Intelligent advertisement playing terminal and carousel control method thereof

Abstract

The invention relates to the technical field of digital advertisement putting, in particular to an intelligent advertisement playing terminal and a carousel control method thereof, and the method comprises the steps: collecting real-time environment parameters and real-time crowd characteristic parameters of an area where the advertisement playing terminal is located; inputting the attention concentration degree and the environment interference degree into a preset carousel strategy model, calculating the dynamic playing weight of each to-be-played advertisement in an advertisement library, and generating an initial carousel list according to the weight from high to low; acquiring historical playing data of each advertisement in the initial carousel list, optimizing the initial carousel list according to the historical playing data, generating a target carousel list, and controlling the terminal to play; and collecting playing effect data of the current carousel period, and adjusting parameters of the carousel strategy model according to the playing effect data. According to the invention, the pertinence and real-time response capability of advertisement playing are improved, and the dynamic optimization and adaptive adjustment of the carousel strategy are realized.

Description

Technical Field

[0001] This invention relates to the field of digital advertising delivery technology, and in particular to an intelligent advertising playback terminal and its carousel control method. Background Technology

[0002] The current outdoor advertising landscape suffers from numerous limitations, primarily stemming from the limited functionality of advertising display terminals and the rigidity of their rotation control strategies. Existing terminals mostly employ a linear, looping playback of pre-set lists, playing ads in a fixed order and with fixed weights regardless of whether anyone is watching or what the audience's characteristics are. This blind playback model leads to a severe disconnect between ad content and the current audience; high-value ads may idle during off-peak hours, while ads matching audience interests fail to receive priority display, resulting in low screen resource utilization. Furthermore, advertising display terminals lack the ability to differentiate their response to various advertising needs. Small-scale, targeted advertising for individual users, regional coverage for small and medium-sized businesses, and comprehensive promotional campaigns for large enterprises cannot be differentiated at the terminal level through adaptive rotation strategies. The diverse needs of different users cannot be fairly and efficiently guaranteed at the terminal level. In addition, advertising display terminals cannot collect audience interaction data, making it difficult to quantify and evaluate campaign effectiveness. Refined models based on second-level pricing for playback duration and week-level pricing for campaign cycles lack a data foundation. Poor coordination between the terminal and the back-end system, lack of closed-loop management from demand issuance to effect feedback, difficulty in ensuring playback stability during network interruptions, and inability to promptly self-heal or report fault conditions. Therefore, there is an urgent need for an intelligent advertising playback terminal and its rotation control method capable of environmental perception, dynamic rotation, data feedback, and strategy self-adaptation to improve advertising resource utilization and delivery effectiveness. Summary of the Invention

[0003] To address this issue, the present invention provides an intelligent advertising playback terminal and its carousel control method, which overcomes the problems in the prior art where the lack of a dynamic carousel adjustment mechanism based on real-time audience characteristics leads to low matching between advertising playback and audience and low screen resource utilization.

[0004] To achieve the above objectives, the present invention provides a method for controlling the carousel playback of intelligent advertising terminals, comprising: Step S1: Collect real-time environmental parameters and real-time crowd characteristic parameters of the area where the advertising playback terminal is located. The real-time environmental parameters include at least the ambient illuminance index and the ambient decibel index. The real-time crowd characteristic parameters include at least the number of people standing, the age distribution ratio, and the average gaze duration. Step S2: Calculate the attention concentration level for the current time period based on the real-time crowd characteristic parameters, and calculate the environmental interference level for the current time period based on the real-time environmental parameters; Step S3: Input the attention concentration and the environmental interference into the preset carousel strategy model, calculate the dynamic playback weight of each advertisement to be played in the advertisement library, and generate an initial carousel list according to the weight from high to low. Step S4: Obtain historical playback data of each advertisement in the initial carousel list. The historical playback data includes historical average viewing time and historical interaction rate. Optimize the initial carousel list based on the historical playback data to generate a target carousel list and control the terminal playback. Step S5: Collect playback effect data for the current carousel cycle. Based on the comparison results between the playback effect data and historical playback effect data, adjust the parameters of the carousel strategy model, including increasing the preset intervention intensity coefficient of the optimization step in step S4, or adjusting the preset number of candidate ad sets.

[0005] Furthermore, the attention concentration in step S2 is determined by the gaze duration coefficient and the number of people coefficient, wherein the gaze duration coefficient is the ratio of the average gaze duration to the preset gaze duration, and the number of people coefficient is the ratio of the number of people standing to the historical peak number of people standing.

[0006] Furthermore, the environmental interference level in step S2 is determined by the luminance interference coefficient and the volume interference coefficient, wherein the luminance interference coefficient is positively correlated with the ambient illuminance index, and the volume interference coefficient is positively correlated with the ambient decibel index.

[0007] Furthermore, the process of obtaining the dynamic playback weight in step S3 includes: Obtain the tag attributes of each advertisement in the advertisement library, including the appropriate age range and the appropriate scene tag; Calculate the age matching degree based on the age distribution ratio and the matching degree between each advertisement's target age range; Calculate the scene matching degree based on the matching rules between the environmental interference degree and the scene matching tags of each advertisement; The age matching degree, the scene matching degree, and the attention concentration degree are weighted and summed to obtain the dynamic playback weight of each advertisement.

[0008] Furthermore, the process of generating the target carousel list in step S4 includes: Obtain a preset number of advertisements from the initial carousel list before sorting, and form a candidate advertisement set; Calculate the gaze deviation between the historical average gaze duration and the preset gaze duration for each advertisement in the candidate advertisement set; Calculate the interaction deviation between the historical average interaction rate and the preset interaction rate for each advertisement in the candidate advertisement set; The gaze deviation value and the interaction deviation value are corrected by a preset intervention intensity coefficient; The advertisements in the candidate ad set are reordered based on the corrected gaze deviation value and the interaction deviation value to generate the target carousel list.

[0009] Furthermore, the method for reordering based on gaze deviation value and interaction deviation value in step S4 includes: When the gaze deviation value is greater than the preset deviation value and the interaction deviation value is greater than the preset deviation value, the original order of the advertisement in the candidate advertisement set is maintained; When the gaze deviation value is greater than the preset deviation value but the interaction deviation value is less than the preset deviation value, the advertisement is moved backward by a preset number of positions in the candidate advertisement set. When the gaze deviation value is less than the preset deviation value, the advertisement is removed from the candidate advertisement set.

[0010] Furthermore, adjusting the model parameters based on the playback effect data in step S5 includes: Calculate the rate of change of the average fixation duration in the current carousel cycle compared to the average fixation duration in the previous carousel cycle; Calculate the rate of change of the average interaction rate in the current carousel cycle compared to the average interaction rate in the previous carousel cycle; When the rate of change of gaze duration is less than the preset rate of change of gaze duration or the rate of change of interaction rate is less than the preset rate of change of interaction rate, it is determined that the playback effect does not meet expectations, and the preset intervention intensity coefficient of the optimization step in step S4 is increased. When the rate of change of gaze duration is greater than or equal to the preset rate of change of gaze duration and the rate of change of interaction rate is greater than or equal to the preset rate of change of interaction rate, it is determined that the playback effect has reached the expectation, and the preset number of candidate ad sets is reduced.

[0011] Furthermore, the method for increasing the preset intervention intensity coefficient in the optimization step S4 includes: Calculate the absolute value of the difference between the fixation duration change rate and the preset fixation duration change rate, and record it as the fixation duration change rate difference; Calculate the absolute value of the difference between the interaction rate change rate and the preset interaction rate change rate, and record it as the interaction rate change rate difference. The maximum value between the difference in the rate of change of gaze duration and the difference in the rate of change of interaction rate is taken as the degree of effect deviation; The adjustment range of the preset intervention intensity coefficient is positively correlated with the degree of effect deviation.

[0012] Furthermore, the reduction in the preset number of candidate advertisements is negatively correlated with the degree of effect deviation.

[0013] The present invention also provides an intelligent advertising playback terminal, comprising: The multimodal sensing module is used to collect environmental parameters and crowd characteristic parameters of the area where the terminal is located in real time; A carousel strategy generation module, which is connected to the multimodal perception module, is used to obtain attention concentration and environmental interference based on the real-time crowd feature parameters and the real-time environment parameters, input the attention concentration and environmental interference into a preset carousel strategy model to generate an initial carousel list, and optimize the initial carousel list based on the historical playback data of each advertisement to generate a target carousel list. A playback control module, which is connected to the carousel strategy generation module, is used to control the terminal to play and switch advertisements in sequence according to the target carousel list; The effect feedback module is connected to the playback control module and the carousel strategy generation module respectively. It is used to collect the playback effect data of the current carousel cycle and adjust the parameters of the carousel strategy model based on the comparison results of the playback effect data and the historical playback effect data. The advertising placement management module is connected to the playback control module and the effect feedback module, respectively, and is used to enable advertisers to place ads independently, manage ad removal and removal, monitor playback, and settle fees.

[0014] Furthermore, the advertising delivery management module includes: The terminal binding unit is used to receive binding requests initiated by advertisers by scanning the QR code of the advertising machine terminal, and to establish the association between the advertiser and the target advertising machine terminal. The advertising delivery unit is used to receive advertising materials uploaded by advertisers and configure the delivery time period. After successful payment, the advertising materials are sent to the target advertising machine terminal according to the delivery time period, and the advertising materials are automatically removed after the delivery time period expires. The order settlement unit is used to generate the order amount based on the advertiser's selected ad placement period and duration, complete the online payment, and generate a detailed expense statement and electronic invoice based on the actual ad playback data. The operation monitoring unit is used to obtain the advertising playback status information of the target advertising machine terminal in real time and push the playback status information to the advertiser's mobile terminal.

[0015] Compared with existing technologies, the beneficial effects of this invention are that it can accurately quantify the attention concentration and environmental interference of the current time period by collecting multimodal perception data such as the ambient illuminance index, ambient decibel index, number of people standing around, age distribution, and average gaze duration in the area where the advertising playback terminal is located in real time. Based on this, the above real-time feature parameters are input into the carousel strategy model to calculate the dynamic playback weight of each advertisement to be played. This allows the advertisement sorting to adaptively match the audience state and environmental conditions of the current scene, avoiding the defects of traditional carousel methods where the advertisement content is out of touch with the on-site audience and the playback strategy is rigid and fixed. This significantly improves the real-time performance, targeting, and scene adaptability of advertising.

[0016] Furthermore, by comparing the average gaze duration change rate and average interaction rate change rate of the current carousel cycle with those of the previous carousel cycle, the system can automatically determine whether the playback effect meets expectations. When the effect does not meet expectations, the system calculates the degree of effect deviation and increases the preset intervention intensity coefficient accordingly to strengthen the correction effect of historical high-quality content on the carousel ranking; when the effect meets or exceeds expectations, the preset number of candidate ad sets is reduced to simplify calculations and improve carousel efficiency. Attached Figure Description

[0017] Figure 1 This is a flowchart of the intelligent advertising playback terminal carousel control method according to an embodiment of the present invention; Figure 2 This is a flowchart illustrating the process of obtaining dynamic playback weights in an embodiment of the present invention; Figure 3 A flowchart for generating a target carousel list according to an embodiment of the present invention; Figure 4 This is a schematic diagram of the module connection of the intelligent advertising playback terminal according to an embodiment of the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0019] It should be noted that the data in this embodiment are all derived from a comprehensive analysis and evaluation of historical test data and corresponding historical test results from the three months prior to this test. Those skilled in the art will understand that the determination of the above-mentioned parameters for any single item in this invention can be achieved by selecting the value with the highest percentage based on the data distribution as the preset standard parameter, using weighted summation to obtain the value as the preset standard parameter, substituting each historical data point into a specific formula and using the value obtained from that formula as the preset standard parameter, or other selection methods, as long as the invention can clearly define different specific situations in the single-item judgment process through the obtained values.

[0020] Please see Figure 1 The diagram shows a flowchart of a smart advertising playback terminal carousel control method according to an embodiment of the present invention; the smart advertising playback terminal carousel control method according to an embodiment of the present invention includes: Step S1: Collect real-time environmental parameters and real-time crowd characteristic parameters of the area where the advertising playback terminal is located. The real-time environmental parameters include at least the ambient illuminance index and the ambient decibel index. The real-time crowd characteristic parameters include at least the number of people standing, the age distribution ratio, and the average gaze duration. Step S2: Calculate the attention concentration level for the current time period based on the real-time crowd characteristic parameters, and calculate the environmental interference level for the current time period based on the real-time environmental parameters; Step S3: Input the attention concentration and the environmental interference into the preset carousel strategy model, calculate the dynamic playback weight of each advertisement to be played in the advertisement library, and generate an initial carousel list according to the weight from high to low. Step S4: Obtain historical playback data of each advertisement in the initial carousel list. The historical playback data includes historical average viewing time and historical interaction rate. Optimize the initial carousel list based on the historical playback data to generate a target carousel list and control the terminal playback. Step S5: Collect playback effect data for the current carousel cycle. Based on the comparison results between the playback effect data and historical playback effect data, adjust the parameters of the carousel strategy model, including increasing the preset intervention intensity coefficient of the optimization step in step S4, or adjusting the preset number of candidate ad sets.

[0021] Specifically, in step S1, the ambient illuminance index is obtained through the photosensitive sensor built into the terminal device, which represents the brightness around the playback screen and is used to determine whether it affects the visual presentation effect of the advertising content; the ambient decibel index is obtained through a microphone or noise sensor to collect the volume and noise level around the terminal, which is used to assess whether the current environment is quiet and thus determine the feasibility of playing the audio advertisement; the number of people stopping is counted by the terminal's camera combined with a human body detection algorithm to count the number of people who stay in front of the screen for more than a preset time (e.g., 3 seconds); the age distribution ratio is estimated by facial recognition technology or skeletal body shape analysis to obtain the age structure of the people stopping; the average gaze duration is calculated by eye tracking or head posture estimation technology to calculate the average duration for all people stopping to look at the screen.

[0022] Specifically, the attention concentration in step S2 is determined by a gaze duration coefficient and a number of people coefficient, where the gaze duration coefficient is the ratio of the average gaze duration to the preset gaze duration, and the number of people coefficient is the ratio of the number of people standing still to the historical peak number of people standing still. The attention concentration is calculated using the following formula: , In the formula, For attention concentration, As the first weighting coefficient, This is the fixation duration coefficient. This is the second weighting coefficient. The first weighting coefficient is 0.6, the preset gaze duration is 5s, and the second weighting coefficient is 0.4.

[0023] Understandably, the gaze duration coefficient reflects the depth of an individual's attraction to the advertising content. In an advertising scenario, the longer a pedestrian's gaze lingers on the screen after stopping, the more effectively the advertising content captures that individual's visual focus and maintains their interest. The headcount coefficient reflects the breadth of the advertisement's influence during the current time period. The more people stop to watch, the wider the reach of the advertisement at that time and location, indicating higher communication value. By weighting and integrating these two coefficients, we can comprehensively and objectively quantify the overall level of audience attention to the advertising content during the current time period, providing accurate data support for subsequent dynamic adjustments to advertising playback weights.

[0024] Specifically, the environmental interference level in step S2 is determined by a luminance interference coefficient and a volume interference coefficient, wherein the luminance interference coefficient is positively correlated with the ambient illuminance index, and the volume interference coefficient is positively correlated with the ambient decibel index. The environmental interference level is calculated using the following formula: ， In the formula, For environmental interference, The third weighting coefficient, The brightness interference coefficient, It is the fourth weighting coefficient. The volume interference coefficient is the ratio of the ambient illuminance index to the preset illuminance index, and the volume interference coefficient is the ratio of the ambient decibel index to the preset decibel index. In this invention, the third weighting coefficient is 0.5, the fourth weighting coefficient is 0.5, and the preset illuminance index is 1000 lux, which corresponds to the typical illuminance of a strong outdoor direct sunlight environment. When the ambient illuminance exceeds this value, the screen reflection is severe, and the visual interference is significantly enhanced. The preset decibel index is 70 dB, which corresponds to the average noise level in a busy commercial area or traffic artery. When the ambient decibel exceeds this value, the advertising audio content will be difficult to perceive clearly.

[0025] Understandably, the brightness interference coefficient, by quantifying the impact of ambient light on screen visibility, reflects the challenges faced by the visual presentation of advertisements; the volume interference coefficient, by quantifying the masking effect of ambient noise on sound propagation, reflects the obstacles faced by the audio reach of advertisements. By weighted and fused together, the overall interference level of the current environment on advertisement playback can be comprehensively and objectively assessed. This allows the carousel strategy model to dynamically adjust the playback weights of different types of advertisements in subsequent steps, thereby optimizing the playback effect.

[0026] Please see Figure 2 The diagram shows a flowchart of obtaining dynamic playback weights according to an embodiment of the present invention; the process of obtaining dynamic playback weights in step S3 includes: Step S31: Obtain the tag attributes of each advertisement in the advertisement library, wherein the tag attributes include the appropriate age range and the appropriate scene tag; Step S32: Calculate the age matching degree based on the matching degree between the age distribution ratio and the age range suitable for each advertisement; Step S33: Calculate the scene matching degree according to the matching rules between the environmental interference degree and the scene matching tags of each advertisement; Step S34: The age matching degree, the scene matching degree, and the attention concentration degree are weighted and summed to obtain the dynamic playback weight of each advertisement.

[0027] Specifically, the suitable age range is used to characterize the age group of the target audience of the advertisement, and the suitable scene tag is used to characterize the type of environment in which the advertisement is suitable to be played.

[0028] Specifically, the age matching degree is the proportion of people in the current crowd who are within the age range that the advertisement is suitable for. The higher the proportion, the greater the proportion of the current crowd who belong to the target audience of the advertisement, and the higher the age matching degree; conversely, the lower the matching degree.

[0029] Specifically, the matching rules are not limited to any particular type, and those skilled in the art can make adaptive adjustments according to actual application scenarios. As a preferred embodiment, the matching rules adopt a range-based determination method. That is, a suitable environmental interference range is pre-set for the appropriate scene tag of each advertisement. When the real-time calculated environmental interference falls within this range, it is determined to be a scene match, and the scene match score is set to a higher value; when the environmental interference exceeds this range, it is determined to be a scene mismatch or partial match, and the scene match score is reduced accordingly.

[0030] Specifically, the age matching degree, scene matching degree, and attention concentration of each advertisement are multiplied by their respective weighting coefficients, and the three products are added together. The sum is the dynamic playback weight of the advertisement in the current time period. In this embodiment of the invention, the weighting coefficient for the age matching degree is set to 0.3, the weighting coefficient for the scene matching degree is 0.3, and the weighting coefficient for the attention concentration is 0.4.

[0031] Understandably, this invention selects age matching, scene matching, and attention concentration as the core parameters for calculating dynamic playback weights, based on a comprehensive consideration of three dimensions in ad playback decisions: audience fit, environmental adaptability, and real-time attention intensity. Age matching assesses the relevance of the ad content to the current audience, scene matching determines the suitability of the current environment for ad playback, and attention concentration quantifies the audience's attention intensity at the current time. Weighting and integrating these three factors comprehensively reflects the overall playback value of each ad at the current moment.

[0032] Please see Figure 3 The diagram shows a flowchart of generating a target carousel list according to an embodiment of the present invention; the process of generating the target carousel list in step S4 includes: Step S41: Obtain a preset number of advertisements from the initial carousel list before sorting to form a candidate advertisement set; Step S42: Calculate the gaze deviation value between the historical average gaze duration and the preset gaze duration for each advertisement in the candidate advertisement set; Step S43: Calculate the interaction deviation between the historical average interaction rate and the preset interaction rate of each advertisement in the candidate advertisement set. Step S44: Correct the gaze deviation value and the interaction deviation value by using a preset intervention intensity coefficient; Step S45: Reorder the advertisements in the candidate advertisement set according to the corrected gaze deviation value and the interaction deviation value to generate the target carousel list.

[0033] Specifically, the fixation deviation value = historical average fixation duration / preset fixation duration, and the interaction deviation value = historical average interaction rate / preset interaction rate.

[0034] Specifically, the historical average gaze duration refers to the average gaze duration obtained from several past playbacks of the advertisement, used to characterize the long-term appeal level of the advertisement content to the audience; the historical average interaction rate refers to the proportion of viewers who interact with the advertisement by scanning the screen QR code or clicking the screen during the advertisement playback period. The historical average interaction rate = number of interaction behaviors during the advertisement playback period / number of people stopping during the advertisement playback period. Multiple clicks within a preset time window are only counted as one valid interaction behavior. The preset time window is, for example, 5 seconds, but is not specifically limited.

[0035] Specifically, the preset intervention intensity coefficient is used to control the degree of influence of historical playback data on the reordering process. The fixation deviation value and the interaction deviation value are multiplied by the preset intervention intensity coefficient to obtain the corrected fixation deviation value and the corrected interaction deviation value.

[0036] Specifically, for each ad in the candidate ad set, its corrected gaze deviation value and corrected interaction deviation value are summed to obtain a comprehensive deviation index. The candidate ads are then reordered according to their comprehensive deviation index from high to low, with higher-scoring ads appearing higher in the target carousel list.

[0037] In this embodiment of the invention, the preset gaze duration is 5 seconds, the preset interaction rate is 10%, and the preset intervention intensity coefficient is 0.8. However, the above values ​​are not limited to these, and those skilled in the art can adjust the above values ​​according to actual needs.

[0038] Specifically, the method for reordering based on gaze deviation value and interaction deviation value in step S4 includes: When the gaze deviation value is greater than the preset deviation value and the interaction deviation value is greater than the preset deviation value, the original order of the advertisement in the candidate advertisement set is maintained; When the gaze deviation value is greater than the preset deviation value but the interaction deviation value is less than the preset deviation value, the advertisement is moved backward by a preset number of positions in the candidate advertisement set. When the gaze deviation value is less than the preset deviation value, the advertisement is removed from the candidate advertisement set.

[0039] In this embodiment of the invention, the preset deviation value is 1, and the preset ranking value is 50% of the number of advertisements in the candidate advertisement set.

[0040] Understandably, view duration is a fundamental threshold for ad reach. If historical data shows that an ad fails to effectively attract viewer attention, the probability of subsequent interaction is extremely low, and continuing to keep it in the candidate set will crowd out playback resources for other high-quality ads. Through hard filtering, the candidate ad set can be purified, ensuring that all ads in the carousel have basic audience appeal.

[0041] Specifically, step S5, which involves adjusting the model parameters based on the playback effect data, includes: Calculate the rate of change of the average fixation duration in the current carousel cycle compared to the average fixation duration in the previous carousel cycle; Calculate the rate of change of the average interaction rate in the current carousel cycle compared to the average interaction rate in the previous carousel cycle; When the rate of change of gaze duration is less than the preset rate of change of gaze duration or the rate of change of interaction rate is less than the preset rate of change of interaction rate, it is determined that the playback effect does not meet expectations, and the preset intervention intensity coefficient of the optimization step in step S4 is increased. When the rate of change of gaze duration is greater than or equal to the preset rate of change of gaze duration and the rate of change of interaction rate is greater than or equal to the preset rate of change of interaction rate, it is determined that the playback effect has reached the expectation, and the preset number of candidate ad sets is reduced.

[0042] Specifically, the rate of change in gaze duration = (average gaze duration in the current cycle - average gaze duration in the previous cycle) / average gaze duration in the previous cycle, and the rate of change in interaction rate = (average interaction rate in the current cycle - average interaction rate in the previous cycle) / average interaction rate in the previous cycle.

[0043] In this embodiment of the invention, the preset gaze duration change rate is 2% and the preset interaction rate change rate is 2%, but the above values ​​are not limited to these. Those skilled in the art can adjust the above values ​​according to actual needs.

[0044] Understandably, the rates of change in gaze duration and interaction rate can be negative; gaze duration and interaction rate represent the attractiveness and conversion rate of an ad, respectively, and a decline in either dimension indicates a weakness in the current playback strategy. To ensure that the overall performance of the system does not show significant defects, an intervention enhancement mechanism is triggered if any indicator fails to reach the expected threshold. This mechanism increases the preset intervention intensity coefficient in step S4, strengthening the reliance on historical high-performing data to correct deviations in a timely manner. Conversely, when both indicators show positive growth, the system determines that the current optimization strategy is not only effective but also has significant improvement potential. Therefore, by reducing the preset number of candidate ad sets, the system narrows the selection range, improves computational efficiency, and allocates more resources to validated high-performing ads.

[0045] Specifically, the method for increasing the preset intervention intensity coefficient in the optimization step S4 includes: Calculate the absolute value of the difference between the fixation duration change rate and the preset fixation duration change rate, and record it as the fixation duration change rate difference; Calculate the absolute value of the difference between the interaction rate change rate and the preset interaction rate change rate, and record it as the interaction rate change rate difference. The maximum value between the difference in the rate of change of gaze duration and the difference in the rate of change of interaction rate is taken as the degree of effect deviation; The adjustment range of the preset intervention intensity coefficient is positively correlated with the degree of effect deviation.

[0046] In this embodiment of the invention, the adjustment range of the preset intervention intensity coefficient is 0.5 to 1.0. The adjustment method of the preset intervention intensity coefficient is not limited, as long as the adjustment range of the preset intervention intensity coefficient is positively correlated with the degree of effect deviation.

[0047] Specifically, the reduction in the preset number of candidate ads is negatively correlated with the degree of performance deviation.

[0048] In this embodiment of the invention, the initial preset quantity is not limited, and the reduction of the preset quantity is at most 20% of the initial preset quantity. The specific method of adjusting the preset quantity of the candidate advertisement set is not limited, as long as the reduced preset quantity is not less than 80% of the initial preset quantity, and the reduction magnitude is negatively correlated with the degree of effect deviation.

[0049] Please see Figure 4 The diagram shown is a schematic diagram of the module connection of the intelligent advertising playback terminal according to an embodiment of the present invention; the intelligent advertising playback terminal according to an embodiment of the present invention includes: The multimodal sensing module is used to collect environmental parameters and crowd characteristic parameters of the area where the terminal is located in real time; A carousel strategy generation module, which is connected to the multimodal perception module, is used to obtain attention concentration and environmental interference based on the real-time crowd feature parameters and the real-time environment parameters, input the attention concentration and environmental interference into a preset carousel strategy model to generate an initial carousel list, and optimize the initial carousel list based on the historical playback data of each advertisement to generate a target carousel list. A playback control module, which is connected to the carousel strategy generation module, is used to control the terminal to play and switch advertisements in sequence according to the target carousel list; The effect feedback module is connected to the playback control module and the carousel strategy generation module respectively. It is used to collect the playback effect data of the current carousel cycle and adjust the parameters of the carousel strategy model based on the comparison results of the playback effect data and the historical playback effect data. The advertising placement management module is connected to the playback control module and the effect feedback module, respectively, and is used to enable advertisers to place ads independently, manage ad removal and removal, monitor playback, and settle fees.

[0050] Specifically, the advertising delivery management module includes: The terminal binding unit is used to receive binding requests initiated by advertisers by scanning the QR code of the advertising machine terminal, and to establish the association between the advertiser and the target advertising machine terminal. The advertising delivery unit is used to receive advertising materials uploaded by advertisers and configure the delivery time period. After successful payment, the advertising materials are sent to the target advertising machine terminal according to the delivery time period, and the advertising materials are automatically removed after the delivery time period expires. The order settlement unit is used to generate the order amount based on the advertiser's selected ad placement period and duration, complete the online payment, and generate a detailed expense statement and electronic invoice based on the actual ad playback data. The operation monitoring unit is used to obtain the advertising playback status information of the target advertising machine terminal in real time and push the playback status information to the advertiser's mobile terminal.

[0051] Specifically, the advertising delivery management module also includes a profile analysis unit, which is connected to the multimodal perception module and the effect feedback module. It is used to receive real-time audience feature parameters uploaded by each advertising playback terminal, perform statistical analysis on the age distribution, gender distribution and dwell time of the audience in different terminals, different time periods and different regions, generate an audience profile analysis report, and push the audience profile analysis report to the mobile terminal of the corresponding advertiser.

[0052] Specifically, the advertising placement management module also includes an effect evaluation unit, which is connected to the profile analysis unit and the order settlement unit respectively. It is used to calculate the input-output ratio of advertising placement based on the number of people stopping during the advertising playback, the average viewing time, the interaction rate and the actual advertising fee paid, generate an effect evaluation report and push it to the advertiser's mobile terminal.

[0053] Those skilled in the art should understand that the discussion of any of the above embodiments is merely exemplary and is not intended to imply that the scope of the invention (including the claims) is limited to these examples; within the framework of the invention, the technical features of the above embodiments or different embodiments can also be combined, the steps can be implemented in any order, and there are many other variations of the different aspects of the invention as described above, which are not provided in the details for the sake of brevity.

[0054] This invention is intended to cover all such substitutions, modifications, and variations that fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for controlling the carousel playback of intelligent advertising terminals, characterized in that, include: Step S1: Collect real-time environmental parameters and real-time crowd characteristic parameters of the area where the advertising playback terminal is located. The real-time environmental parameters include at least the ambient illuminance index and the ambient decibel index. The real-time crowd characteristic parameters include at least the number of people standing, the age distribution ratio, and the average gaze duration. Step S2: Calculate the attention concentration level for the current time period based on the real-time crowd characteristic parameters, and calculate the environmental interference level for the current time period based on the real-time environmental parameters; Step S3: Input the attention concentration and the environmental interference into the preset carousel strategy model, calculate the dynamic playback weight of each advertisement to be played in the advertisement library, and generate an initial carousel list according to the weight from high to low. Step S4: Obtain historical playback data of each advertisement in the initial carousel list. The historical playback data includes historical average viewing time and historical interaction rate. Optimize the initial carousel list based on the historical playback data to generate a target carousel list and control the terminal playback. Step S5: Collect playback effect data for the current carousel cycle. Based on the comparison results between the playback effect data and historical playback effect data, adjust the parameters of the carousel strategy model, including increasing the preset intervention intensity coefficient of the optimization step in step S4, or adjusting the preset number of candidate ad sets.

2. The intelligent advertising playback terminal carousel control method according to claim 1, characterized in that, The attention concentration in step S2 is determined by the gaze duration coefficient and the number of people coefficient. The gaze duration coefficient is the ratio of the average gaze duration to the preset gaze duration, and the number of people coefficient is the ratio of the number of people standing to the historical peak number of people standing.

3. The intelligent advertising playback terminal carousel control method according to claim 2, characterized in that, The environmental interference level in step S2 is determined by the luminance interference coefficient and the volume interference coefficient, wherein the luminance interference coefficient is positively correlated with the ambient illuminance index and the volume interference coefficient is positively correlated with the ambient decibel index.

4. The intelligent advertising playback terminal carousel control method according to claim 3, characterized in that, The process of obtaining the dynamic playback weight in step S3 includes: Obtain the tag attributes of each advertisement in the advertisement library, including the appropriate age range and the appropriate scene tag; Calculate the age matching degree based on the age distribution ratio and the matching degree between each advertisement's target age range; Calculate the scene matching degree based on the matching rules between the environmental interference degree and the scene matching tags of each advertisement; The age matching degree, the scene matching degree, and the attention concentration degree are weighted and summed to obtain the dynamic playback weight of each advertisement.

5. The intelligent advertising playback terminal carousel control method according to claim 4, characterized in that, The process of generating the target carousel list in step S4 includes: Obtain a preset number of advertisements from the initial carousel list before sorting, and form a candidate advertisement set; Calculate the gaze deviation between the historical average gaze duration and the preset gaze duration for each advertisement in the candidate advertisement set; Calculate the interaction deviation between the historical average interaction rate and the preset interaction rate for each advertisement in the candidate advertisement set; The gaze deviation value and the interaction deviation value are corrected by a preset intervention intensity coefficient; The advertisements in the candidate ad set are reordered based on the corrected gaze deviation value and the interaction deviation value to generate the target carousel list.

6. The intelligent advertising playback terminal carousel control method according to claim 5, characterized in that, The method for reordering based on gaze deviation value and interaction deviation value in step S4 includes: When the gaze deviation value is greater than the preset deviation value and the interaction deviation value is greater than the preset deviation value, the original order of the advertisement in the candidate advertisement set is maintained; When the gaze deviation value is greater than the preset deviation value but the interaction deviation value is less than the preset deviation value, the advertisement is moved backward by a preset number of positions in the candidate advertisement set. When the gaze deviation value is less than the preset deviation value, the advertisement is removed from the candidate advertisement set.

7. The intelligent advertising playback terminal carousel control method according to claim 6, characterized in that, The step S5, which involves adjusting the model parameters based on the playback effect data, includes: Calculate the rate of change of the average fixation duration in the current carousel cycle compared to the average fixation duration in the previous carousel cycle; Calculate the rate of change of the average interaction rate in the current carousel cycle compared to the average interaction rate in the previous carousel cycle; When the rate of change of gaze duration is less than the preset rate of change of gaze duration or the rate of change of interaction rate is less than the preset rate of change of interaction rate, it is determined that the playback effect does not meet expectations, and the preset intervention intensity coefficient of the optimization step in step S4 is increased. When the rate of change of gaze duration is greater than or equal to the preset rate of change of gaze duration and the rate of change of interaction rate is greater than or equal to the preset rate of change of interaction rate, it is determined that the playback effect has reached the expectation, and the preset number of candidate ad sets is reduced.

8. The intelligent advertising playback terminal carousel control method according to claim 7, characterized in that, The methods for increasing the preset intervention intensity coefficient in the optimization step S4 include: Calculate the absolute value of the difference between the fixation duration change rate and the preset fixation duration change rate, and record it as the fixation duration change rate difference; Calculate the absolute value of the difference between the interaction rate change rate and the preset interaction rate change rate, and record it as the interaction rate change rate difference. The maximum value between the difference in the rate of change of gaze duration and the difference in the rate of change of interaction rate is taken as the degree of effect deviation; The adjustment range of the preset intervention intensity coefficient is positively correlated with the degree of effect deviation.

9. A smart advertising playback terminal, applied to the smart advertising playback terminal carousel control method according to any one of claims 1-8, characterized in that, include: The multimodal sensing module is used to collect environmental parameters and crowd characteristic parameters of the area where the terminal is located in real time; A carousel strategy generation module, which is connected to the multimodal perception module, is used to obtain attention concentration and environmental interference based on the real-time crowd feature parameters and the real-time environment parameters, input the attention concentration and environmental interference into a preset carousel strategy model to generate an initial carousel list, and optimize the initial carousel list based on the historical playback data of each advertisement to generate a target carousel list. A playback control module, which is connected to the carousel strategy generation module, is used to control the terminal to play and switch advertisements in sequence according to the target carousel list; The effect feedback module is connected to the playback control module and the carousel strategy generation module respectively. It is used to collect the playback effect data of the current carousel cycle and adjust the parameters of the carousel strategy model based on the comparison results of the playback effect data and the historical playback effect data. The advertising placement management module is connected to the playback control module and the effect feedback module, respectively, and is used to enable advertisers to place ads independently, manage ad removal and removal, monitor playback, and settle fees.

10. The intelligent advertising playback terminal according to claim 9, characterized in that, The advertising delivery management module includes: The terminal binding unit is used to receive binding requests initiated by advertisers by scanning the QR code of the advertising machine terminal, and to establish the association between the advertiser and the target advertising machine terminal. The advertising delivery unit is used to receive advertising materials uploaded by advertisers and configure the delivery time period. After successful payment, the advertising materials are sent to the target advertising machine terminal according to the delivery time period, and the advertising materials are automatically removed after the delivery time period expires. The order settlement unit is used to generate the order amount based on the advertiser's selected ad placement period and duration, complete the online payment, and generate a detailed expense statement and electronic invoice based on the actual ad playback data. The monitoring unit is used to obtain the advertising playback status information of the target advertising machine terminal in real time and push the playback status information to the advertiser's mobile terminal.

Images