and vertical height of the screen) and pixel density parameters (representing the number of pixels per inch of the terminal screen) of each terminal through the system information interface of the terminal application. Based on the obtained terminal screen size and pixel density, the platform performs normalization processing: First, the horizontal length and vertical height of the terminal screen are converted into pixel units. The conversion method is to calculate the number of pixels per millimeter based on the pixel density (i.e., the number of pixels per millimeter = pixel density / 25.4, where 25.4 is the number of pixels per inch). First, multiply the screen's horizontal length (in meters) by the number of pixels per millimeter to obtain the total horizontal pixel count. Then, multiply the screen's vertical height (in millimeters) by the number of pixels per millimeter to obtain the total vertical pixel count. Second, establish a two-dimensional coordinate system within the screen, with the horizontal axis as the x-coordinate and the vertical axis as the y-coordinate. Subtract half of the total horizontal pixel count from the x-coordinate of the pixel center of each online touchpoint to obtain the horizontal pixel offset relative to the origin. Subtract half of the total vertical pixel count from the y-coordinate of the pixel center of each online touchpoint. The first step involves obtaining the pixel offset in the vertical direction relative to the origin. The second step divides the pixel offset in the horizontal direction relative to the origin by half the total number of horizontal pixels on the screen to obtain the normalized horizontal position of the online touchpoint. Similarly, the third step divides the pixel offset in the vertical direction relative to the origin by half the total number of vertical pixels on the screen to obtain the normalized vertical position of the online touchpoint. This normalization process ensures that the values ​​of both the normalized horizontal and vertical positions fall within the range of [-1, 1], thus standardizing the online touchpoint position parameters for terminals of different sizes and pixel densities. This operation converts the heterogeneous in-screen layout parameters of online terminals into position data of a unified scale, eliminating the influence of terminal hardware differences on the representation of touchpoint positions. It provides standardized position input for subsequently mapping the online and offline touchpoint spatial positions to a gravity-constrained natural line-of-sight coordinate system.

[0063] Specifically, the platform invokes the campaign log storage module to extract complete campaign log data for each touchpoint within the current marketing cycle. This campaign log data includes basic data such as the campaign period, campaign duration, number of impressions, number of clicks, and number of users reached for each touchpoint. Simultaneously, the platform initiates a reach statistics analysis process to clean the aforementioned campaign log data, removing invalid data (such as duplicate impression records, abnormal click records, and reach records from non-target audiences) to obtain valid reach data. Using a weighted summation method, the number of impressions and the number of users reached are selected as core statistical indicators, and preset parameters are assigned to each indicator. The weights (with the number of impressions as the weight of 0.6 and the number of users reached as the weight of 0.4, and these preset weights can be adaptively adjusted according to the marketing scenario requirements) are first normalized to the number of impressions and the number of users reached for each touchpoint (normalized to the [0,1] interval). Then, the normalized number of impressions is multiplied by the corresponding weight, and the normalized number of users reached is multiplied by the corresponding weight to obtain the initial exposure weight of each touchpoint. Finally, the initial exposure weights of all touchpoints are summed, and the initial exposure weight of each touchpoint is divided by the total weight to obtain the final nominal exposure weight, so that the sum of the nominal exposure weights of all touchpoints is 1.

[0064] Specifically, touchpoint record vectors are constructed according to preset data encapsulation rules: For each offline touchpoint, a five-dimensional data vector is encapsulated in the order of vertical height difference, horizontal distance, 0, 0, and nominal exposure weight (where the normalized horizontal and vertical positions specific to online touchpoints are filled with 0 to distinguish offline touchpoint types); for each online touchpoint, a five-dimensional data vector is encapsulated in the order of 0, 0, normalized horizontal position, normalized vertical position, and nominal exposure weight (where the vertical height difference and horizontal distance specific to offline touchpoints are filled with 0 to distinguish online touchpoint types); after encapsulation, a unique touchpoint identifier label is added to each touchpoint record vector, forming a complete touchpoint record vector containing the touchpoint identifier and the five-dimensional data vector. This operation can integrate scattered heterogeneous online and offline touchpoint data into vector data with a unified structure, achieving standardized encapsulation of cross-type touchpoint data and ensuring the uniformity and efficiency of subsequent data processing workflows.

[0065] The line-of-sight statistics module is used to calculate the vertical line-of-sight deflection angle based on the touch point recording vector, and to calculate the robustness of the comfortable line-of-sight center angle and the vertical line-of-sight deflection angle based on the vertical line-of-sight deflection angle and the nominal exposure weight.

[0066] Specifically, the gaze statistics module transforms the raw data in the touchpoint recording vector, which only reflects geometric position and exposure intensity, into statistical quantities that characterize the natural gaze behavior of the target audience. The vertical gaze deviation angle quantifies the vertical deviation of each touchpoint relative to the natural gaze direction of the target audience. The comfortable gaze center angle reflects the vertical center position where the target audience's gaze is most likely to converge under the current nominal exposure structure. The robust dispersion of the vertical gaze deviation angle measures the overall dispersion of each touchpoint around this center angle and reduces the interference of extreme touchpoints. By performing unified calculations of these three quantities within this module, the platform can directly use the gaze center and dispersion to describe the matching relationship between touchpoints and natural gaze when subsequently calculating the gaze reachable contribution coefficient and exposure quota. This explicitly reflects the reliance on natural gaze patterns during the resource allocation phase.

[0067] In embodiments of the present invention, the vertical viewing angle is calculated based on the touch point recording vector, and the robust dispersion of the comfortable viewing center angle and the vertical viewing angle is calculated based on the vertical viewing angle and the nominal exposure weight, including:

[0068] Obtain the touch point recording vector for each touch point. The touch point recording vector includes the vertical height difference and horizontal distance of the offline touch points, the normalized horizontal position and normalized vertical position of the online touch points, and the nominal exposure weight.

[0069] The angle between the vertical height difference and the horizontal distance is calculated to obtain the vertical line-of-sight angle of the offline contact point;

[0070] The screen size and pixel density of the terminal are converted to obtain the in-screen normalized scale;

[0071] The angle between the normalized vertical position and the screen-in-screen normalized scale is calculated to obtain the vertical line-of-sight angle of the online touch point.

[0072] The comfortable viewing center angle is obtained by weighting the vertical viewing angle of each contact point with the nominal exposure weight;

[0073] The median absolute deviation of the vertical line of sight angle at each contact point is calculated to obtain the robust dispersion of the vertical line of sight angle at each contact point.

[0074] Specifically, the vertical height difference and horizontal distance data corresponding to each offline touchpoint are extracted from the constructed touchpoint recording vector. The arctangent function is used to calculate the angle between the vertical height difference and the horizontal distance of each offline touchpoint. The specific calculation logic is as follows: taking the vertical height difference as the opposite side of a right triangle and the horizontal distance as the adjacent side of a right triangle, the arctangent function is used to calculate the angle between the opposite side and the adjacent side of the right triangle. This angle is the vertical line-of-sight angle of the offline touchpoint. After the calculation is completed, the angle result is standardized and converted into angles. When the angle result is positive, it indicates that the offline touchpoint is above the comfortable line of sight of the target audience, and the user needs to look up to observe it. When the angle result is negative, it indicates that the offline touchpoint is below the comfortable line of sight of the target audience, and the user needs to look down to observe it.

[0075] Specifically, the terminal information storage module is invoked to extract the screen size parameters (specifically, the physical length of the screen in the vertical direction, in millimeters) and pixel density parameters (in pixels per inch) of the terminal to which each online touchpoint belongs. Then, two unit conversion processes are performed: First, the physical length of the screen in the vertical direction is converted from millimeters to inches, based on the formula 1 inch equals 25.4 millimeters, i.e., screen vertical physical length (inches) = screen vertical physical length (millimeters) / 25.4; Second, the total number of pixels in the vertical direction of the screen is calculated based on the pixel density, with the calculation logic being: total number of pixels in the vertical direction of the screen = screen vertical physical length (inches) × pixel density (pixels per inch); Finally, half of the total number of pixels in the vertical direction of the screen is used as the screen-intra-screen normalization scale, i.e., screen-intra-screen normalization scale = total number of pixels in the vertical direction of the screen / 2. This operation establishes a bridge between the physical size of the terminal screen and the pixel size, generating a unified scale benchmark. It transforms the normalized vertical position of the online touchpoint into an equivalent viewing offset angle, eliminating the influence of differences in screen size and pixel density between different terminals on the calculation of the viewing parameters of the online touchpoint.

[0076] Specifically, the normalized vertical position data of each online touchpoint is extracted from the touchpoint recording vector. The corresponding in-screen normalized scale data of the terminal is extracted from the above conversion steps to ensure a one-to-one correspondence between online touchpoints and in-screen normalized scales. The equivalent pixel offset of the online touchpoint in the vertical direction of the screen is calculated using the logic: Equivalent pixel offset = Normalized vertical position × In-screen normalized scale. This equivalent pixel offset represents the vertical pixel offset distance of the online touchpoint relative to the center of the screen. Next, the arctangent function is used to calculate the angle between the equivalent pixel offset and the in-screen normalized scale. Specifically, the equivalent pixel offset is considered as the opposite side of a right triangle, and the in-screen normalized scale as the adjacent side. The included angle is calculated using the arctangent function, and this included angle is the vertical viewing angle of the online touchpoint. Finally, the angle result is standardized and converted into angles. A positive angle result indicates that the online touchpoint is located above the center of the screen, requiring the user to look up; a negative angle result indicates that the online touchpoint is located below the center of the screen, requiring the user to look down.

[0077] Specifically, the vertical gaze angle data and corresponding nominal exposure weight data of all touchpoints (including online and offline touchpoints) are extracted to construct angle-weight data pairs. Data pairs with a nominal exposure weight of 0 or abnormal vertical gaze angles (such as exceeding ±90°, i.e., exceeding the normal human gaze deviation range) are removed. Then, a weighted summation operation is performed: for each valid data pair, the product of the vertical gaze angle and the corresponding nominal exposure weight is calculated, and the products of all data pairs are summed to obtain a weighted sum. Simultaneously, the nominal exposure weights of all valid data pairs are summed to obtain a weighted sum. Finally, the weighted sum is divided by the weighted sum to obtain the comfortable gaze center angle. This operation adaptively estimates the comfortable gaze reference position of the target audience based on the actual reach intensity of each touchpoint within the current marketing cycle, avoiding insufficient scene adaptability caused by using fixed empirical thresholds. It provides core reference parameters for subsequently judging the degree of gaze deviation at each touchpoint and calculating the gaze gravity mismatch.

[0078] Specifically, vertical line-of-sight angle data from all valid touchpoints are collected to form an angle dataset. Then, the median of this angle dataset is calculated: the angle data are sorted in ascending order; if the number of data points is odd, the middle data point is used as the median; if the number of data points is even, the arithmetic mean of the two middle data points is used as the median. Next, the absolute value of the difference between each angle data point and the median is calculated to obtain the absolute deviation dataset. Then, the median of the absolute deviation dataset is calculated to obtain the deviation median. Finally, the deviation median is multiplied by a robustness coefficient of 1.4826, and the result is the robust dispersion of the vertical line-of-sight angle. This operation quantifies the dispersion of the vertical line-of-sight angle distribution of all current touchpoints. Using the median absolute deviation calculation effectively resists interference from abnormal angle data, providing a dispersion parameter that reflects the true characteristics of the angle distribution for subsequent mismatch modulation scale calculations, ensuring the accuracy of the subsequent line-of-sight reach contribution coefficient calculation.

[0079] The reachability contribution module is used to calculate the reachability contribution coefficient of the line of sight based on the vertical line of sight deflection angle, the comfortable line of sight center angle, and the robust dispersion of the vertical line of sight deflection angle.

[0080] Specifically, the reachability contribution module transforms the line-of-sight statistics obtained in the previous stage into weighting factors that can directly participate in resource allocation calculations. The vertical line-of-sight deflection angle represents the degree of deviation of each touchpoint from the comfortable line-of-sight center angle, and the robustness dispersion of the vertical line-of-sight deflection angle represents the dispersion level of the overall line-of-sight distribution. Together, they determine the range that can be considered an area easily reached by the line of sight. By constructing a line-of-sight reachability contribution coefficient in this module, each touchpoint corresponds to a reachability weight that continuously changes with the degree of deviation. When allocating the total exposure quota, the subsequent exposure quota module can multiply the nominal exposure weight by the line-of-sight reachability contribution coefficient, thereby giving more exposure resources to touchpoints that are more in line with the natural line-of-sight pattern. In this way, the impact of line-of-sight gravity deflection mismatch on the resource allocation results is explicitly reflected in the calculation link without changing the data structure.

[0081] In embodiments of the present invention, the achievable contribution coefficient of the line of sight is calculated based on the vertical line of sight deflection angle, the comfortable line of sight center angle, and the robust dispersion of the vertical line of sight deflection angle, including:

[0082] Calculate the absolute difference between the vertical line-of-sight angle and the center angle of the comfortable line-of-sight at each contact point, and perform a weighted average calculation on the absolute difference based on the nominal exposure weight to obtain the line-of-sight gravity mismatch.

[0083] The mismatch modulation scale is obtained by adding the robust dispersion of the vertical line of sight deflection angle to the line of sight gravity skew mismatch.

[0084] Based on the mismatch modulation scale, the absolute difference between the vertical line-of-sight deflection angle of each contact point and the center angle of the comfortable line-of-sight is subjected to exponential decay mapping to obtain the line-of-sight reachable contribution coefficient.

[0085] Specifically, the vertical line-of-sight angles of all valid touchpoints and the adaptively estimated comfortable line-of-sight center angle are extracted from the output of the line-of-sight statistics module. Simultaneously, the nominal exposure weights corresponding to each valid touchpoint are extracted from the touchpoint recording vector, ensuring that the vertical line-of-sight angle, the comfortable line-of-sight center angle, and the nominal exposure weights are all associated with the same touchpoint identifier. For each valid touchpoint, the absolute difference between its vertical line-of-sight angle and the comfortable line-of-sight center angle is calculated. This absolute difference directly represents the degree of deviation of a single touchpoint from the target group's comfortable line-of-sight benchmark; the larger the difference, the farther the touchpoint deviates from the comfortable viewing area. Then, a weighted average calculation is performed: first, the absolute difference of each valid touchpoint is multiplied by the nominal exposure weight; then, the product of these products for all valid touchpoints is summed to obtain the weighted sum of differences; and finally, the nominal exposure weights of all valid touchpoints are summed to obtain the weighted sum. Finally, the weighted sum of differences is divided by the weighted sum to obtain the line-of-sight gravity mismatch, which ranges from 0° to 90°. A larger value indicates a more severe deviation of the overall exposure weights of all current touchpoints from the comfortable line-of-sight. By weighting the visual deviation of individual touchpoints using nominal exposure weights, the overall mismatch of multi-touchpoint exposure resources in the visual dimension is quantified, providing a core quantitative basis for the adjustment of subsequent resource allocation.

[0086] Specifically, the robust dispersion of the vertical viewing angle output by the viewing statistics module and the viewing gravity skew mismatch calculated above are extracted. Then, an addition operation is directly performed to add the values ​​of the robust dispersion and the viewing gravity skew mismatch to obtain the mismatch modulation scale. The distribution dispersion characteristics of the contact point viewing angle are fused with the overall mismatch characteristics of multiple contact points to form a unified modulation benchmark. The robust dispersion reflects the natural distribution differences of the viewing angles of each contact point, and the viewing gravity skew mismatch reflects the overall degree of deviation. The superposition of the two can enable the subsequent exponential decay mapping to adapt to the distribution differences of the contact point angles in different scenarios and respond to the severity of the overall mismatch, balancing the sensitivity and stability of resource allocation.

[0087] Specifically, the absolute difference between the vertical line-of-sight angle and the comfortable line-of-sight center angle of each effective contact point, as well as the newly obtained mismatch modulation scale, are extracted. Then, for each effective contact point, an exponential decay mapping function is constructed and calculated. The specific logic is as follows: using the natural constant e as the base, the ratio of the absolute difference of a single contact point to the mismatch modulation scale is calculated. The negative number of this ratio is used as the exponent, and the result of the exponential function is calculated, which is the line-of-sight reach contribution coefficient of the corresponding contact point. The core logic of the exponential decay mapping is: the smaller the absolute difference of a single contact point (i.e., the closer it is to the comfortable line of sight), the smaller the ratio of the absolute difference to the mismatch modulation scale, the closer its negative value is to 0, and the closer the exponential function result is to 1, indicating that the line-of-sight reach of that contact point is stronger and its contribution to effective attention is greater; conversely, the larger the absolute difference, the closer the exponential function result is to 0, indicating that the line-of-sight reach of that contact point is weaker and its contribution is smaller. Through this mapping operation, the range of the line-of-sight reach contribution coefficient strictly falls within the (0,1) interval, realizing the quantitative standardization of the reachability of contact points with different degrees of deviation.

[0088] The exposure quota module is used to obtain the total exposure quota for the current marketing cycle and determine the exposure quota for each touchpoint based on the total exposure quota, nominal exposure weight, and visibility reach contribution coefficient.

[0089] In embodiments of the present invention, the exposure quota for each touchpoint is determined based on the total exposure quota, nominal exposure weight, and line-of-sight reach contribution coefficient, wherein the formula for calculating the exposure quota is:

[0090] ;

[0091] In the formula, For the first Exposure quota per touchpoint Total exposure quota, For the first Nominal exposure weight per touchpoint For the first The line of sight at each contact point can contribute to the coefficient. For the set of contact points, For index variables, This is the result of summing the products of the nominal exposure weights and the line-of-sight reach contribution coefficients for all touchpoints in the touchpoint set;

[0092] Specifically, the total exposure quota for the current marketing cycle refers to the upper limit of the total amount of exposure resources that the digital integrated marketing service platform can allocate to all online and offline marketing touchpoints within a single continuous marketing time unit set by the platform. The exposure quota, on the other hand, refers to the amount of available exposure resources that the digital integrated marketing service platform allocates to a particular touchpoint within the current marketing cycle.

[0093] The calculation of exposure quotas is essentially a process of allocating resources according to weighted proportions under the premise of total conservation. The total exposure quota corresponds to the total amount of fixed exposure resources that can be allocated within a marketing cycle. The nominal exposure weight describes the basic importance of touchpoints in historical or planned campaigns. The visual reach contribution coefficient describes the effectiveness of touchpoints in aligning with the natural line of sight. The product of the two can be understood as the comprehensive contribution intensity of touchpoints to the overall exposure resources. Summing the contribution intensity of all touchpoints is equivalent to obtaining the total contribution of the entire touchpoint set. Then, the ratio of the contribution intensity of a single touchpoint to this total contribution is used as the allocation coefficient, so that the exposure quota of a touchpoint is equal to the total exposure quota multiplied by the allocation coefficient. This ensures that the sum of the exposure quotas of all touchpoints equals the total exposure quota, and that touchpoints with greater contributions receive higher exposure quotas. This weighted allocation method reflects an allocation rule that takes into account both the conservation of overall resources and the local contribution.

[0094] The instruction generation module is used to generate cross-channel delivery instruction sets based on exposure quotas;

[0095] In an embodiment of the present invention, a cross-channel delivery instruction set is generated based on the exposure quota, including:

[0096] Obtain the unit resource cost of each touchpoint, and based on the ratio of the exposure quota of each touchpoint to the unit resource cost, obtain the instruction quantity of each touchpoint;

[0097] A cross-channel delivery instruction set is constructed based on the instruction quantity of each touchpoint. The cross-channel delivery instruction set includes touchpoint identifiers and instruction quantities.

[0098] Specifically, unit resource cost refers to the cost required to execute a basic delivery unit for a specific touchpoint within its respective channel. This basic delivery unit can be a content display, a fixed-duration playback, or a response to an ad request. The instruction quantity is the number of executable delivery units obtained by dividing the exposure quota by the unit resource cost. It is used to guide the channel control system on how many basic delivery operations to actually execute at this touchpoint. The cross-channel delivery instruction set is a set of instructions formed by the platform after integrating the instruction quantities of all touchpoints, which is used to issue to different online and offline channel control systems. Each instruction consists of a touchpoint identifier and an instruction quantity. The touchpoint identifier is used to uniquely identify the identity of a specific touchpoint in the platform and its corresponding channel position. The instruction quantity is used to indicate the number of basic delivery units that the touchpoint should execute in the current marketing cycle, thereby realizing cross-channel unified scheduling and execution control based on exposure quotas.

[0099] Specifically, when obtaining the unit resource cost for each touchpoint, the unit resource cost is first defined as the cost required for a single touchpoint to provide one unit of exposure resource (such as the unit playback cost for offline touchpoints and the unit display cost for online touchpoints). Its unit of measurement must be consistent with the total exposure quota and the unit of the exposure quota. Data is collected through multi-channel collaboration. Online touchpoints connect with the data interface of their respective terminal applications (APP, information flow platform, etc.) to extract the original cost data recorded in their advertising backend that matches the exposure quota measurement dimension (such as the unit display cost when the exposure quota is the number of impressions). Offline touchpoints collect fixed costs (equipment rental costs, electricity costs, maintenance costs, etc.) and variable costs (linked to playback duration / number of impressions) provided by the operator through the platform's offline resource management module. Combined with the total exposure volume of the touchpoint in the historical period, the unit resource cost is calculated by dividing the sum of the fixed cost and the variable cost by the total historical exposure volume. Extract the exposure quota and corresponding unit resource cost of each valid touchpoint, establish a one-to-one correspondence through touchpoint identification, perform a division operation on each valid touchpoint, divide the exposure quota by the unit resource cost to obtain the initial instruction quantity, round the initial instruction quantity to the nearest integer, if the instruction quantity after rounding is 0 and the exposure quota is not 0, mark the touchpoint as a low cost-performance touchpoint, and reallocate its exposure quota to the same type of high cost-performance touchpoint (low unit resource cost, high line-of-sight reach contribution coefficient) and recalculate the instruction quantity of the relevant touchpoints. Collect the unique identifiers and corresponding integer instruction quantities of all valid touchpoints, construct a set of touchpoint identifier-instruction quantity data pairs, and then classify and sort them according to the hierarchical order of touchpoint type (online / offline), touchpoint platform / region, and instruction quantity size. Supplement with auxiliary information for delivery (such as peak time priority for offline touchpoints, matching user activity time periods for online touchpoints, content adaptation requirements such as online touchpoint material size specifications and offline touchpoint playback format requirements). Finally, combine the touchpoint identifier, instruction quantity, and auxiliary information and encapsulate them according to the platform's preset format (such as JSON or XML) to form a cross-channel delivery instruction set with core fields including the unique touchpoint identifier, execution instruction quantity, delivery time period, and content requirements, ensuring that online and offline delivery executors can directly parse and execute it.

[0100] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A digital integrated marketing service platform, characterized in that, include: The touch point recording vector module is used to calculate the touch point recording vector of each touch point in the touch point set. The touch point recording vector includes the vertical height difference, horizontal distance, normalized vertical position, normalized horizontal position, and nominal exposure weight. The line-of-sight statistics module is used to calculate the vertical line-of-sight deflection angle based on the touch point recording vector, and to calculate the robustness of the comfortable line-of-sight center angle and the vertical line-of-sight deflection angle based on the vertical line-of-sight deflection angle and the nominal exposure weight. The reachability contribution module is used to calculate the reachability contribution coefficient of the line of sight based on the vertical line of sight deflection angle, the comfortable line of sight center angle, and the robust dispersion of the vertical line of sight deflection angle. The exposure quota module is used to obtain the total exposure quota for the current marketing cycle and determine the exposure quota for each touchpoint based on the total exposure quota, nominal exposure weight, and visibility reach contribution coefficient. The instruction generation module is used to generate cross-channel delivery instruction sets based on exposure quotas.

2. The digital integrated marketing service platform according to claim 1, characterized in that, Calculate the contact record vector for each contact in the contact set, including: Obtain a set of touchpoints, which includes offline touchpoints and online touchpoints; Acquire data on the installation height and horizontal position of offline touchpoints, as well as the average eye height of the target population; The vertical height difference of the offline contact point is calculated based on the installation height and the average eye height data of the target population. The horizontal distance to the offline contact point is calculated based on the horizontal position and the reference line-of-sight position of the target group; Obtain the in-screen layout parameters of the online touchpoints, and normalize the in-screen layout parameters based on the terminal screen size and pixel density to obtain the normalized horizontal and vertical positions of the online touchpoints. Nominal exposure weights are generated based on campaign logs and reach statistics; The vertical height difference of the lower touch point, the horizontal distance of the lower touch point, the normalized horizontal position of the upper touch point, the normalized vertical position of the upper touch point, and the nominal exposure weight are used to form the touch point recording vector.

3. The digital integrated marketing service platform according to claim 1, characterized in that, Calculate the vertical line-of-sight angle based on the contact point recording vector, including: Obtain the touch point recording vector for each touch point. The touch point recording vector includes the vertical height difference and horizontal distance of the offline touch points, the normalized horizontal position and normalized vertical position of the online touch points, and the nominal exposure weight. The angle between the vertical height difference and the horizontal distance is calculated to obtain the vertical line-of-sight angle of the offline contact point; The screen size and pixel density of the terminal are converted to obtain the in-screen normalized scale; The angle between the normalized vertical position and the screen normalized scale is calculated to obtain the vertical line-of-sight angle of the online touch point.

4. The digital integrated marketing service platform according to claim 1, characterized in that, Calculate the comfortable line-of-sight center angle, including: The comfortable viewing center angle is obtained by weighting the vertical viewing angle of each contact point with the nominal exposure weight.

5. The digital integrated marketing service platform according to claim 1, characterized in that, Calculate the robust dispersion of the vertical line-of-sight deflection angle, including: The median absolute deviation of the vertical line-of-sight angle at each contact point is calculated to obtain the robust dispersion of the vertical line-of-sight angle at each contact point.

6. The digital integrated marketing service platform according to claim 1, characterized in that, The reachability contribution coefficient is calculated based on the vertical line-of-sight deflection angle, the comfortable line-of-sight center angle, and the robust dispersion of the vertical line-of-sight deflection angle, including: Calculate the absolute difference between the vertical line-of-sight angle and the center angle of the comfortable line-of-sight at each contact point, and perform a weighted average calculation on the absolute difference based on the nominal exposure weight to obtain the line-of-sight gravity mismatch. The mismatch modulation scale is obtained by adding the robust dispersion of the vertical line of sight deflection angle to the line of sight gravity skew mismatch. Based on the mismatch modulation scale, the absolute difference between the vertical line-of-sight deflection angle of each contact point and the center angle of the comfortable line-of-sight is subjected to exponential decay mapping to obtain the line-of-sight reachable contribution coefficient.

7. The digital integrated marketing service platform according to claim 1, characterized in that, The exposure quota for each touchpoint is determined based on the total exposure quota, nominal exposure weight, and line-of-sight reach contribution coefficient. The formula for calculating the exposure quota is as follows: ； In the formula, For the first Exposure quota per touchpoint Total exposure quota, For the first Nominal exposure weight per touchpoint For the first The line of sight at each contact point can contribute to the coefficient. For the set of contact points, For index variables, This is the result of summing the products of the nominal exposure weights and the reachable contribution coefficients of all touchpoints in the touchpoint set.

8. The digital integrated marketing service platform according to claim 1, characterized in that, Generate a cross-channel delivery instruction set based on the exposure quota, including: Obtain the unit resource cost of each touchpoint, and based on the ratio of the exposure quota of each touchpoint to the unit resource cost, obtain the instruction quantity of each touchpoint; A cross-channel delivery instruction set is constructed based on the instruction quantity of each touchpoint. The cross-channel delivery instruction set includes touchpoint identifiers and instruction quantities.

Images