Advertisement accurate putting method based on rural new media

Abstract

The invention provides a precise advertisement putting method based on rural new media, and belongs to the technical field of rural new media advertisements. The method comprises the following specific steps: step 1, collecting user behavior data in each platform, and generating a first data set, a second data set and a third data set; step 2, acquiring an interaction coefficient # imgabs0 #, a public medium influence coefficient # imgabs1 # and a behavior-driven simulcast coefficient # imgabs2 #; step 3, obtaining a user social interaction cluster income coefficient # imgabs3 #, a public medium propaganda income coefficient # imgabs4 # and a simulcast income coefficient # imgabs5 #; and step 4, obtaining a first prediction transformation coefficient # imgabs6 #, a second prediction transformation coefficient # imgabs7 # and a third prediction transformation coefficient # imgabs8 #. According to the method, the advertisement conversion rate can be calculated, the advertisement watching duration of resident people can reach 9 seconds, the playing of auxiliary advertisements can be expanded, the advertisements can be accurately put, and the propaganda effect on villages is improved.

Description

Technical Field

[0001] The present invention relates to the field of rural new media advertising technology, and in particular to a method for precise advertising delivery based on rural new media. Background Art

[0002] With the development of the economy, the development of the rural economy and brand building have gradually become the focus of social attention. Advertising, as an important means to promote the display of rural agricultural products, has played a key role in rural development. In recent years, with the rapid popularization of mobile Internet and social platforms, rural advertising has gradually shifted from traditional offline flyers to new media platforms represented by social media, short videos, live broadcast platforms and content communities. This transformation not only expands the scope of advertising dissemination, but also improves the accuracy and interactivity of advertising in reaching users.

[0003] At present, rural agricultural product advertisements are promoted through various media platforms. Although the number of views, likes and shares can be obtained from various media platforms, these behaviors cannot accurately know the conversion rate of advertising investment. For rural agricultural product advertising, in order to increase the profitability of the countryside, it is necessary to know the conversion rate of advertisements placed on different media platforms. Only when the conversion rate is clear can the advertisements be placed accurately to obtain the maximum benefit. In addition, the content of rural advertisements is currently single and there is no additional suitable advertising content, which leads to limited benefits and the inability to expand corresponding rural advertisements.

[0004] In summary, it is particularly important to know the conversion rate of rural advertising on different platforms so that advertising can be delivered accurately. In addition, the expansion of rural advertising can increase revenue. Therefore, a method of accurate advertising delivery based on rural new media is proposed to solve this problem. Summary of the Invention

[0005] In order to make up for the above deficiencies, the present invention provides a method for precise advertising delivery based on rural new media that overcomes the above technical problems or at least partially solves the above problems.

[0006] The present invention is achieved in that: The present invention provides a method for accurately delivering advertisements based on rural new media, and the specific steps include: Step 1: Build a multi-type target rural new media advertising platform, place the target rural advertisements on the new media platform, and collect user behavior data from each platform to generate the first data set, the second data set, and the third data set; Step 2: Establish a user behavior model and input the first, second, and third data sets into the analysis to obtain the interaction coefficients , public media influence coefficient and behavior-driven hookup coefficient ; Step 3: Establish a user analysis model. After inputting data analysis, continue to collect advertising conversion rates on each platform to obtain the user social interaction cluster profit coefficient. , Public media publicity income coefficient and simulcast revenue coefficient ; Step 4: The interaction coefficient Cluster profit coefficient of social interaction with users The public media influence coefficient Coefficient of return on public media publicity Associated with the simulcast revenue coefficient and behavior-driven hookup coefficients Associated, the first predicted conversion coefficient is obtained respectively , the second predicted conversion coefficient and the third predicted conversion coefficient , evaluate them respectively, and generate corresponding advertising delivery strategies.

[0007] In a preferred embodiment, step one comprises: S11. Use Python to obtain the number of times the i-th user sends the target village-related words in the user social interaction cluster. , using the user social interaction cluster open platform to detect the number of message forwarding, detect the number of times the i-th user forwards the message, and obtain the number of times the i-th user forwards the message , using the timestamp in the user social interaction cluster, record the time when each message of the i-th user is sent, and use the time difference to obtain the interval time between the i-th user's message replies By setting different page number links on the chat page and displaying different page number links, we can get the page browsing times of the i-th user. , establish the first data set; S12: Establish a target area at a position 1m-6m in front of the target display in the park, and collect the number of people in the target area by installing an AI camera on the target display to obtain the number of people in the target area. , and based on the AI ​​camera combined with the time difference, obtain the residence time of people in the target area , through the QR code platform to count the number of scans in real time, and obtain the number of scans , establish a second data set; S13, using a voice-to-text converter to recognize and convert the voice of the video of the initial rural advertisement into text, and extract the number of keywords from the text , use a timer to detect the time when the resident crowd watches the first rural advertisement and obtain the viewing time , establish the third data set.

[0008] In a preferred embodiment, step 2 includes: S21, for extracting from the first data set the number of times the i-th user sends the target village-related words in the user social interaction cluster , the number of times the i-th user forwards the message The interval between the reply to the message of the i-th user , the number of page views by the i-th user , and the interaction coefficient is obtained by calculating .

[0009] In a preferred embodiment, step 2 further comprises: S22, for extracting the number of people in the target area from the second data set , the length of time people stay in the target area and number of scans , and calculate the public media influence coefficient .

[0010] In a preferred embodiment, step 2 further comprises: S23, by presetting the public media influence threshold A, the public media influence threshold A and the public media influence coefficient Compare and obtain the first phase evaluation results, including: when When the value is greater than A, it indicates that the results of rural advertising on public media are satisfactory, and rural advertising on public media should continue to be placed; when When ≤A, it means that the results of rural advertising on public media are unqualified, and the first correction strategy is generated, including: guiding the number of QR code scans in the video to increase by 10%-20%, and increasing the frequency of the detected people's residence time in the target area by 5%-8%. When it is detected that the people stay in the target area for more than 9s, the play instruction is triggered, and the rural advertising content is adjusted to increase the attraction rate by 10%-30%.

[0011] In a preferred embodiment, the play instruction includes: Used to extract the number of keywords from the third dataset and viewing time , and combined with the number of scans , the behavior-driven simulcast coefficient is obtained by calculating ; By presetting the behavior-driven simulcast threshold Z and combining the behavior-driven simulcast threshold Z with the behavior-driven simulcast coefficient Compare and generate the second-stage evaluation results, including: when >Z, indicating that the user behavior triggered by the target rural advertisement meets the triggering conditions, which in turn leads to the playback of the secondary advertisement. When the expected result of the secondary advertisement is greater than the preset evaluation threshold, the system automatically executes the continuous playback of the secondary advertisement; when ≤Z indicates that although the current rural advertisement leads to a sub-advertisement, the expected result of the sub-advertisement is not up to standard. A second correction strategy is generated, including: increasing the content of the sub-advertisement to increase its relevance to the content of the initial rural advertisement by 70%-80%, adjusting the jump logic of the sub-advertisement, optimizing the code scanning path, reducing the process by 20%-25%, and improving the accuracy of keyword recognition for the initial rural advertisement by 80%-90%. The sub-advertisement used for jump should echo the initial rural advertisement.

[0012] In a preferred embodiment, step three includes: S31, used to calculate the transaction amount of the i-th user on the user social interaction cluster platform Collect and build a data set of total revenue of user social interaction cluster platform; S32, used to calculate the transaction amount of the i-th user on the public media promotion platform Collect and construct a data set of total revenue of public media platforms; S33, used to calculate the transaction volume generated by behavior-driven simulcasting Collect and build a data set of behavior-driven simulcast revenue; S33, based on the transaction amount of the i-th user in the user social interaction cluster total income dataset on the live broadcast platform , calculate the user social interaction cluster platform profit coefficient ; S34, based on the transaction amount of the i-th user in the live broadcast platform in the public media total revenue dataset , calculate the public media revenue coefficient ; S34. Extracting transaction amounts from behavior-driven simulcast revenue data sets , calculate the simulcast revenue coefficient .

[0013] In a preferred embodiment, step 4 includes: S41, the interaction coefficient Cluster profit coefficient of social interaction with users The first predicted conversion coefficient is calculated by the following formula ; S42, for presetting a user social interaction cluster platform income threshold B, and comparing the user social interaction cluster platform income threshold B with the second predicted conversion coefficient Perform a comparison to generate a second evaluation result, including: when When the value is greater than B, it indicates that the conversion rate of the target rural advertising investment on the user social interaction cluster platform is qualified, and the current advertising volume in the user social interaction cluster is maintained; when When ≤B, it means that the conversion rate of the target rural advertising investment in the user social interaction cluster platform is less than the preset conversion rate, and the conversion rate is unqualified. The second strategy is generated, including: adjusting the video content, optimizing the existing advertising content, introducing distinctive rural plantations, increasing the number of users attracted by 20%-30%, taking advantage of time periods with frequent social interactions to increase the amount of advertising by 50%-60%, and reducing the amount of advertising by 20%-40% during time periods with sparse social interactions.

[0014] In a preferred embodiment, step 4 further comprises: S43. The public media influence coefficient Coefficient of return on public media publicity The second predicted conversion coefficient is calculated by the following formula ; S44, for presetting a public media platform revenue threshold C, and comparing the public media platform revenue threshold C with the third predicted conversion coefficient Perform comparisons to generate third evaluation results, including: when When it is greater than C, it means that the conversion rate of the target rural advertising investment on the public media platform is qualified, and the current amount of advertising investment on the public media platform is maintained; when When ≤C, it means that the conversion rate of the target rural advertising investment on the live broadcast platform is lower than the expected conversion rate, and the conversion rate is unqualified. The third strategy is generated, including: reducing the frequency of target rural advertising on public media platforms by 15%-35%, and eliminating user groups with a click-through rate below 30% of the average level, increasing advertising content, combining rural geographical characteristics and characteristic agricultural products, launching novel advertising content, and increasing the dissemination rate by 40%-50%.

[0015] In a preferred embodiment, step 4 further comprises: S45, will be broadcast revenue coefficient and behavior-driven hookup coefficients The third predicted conversion coefficient is obtained by the following formula ; S46: By presetting the behavior-driven simulcast revenue threshold V, the behavior-driven simulcast revenue threshold V is combined with the third predicted conversion coefficient. Perform a comparison to generate a fourth evaluation result, including: when When it is greater than V, it indicates that the expected revenue generated by the simulcast is qualified, and the secondary advertisements and monitoring will continue; when When ≤V, it indicates that the expected revenue generated by the simulcast is not up to standard, and the fourth strategy is generated, including: adjusting the content of the secondary advertisements, increasing the amount of content related to the initial rural advertisements by 30%-50%, reducing the delivery range of secondary advertisements by 10%-15%, and accurately delivering them to the people in need, increasing the exposure ratio to high-potential groups by 10%-20%.

[0016] The method for precise advertising delivery based on rural new media provided by the present invention has the following beneficial effects: 1. Through precise data analysis and behavioral modeling, efficient matching of advertising and optimal allocation of resources are achieved, which significantly improves the conversion rate and return on investment of advertising, helping local businesses to achieve higher profits at a lower cost. In addition, the present invention establishes a multi-factor prediction conversion model to associate user interaction with revenue results, media influence with media returns, and simulcast behavior with revenue feedback, forming the first, second, and third predicted conversion coefficients, realizing the performance evaluation of various advertising delivery nodes, and then driving the automatic optimization of strategies, improving the efficiency of advertising budget utilization, and accurately stimulating users' attention by detecting the resident population in the initial advertisement and triggering the video playback instruction when the resident time exceeds 9 seconds, effectively promoting the promotion of rural agricultural products.

[0017] 2. By building multiple types of rural new media advertising platforms, including social interaction cluster platforms, public media platforms and video syndicated platforms, we can achieve multi-dimensional delivery and coverage of target rural advertising content, enhance the depth and breadth of advertising content dissemination, and significantly improve user contact frequency and interaction rate. We will use three types of data sets (user interaction behavior data, public media response data, and syndicated behavior data) to establish a user behavior modeling system, accurately quantify the interaction coefficient, public media influence coefficient and behavior-driven syndicated coefficient, provide high-quality input for subsequent conversion rate prediction, and improve the scientific nature and traceability of behavior judgment. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. It should be understood that the following drawings only illustrate certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without paying any creative work.

[0019] Figure 1 It is a flow chart provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0020] To make the purpose, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0021] Example 1, with reference to Figure 1 The present invention provides a technical solution: a method for accurately delivering advertisements based on rural new media, the specific steps of which include: Step 1: Build a multi-type target rural new media advertising platform, place the target rural advertisements on the new media platform, and collect user behavior data from each platform to generate the first data set, the second data set, and the third data set; Step 2: Establish a user behavior model and input the first, second, and third data sets into the analysis to obtain the interaction coefficients , public media influence coefficient and behavior-driven hookup coefficient ; Step 3: Establish a user analysis model. After inputting data analysis, continue to collect advertising conversion rates on each platform to obtain the user social interaction cluster profit coefficient. , Public media publicity income coefficient and simulcast revenue coefficient ; Step 4: The interaction coefficient Cluster profit coefficient of social interaction with users The public media influence coefficient Coefficient of return on public media publicity Associated with the simulcast revenue coefficient and behavior-driven hookup coefficients Associated, the first predicted conversion coefficient is obtained respectively , the second predicted conversion coefficient and the third predicted conversion coefficient , evaluate them respectively, and generate corresponding advertising delivery strategies.

[0022] In this example, the method demonstrated significant, multi-dimensional benefits in practical application. First, at the economic level, through precise data analysis and behavioral modeling, it enabled efficient advertising matching and optimized resource allocation, significantly improving advertising conversion rates and return on investment, helping local businesses achieve higher profits at a lower cost. Furthermore, the method deeply tapped into the consumption potential of rural markets, driving agricultural product sales growth and injecting vitality into the local economy.

[0023] Secondly, at the social level, this method has effectively promoted rural informatization and the popularization of new media, increasing rural users' digital engagement and media literacy, and further bridging the urban-rural digital divide. Through the mining and feedback mechanisms of social interaction data, it enhances rural users' voice and sense of participation in information dissemination. Furthermore, this method can also facilitate the dissemination of rural culture and the promotion of local brands, inspiring rural cultural confidence and a sense of local identity.

[0024] Finally, from a long-term development perspective, this approach can foster the cultivation of rural new media talent, stimulate the local content creation ecosystem, and foster a new model for the rural digital economy. Furthermore, the resulting data and model systems can serve a wider range of application scenarios in the future, including rural governance, government information dissemination, and public service optimization, providing solid support for the development of digital villages.

[0025] Example 2: This example is an explanation of Example 1. Please refer to Figure 1 ,Specifically, step one includes: S11. Use Python to obtain the number of times the i-th user sends the target village-related words in the user social interaction cluster. , using the user social interaction cluster open platform to detect the number of message forwarding, detect the number of times the i-th user forwards the message, and obtain the number of times the i-th user forwards the message , using the timestamp in the user social interaction cluster, record the time when each message of the i-th user is sent, and use the time difference to obtain the interval time between the i-th user's message replies By setting different page number links on the chat page and displaying different page number links, we can get the page browsing times of the i-th user. , establish the first data set; S12: Establish a target area at a position 1m-6m in front of the target display in the park, and collect the number of people in the target area by installing an AI camera on the target display to obtain the number of people in the target area. , and based on the AI ​​camera combined with the time difference, obtain the residence time of people in the target area , through the QR code platform to count the number of scans in real time, and obtain the number of scans , establish a second data set; S13, using a voice-to-text converter to recognize and convert the voice of the video of the first rural advertisement into text, and extract the number of keywords from the text , use a timer to detect the time when the resident crowd watches the first rural advertisement and obtain the viewing time , establish the third data set.

[0026] In this embodiment, the technical implementation of S11, S12, and S13 further enriches and enhances the practical application value of the method for precise rural new media advertising delivery. In terms of user behavior data collection, S11 uses Python scripts to track user behavior in social interaction clusters in detail, including key indicators such as keyword sending frequency, message forwarding times, reply time intervals, and page browsing times, thereby constructing a high-dimensional, time-series first dataset. This precise behavior collection method helps identify active user groups and potential interest preferences, providing solid data support for targeted advertising delivery.

[0027] In terms of real-world perception, the S12 leverages AI vision technology to monitor the flow and resident behavior of people within a target area of ​​1 to 6 meters in front of the advertising screen. This data, combined with interaction data from the QR code scanning platform, effectively builds a secondary dataset. This data not only truly reflects the offline audience's exposure to and interest in the ads, but also provides a quantitative basis for evaluating the effectiveness of public media communication, significantly enhancing the digital evaluation capabilities of traditional outdoor advertising.

[0028] In the ad content understanding and feedback phase, S13 leverages Google speech recognition technology to transcribe the audio content of video ads into text, extract keywords, and combine this with viewing time statistics to construct a third dataset. This allows for quantifiable measurement of the comprehensibility and appeal of ad content, providing crucial insights for subsequent creative optimization and emotional adjustment.

[0029] Example 3 This embodiment is explained in Example 1, please refer to Figure 1 Specifically, step 2 includes: S21, for extracting from the first data set the number of times the i-th user sends the target village-related words in the user social interaction cluster , the number of times the i-th user forwards the message The interval between the reply to the message of the i-th user , the number of page views by the i-th user , after dimensionless processing, we can get 、 、 and ;In the formula It is represented as the normalized value of the number of times the i-th user sends the target village-related words in the user social interaction cluster. It is represented as the normalized value of the number of times the i-th user forwards the message. It is represented as the normalized value of the reply interval of the i-th user message, It is represented as the normalized value of the number of page views of the i-th user. Indicates the maximum number of times the target village-related words are sent in the user social interaction cluster, obtained using Python , It is expressed as the maximum number of message forwarding times. The number of message forwarding times is detected by the user social interaction cluster open platform. , It is expressed as the maximum message reply interval. By using the timestamp in the user social interaction cluster, the sending time of each message is recorded to obtain , It represents the maximum number of page views. Set different page number links for the chat page. By displaying different page number links, you can get ; The interaction coefficient is calculated by the following formula ; Where M represents the total number of users, Expressed as a weight coefficient.

[0030] set up is 5, 1 is 0.25, 2 is 0.25, 3 is 0.25, 4 is 0.25, is 50, is 60, is 100, is 80; The following is an example table of interaction coefficients, see Table 1

[0031] In this embodiment, the interaction coefficient constructed in step S21 This computational method further enhances the accuracy and practicality of the rural new media advertising system in understanding and modeling user behavior. This step accurately extracts the i-th user's keyword sending frequency, message forwarding frequency, reply interval, and page-turning behavior from the first dataset. Using dimensionless normalization, the raw heterogeneous data is unified into comparable standardized metrics. This approach effectively avoids model drift caused by data scale differences, making user behavior variables more versatile and stable during interaction modeling.

[0032] By setting appropriate maximum benchmarks and combining them with weighted coefficient design, the system can dynamically adapt to the varying behavioral intensities of different user groups, achieving a more nuanced portrayal of user behavior. The resulting interaction coefficient not only reflects the user's comprehensive ability to interact with advertising content but also provides a quantitative basis for subsequent personalized advertising push, social influence ranking, and high-value user identification. This fine-grained data modeling approach is particularly important in the rural new media environment, effectively complementing the traditional advertising system's lack of understanding of users outside of urban centers.

[0033] Example 4 This embodiment is explained in Example 1, please refer to Figure 1 , specifically, Step 2 also includes: S22, for extracting the number of people in the target area from the second data set , the length of time people stay in the target area and number of scans , after dimensionless processing, we can obtain 、 Where It is expressed as the normalized value of the number of people in the target area. It is expressed as the normalized value of the residence time of people in the target area. It is expressed as the normalized value of the number of scans. Indicates the maximum number of people in the target area. Use AI smart camera to identify the number of people in the target area and obtain , It is expressed as the maximum residence time of the number of people in the target area. It is obtained by tracking the time from the individual entering the target area to leaving the target area based on AI smart cameras combined with behavior recognition. , Expressed as the maximum number of code scans, the QR code software is used to count the number of code scans per unit time and obtain ; The public media influence coefficient is calculated using the following formula ; Where, Expressed as a weight coefficient.

[0034] Set the maximum number of people in the target area to 200, the maximum stay time in the target area to 200 seconds, and the maximum number of code scans to 6; The following are the public media impact coefficients The sample table is shown in Table 2:

[0035] In this embodiment, through the implementation of S22, the system achieved a significant breakthrough in the quantitative evaluation of public media communication effectiveness. This step leverages AI-powered intelligent cameras to identify and track the real-time behavior of people in the target area, accurately collecting data on actual audience numbers, dwell time, and code scanning behavior before the ad. Dimensionless normalization eliminates scale differences between different indicators, constructing a standardized dataset with universal explanatory power. This approach significantly improves the quantification and data credibility of offline advertising exposure.

[0036] By constructing a public media impact coefficient, we unify the modeling of an ad's audience reach (number of people), depth of engagement (dwell time), and conversion intent (number of scans), effectively reflecting the true reach and appeal of ads in crowded locations. This coefficient is not only calculable and dynamically adjustable in real time, but also serves as a crucial reference for optimizing offline ad placement, exposure frequency, and content format.

[0037] Example 5 This embodiment is explained in Example 1, please refer to Figure 1 , specifically, Step 2 also includes: S23: By presetting the public media influence threshold A and using the actual conversions brought about in multiple past advertising delivery scenarios, the public media influence threshold A is obtained, and the public media influence threshold A is compared with the public media influence coefficient. Compare and obtain the first phase evaluation results, including: when When the value is greater than A, it indicates that the results of rural advertising on public media are satisfactory, and rural advertising on public media should continue to be placed; when When ≤A, it means that the results of rural advertising on public media are unqualified, and the first correction strategy is generated, including: guiding the number of QR code scans in the video to increase by 10%-20%, and increasing the frequency of the detected people's residence time in the target area by 5%-8%. When it is detected that the people stay in the target area for more than 9s, the play instruction is triggered, and the rural advertising content is adjusted to increase the attraction rate by 10%-30%.

[0038] The following are the public media impact threshold A and the public media impact coefficient A comparison chart is shown in Table 3:

[0039] In this embodiment, the public media impact threshold assessment mechanism introduced in step S23 significantly enhances the dynamic feedback capabilities and adaptive optimization level of the rural advertising system. This mechanism forms a public media impact threshold A based on historical advertising data, which serves as a benchmark for evaluating the effectiveness of current advertising dissemination, thereby enabling quantitative judgment of current advertising performance. When the assessment result is above the threshold, the system automatically confirms that the advertising dissemination effect on the public media platform has met the requirements and maintains the current delivery strategy. When the assessment result is below the threshold, the system immediately triggers the first-stage correction strategy and enters the intelligent optimization process.

[0040] The positive strategy fully demonstrates the system's refined control capabilities and conversion-oriented awareness. For example, by increasing the intensity of QR code scanning prompts by 10%–20%, user engagement is enhanced; while the frequency of viewer dwell time is moderately extended by 5%–8% to enhance the immersive experience of the ad; and when the dwell time exceeds 9 seconds, a video playback command is triggered to precisely stimulate user attention. Furthermore, targeted adjustments are made to the ad content to increase overall appeal by 10%–30%. This complete mechanism fully embodies the closed-loop path of "data feedback – behavioral analysis – content optimization," providing intelligent, efficient, and real-time optimization for rural advertising.

[0041] Example 6 This embodiment is explained in Example 1, please refer to Figure 1 , specifically, Playback instructions include: Used to extract the number of keywords from the third dataset and viewing time , and combined with the number of scans , after dimensionless processing, the behavior-driven simulcast coefficient is obtained by the following formula ; Where, 、 and is the weight coefficient; set up is 0.4, is 0.3, is 0.3; Here are the behavior-driven hookup coefficients The sample table is shown in Table 4:

[0042] By presetting the behavior-driven simulcast threshold Z, training the model through historical label data, and finding the optimal threshold, the behavior-driven simulcast threshold Z is obtained, and the behavior-driven simulcast threshold Z is compared with the behavior-driven simulcast coefficient Compare and generate the second-stage evaluation results, including: when >Z, indicating that the user behavior triggered by the target rural advertisement meets the triggering conditions, which in turn leads to the playback of the secondary advertisement. When the expected result of the secondary advertisement is greater than the preset evaluation threshold, the system automatically executes the continuous playback of the secondary advertisement; when ≤Z indicates that although the current rural advertisement leads to a sub-advertisement, the expected result of the sub-advertisement is not up to standard. A second correction strategy is generated, including: increasing the content of the sub-advertisement to increase its relevance to the content of the initial rural advertisement by 70%-80%, adjusting the jump logic of the sub-advertisement, optimizing the code scanning path, reducing the process by 20%-25%, and improving the accuracy of keyword recognition for the initial rural advertisement by 80%-90%. The sub-advertisement used for jump should echo the initial rural advertisement.

[0043] The following are the behavior-driven simulcast threshold Z and the behavior-driven simulcast coefficient For a comparison example, see Table 5:

[0044] In this embodiment, the system implements highly intelligent and responsive content decision-making within the advertising simulcast strategy through an evaluation mechanism based on the behavior-driven simulcast coefficient and a preset simulcast threshold Z. Specifically, this mechanism extracts the number of keywords, viewing duration, and code scanning behavior from a third dataset, performs dimensionless processing, and then calculates the behavior-driven simulcast coefficient. This coefficient quantifies the behavioral driving effect of the initial ad leading to the secondary ad. Using the simulcast threshold Z, trained with historical ad label data, as an evaluation criterion, the system is able to dynamically assess the actual effectiveness of the simulcast strategy.

[0045] When the behavior-driven simulcast coefficient exceeds threshold Z, the system determines that the current initial rural ad has successfully guided the secondary ad and that the secondary ad has achieved expected performance, thus continuing the subsequent simulcast logic. Conversely, when the simulcast coefficient falls below threshold Z, the system immediately initiates the second-stage correction strategy, demonstrating extremely high adaptive adjustment capabilities and user conversion guidance. The correction strategy includes three core optimization methods: First, the semantic and visual level of the secondary ad content is strengthened, increasing its relevance to the initial ad by 70%-80%, achieving a closed-loop content logic and enhancing memory linkage; second, the jump process of the secondary ad is compressed and optimized, reducing user operation costs by 20%-25% and improving conversion efficiency; third, the keyword recognition accuracy in the initial ad is optimized to 80%-90%, ensuring the accurate triggering of the secondary ad and improving the consistency of the user experience.

[0046] This system intelligently identifies the time when viewers are watching the initial ad, specifically setting a "nine-second dwell time" as a key trigger threshold. This system establishes an automated playback mechanism based on behavioral immersion. The core of this mechanism lies in accurately capturing moments of deep user attention and triggering playback accordingly, achieving a seamless transition between the initial ad and the secondary ad. This "behavior-triggered" simulcast logic significantly improves the responsiveness of ad content and the continuity of the conversion chain, effectively reducing the risk of users being disconnected during the ad process and attracting interested viewers.

[0047] Example 7 This embodiment is explained in Example 1, please refer to Figure 1 , specifically, Step three includes: S31, used to calculate the transaction amount of the i-th user on the user social interaction cluster platform Collect and build a data set of total revenue of user social interaction cluster platform; S32, used to calculate the transaction amount of the i-th user on the public media promotion platform Collect and construct a data set of total revenue of public media platforms; S33, used to calculate the transaction volume generated by behavior-driven simulcasting Collect and build a data set of behavior-driven simulcast revenue; S33, based on the transaction amount of the i-th user in the user social interaction cluster total income dataset on the live broadcast platform , calculate the user social interaction cluster platform profit coefficient , where It is represented by the maximum transaction amount of the user social interaction cluster, Expressed as weight coefficient; set up is 1, is 5, is 10000; The following is an example table of the revenue coefficient of the user social interaction cluster platform, as shown in Table 6:

[0048] S34, based on the transaction amount of the i-th user in the live broadcast platform in the public media total revenue dataset , calculate the public media revenue coefficient , where It is represented by the maximum transaction amount of the Park Ring Display Platform. Expressed as weight coefficient; set up is 1, is 5, is 10000; The following is an example table of public media revenue coefficients, see Table 7:

[0049] S35. Extracting transaction amounts from behavior-driven simulcast revenue data sets , calculate the simulcast revenue coefficient , , where It is expressed as the maximum value of simulcast revenue, is the weight coefficient; set up is 1, is 6000; The following is an example table of simulcast revenue coefficients, see Table 8:

[0050] In this embodiment, the system achieves comprehensive integration and quantitative modeling of multi-source advertising revenue data, significantly enhancing the ability to track the economic value of advertising and the feedback mechanism for conversion effects. Specifically, the system collects user transactions on social interaction cluster platforms, public media promotion platforms, and behavior-driven syndicated broadcasts, constructing three revenue data sets. This design enables advertising effectiveness to move beyond superficial metrics such as clicks, views, and scans, delving deeper into the final transaction behavior, establishing a closed-loop chain from behavioral triggers to actual revenue.

[0051] In terms of data processing, the system uses dimensionless normalization to standardize revenue data across different users and platforms, calculating revenue coefficients for user social interaction cluster platforms, public media platforms, and simulcast platforms. Each coefficient is calibrated based on the corresponding platform's maximum transaction volume, and a weighting parameter is introduced to reflect the platform's relative influence in the overall conversion path, ensuring comparability and strategic guidance.

[0052] The construction of this revenue coefficient model not only provides a reliable basis for the refinement of subsequent advertising strategies, but also provides quantitative criteria and hierarchical comparison capabilities for evaluating the effectiveness of advertising campaigns. The system can identify the actual economic returns generated by different channels, content formats, or user behavior paths, and dynamically adjust resource allocation, such as increasing advertising frequency on high-yield social platforms or strengthening content integration strategies for outstanding content within syndicated channels.

[0053] Example 8 This embodiment is explained in Example 1, please refer to Figure 1 , specifically, Step 4 includes: S41, the interaction coefficient Cluster profit coefficient of social interaction with users The first predicted conversion coefficient is calculated by the following formula , Where, Expressed as weight coefficient, Expressed as the maximum value of the interaction coefficient, It is expressed as the maximum benefit of the user social interaction cluster, obtained by recording data from the user social interaction cluster platform. and ; set up is 0.6, is 1.0, is 0.9; The following is an example table of the first predicted conversion coefficient, see Table 9:

[0054] S42 is used to preset the user social interaction cluster platform revenue threshold B. Here, the user social interaction cluster platform revenue threshold is a reference brick used to judge whether the advertisement is placed on the platform up to standard. It is derived from historical data analysis. By obtaining the minimum value of historical data, the user social interaction cluster platform revenue threshold B is obtained, and the user social interaction cluster platform revenue threshold B is compared with the second predicted conversion coefficient. Perform a comparison to generate a second evaluation result, including: when When the value is greater than B, it indicates that the conversion rate of the target rural advertising investment on the user social interaction cluster platform is qualified, and the current advertising volume in the user social interaction cluster is maintained; when When ≤B, it means that the conversion rate of the target rural advertising investment in the user social interaction cluster platform is less than the preset conversion rate, and the conversion rate is unqualified. The second strategy is generated, including: adjusting the video content, optimizing the existing advertising content, introducing distinctive rural plantations, increasing the number of users attracted by 20%-30%, taking advantage of time periods with frequent social interactions to increase the amount of advertising by 50%-60%, and reducing the amount of advertising by 20%-40% during time periods with sparse social interactions.

[0055] Set the user social interaction cluster platform income threshold B to 0.6; 0.67>0.6, indicating that the conversion rate of target rural advertising investment on the user social interaction cluster platform is qualified.

[0056] In this embodiment, step four constructs a first predicted conversion coefficient by correlating the interaction coefficient with the user social interaction cluster revenue coefficient, achieving a deep fusion analysis of user behavior data and economic benefits. This fusion not only improves the accuracy of conversion predictions but also enables forward-looking predictions of advertising effectiveness, providing a scientific basis for the formulation of subsequent advertising strategies. By mining and normalizing historical data from the user social interaction cluster platform, the system can dynamically capture conversion trends and behavioral changes, forming an accurate conversion rate evaluation system.

[0057] The preset user social interaction cluster platform revenue threshold B serves as a key judgment criterion, further strengthening the quantitative control of advertising effectiveness. When the predicted conversion coefficient exceeds the threshold, the system confirms the effectiveness of the current rural advertising strategy and maintains stable delivery to ensure the rational use of resources. When the conversion coefficient falls below the threshold, the system quickly triggers a second strategy revision plan to achieve intelligent optimization of advertising content and delivery rhythm. Specifically, this includes targeted adjustments to video content, incorporating local elements such as distinctive rural plantations to increase the appeal of ads by 20%-30%. Ad delivery is increased by 50%-60% during peak user social interaction periods, while moderately reducing it by 20%-40% during low interaction periods, achieving dynamic matching of delivery resources and maximizing conversion effectiveness.

[0058] Example 9 This embodiment is explained in Example 1, please refer to Figure 1 , specifically, Step 4 also includes: S43. The public media influence coefficient Coefficient of return on public media publicity The second predicted conversion coefficient is calculated by the following formula , Where, Expressed as weight coefficient, Expressed as the maximum value of the public media influence coefficient, It is expressed as the maximum value of public media publicity income, obtained by obtaining the background data of the public media device. and ; set up is 0.5, is 1.0, is 0.85,; The following is an example table of the second predicted conversion coefficient, see Table 10:

[0059] S44: Preset the public media platform revenue threshold C, infer the minimum conversion rate based on the estimated number of exposures and the number of residents, obtain the public media platform revenue threshold C, and compare the public media platform revenue threshold C with the third predicted conversion coefficient. Perform comparisons to generate third evaluation results, including: when When it is greater than C, it means that the conversion rate of the target rural advertising investment on the public media platform is qualified, and the current amount of advertising investment on the public media platform is maintained; when When ≤C, it means that the conversion rate of the target rural advertising investment on the live broadcast platform is lower than the expected conversion rate, and the conversion rate is unqualified. The third strategy is generated, including: reducing the frequency of target rural advertising on public media platforms by 15%-35%, and eliminating user groups with a click-through rate below 30% of the average level, increasing advertising content, combining rural geographical characteristics and characteristic agricultural products, launching novel advertising content, and increasing the dissemination rate by 40%-50%.

[0060] Set the public media platform revenue threshold C to 1; Then 0.82<1, indicating failure.

[0061] In this embodiment, step four constructs a second predicted conversion coefficient by associating the public media influence coefficient with the public media publicity revenue coefficient, thereby achieving a scientific evaluation of the advertising effect of the public media platform. Based on background data extraction and dimensionless normalization processing, the system can accurately reflect the influence and actual revenue of advertising on public media, and then complete the dynamic judgment of the delivery effect through the preset public media platform revenue threshold C. When the delivery effect is qualified, the current advertising frequency is maintained to ensure stable resource output; when the conversion rate is lower than the threshold, the system automatically generates an optimization strategy to moderately reduce the delivery frequency by 15%-35%, eliminate low click-through rate user groups, and combine rural regional characteristics with characteristic agricultural products to launch novel and more attractive advertising content, significantly increasing the dissemination rate by 40%-50%. This strategy effectively avoids waste of resources, while stimulating user interest and promoting the precise dissemination of rural brands.

[0062] Example 10 This embodiment is explained in Example 1, please refer to Figure 1 , specifically, Step 4 also includes: S45, will be broadcast revenue coefficient and behavior-driven hookup coefficients The third predicted conversion coefficient is obtained by the following formula , Where, Expressed as the maximum value of the behavior-driven simulcast coefficient, Expressed as the maximum value of the simulcast revenue coefficient, Expressed as weight coefficient; set up is 0.6, is 39.3, is 1.0; The following is an example table of the third predicted conversion coefficient, see Table 11:

[0063] S46: By presetting the behavior-driven simulcast revenue threshold V, extract a batch of revenue values ​​from the historical data, remove the maximum and minimum values, and then take the average value to obtain the behavior-driven simulcast revenue threshold V. The behavior-driven simulcast revenue threshold V is combined with the third predicted conversion coefficient. Perform a comparison to generate a fourth evaluation result, including: when When it is greater than V, it indicates that the expected revenue generated by the simulcast is qualified, and the secondary advertisements and monitoring will continue; when When ≤V, it indicates that the expected revenue generated by the simulcast is not up to standard, and the fourth strategy is generated, including: adjusting the content of the secondary advertisements, increasing the amount of content related to the initial rural advertisements by 30%-50%, reducing the delivery range of secondary advertisements by 10%-15%, and accurately delivering them to the people in need, increasing the exposure ratio to high-potential groups by 10%-20%.

[0064] The following are the behavior-driven simulcast revenue threshold V and the third predicted conversion coefficient For a comparison example, see Table 12:

[0065] In this embodiment, in the behavior-driven simulcast link, the system calculates the third predicted conversion coefficient by associating the behavior-driven simulcast coefficient with the simulcast revenue coefficient, and accurately measures the revenue performance of the secondary advertisement. Combined with the average revenue threshold V after removing the extreme value of historical data, a robust evaluation of the simulcast delivery effect is achieved. When the simulcast revenue is qualified, the system continues to maintain the delivery and monitoring of secondary advertisements to ensure the continuity of content linkage and maximize the effect; when the simulcast revenue does not meet the standard, the system initiates a correction plan by increasing the relevance of the secondary advertisement to the primary advertisement content by 30%-50%, narrowing the delivery range by 10%-15%, and accurately delivering it to high-potential groups, increasing the exposure ratio of this group by 10%-20%, thereby improving the accuracy and conversion rate of advertising reach.

[0066] The above formulas are obtained by collecting a large amount of data and performing software simulation, and a formula close to the actual value is selected. The coefficients in the formula are set by those skilled in the art according to actual conditions. The above is only a preferred specific implementation method of the present invention, but the protection scope of the present invention is not limited to this. Any technician familiar with this technical field, within the technical scope disclosed by the present invention, can make equivalent replacements or changes based on the technical solution and inventive concept of the present invention, which should be covered by the protection scope of the present invention.

Claims

1. The method of precise advertising delivery based on rural new media is characterized by: The specific steps include: Step 1: Build a multi-type target rural new media advertising platform, place the target rural advertisements on the new media platform, and collect user behavior data from each platform to generate the first data set, the second data set, and the third data set; Step 2: Establish a user behavior model and input the first, second, and third data sets into the analysis to obtain the interaction coefficients , public media influence coefficient and behavior-driven hookup coefficient ; Step 3: Establish a user analysis model. After inputting data analysis, continue to collect advertising conversion rates on each platform to obtain the user social interaction cluster profit coefficient. , Public media publicity income coefficient and simulcast revenue coefficient ; Step 4: The interaction coefficient Cluster profit coefficient of social interaction with users The public media influence coefficient Coefficient of return on public media publicity Associated with the simulcast revenue coefficient and behavior-driven hookup coefficients Associated, the first predicted conversion coefficient is obtained respectively , the second predicted conversion coefficient and the third predicted conversion coefficient , evaluate them respectively, and generate corresponding advertising delivery strategies.

2. The method for precise advertising delivery based on rural new media according to claim 1, characterized in that: Step one includes: S11. Use Python to obtain the number of times the i-th user sends the target village-related words in the user social interaction cluster. , using the user social interaction cluster open platform to detect the number of message forwarding, detect the number of times the i-th user forwards the message, and obtain the number of times the i-th user forwards the message , using the timestamp in the user social interaction cluster, record the time when each message of the i-th user is sent, and use the time difference to obtain the interval time between the i-th user's message replies By setting different page number links on the chat page and displaying different page number links, we can get the page browsing times of the i-th user. , establish the first data set; S12: Establish a target area at a position 1m-6m in front of the target display in the park, and collect the number of people in the target area by installing an AI camera on the target display to obtain the number of people in the target area. , and based on the AI ​​camera combined with the time difference, obtain the residence time of people in the target area , through the QR code platform to count the number of scans in real time, and obtain the number of scans , establish a second data set; S13, using a voice-to-text converter to recognize and convert the voice of the video of the initial rural advertisement into text, and extract the number of keywords from the text , use a timer to detect the time when the resident crowd watches the first rural advertisement and obtain the viewing time , establish the third data set.

3. The method for precise advertising delivery based on rural new media according to claim 2, characterized in that: Step 2 includes: S21, for extracting from the first data set the number of times the i-th user sends the target village-related words in the user social interaction cluster , the number of times the i-th user forwards the message The interval between the reply to the message of the i-th user , the number of page views by the i-th user , and the interaction coefficient is obtained by calculating .

4. The method for precise advertising delivery based on rural new media according to claim 3 is characterized by: Step 2 also includes: S22, for extracting the number of people in the target area from the second data set , the length of time people stay in the target area and number of scans , and calculate the public media influence coefficient .

5. The method for precise advertising delivery based on rural new media according to claim 4 is characterized in that: Step 2 also includes: S23, by presetting the public media influence threshold A, the public media influence threshold A and the public media influence coefficient Compare and obtain the first phase evaluation results, including when When the value is greater than A, it indicates that the results of rural advertising on public media are satisfactory, and rural advertising on public media should continue to be placed; when When ≤A, it means that the results of rural advertising on public media are unqualified, and the first correction strategy is generated, including: guiding the number of QR code scans in the video to increase by 10%-20%, and increasing the frequency of the detected people's residence time in the target area by 5%-8%. When it is detected that the people stay in the target area for more than 9s, the play instruction is triggered, and the rural advertising content is adjusted to increase the attraction rate by 10%-30%.

6. The method for precise advertising delivery based on rural new media according to claim 5, characterized in that: Playback instructions include: Used to extract the number of keywords from the third dataset and viewing time , and combined with the number of scans , the behavior-driven simulcast coefficient is obtained by calculating ; By presetting the behavior-driven simulcast threshold Z and combining the behavior-driven simulcast threshold Z with the behavior-driven simulcast coefficient Compare and generate the second-stage evaluation results, including: when >Z, indicating that the user behavior triggered by the target rural advertisement meets the triggering conditions, which in turn leads to the playback of the secondary advertisement. When the expected result of the secondary advertisement delivery is greater than the preset evaluation threshold, the system automatically executes the continuous delivery of the secondary advertisement; when ≤Z indicates that although the current rural advertisement leads to a sub-advertisement, the expected result of the sub-advertisement is not up to standard. A second correction strategy is generated, including: increasing the content of the sub-advertisement to increase its relevance to the content of the initial rural advertisement by 70%-80%, adjusting the jump logic of the sub-advertisement, optimizing the code scanning path, reducing the process by 20%-25%, and improving the accuracy of keyword recognition for the initial rural advertisement by 80%-90%. The sub-advertisement used for jump should echo the initial rural advertisement.

7. The method for accurate advertising delivery based on rural new media according to claim 6, characterized in that: Step three includes: S31, used to calculate the transaction amount of the i-th user on the user social interaction cluster platform Collect and build a data set of total revenue of user social interaction cluster platform; S32, used to calculate the transaction amount of the i-th user on the public media promotion platform Collect and construct a data set of total revenue of public media platforms; S33, used to calculate the transaction volume generated by behavior-driven simulcasting Collect and build a data set of behavior-driven simulcast revenue; S33, based on the transaction amount of the i-th user in the user social interaction cluster total income dataset on the live broadcast platform , calculate the user social interaction cluster platform profit coefficient ; S34, based on the transaction amount of the i-th user in the live broadcast platform in the public media total revenue dataset , calculate the public media revenue coefficient ; S35. Extracting transaction amounts from behavior-driven simulcast revenue data sets , calculate the simulcast revenue coefficient .

8. The method for precise advertising delivery based on rural new media according to claim 7 is characterized by: Step 4 includes: S41, the interaction coefficient Cluster profit coefficient of social interaction with users The first predicted conversion coefficient is obtained by calculation ; S42, for presetting a user social interaction cluster platform income threshold B, and comparing the user social interaction cluster platform income threshold B with the first predicted conversion coefficient Perform a comparison to generate a second evaluation result, including: when When the value is greater than B, it indicates that the conversion rate of the target rural advertising investment on the user social interaction cluster platform is qualified, and the current advertising volume in the user social interaction cluster is maintained; when When ≤B, it means that the conversion rate of the target rural advertising investment in the user social interaction cluster platform is less than the preset conversion rate, and the conversion rate is unqualified. The second strategy is generated, including: adjusting the video content, optimizing the existing advertising content, introducing distinctive rural plantations, increasing the number of users attracted by 20%-30%, taking advantage of time periods with frequent social interactions to increase the amount of advertising by 50%-60%, and reducing the amount of advertising by 20%-40% during time periods with sparse social interactions.

9. The method for precise advertising delivery based on rural new media according to claim 8 is characterized by: Step 4 also includes: S43. The public media influence coefficient Coefficient of return on public media publicity The second predicted conversion coefficient is obtained by calculation ; S44, for presetting a public media platform revenue threshold C, and comparing the public media platform revenue threshold C with the third predicted conversion coefficient Perform comparisons to generate third evaluation results, including: when When it is greater than C, it means that the conversion rate of the target rural advertising investment on the public media platform is qualified, and the current amount of advertising investment on the public media platform is maintained; when When ≤C, it means that the conversion rate of the target rural advertising investment on the live broadcast platform is lower than the expected conversion rate, and the conversion rate is unqualified. The third strategy is generated, including: reducing the frequency of target rural advertising on public media platforms by 15%-35%, and eliminating user groups with a click-through rate below 30% of the average level, increasing advertising content, combining rural geographical characteristics and characteristic agricultural products, launching novel advertising content, and increasing the dissemination rate by 40%-50%.

10. The method for accurate advertising delivery based on rural new media according to claim 9 is characterized in that: Step 4 also includes: S45, will be broadcast revenue coefficient and behavior-driven hookup coefficients The third predicted conversion coefficient is obtained by calculation ; S46: By presetting the behavior-driven simulcast revenue threshold V, the behavior-driven simulcast revenue threshold V is combined with the third predicted conversion coefficient. Perform a comparison to generate a fourth evaluation result, including: when When it is greater than V, it indicates that the expected revenue generated by the simulcast is qualified, and the secondary advertisements and monitoring will continue; when When ≤V, it indicates that the expected revenue generated by the simulcast is not up to standard, and the fourth strategy is generated, including: adjusting the content of the secondary advertisements, increasing the amount of content related to the initial rural advertisements by 30%-50%, reducing the delivery range of secondary advertisements by 10%-15%, and accurately delivering them to the people in need, increasing the exposure ratio to high-potential groups by 10%-20%.

Images