An interactive method, system and promotional cart with technical popularization

Abstract

The application belongs to the technical popularization technical field, and provides an interactive method, system and propaganda cart with technical popularization, which comprises the following steps: adopting a K-means clustering algorithm, performing clustering analysis on user data according to a preset propaganda type, and outputting a clustering cluster; based on any clustering cluster, determining a popular time period by performing time sequence statistical analysis on user historical behavior data; acquiring an existing propaganda type, comparing and analyzing the existing propaganda type with the preset propaganda type, judging whether there is a lack of the existing propaganda type compared with the preset propaganda type, and generating a first popularization sequence list; if there is a lack, performing feature similarity analysis on the missing propaganda type and the non-missing propaganda type, identifying similar propaganda types in the non-missing propaganda type, and generating a second popularization sequence list; and the application tilts key resources to high-potential user clusters and optimal popularization time periods, thereby reducing popularization cost and improving popularization cost performance.

Description

Technical Field

[0001] This invention belongs to the field of industrial internet platform technology, specifically an interactive method, system, and promotional cart for technology promotion. Background Technology

[0002] With the rapid development of digital technology, technology promotion has become an important means for enterprises to enhance their core competitiveness and expand their technological influence. As a key carrier connecting the promotion subject and the target users, the rationality and efficiency of technology promotion interaction methods directly determine the effect of technology promotion and the efficiency of resource utilization.

[0003] Currently, in terms of timing for promotions, existing promotional interaction methods lack in-depth time-series analysis of user behavior data, making it impossible to accurately identify peak user activity periods. They rely heavily on manual experience to set push times, leading to unreasonable timing of promotional content delivery. This approach not only fails to increase the immediate exposure of promotional content and effectively increase the probability of user clicks, views, and interactions, thus affecting promotional results, but also easily pushes content during periods of low user activity, resulting in a serious waste of promotional resources (human, material, and financial). Furthermore, it violates user behavior habits and reduces user experience. In addition, existing methods do not use peak time periods as the core sorting standard for promotion sequences, failing to optimize the promotion rhythm based on user activity patterns, leading to a chaotic promotion process and further hindering the improvement of promotional effectiveness.

[0004] Regarding fault tolerance and content adaptation in the promotion process, existing technical promotion interaction methods lack a robust anomaly response mechanism. When unexpected business situations arise, such as data collection link failures or promotional content not being online, resulting in the absence of preset promotional types, the methods cannot promptly identify and match alternative promotional types with similar characteristics to the missing type. This often leads to the entire technical promotion process stalling, reducing its stability and usability. Furthermore, existing methods fail to organically combine popular time slot ranking with promotional content preferences. Even if the promotion process is forced to proceed, it is difficult to guarantee the matching degree between promotional content and user preferences, further weakening the targeting and effectiveness of technical promotion and failing to fully realize its value.

[0005] Therefore, the present invention provides an interactive method, system, and promotional cart for technology promotion. Summary of the Invention

[0006] In order to overcome the shortcomings of the prior art, at least one technical problem raised in the background art is solved.

[0007] The technical solution adopted by this invention to solve its technical problem is: 1. An interactive method for technology promotion, characterized in that it includes: Collect historical behavioral data of users when interacting with the advertising platform, and perform data preprocessing, including data cleaning, noise reduction, and missing value imputation; Based on preprocessed historical behavior data, the K-means clustering algorithm is used to perform cluster analysis on user data according to the preset advertising type, and output clusters. Based on any cluster, popular time periods are determined by performing time-series statistical analysis on users' historical behavior data; Obtain existing promotional types and compare them with preset promotional types to determine if there are any missing promotional types. Generate a first promotional sequence list. If there are missing promotional types, perform feature similarity analysis between the missing promotional types and the non-missing promotional types to identify similar promotional types among the non-missing promotional types and generate a second promotional sequence list.

[0008] Preferably, the process of outputting clusters is as follows: Historical behavioral data includes content click frequency, browsing duration, and social interaction frequency. The K-value is determined based on the preset number of promotional types. Initialization: Randomly select K feature vectors as initial cluster centers, corresponding to the initial features of the 3 types of advertising; Sample allocation: The distance from each user to the K centers is calculated using the Euclidean distance formula, and the user is assigned to the cluster of the nearest center; Cluster update: Calculate the mean of all user features within each cluster and use it as the new cluster center; Iterative convergence: Repeat the sample assignment and center update operations until the cluster centers no longer change significantly; After convergence, K clusters are obtained. Combining the historical behavior data of users within the clusters, the clusters are matched with the preset advertising types. Finally, the user list and corresponding advertising type of each cluster are output.

[0009] Preferably, the process of determining popular time periods is as follows: The user's historical behavior data is sorted in chronological order, and the 24 hours are divided into several analysis periods with equal time intervals. By analyzing and processing the user's historical behavior data within the analysis period, the interaction representation value of the analysis period is determined. If the interaction representation value is greater than or equal to the interaction representation threshold, the corresponding analysis period is recorded as a suspected hot period. Sort the interaction values ​​of suspected popular time periods in descending order, and extract the suspected popular time period with the highest order, which is recorded as the popular time period corresponding to the promotion type.

[0010] Preferably, the process for determining the interaction representation value of the analysis period is as follows: The click frequency within the statistical analysis period is then compared with a preset standard click frequency value to obtain the click frequency characteristic value for the analysis period. The browsing duration corresponding to each click within the analysis period is summed, and then compared with the duration corresponding to the analysis period to obtain the browsing duration feature value of the analysis period. The frequency of social interactions within the statistical analysis period is then compared with a preset standard value for the frequency of social interactions to obtain the characteristic value of the frequency of social interactions during the analysis period. The click frequency feature value, browsing duration feature value, and social interaction frequency feature value of the analysis period are weighted and summed to obtain the interaction characterization value of the analysis period.

[0011] Preferably, the process of determining whether there are any missing elements compared to the preset promotional types is as follows: All preset advertising types are integrated into a preset advertising type sequence. Based on any existing advertising type, the preset advertising type sequence is traversed. If there is an identical existing advertising type in the preset advertising type sequence, the corresponding preset advertising type is recorded as a non-missing advertising type; if there is no identical existing advertising type in the preset advertising type sequence, the corresponding preset advertising type is recorded as a missing advertising type. The number of missing promotional types is counted. If the number of missing promotional types is greater than or equal to 1, it means that there are missing promotional types compared with the preset promotional types, and a missing signal is generated; otherwise, it means that there are no missing promotional types compared with the preset promotional types.

[0012] Preferably, the process of generating the first promotion sequence list is as follows: Since there are no missing promotional types compared to the preset types, all promotional types are sorted according to the time sequence of peak periods to obtain a promotional sequence list. Based on the promotional sequence list, the corresponding promotional types are set for peak periods and promotions are carried out. Preferably, the process of performing feature similarity analysis based on missing and non-missing advertising types is as follows: Based on the missing signal, the popular time periods of the missing promotional type and the corresponding popular time periods of the non-missing promotional type are extracted. If the popular time period of the missing promotional type is the same as the popular time period of the non-missing promotional type, the corresponding non-missing promotional type is recorded as the type to be analyzed. Extract the click frequency feature value, browsing duration feature value, and social interaction frequency feature value corresponding to the popular time periods of the type to be analyzed. Use the cosine similarity formula to calculate the click frequency feature value, browsing duration feature value, and social interaction frequency feature value corresponding to the popular time periods of the type to be analyzed and the missing promotional type to obtain the type similarity.

[0013] Preferably, the process of generating the second promotion sequence list is as follows: If the type similarity is greater than or equal to the type similarity threshold, then during the peak period, the missing promotional type is replaced with the type to be analyzed. Based on the replaced promotional type, all promotional types are sorted according to the time order of the peak period to obtain the second promotional sequence list.

[0014] Secondly, an interactive system for technology promotion, comprising: Data preprocessing module: Collects historical behavioral data of users when interacting with the promotional platform and performs data preprocessing, including data cleaning, noise reduction, and missing value imputation; Clustering analysis module: Based on preprocessed historical behavior data, the K-means clustering algorithm is used to perform clustering analysis on user data according to the preset advertising type, and output clusters; Popular time period determination module: Based on any cluster, popular time periods are determined by performing time-series statistical analysis on users' historical behavior data; Promotion sequence list generation module: Obtain existing promotion types, compare and analyze them with preset promotion types, determine whether there are any missing promotion types compared with preset promotion types, and generate the first promotion sequence list; if there are missing promotion types, perform feature similarity analysis based on the missing promotion types and the non-missing promotion types, identify similar promotion types among the non-missing promotion types, and generate the second promotion sequence list.

[0015] Thirdly, a promotional cart for technology promotion includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus. Memory, used to store computer programs; A processor is used to execute methods and steps when executing programs stored in memory.

[0016] The beneficial effects of this invention are as follows: 1. Pushing promotional content during peak user activity periods can increase the instant exposure of the promotional content, increase the probability of user clicks, views, and interactions, and improve the promotion effect; it aligns with user behavior habits: promoting based on users' actual active periods avoids the waste of resources caused by pushing during inactive periods, improving user experience; peak time periods serve as the core sorting standard for the promotion sequence, ensuring that promotional content is pushed at the optimal time, further optimizing the promotion rhythm.

[0017] 2. Optimize the promotion rhythm based on popular time periods to ensure that promotional content is pushed out at the optimal time. At the same time, ensure the matching degree of promotional content with user preferences by replacing similar types, and further improve the promotion effect. Provide reasonable solutions for unexpected situations in actual business such as collection link failures and content not being online, so as to avoid the entire promotion process from being stopped due to local problems, and improve the stability and usability of the process. Attached Figure Description

[0018] The invention will now be further described with reference to the accompanying drawings.

[0019] Figure 1 This is a flowchart illustrating the steps of an interactive method for promoting technology according to an embodiment of the present invention; Figure 2 This is a system block diagram of an interactive system with technical promotion according to an embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of a promotional cart for technology dissemination according to an embodiment of the present invention; 301. Processor; 302. Communication interface; 303. Memory; 304. Communication bus. Detailed Implementation

[0020] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0021] Example 1 Please see Figure 1 As shown in the figure, an interactive method for technology promotion according to an embodiment of the present invention includes the following steps: Step 1: Collect historical behavioral data of users when interacting with the promotional platform, and perform data preprocessing, including data cleaning, noise reduction, and missing value imputation; Historical behavioral data includes content click frequency, browsing duration, and social interaction frequency; Before data collection, standardized data fields are preset, including user unique identifier, behavior type, behavior occurrence time, behavior parameters, device information, and network status; During the data collection process, the data is encapsulated in real time and uniformly converted into JSON format to form the original behavioral dataset. Basic cleaning of the original behavior dataset is performed, specifically including: Duplicate data removal: Based on a combined primary key of user unique identifier, behavior occurrence time, and behavior type, duplicate collected behavior data is deleted; Format standardization validation: Validate the data type of each field, convert the time field to timestamp format, convert the behavior parameter field to key-value pair format, and mark data that does not meet the format requirements as data to be corrected. Based on behavioral characteristics and statistical rules, the cleaned data is subjected to layered noise reduction, specifically including: Rule-based noise filtering: Removes accidental touch noise, such as repeated clicks within a short period of time or invalid browsing behavior with a dwell time shorter than a preset dwell time threshold; removes abnormal traffic noise, such as batch data requests not actively triggered by the user. Statistical noise filtering: Calculate the mean and standard deviation of behavioral parameters, and remove extreme behavioral data that deviates from the mean by more than 3 times the standard deviation. For missing fields in the denoised data, a multi-dimensional adaptive imputation strategy is adopted, which includes: Missing data classification: Missing fields are classified into critical missing data and non-critical missing data. Critical missing data includes user unique identifiers and behavior occurrence time, and such missing data is directly removed. Non-critical missing data includes behavior parameters, device information, and network status, and such data is filled in. Dimensional adaptation filling: Fill by user dimension: For the same user, missing fields are filled with the mean or median of the corresponding field in the user's historical behavior; Fill by behavior type: Under the same behavior type, missing fields are filled with the statistical feature values ​​of the fields corresponding to that behavior type; Step 2: Based on the preprocessed historical behavior data, the K-means clustering algorithm is used to perform cluster analysis on the user data according to the preset advertising type, and the clusters are output. The types of promotional materials include, but are not limited to, technology, environmental protection, and healthcare. The K-means clustering algorithm is used to perform cluster analysis on user data according to the preset advertising type, and the clusters are output. Specifically: The K value (K=3) is determined based on the preset number of advertising types (such as 3 types: technology, environmental protection, and medical health). That is, users are to be grouped into 3 clusters, and each cluster corresponds to one type of advertising. Initialization: Randomly select K feature vectors as initial cluster centers, corresponding to the initial features of the 3 types of advertising; Sample allocation: The distance from each user to the K centers is calculated using the Euclidean distance formula, and the user is assigned to the cluster of the nearest center; Cluster update: Calculate the mean of all user features within each cluster and use it as the new cluster center; Iterative convergence: Repeat the sample assignment and center update operations until the cluster centers no longer change significantly; After convergence, K clusters are obtained. Combining the historical behavior data of users within the clusters (e.g., users in cluster 1 have much higher interaction with technology-related users than others), the clusters are matched to the preset advertising type (cluster 1 → technology-related cluster). Finally, the user list and corresponding advertising type of each cluster are output. Step 3: Based on any cluster, determine the peak time periods by performing time-series statistical analysis on users' historical behavior data; The user's historical behavior data is sorted in chronological order, and the 24 hours are divided into several analysis periods with equal time intervals. The click frequency within the analysis period is counted and then compared with the preset click frequency standard value to obtain the click frequency feature value of the analysis period. The browsing duration corresponding to each click within the analysis period is summed, and then compared with the duration corresponding to the analysis period to obtain the browsing duration feature value of the analysis period. The frequency of social interactions within the statistical analysis period is then compared with a preset standard value for the frequency of social interactions to obtain the characteristic value of the frequency of social interactions during the analysis period. It should be noted that the standard values ​​for click frequency and social interaction frequency were set by those skilled in the art based on historical experience. Using the analytic hierarchy process (AHP), weight ratios were set for the click frequency feature value, browsing duration feature value, and social interaction frequency feature value of the analysis period. The click frequency feature value, browsing duration feature value, and social interaction frequency feature value of the analysis period were then weighted and summed to obtain the interaction representation value of the analysis period. If the interaction representation value is greater than or equal to the interaction representation threshold, the corresponding analysis period is recorded as a suspected hot period. If the interaction representation value is less than the interaction representation threshold, the corresponding analysis period is recorded as a non-popular period. Sort the interaction representation values ​​of the suspected popular time periods in descending order, and extract the suspected popular time period with the highest order, and record it as the popular time period corresponding to the promotion type. The purpose of identifying peak time periods in this step is: Function 1: Pushing promotional content during peak user activity periods can increase the instant exposure of the promotional content, increase the probability of user clicks, views, and interactions, and improve the promotion effect; Function 2: Aligning with user behavior habits: Promotions are based on users' actual active periods, avoiding the waste of resources caused by pushing during inactive periods and improving user experience; Function 3: Popular time slots serve as the core sorting standard for the promotion sequence, ensuring that promotional content is pushed out at the optimal time, further optimizing the promotion rhythm; Step 4: Obtain existing promotional types and compare them with preset promotional types to determine if there are any missing promotional types. Generate the first promotional sequence list. If there are any missing promotional types, perform feature similarity analysis between the missing promotional types and the non-missing promotional types to identify similar promotional types among the non-missing promotional types and generate the second promotional sequence list. It should be noted that the absence of corresponding content for a promotional type (such as not releasing technology-related promotional materials), or a failure in the data collection process (such as the lack of deployment of tracking points for promotional content or failure to report data), will all result in the inability to collect effective behavioral data and thus lead to the absence of promotional types. All preset advertising types are integrated into a preset advertising type sequence. Based on any existing advertising type, the preset advertising type sequence is traversed. If there is an identical existing advertising type in the preset advertising type sequence, the corresponding preset advertising type is recorded as a non-missing advertising type; if there is no identical existing advertising type in the preset advertising type sequence, the corresponding preset advertising type is recorded as a missing advertising type. The number of missing promotional types is counted. If the number of missing promotional types is greater than or equal to 1, it means that there are missing promotional types compared with the preset promotional types, and a missing signal is generated. If the number of missing promotional types is less than 1, it means that there are no missing promotional types compared with the preset promotional types. All promotional types are sorted according to the time order of popular time periods to obtain the first promotional sequence list. Based on the first promotional sequence list, the corresponding promotional types are set during popular time periods and promotion is carried out. Based on the missing signal, the popular time periods of the missing promotional type and the corresponding popular time periods of the non-missing promotional type are extracted. If the popular time periods of the missing promotional type are the same as the popular time periods of the non-missing promotional type, the corresponding non-missing promotional type is recorded as the type to be analyzed; if the popular time periods of the missing promotional type are the same as the popular time periods of the non-missing promotional type, the corresponding non-missing promotional type is recorded as the mismatch type and no processing is performed. Extract the click frequency feature value, browsing duration feature value, and social interaction frequency feature value corresponding to the popular time period of the type to be analyzed. Use the cosine similarity formula to calculate the click frequency feature value, browsing duration feature value, and social interaction frequency feature value corresponding to the popular time period of the type to be analyzed and the missing promotional type to obtain the type similarity. If the type similarity is greater than or equal to the type similarity threshold, then during the peak period, the missing promotional type is replaced with the type to be analyzed. Based on the replaced promotional type, all promotional types are sorted according to the time order of the peak period to obtain the second promotional sequence list. Based on the second promotional sequence list, the corresponding promotional type is set during the peak period for promotion. If the type similarity is less than the type similarity threshold, no action is taken. The purpose of identifying similar advertising types in this step is: Function 1: Based on popular time periods, optimize the promotion rhythm and ensure that the promotional content is pushed at the optimal time. At the same time, by replacing similar types, ensure the matching degree of promotional content with user preferences and further improve the promotion effect; Function 2: To provide reasonable solutions (similar type replacement) for unexpected situations in actual business, such as failure of the data collection link or failure of content to go online, so as to avoid the entire promotion process from being stopped due to local problems and improve the stability and usability of the process; The technical solution of this invention is as follows: Collect historical behavioral data of users interacting with an advertising platform, and perform data preprocessing, including data cleaning, noise reduction, and missing value imputation; based on the preprocessed historical behavioral data, use the K-means clustering algorithm to cluster the user data according to preset advertising types, and output clusters; based on any cluster, determine popular time periods by performing time-series statistical analysis on the user's historical behavioral data; obtain existing advertising types, compare and analyze them with preset advertising types, determine whether there are any missing advertising types compared to the preset advertising types, and generate a first promotion sequence list; if missing, perform feature similarity analysis based on the missing advertising types and the non-missing advertising types, identify similar advertising types among the non-missing advertising types, and generate a second promotion sequence list; this invention reduces promotion costs and improves the cost-effectiveness of promotion by allocating key resources to high-potential user clusters and optimal promotion time periods; at the same time, all operations revolve around user behavioral data, which can provide reliable data analysis support for the formulation, adjustment, and iteration of advertising platform promotion strategies, and help continuously optimize promotion strategies.

[0022] Example 2 Based on the same inventive concept as the interactive method with technical promotion in the foregoing embodiments, such as Figure 2 As shown, this application provides an interactive system with technological promotion capabilities, wherein the system specifically includes: Data preprocessing module: Collects historical behavioral data of users when interacting with the promotional platform and performs data preprocessing, including data cleaning, noise reduction, and missing value imputation; Clustering analysis module: Based on preprocessed historical behavior data, the K-means clustering algorithm is used to perform clustering analysis on user data according to the preset advertising type, and output clusters; Popular time period determination module: Based on any cluster, popular time periods are determined by performing time-series statistical analysis on users' historical behavior data; Promotion sequence list generation module: Obtain existing promotion types, compare and analyze them with preset promotion types, determine whether there are any missing promotion types compared with preset promotion types, and generate the first promotion sequence list; if there are missing promotion types, perform feature similarity analysis based on the missing promotion types and the non-missing promotion types, identify similar promotion types among the non-missing promotion types, and generate the second promotion sequence list.

[0023] Example 3 Please see Figure 3As shown, the present invention is an electronic device, including a processor 301, a communication interface 302, a memory 303 and a communication bus 304, wherein the processor 301, the communication interface 302 and the memory 303 communicate with each other through the communication bus 304. Memory 303 is used to store computer programs; When processor 301 executes a program stored in memory 303, it performs the following steps: Step 1: Collect historical behavioral data of users when interacting with the promotional platform, and perform data preprocessing, including data cleaning, noise reduction, and missing value imputation; Step 2: Based on the preprocessed historical behavior data, the K-means clustering algorithm is used to perform cluster analysis on the user data according to the preset advertising type, and the clusters are output. Step 3: Based on any cluster, determine the peak time periods by performing time-series statistical analysis on users' historical behavior data; Step 4: Obtain existing promotional types and compare them with preset promotional types to determine if there are any missing promotional types. Generate a first promotional sequence list. If there are missing promotional types, perform feature similarity analysis between the missing promotional types and the non-missing promotional types to identify similar promotional types among the non-missing promotional types and generate a second promotional sequence list.

[0024] The communication bus mentioned in the above electronic devices can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus can be divided into address bus, data bus, control bus, etc. For ease of illustration, only one thick line is used to represent it in the diagram, but this does not indicate that there is only one bus or one type of bus.

[0025] The communication interface is used for communication between the aforementioned electronic devices and other devices.

[0026] The memory may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage device.

[0027] Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor.

[0028] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc. It can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0029] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. An interactive method for technology promotion, characterized in that: include: Collect historical behavioral data of users when interacting with the advertising platform, and perform data preprocessing, including data cleaning, noise reduction, and missing value imputation; Based on preprocessed historical behavior data, the K-means clustering algorithm is used to perform cluster analysis on user data according to the preset advertising type, and output clusters. Based on any cluster, popular time periods are determined by performing time-series statistical analysis on users' historical behavior data; Obtain existing promotional types and compare them with preset promotional types to determine if there are any missing promotional types. Generate a first promotional sequence list. If there are missing promotional types, perform feature similarity analysis between the missing promotional types and the non-missing promotional types to identify similar promotional types among the non-missing promotional types and generate a second promotional sequence list.

2. The interactive method for technology promotion according to claim 1, characterized in that: The process of outputting clusters is as follows: Historical behavioral data includes content click frequency, browsing duration, and social interaction frequency. The K-value is determined based on the preset number of promotional types. Initialization: Randomly select K feature vectors as initial cluster centers, corresponding to the initial features of the 3 types of advertising; Sample allocation: The distance from each user to the K centers is calculated using the Euclidean distance formula, and the user is assigned to the cluster of the nearest center; Cluster update: Calculate the mean of all user features within each cluster and use it as the new cluster center; Iterative convergence: Repeat the sample assignment and center update operations until the cluster centers no longer change significantly; After convergence, K clusters are obtained. Combining the historical behavior data of users within the clusters, the clusters are matched with the preset advertising types. Finally, the user list and corresponding advertising type of each cluster are output.

3. The interactive method for technology promotion according to claim 2, characterized in that: The process for determining popular time periods is as follows: The user's historical behavior data is sorted in chronological order, and the 24 hours are divided into several analysis periods with equal time intervals. By analyzing and processing the user's historical behavior data within the analysis period, the interaction representation value of the analysis period is determined. If the interaction representation value is greater than or equal to the interaction representation threshold, the corresponding analysis period is recorded as a suspected hot period. Sort the interaction values ​​of suspected popular time periods in descending order, and extract the suspected popular time period with the highest order, which is recorded as the popular time period corresponding to the promotion type.

4. The interactive method for technology promotion according to claim 3, characterized in that: The process for determining the interaction representation value for the analysis period is as follows: The click frequency within the statistical analysis period is then compared with a preset standard click frequency value to obtain the click frequency characteristic value for the analysis period. The browsing duration corresponding to each click within the analysis period is summed, and then compared with the duration corresponding to the analysis period to obtain the browsing duration feature value of the analysis period. The frequency of social interactions within the statistical analysis period is then compared with a preset standard value for the frequency of social interactions to obtain the characteristic value of the frequency of social interactions during the analysis period. The click frequency feature value, browsing duration feature value, and social interaction frequency feature value of the analysis period are weighted and summed to obtain the interaction characterization value of the analysis period.

5. The interactive method for technology promotion according to claim 4, characterized in that: The process of determining whether there are any missing promotional types compared to the preset promotional types is as follows: All preset advertising types are integrated into a preset advertising type sequence. Based on any existing advertising type, the preset advertising type sequence is traversed. If the same existing advertising type exists in the preset advertising type sequence, the corresponding preset advertising type is recorded as a non-missing advertising type. If there is no existing promotional type in the preset promotional type sequence, the corresponding preset promotional type will be recorded as a missing promotional type. The number of missing promotional types is counted. If the number of missing promotional types is greater than or equal to 1, it means that there are missing promotional types compared with the preset promotional types, and a missing signal is generated; otherwise, it means that there are no missing promotional types compared with the preset promotional types.

6. The interactive method for technology promotion according to claim 5, characterized in that: The process of generating the first promotion sequence list is as follows: Since there are no missing promotional types compared to the preset promotional types, all promotional types are sorted according to the time sequence of popular periods to obtain a promotional sequence list. Based on the promotional sequence list, the corresponding promotional types are set for popular periods and promotions are carried out.

7. The interactive method for technology promotion according to claim 5, characterized in that: The process of performing feature similarity analysis based on missing and non-missing advertising types is as follows: Based on the missing signal, the popular time periods of the missing promotional type and the corresponding popular time periods of the non-missing promotional type are extracted. If the popular time period of the missing promotional type is the same as the popular time period of the non-missing promotional type, the corresponding non-missing promotional type is recorded as the type to be analyzed. Extract the click frequency feature value, browsing duration feature value, and social interaction frequency feature value corresponding to the popular time periods of the type to be analyzed. Use the cosine similarity formula to calculate the click frequency feature value, browsing duration feature value, and social interaction frequency feature value corresponding to the popular time periods of the type to be analyzed and the missing promotional type to obtain the type similarity.

8. The interactive method for technology promotion according to claim 7, characterized in that: The process of generating the second promotion sequence list is as follows: If the type similarity is greater than or equal to the type similarity threshold, then during the peak period, the missing promotional type is replaced with the type to be analyzed. Based on the replaced promotional type, all promotional types are sorted according to the time order of the peak period to obtain the second promotional sequence list.

9. An interactive system for technology promotion, characterized in that, The system is used to perform the method according to any one of claims 1-8, the system comprising: Data preprocessing module: Collects historical behavioral data of users when interacting with the promotional platform and performs data preprocessing, including data cleaning, noise reduction, and missing value imputation; Clustering analysis module: Based on preprocessed historical behavior data, the K-means clustering algorithm is used to perform clustering analysis on user data according to the preset advertising type, and output clusters; Popular time period determination module: Based on any cluster, popular time periods are determined by performing time-series statistical analysis on users' historical behavior data; Promotion sequence list generation module: Obtain existing promotion types, compare and analyze them with preset promotion types, determine whether there are any missing promotion types compared with preset promotion types, and generate the first promotion sequence list; if there are missing promotion types, perform feature similarity analysis based on the missing promotion types and the non-missing promotion types, identify similar promotion types among the non-missing promotion types, and generate the second promotion sequence list.

10. A promotional cart for technology dissemination, characterized in that: It includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; Memory, used to store computer programs; A processor, when executing a program stored in memory, implements the steps of the method described in any one of claims 1-9.

Images