AI-based multi-media labor education content dynamic generation and distribution system and method

Abstract

The application relates to the technical field of education content generation and distribution, and discloses an AI-based multimedia labor education content dynamic generation and distribution system and method, which comprises a response module, a feedback module and an adjustment module. The method corresponds to the system. The application realizes dynamic and intelligent closed-loop iteration of labor education content generation and distribution, completes quantitative setting of parameters to make the propagation efficiency determination more accurate, provides reliable data support for subsequent analysis, improves the precision of potential audience feature extraction through a suboptimal data screening method of hierarchical clustering, makes the content and platform adjustment more targeted, integrates student group linkage behavior characteristics, makes the propagation combination highly adapt to the propagation and behavior characteristics of the student group, and effectively improves the group reach rate, participation and collaborative propagation effect of the labor education content.

Description

Technical Field

[0001] This application relates to the field of educational content generation and distribution technology, specifically an AI-based system and method for dynamic generation and distribution of converged media labor education content. Background Technology

[0002] In the current integration of media convergence technology with labor education, the systems for generating and distributing labor education content still suffer from numerous technical deficiencies, making it difficult to adapt to the actual needs of labor education. Existing systems often fail to generate suitable text, image, and video content tailored to labor education objectives. The combination of dissemination platforms and content media formats lacks a rational basis, easily leading to mismatches between content and dissemination channels. Furthermore, the systems only perform basic data collection after dissemination, without refined data filtering and mining, failing to identify effective data that reflects potential audiences. Optimization relies solely on full or optimal feedback data, resulting in a lack of targeted optimization. In addition, existing systems lack designs for adjusting dissemination combinations based on the characteristics of potential audiences, failing to achieve dynamic adaptation and optimization between labor education content and dissemination platforms. This hinders the formation of a complete closed loop of content generation, distribution, feedback, and adjustment, ultimately affecting the achievement of labor education objectives and compromising both dissemination efficiency and educational effectiveness. Summary of the Invention

[0003] The purpose of this application is to provide an AI-based system and method for dynamic generation and distribution of labor education content in converged media, in order to solve the technical problems in the existing technology that lack the design of adjusting the dissemination combination based on the characteristics of potential audience groups, cannot achieve dynamic adaptation and optimization of labor education content and dissemination platforms, and are difficult to form a complete closed loop of content generation, distribution, feedback and adjustment, ultimately affecting the achievement of labor education goals and making it difficult to guarantee dissemination efficiency and educational effectiveness.

[0004] To achieve the above objectives, this application provides an AI-based system for dynamically generating and distributing converged media labor education content, the system comprising:

[0005] The response module is configured to: respond to the input labor education objectives and output a communication combination, which includes labor education content and a communication platform; wherein: the labor education content is based on the labor education objectives and is obtained through AI content generation technology, and its media forms include text and / or video;

[0006] The feedback module is configured to: collect feedback data corresponding to different communication combinations, which is used to characterize the achievement of labor education goals; and filter the feedback data based on a preset threshold range to obtain suboptimal feedback data, which is used to map potential audience groups.

[0007] The adjustment module is configured to: identify potential audience groups based on suboptimal feedback data, and adjust the communication mix based on these potential audience groups to optimize the dynamic generation and distribution of labor education content; wherein, adjusting the communication mix includes adjusting labor education content and / or the communication platform.

[0008] Preferably, the operation of the response module includes:

[0009] The labor education objectives are obtained, including educational objectives and practical objectives, and the labor education objectives are set with corresponding completion parameters, which are used to quantify the achievement of the labor education objectives;

[0010] Based on the aforementioned labor education objectives, a pre-set AI content generation platform is invoked to output labor education content;

[0011] Based on the media format of the labor education content and the completion parameters, determine the dissemination platform;

[0012] The labor education content and corresponding dissemination platforms are combined to output the dissemination combination.

[0013] Preferably, the operation of the feedback module includes:

[0014] Collect feedback data corresponding to different propagation combinations and determine the corresponding completion status; wherein the collection of the feedback data corresponds to the completion parameter;

[0015] The feedback data is sorted based on the completion status, and the target completion level is divided according to the completion parameters of the labor education goals.

[0016] The feedback data that falls into the first and last tiers of the target completion level are removed. The remaining feedback data corresponding to the target completion level are then used to filter out the second-best feedback data that falls into the second-best tier.

[0017] Preferably, the operation of the adjustment module includes:

[0018] Receive suboptimal feedback data and extract the potential audience characteristics associated with the data. The potential audience characteristics include at least the audience's preference for media forms, focus of content interest, interactive behavior habits, and cognitive acceptance.

[0019] The content of labor education is adjusted based on the characteristics of the potential audience, without adding any new labor education objectives;

[0020] Simultaneously adjust the communication platform based on the characteristics of the potential audience, including at least switching to a communication platform where the potential audience is more active;

[0021] The combined and adjusted labor education content and dissemination platform will output new dissemination combinations, optimize the dynamic generation and distribution of labor education content, and await the next round of feedback data collection and iteration.

[0022] Preferably, the completion parameter is a collection indicator used to quantify the efficiency of labor education dissemination;

[0023] For educational objectives, the data collection metrics should include at least the number of valid views and the percentage of time spent on the content.

[0024] For practical objectives, the data collection indicators should include at least the number of participants and the number of records of group participation.

[0025] Preferably, the filtering of the suboptimal feedback data includes:

[0026] Input the completion parameters corresponding to the labor education goals into the preset large language model, and construct the first prompt word through the large language model. The first prompt word is used to clarify the type of feedback data collection.

[0027] Based on the first prompt word, feedback data generated by different propagation combinations on the corresponding propagation platforms is collected. The collected feedback data is correlated with the completion parameters to obtain the completion status of each propagation combination.

[0028] Input the completion status into the large language model, and construct a second cue word through the large language model. The second cue word is used to define the audience feature dimensions corresponding to different completion statuses.

[0029] Based on the audience characteristic dimensions defined by the second prompt word, cluster analysis was performed on the feedback data to obtain multiple feedback data clusters corresponding to different audience groups;

[0030] The feedback data clusters within each suboptimal level are sorted independently according to their completion status. The data clusters at the beginning and end of the sorting within each suboptimal level are removed, and the feedback data corresponding to the remaining data clusters are defined as suboptimal feedback data.

[0031] Preferably, the characteristics of the potential audience also include student group linkage behavior characteristics, which include at least: student group gathering scene identifiers, including at least dormitories, classes or clubs; synchronous access time of the group; and group-related interaction trajectories, including at least the records of multiple people in the same group clicking and forwarding in succession.

[0032] The adjustment module adjusts the dissemination combination based on the characteristics of student group interaction behavior. Specifically, it adds group collaboration guidance when generating labor education content, pushes content in a targeted manner according to the time period when the group gathers, and opens a group sharing portal on the dissemination platform.

[0033] To achieve the above objectives, this application also provides an AI-based method for dynamically generating and distributing converged media labor education content, applied to the AI-based converged media labor education content dynamic generation and distribution system described above. The method includes:

[0034] In response to the input labor education goals, an output communication mix is ​​generated, which includes labor education content and a communication platform; wherein: the labor education content is based on the labor education goals, obtained through AI content generation technology, and its media forms include text and / or video;

[0035] Feedback data corresponding to different communication combinations is collected, and this feedback data is used to characterize the achievement of labor education goals; based on a preset threshold range, suboptimal feedback data is obtained from the feedback data, and this suboptimal feedback data is used to map potential audience groups.

[0036] Based on suboptimal feedback data, potential audience groups are identified, and the communication mix is ​​adjusted based on these potential audience groups to optimize the dynamic generation and distribution of labor education content; wherein, adjusting the communication mix includes adjusting labor education content and / or communication platforms.

[0037] To achieve the above objectives, this application also provides a computer device for dynamically generating and distributing converged media labor education content based on AI, including at least one processor, at least one memory, and a data bus;

[0038] The processor and the memory communicate with each other via the data bus;

[0039] The memory stores program instructions that can be executed by the processor, which calls the program instructions to execute the AI-based method for dynamically generating and distributing converged media labor education content as described above.

[0040] To achieve the above objectives, this application also provides a storage medium storing a computer program thereon, which, when executed by a processor, implements the AI-based method for dynamically generating and distributing converged media labor education content as described above.

[0041] Beneficial effects: The AI-based integrated media labor education content dynamic generation and distribution system and method of this application realizes dynamic and intelligent closed-loop iteration of labor education content generation and distribution. The quantitative setting of parameters makes the judgment of dissemination efficiency more accurate and provides reliable data support for subsequent analysis. The suboptimal data screening method of hierarchical clustering improves the accuracy of potential audience feature extraction and makes the adjustment of content and platform more targeted. The integration of student group linkage behavior characteristics makes the dissemination combination highly adapted to the dissemination and behavior characteristics of the student group, effectively improving the group reach, participation and collaborative dissemination effect of labor education content. Attached Figure Description

[0042] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0043] Figure 1 A structural block diagram of an AI-based integrated media labor education content dynamic generation and distribution system provided in this application embodiment;

[0044] Figure 2 A flowchart illustrating the operation of the response module provided in this application embodiment;

[0045] Figure 3 A flowchart illustrating the operation of the feedback module provided in this embodiment of the application;

[0046] Figure 4 A flowchart illustrating the operation of the response module provided in this application embodiment;

[0047] Figure 5 A flowchart illustrating the filtering of suboptimal feedback data provided in the embodiments of this application;

[0048] Figure 6 A flowchart illustrating the AI-based method for dynamically generating and distributing converged media labor education content, as provided in this application embodiment.

[0049] The implementation, functional features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0050] The technical solutions in the embodiments of this application will be clearly and completely described below. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0051] In this document, the term "comprising" is intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0052] Reference Figure 1 , Figure 1 This is a structural block diagram of an AI-based integrated media labor education content dynamic generation and distribution system provided in an embodiment of this application.

[0053] like Figure 1 As shown in the figure, this embodiment discloses an AI-based system for dynamically generating and distributing converged media labor education content, including:

[0054] The response module 10 is configured to: respond to the input labor education objectives and output a communication combination, which includes labor education content and a communication platform; wherein: the labor education content is based on the labor education objectives and is obtained through AI content generation technology, and its media form includes text and / or video.

[0055] Reference Figure 2 , Figure 2 A flowchart illustrating the operation of the response module provided in this embodiment of the application.

[0056] Specifically, such as Figure 2 As shown, the operation of the response module includes:

[0057] S101: Obtain labor education objectives. Labor education objectives include educational objectives and practical objectives. Each labor education objective has corresponding completion parameters, which are used to quantify the achievement of the labor education objectives.

[0058] In this specific application, the response module 10 obtains the labor education goals set by the staff through the existing education terminal input interface. These goals are clearly divided into educational goals and practical goals, and corresponding completion parameters are configured for each type of goal. These completion parameters serve as core indicators for quantifying the achievement of labor education goals and are simultaneously entered into the system database. In a specific example, the educational goal might be set as popularizing daily labor norms and basic knowledge of professional labor skills, with completion parameters including effective content views and content dwell time percentage. The practical goal might be set as conducting campus volunteer services and community public welfare volunteer activities, with completion parameters including the actual number of participants in volunteer activities and the number of volunteer group service records. A unique association is established between the completion parameters and the corresponding labor education goal, clearly defining quantitative standards and providing a unified and reusable quantitative reference for subsequent AI content generation, dissemination platform matching, and feedback data collection.

[0059] S102: Based on the goals of labor education, call the preset AI content generation platform to output labor education content.

[0060] In the specific application of this embodiment, the response module 10 first retrieves the recorded general labor education goals and related information from the system data platform, and then sends a goal generation instruction to the preset AI content generation platform. This platform is equipped with existing AI text and image generation and short video generation technologies, and can output labor education content adapted to the media format according to the type of education goal. For educational goals that popularize daily labor norms and basic knowledge of professional labor literacy, the platform generates supporting popular science texts and animated short videos; for practical goals that carry out campus volunteer services and community public welfare labor, the platform generates volunteer activity process guidance texts and short videos for practical teaching of public welfare labor. The generated content all meet the general teaching requirements of labor education in higher vocational colleges, and after being processed in a standardized format, it is uniformly output to the system content resource library for subsequent steps to call.

[0061] S103: Determine the dissemination platform based on the media format and completion parameters of labor education content.

[0062] In the specific application of this embodiment, the response module 10 first retrieves the labor education content generated in S102 from the system content resource library, clarifies its media format (such as text, images, and short videos), and then retrieves the completion parameters set in S101 from the data platform. Combining these two parameters, it comprehensively matches and adapts the dissemination platform to ensure that the dissemination platform can adapt to both the content media format and accurately collect the feedback data corresponding to the completion parameters. In a specific example, for educational goals such as popular science text and animated short videos, to collect effective play counts and dwell time percentages, the vocational college campus administration WeChat official account (adapted to text push) and the Douyin campus official account (adapted to short video dissemination, facilitating the statistics of play counts and dwell time) are selected. For practical goals such as volunteer process text and practical short videos, to collect the number of participants and linkage service records, the campus volunteer service mini-program (adapted to process push and registration statistics) and the class enterprise WeChat group (adapted to short video forwarding and linkage record retention) are selected. Finally, the dissemination platform corresponding to each content is determined, forming a preliminary matching relationship.

[0063] S104: Combine labor education content with corresponding dissemination platforms to output dissemination combinations.

[0064] In the specific application of this embodiment, the response module 10 first synchronously retrieves the labor education content generated in S102 and the corresponding dissemination platform matched in S103 from the system content resource library and platform matching library. A standardized dissemination combination is constructed according to the unique "content-platform" adaptation relationship. This combination is the core basic unit for subsequent system distribution and feedback data collection, and can be represented by a mathematical expression as: Dissemination Combination ,in, For the first One aspect of labor education, for Corresponding dissemination platforms. In a specific example, daily labor regulations popular science articles and the campus academic affairs WeChat official account form one dissemination combination; vocational labor literacy animated short videos and the official campus Douyin account form another; volunteer activity flowcharts and texts and the campus volunteer service mini-program form a third; and public welfare labor practice short videos and class WeChat groups form yet another. All dissemination combinations are uniformly entered into the system distribution database, and the associated labor education goals and completion parameters are marked, providing a basis for subsequent accurate distribution and targeted collection of feedback data.

[0065] Feedback module 20 is configured to: collect feedback data corresponding to different communication combinations, which is used to characterize the achievement of labor education goals; and filter the feedback data based on a preset threshold range to obtain suboptimal feedback data, which is used to map potential audience groups.

[0066] Reference Figure 3 , Figure 3 A flowchart illustrating the operation of the feedback module provided in this embodiment of the application.

[0067] Specifically, such as Figure 3 As shown, the operation of the feedback module includes:

[0068] S201: Collect feedback data corresponding to different propagation combinations and determine the corresponding completion status; wherein, the collection of feedback data corresponds to the completion parameters.

[0069] In the specific application of this embodiment, the feedback module 20 first retrieves all output communication combinations from the system distribution library, along with the completion parameters of the labor education goals associated with each combination, establishing a one-to-one correspondence between the feedback data collection dimensions and the completion parameters. Then, it collects feedback data for different communication combinations through the backend data interfaces of each communication platform. In a specific example, for educational communication combinations, it collects the effective readership and content dwell time percentage of daily labor norms articles from the campus academic affairs WeChat official account, and the effective play count and viewing dwell time percentage of short videos on professional labor literacy from the Douyin campus official account. For practical communication combinations, it collects the actual number of participants and the number of group linkage service records corresponding to volunteer process flowcharts from the campus volunteer service mini-program, and the number of volunteer participants and linkage service check-in records corresponding to public welfare practical short videos from the class enterprise WeChat group. After collection, the data is cleaned and deduplicated, and the goal completion status of each communication combination is calculated based on the quantitative standards of the completion parameters, and then synchronously entered into the system feedback database.

[0070] S202: Sort the feedback data based on the completion status, and classify the target completion level corresponding to the feedback data in combination with the completion parameters of labor education goals.

[0071] In the specific application of this embodiment, the feedback module 20 first retrieves the completion status and corresponding feedback data of each propagation combination from the feedback database, sorts them globally from high to low according to the comprehensive value of the completion status, and then combines the labor education goal completion parameters set in S101, presets multiple quantitative thresholds, and divides the goal completion level corresponding to the feedback data.

[0072] In the specific application of this embodiment, the target completion level can be characterized as: ,in: The overall value of the communication portfolio's goal achievement. This is a grading function. In a specific example, the grading function divides data into five levels based on the completion rate of parameters: ≥80% is grade G1, 60% ≤ <80% is grade G2, 40% ≤ <60% is grade G3, 20% ≤ <40% is grade G4, and <20% is grade G5. For example, a 75% completion rate for the campus administration WeChat official account's text and image combination is classified as grade G2, and a 50% completion rate for the class enterprise WeChat group's short video combination is classified as grade G3. All communication combinations are labeled with their corresponding grades and simultaneously entered into the system's grading feedback database, providing a grading basis for subsequent selection of suboptimal data.

[0073] S203: Remove feedback data that are at the first and last level of target completion, and filter out the second-best feedback data that are at the second-best level from the remaining feedback data corresponding to the target completion level.

[0074] In the specific application of this embodiment, the feedback module 20 first retrieves the feedback data and labeled target completion levels of all communication combinations from the system's hierarchical feedback library. According to preset filtering rules, it directly removes feedback data at the first (G1) and last (G5) target completion levels. These two types of data correspond to the best and worst communication combinations in terms of target completion effect, and have no reference value for potential audience mining. In a specific example, after removing the G1 level data (≥80% achievement rate) and the G5 level data (<20% achievement rate), the remaining feedback data corresponding to the G2 level (60%≤achievement rate<80%), G3 level (40%≤achievement rate<60%), and G4 level (20%≤achievement rate<40%) are uniformly defined as suboptimal feedback data. For example, G2 level data for campus academic affairs WeChat official account text and image combinations and G3 level data for class enterprise WeChat group short video combinations are included in the scope. All suboptimal feedback data are entered into the system's dedicated library for subsequent adjustment module retrieval and analysis.

[0075] The adjustment module 30 is configured to: identify potential audience groups based on suboptimal feedback data, and adjust the communication mix based on the potential audience groups to optimize the dynamic generation and distribution of labor education content; wherein, adjusting the communication mix includes adjusting the labor education content and / or the communication platform.

[0076] Reference Figure 4 , Figure 4 A flowchart illustrating the operation of the response module provided in this embodiment of the application.

[0077] Specifically, such as Figure 4 As shown, adjusting the module's operation includes:

[0078] S301: Receive suboptimal feedback data and extract the characteristics of potential audience groups associated with the data. The characteristics of potential audience groups include at least the audience's preference for media forms, focus of content interest, interactive behavior habits, and cognitive acceptance.

[0079] In the specific application of this embodiment, the adjustment module 30 first retrieves the filtered feedback data corresponding to G2, G3, and G4 levels from the dedicated database of suboptimal feedback data. Based on the correlation logic between data dimensions and audience behavior, it extracts features from the data to obtain four core characteristics of the potential audience. In a specific example, the extracted features may be as follows: In terms of media format preference, they prefer short video content rather than pure text and image content, adapting to fragmented viewing habits; in terms of content interest focus, they pay more attention to practical skills-based content on campus volunteer services, and have lower attention to purely theoretical labor norms; in terms of interactive behavior habits, they mostly click on content during evening after-school hours, and there is a characteristic of viewing in groups; in terms of cognitive acceptance, they have a high acceptance of simple and straightforward explanations, and a low acceptance of lengthy text descriptions. After the extracted features are structured, they are entered into the system feature database, providing a precise basis for subsequent adjustments to the dissemination combination.

[0080] S302: Adjust the content of labor education based on the characteristics of the potential audience, without adding new labor education objectives.

[0081] In the specific application of this embodiment, the adjustment module 30 first retrieves the potential audience feature set extracted in S301, and simultaneously retrieves the initial labor education goals, strictly adhering to the principle of not adding new labor education goals. Based on the features, the original labor education content is optimized for suitability, and the adjusted labor education content is then... This can be expressed mathematically as: ,in, This refers to the labor education content in the original dissemination package. The target completion level corresponds to the suboptimal feedback data. For the set of characteristics of the potential audience, The mapping function is optimized based on levels and features. In a specific example, considering the audience's preference for short videos, focus on practical volunteer service, and acceptance of straightforward explanations, the original plain text and graphics of daily labor norms and theoretical text and graphics of professional labor literacy are optimized into animated short videos with simplified theoretical content and real-life case studies. The text and graphics of campus volunteer service processes are upgraded to practical demonstration short videos, and lengthy textual explanations are removed. All adjustments are based on the initial labor education goal of "popularizing labor norms and carrying out volunteer activities" and are entered into the updated content resource library of the system.

[0082] In the specific application of this embodiment, To adapt and optimize the mapping function based on the original labor education content, target completion levels, and potential audience characteristics, the process includes four stages: parameter quantification and analysis, feature weight adaptation, level association verification, and target consistency verification. In a specific example, quantified assignment reflects the pattern that level G2 represents the core target with the greatest potential for dissemination and improvement among the suboptimal options. Specifically, it includes:

[0083] When performing parameter quantization analysis, the input parameters are standardized and quantified, and the original labor education content is converted into quantization. Break it down into three quantitative dimensions: media format (e.g., text / short video), content type (e.g., theory / practice), and information density (e.g., text ratio / demonstration ratio); and classify the goal completion level. The potential for improvement is prioritized and quantified into numerical values: G2 (60% ≤ achievement rate < 80%, highest potential for improvement) is assigned a value of 1, G3 (40% ≤ achievement rate < 60%, medium potential) is assigned a value of 2, and G4 (20% ≤ achievement rate < 40%, low potential) is assigned a value of 3. Corresponding weights are also assigned to each level: G2 0.6, G3 0.3, and G4 0.1. The characteristics of the potential audience are then aggregated. The four features of media preference, content interest focus, interactive behavior habits, and cognitive acceptance are assigned weight coefficients of 0.3, 0.3, 0.2, and 0.2 respectively to form a standardized feature vector.

[0084] When performing feature weight adaptation, combine Adjustment weights and eigenvectors, for Differentiated adaptation adjustments will be made, prioritizing resource optimization for G2-tier content. For example: if it is G2-tier (weight 0.6), then... Precise optimization is performed based on features. For example, if the media preference is short videos, the video will be converted to a 1-2 minute edited version. If the practical interest weight is ≥0.3, the proportion of the demonstration clip will be increased to 70%. If it is G3 (weight 0.3), then... Make basic optimizations, such as only adding core practical steps; if it is G4 grade (weight 0.1), only make simplification and adaptation, such as removing unnecessary content and retaining the core points.

[0085] When performing grade association verification, based on The quantitative identification settings are optimized for greater depth, enhancing the efficiency of the G2 level: If (i.e., G2 level), retaining over 90% of the core knowledge points of the original content, and adding 1 to 2 highly relevant interactive nodes (such as embedding practical skills pop-ups in short videos); if (i.e., G3 level), retaining over 70% of the core knowledge points, with only new basic interactive guidance added; if (i.e., G4 format) uses a storyboard-style breakdown to reduce information density, splitting a single piece of content into two sub-segments without adding any additional interactive design.

[0086] When verifying the consistency of objectives, the adjusted labor education content should be considered. ,check Does it only focus on the initial general labor education goals (popularizing daily labor norms and carrying out campus / community volunteer activities), ensuring that no new educational or practical goals are added? After verification, the adjusted version will be output. .

[0087] In one specific example of this embodiment, the original labor education content It is a "purely graphic and textual standard for daily labor practices". (i.e., G2 level, with a compliance rate of 75% and an adjustment weight of 0.6). The feature vector is These correspond to short video preferences, interest in practical volunteer service, evening interaction habits, and acceptance of straightforward explanations, respectively. After mapping, output The standard is to "edit a 1-2 minute animated short video, retaining the core norms of 'dormitory housekeeping and public area cleaning' (92%), and adding real-life campus demonstration clips (72%), removing redundant theoretical text, retaining only 3 core operational steps, and embedding an interactive pop-up window for 'housekeeping tips' at the 1-minute mark of the video"; while the second-best G3 level (50% compliance rate) "professional labor literacy theory and text" is only optimized to "a simplified version of the text and text, supplemented with 2 practical points of professional etiquette", without any in-depth adjustments, fully reflecting the priority improvement strategy of G2 level, and none of the adjustments have added labor education goals.

[0088] S303: Simultaneously adjust the communication platform based on the characteristics of the potential audience, including at least switching to a communication platform where the potential audience is more active.

[0089] In the specific application of this embodiment, the adjustment module 30 synchronously retrieves the matching relationship between the potential audience characteristics extracted in S301 and the original dissemination platform. Based on the core rule of switching to a platform with higher potential audience activity, it makes adaptive adjustments to the dissemination platform to ensure synergistic optimization with the content adjustments in S302. In a specific example, for the potential audience characteristics of the G2 level "Daily Labor Norms" content (preference for short videos, active during evening after-school hours), the original dissemination platform "Campus Academic Affairs Official Account (text and image push, mainly traffic during lunchtime)" is switched to "Douyin Campus Official Account (short video dissemination, evening user activity increased by more than 40%)". For the potential audience characteristics of the G2 level "Campus Volunteer Service" content (habit of group collaboration), the original dissemination platform "Class Enterprise WeChat Group (mainly notification reach)" is switched to "Campus Volunteer Service Mini Program Dynamic Activity Section (supports team registration, joint check-in, daily user interaction increased by 35%)". All adjustments revolve around the initial labor education goals and form a new dissemination combination with the adjusted content, laying the foundation for improving the goal completion rate in subsequent distribution.

[0090] S304: Combine and adjust the labor education content and dissemination platform, output new dissemination combinations, optimize the dynamic generation and distribution of labor education content, and await the next round of feedback data collection and iteration.

[0091] In the specific application of this embodiment, the adjustment module 30 synchronously retrieves the optimized labor education content (S302) and the adjusted dissemination platform (S303) based on potential audience characteristics from the updated content resource library and platform matching library. Following the "adjusted content - high-activity platform" adaptation principle, it constructs a new standardized dissemination combination. In a specific example, a 1-2 minute animated short video of daily labor regulations is combined with the official Douyin campus account, and a short video demonstrating practical campus volunteer services is combined with the dynamic activity section of the campus volunteer service mini-program, forming multiple new dissemination combinations. All new combinations are uniformly entered into the system's optimized distribution library, completing the full-link dynamic optimization of labor education content generation and distribution, and achieving the first round of dissemination combination iteration. The system maintains the normal distribution status of the new combinations, awaiting the targeted collection and analysis of the next round of feedback data, entering a continuous closed-loop iterative optimization cycle, gradually improving the completion rate of labor education goals.

[0092] In existing technologies, there is a lack of unified and targeted data collection indicators for quantifying the efficiency of labor education dissemination, which can easily lead to biases in the judgment of goal completion. Therefore, this embodiment designs a completion parameter feature to solve this problem. This completion parameter is a data collection indicator used to quantify the efficiency of labor education dissemination, and the corresponding core data collection indicators are clearly defined according to the attribute differences between educational and practical goals.

[0093] Specifically, the completed parameters are the data collection indicators used to quantify the efficiency of labor education dissemination;

[0094] For educational objectives, the data collection metrics should include at least the number of valid views and the percentage of time spent on the content.

[0095] For practical objectives, the data collection indicators should include at least the number of participants and the number of records of group participation.

[0096] In this specific application, the completion parameters are core data collection indicators specifically used to quantify the efficiency of labor education dissemination. These are precisely categorized according to the differences in attributes between educational and practical labor education goals. All indicator data is collected from the corresponding dissemination platform's backend, serving as the sole basis for quantifying goal completion. For educational goals aimed at popularizing daily labor norms and basic knowledge of professional labor ethics, the core data collection indicators are the percentage of effective plays and content dwell time. Effective plays exclude invalid data such as idle plays and quick swipes, only counting actual plays with a viewing time ≥ 30 seconds. The content dwell time percentage is the percentage of the audience's actual dwell time to the total content duration. For practical goals involving campus volunteer services and community public service, the core data collection indicators are the number of participants and the number of group participation records. The number of participants excludes duplicate registrations, only counting the actual number of attendees. The number of group participation records refers to valid service check-in records from groups of 3 or more. The quantitative data of these indicators serve as core data support for subsequent goal completion level classification and dissemination combination optimization.

[0097] In existing technologies, the selection of suboptimal feedback data often relies on a single numerical sorting method, lacking correlation analysis of audience characteristics. This can easily lead to insufficient reference value of the selected data, making it difficult to support subsequent precise optimization. This embodiment addresses this problem by combining existing large language models and cluster analysis to optimize the suboptimal feedback data selection logic.

[0098] Reference Figure 5 , Figure 5 A flowchart illustrating the filtering of suboptimal feedback data provided in this application embodiment.

[0099] Specifically, such as Figure 5 As shown, the filtering of suboptimal feedback data includes:

[0100] A10: Input the completion parameters corresponding to the labor education goals into the preset large language model, and construct the first prompt word through the large language model. The first prompt word is used to clarify the type of feedback data collection.

[0101] In the specific application of this embodiment, the feedback module 20 first retrieves the completion parameters set for the labor education goals in S101 from the data platform, standardizes and organizes them, and then inputs them into a preset large language model. Based on the attributes and quantitative requirements of the completion parameters, the large language model constructs a first prompt word to clarify the type of feedback data collection, and defines the collection scope in the prompt word to ensure the accuracy of subsequent data collection. In a specific example, after inputting the completion parameters of the education-type goal, "effective playback volume and content dwell time percentage", into the large language model, the first prompt word generated is "collect effective playback data and user content dwell time percentage data of labor education theory content. Effective playback volume is defined as playback records with a viewing time of ≥30 seconds, and dwell time percentage is the percentage of the user's actual dwell time to the total content duration". After inputting the completion parameters of the practical-type goal, "number of participants in activities and number of group linkage participation records", the first prompt word generated is "collect actual number of participants in labor education volunteer activities and number of group linkage service records. Participants are defined as the actual number of sign-ins, and linkage records are the number of effective check-in records of groups of 3 or more people". Once the first prompt word is generated, it is entered into the system's data collection terminal to serve as a unified standard for subsequent feedback data collection. This avoids invalid data collection from the source and ensures the reference value of subsequent suboptimal data.

[0102] A20: Based on the first prompt word, collect feedback data generated by different propagation combinations on the corresponding propagation platforms, and correlate the collected feedback data with the completion parameters to obtain the completion status of each propagation combination.

[0103] In the specific application of this embodiment, the feedback module 20 first retrieves the first prompt word generated by A10 as a unified collection standard. Through the backend data interface of each propagation platform, it collects feedback data of different propagation combinations in a targeted manner, strictly removes invalid data according to the prompt word criteria, and then performs correlation calculation with the standardized collection data and preset completion parameters to obtain the target completion status of each propagation combination. The mathematical expression of the completion status can be represented as: ,in, For the first The actual values ​​of the collected indicators For the first The item completes the preset parameter values. This represents the total number of indicators. In a specific example, for the labor regulations short video on the official Douyin campus account, the actual number of valid views was 800 (preset value 1000), and the actual dwell time percentage was 70% (preset value 80%), resulting in a completion rate of 75%. For the volunteer service mini-program's volunteer practice content, the actual number of participants was 150 (preset value 200), and the actual number of linked records was 30 (preset value 40), also resulting in a completion rate of 75%. The completion status of all dissemination combinations is annotated with related information and entered into the system feedback database, providing a precise quantitative basis for subsequent level classification.

[0104] A30: Input the completion status into the large language model, and construct a second cue word through the large language model. The second cue word is used to define the audience feature dimensions corresponding to different completion statuses.

[0105] In the specific application of this embodiment, the feedback module 20 first retrieves the completion status of each dissemination combination calculated by A20 from the feedback database. It then standardizes and organizes these completion status datasets into three levels: G2 (60% ≤ completion rate < 80%), G3 (40% ≤ completion rate < 60%), and G4 (20% ≤ completion rate < 40%). These datasets are then associated with the corresponding labor education goal types and input into the large language model. Based on the differences in completion status levels and content attributes, the large language model constructs a second prompt word to define the audience characteristic dimensions, clarifying the association logic between the feature analysis dimensions and the completion status. In a specific example, after inputting the completion status datasets for education and practical categories, the second prompt word is generated as: "For the completion status of labor education content dissemination in levels G2, G3, and G4, analyze audience characteristics from four core dimensions: media form preference, content interest focus, interactive behavior habits, and cognitive acceptance. Each dimension needs to match the actual dissemination data of the corresponding level to define the characteristic differences of the audience at different levels." The second prompt word is entered into the system's feature extraction end as a unified dimensional standard for subsequent cluster analysis.

[0106] A40: Based on the audience characteristic dimensions defined by the second prompt word, cluster analysis is performed on the feedback data to obtain multiple feedback data clusters corresponding to different audience groups.

[0107] In this specific application, the feedback module 20 first retrieves the second prompt word generated by A30 to clarify four audience characteristic analysis dimensions: media form preference, content interest focus, interactive behavior habits, and cognitive acceptance. Then, it retrieves the standardized and cleaned feedback data and uses the K-means clustering algorithm to group the feedback data according to the similarity of each dimension's features. Feedback data with similar features are grouped into the same cluster, forming multiple feedback data clusters corresponding to different audience groups. In a specific example, after cluster analysis, three core data clusters are obtained: Cluster 1 corresponds to an audience group that prefers short videos, focuses on practical content, is active during evening after-school hours, and accepts straightforward explanations; Cluster 2 corresponds to an audience group that accepts both text / images and short videos, balances theory and practice, is active during lunchtime, and has moderate cognitive acceptance; Cluster 3 corresponds to an audience group that prefers pure text / images, focuses on theoretical content, browses during fragmented time slots, and has high text acceptance. Each data cluster is labeled with the associated dissemination combination completion level, providing accurate group-based data support for subsequent sorting by completion status and removing the first and last data clusters.

[0108] A50: Sort each feedback data cluster according to its completion status, remove the data clusters at the top and bottom of the sort, and define the feedback data corresponding to the remaining data clusters as the second-best feedback data.

[0109] In the specific application of this embodiment, the feedback module 20 first retrieves the feedback data clusters obtained by clustering A40 according to the four major audience characteristic dimensions, and completes the hierarchical classification according to each sub-optimal level of G2, G3, and G4. Then, for the feedback data clusters within each level, they are sorted from high to low according to the comprehensive value of completion status. The clusters at the beginning and end of the sorting within each sub-optimal level are removed. Finally, the feedback data of the remaining clusters in each level are integrated and defined as the updated sub-optimal feedback data. In a specific example, within the G2 tier (60% ≤ achievement rate < 80%), three data clusters were identified, with completion rates of 78%, 72%, and 65% respectively. After removing the first and last clusters, the remaining cluster was 72%. Within the G3 tier (40% ≤ achievement rate < 60%), the three data clusters had completion rates of 58%, 52%, and 45%, respectively. After removing the first and last clusters, the remaining cluster was 52%. Within the G4 tier (20% ≤ achievement rate < 40%), the three data clusters had completion rates of 38%, 32%, and 25%, respectively. After removing the first and last clusters, the remaining cluster was 32%. The feedback data corresponding to these retained clusters from these three tiers were integrated to form refined suboptimal feedback data, which was then entered into the system's dedicated suboptimal data library. This provides more accurate data support for subsequent extraction of potential audience characteristics, tailored to each tier's audience.

[0110] Existing technologies, when identifying potential audience characteristics, do not specifically focus on the collaborative behavioral attributes of student groups. This makes it difficult to adapt communication mix adjustments to the collaborative communication characteristics of student groups, thus limiting optimization effectiveness. This embodiment addresses this issue by enriching the dimensions of potential audience characteristics and clarifying the rules for adjusting communication mix based on these characteristics.

[0111] Specifically, the characteristics of the potential audience also include the characteristics of student group interaction behavior, which include at least: the scene of student group gathering, including at least dormitory, class or club; the time period of synchronous access by the group; and the interaction trajectory of the group, including at least the records of multiple people in the same group clicking and forwarding.

[0112] The adjustment module adjusts the dissemination combination based on the characteristics of student group interaction behavior. Specifically, it adds group collaboration guidance when generating labor education content, pushes content in a targeted manner according to the time period when the group gathers, and opens a group sharing portal on the dissemination platform.

[0113] In the specific application of this embodiment, the potential audience characteristics are supplemented with student group collaborative behavior characteristics, specifically including three core features: student group gathering scene identifiers, synchronized access time periods, and group-related interaction trajectories. Gathering scene identifiers cover dormitories, classes, and clubs, while the related interaction trajectories are records of clicks and forwards by multiple individuals within the same group. The adjustment module optimizes the dissemination combination based on these features: for the clustered dormitory and class groups, it adds group collaboration guidance such as "collaborative dormitory tidying" and "class volunteer activity team formation" when generating labor education content; it also performs targeted push notifications on corresponding dissemination platforms based on the peak time of synchronized group access (8-10 PM); and it opens exclusive group sharing portals for classes and clubs on the campus volunteer service mini-program and Douyin campus account, supporting one-click sharing to group communities. This adjustment accurately adapts to the collaborative dissemination characteristics of student groups, effectively improving the group reach and participation of labor education content.

[0114] Based on the above, this embodiment forms a progressive technical solution around an AI-based integrated media labor education content dynamic generation and distribution system. It breaks through the technical bottlenecks of traditional labor education content static generation, single-channel distribution, and extensive feedback optimization. Specifically, it constructs an overall system architecture for AI-driven integrated media labor education content dynamic generation, distribution, and closed-loop iteration, laying the core framework for subsequent technical optimization; it further refines the core steps of dynamic generation, distribution, and optimization of each module of the system, forming a full-link technical system of "generation-distribution-feedback-optimization," solving the problems of existing technology content being disconnected from distribution and lacking a continuous iteration mechanism. Building upon this foundation, the system innovatively sets quantitative dissemination efficiency completion parameters based on differentiated educational and practical objectives, resolving the technical issues of existing technologies lacking unified and accurate data collection indicators for labor education dissemination efficiency and exhibiting biased judgments of completion status. Furthermore, it creatively combines large language model prompt word construction with hierarchical cluster analysis to achieve refined filtering of suboptimal feedback data, overcoming the limitations of traditional single-value sorting and insufficient data reference, making suboptimal data more aligned with actual optimization needs. Finally, it specifically mines the characteristics of student group collaborative behavior and clarifies corresponding dissemination combination adjustment rules, filling the gap in existing technologies that do not consider the collaborative dissemination attributes of student groups in audience characteristic dimensions, achieving precise adaptation to student groups.

[0115] In summary, this embodiment's AI-based dynamic generation and distribution system for labor education content in converged media achieves dynamic and intelligent closed-loop iteration of labor education content generation and distribution. Quantitative parameter settings allow for more accurate assessment of dissemination efficiency, providing reliable data support for subsequent analysis. The hierarchical clustering suboptimal data filtering method improves the accuracy of potential audience feature extraction, making content and platform adjustments more targeted. The integration of student group interaction behavior characteristics ensures that the dissemination combination is highly adapted to the dissemination and behavioral characteristics of the student group, effectively improving the group reach, participation, and collaborative dissemination effect of labor education content. Overall, this embodiment makes the generation of labor education content in the converged media environment more aligned with audience needs, the distribution more matched to the dissemination scenario, and the optimization more data-supported, significantly improving the overall efficiency and target completion rate of labor education dissemination, and realizing intelligent, personalized, and scenario-based dissemination of labor education in converged media scenarios.

[0116] Reference Figure 6 , Figure 6 A flowchart illustrating the AI-based method for dynamically generating and distributing converged media labor education content, as provided in this application embodiment.

[0117] like Figure 6 As shown, this embodiment also discloses an AI-based method for dynamically generating and distributing converged media labor education content, applied to the AI-based converged media labor education content dynamic generation and distribution system described above, including:

[0118] S10: In response to the input labor education objectives, output a communication mix, which includes labor education content and a communication platform; wherein: the labor education content is based on the labor education objectives, obtained through AI content generation technology, and its media forms include text and / or video.

[0119] S20: Collect feedback data corresponding to different communication combinations. This feedback data is used to characterize the achievement of labor education goals. Based on a preset threshold range, the feedback data is filtered to obtain suboptimal feedback data, which is used to map potential audience groups.

[0120] S30: Based on suboptimal feedback data, identify potential audience groups and adjust the communication mix based on these potential audience groups to optimize the dynamic generation and distribution of labor education content; wherein, adjusting the communication mix includes adjusting labor education content and / or communication platforms.

[0121] This embodiment also discloses a computer device for dynamically generating and distributing converged media labor education content based on AI, including at least one processor, at least one memory and a data bus;

[0122] The processor and memory communicate with each other via a data bus;

[0123] The memory stores program instructions that can be executed by the processor, which calls the program instructions to execute the AI-based method for dynamically generating and distributing converged media labor education content as described above.

[0124] This embodiment also discloses a storage medium on which a computer program is stored. When the computer program is executed by a processor, it implements the AI-based method for dynamically generating and distributing converged media labor education content as described above.

[0125] It should be noted that the AI-based method, device, and storage medium for dynamically generating and distributing labor education content in converged media in this embodiment correspond to the aforementioned AI-based system for dynamically generating and distributing labor education content in converged media. Therefore, any content not specifically described in the AI-based method, device, and storage medium for dynamically generating and distributing labor education content in converged media in this embodiment, including but not limited to functional definitions, working principles, and technical effects, can be referred to the descriptions in the aforementioned system for dynamically generating and distributing labor education content in converged media, and will not be repeated here.

[0126] In the embodiments provided in this application, it should be understood that the embodiments described herein can be implemented in hardware, software, firmware, middleware, code, or any suitable combination thereof. For hardware implementation, the processor may be implemented in one or more of the following: application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), processors, controllers, microcontrollers, microprocessors, other electronic units designed to implement the functions described herein, or combinations thereof. For software implementation, some or all of the processes of the embodiments may be performed by a computer program instructing the associated hardware. During implementation, the program may be stored in a computer-readable storage medium or transmitted as one or more instructions or code on a computer-readable storage medium. Computer-readable storage media include computer storage media and communication media, wherein communication media include any medium that facilitates the transmission of a computer program from one place to another. Storage media may be any available medium accessible to a computer. Computer-readable storage media may include, but are not limited to, RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code having the form of instructions or data structures and accessible to a computer.

[0127] Finally, it should be noted that the above description is only a preferred embodiment of this application and is not intended to limit this application. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. An AI-based system for dynamically generating and distributing integrated media labor education content, characterized in that, The system includes: The response module is configured to: respond to the input labor education goal and output a dissemination combination, which includes labor education content and a dissemination platform; wherein: the labor education content is based on the labor education goal and is obtained through AI content generation technology, and its media form includes text and / or video; and the labor education goal is set with corresponding completion parameters, which are used to quantify the achievement of the labor education goal; The feedback module is configured to: collect feedback data corresponding to different communication combinations, which is used to characterize the achievement of labor education goals; for the feedback data, filter it based on a preset threshold range to obtain suboptimal data, and then filter the suboptimal data to obtain suboptimal feedback data, which is used to map potential audiences; the filtering of the suboptimal feedback data includes: Input the completion parameters corresponding to the labor education goals into the preset large language model, and construct the first prompt word through the large language model. The first prompt word is used to clarify the type of feedback data collection. Based on the first prompt word, feedback data generated by different propagation combinations on the corresponding propagation platforms is collected. The collected feedback data is correlated with the completion parameters to obtain the completion status of each propagation combination. Input the completion status into the large language model, and construct a second cue word through the large language model. The second cue word is used to define the audience feature dimensions corresponding to different completion statuses. Based on the audience characteristic dimensions defined by the second prompt word, cluster analysis was performed on the feedback data to obtain multiple feedback data clusters corresponding to different audience groups; The feedback data clusters within each suboptimal level are sorted independently according to their completion status. The data clusters at the beginning and end of the sorting within each suboptimal level are removed, and the feedback data corresponding to the remaining data clusters are defined as suboptimal feedback data. The adjustment module is configured to: identify potential audience groups based on suboptimal feedback data, and adjust the dissemination mix based on these potential audience groups to optimize the dynamic generation and distribution of labor education content; wherein, adjusting the dissemination mix includes adjusting the labor education content and / or the dissemination platform; the operation of the adjustment module includes: Receive suboptimal feedback data and extract the potential audience characteristics associated with the data. The potential audience characteristics include at least the audience's preference for media forms, focus of content interest, and interactive behavior habits. The content of labor education is adjusted based on the characteristics of the potential audience, without adding any new labor education objectives; Simultaneously adjust the communication platform based on the characteristics of the potential audience, including at least switching to a communication platform where the potential audience is more active; The combined and adjusted labor education content and dissemination platform will output new dissemination combinations, optimize the dynamic generation and distribution of labor education content, and await the next round of feedback data collection and iteration.

2. The AI-based integrated media labor education content dynamic generation and distribution system according to claim 1, characterized in that, The operation of the response module includes: The objectives of labor education are determined, including educational objectives and practical objectives. Based on the aforementioned labor education objectives, a pre-set AI content generation platform is invoked to output labor education content; Based on the media format of the labor education content and the completion parameters, determine the dissemination platform; The labor education content and corresponding dissemination platforms are combined to output the dissemination combination.

3. The AI-based dynamic generation and distribution system for converged media labor education content according to claim 2, characterized in that, The operation of the feedback module includes: Collect feedback data corresponding to different propagation combinations and determine the corresponding completion status; wherein the collection of the feedback data corresponds to the completion parameter; The feedback data is sorted based on the completion status, and the target completion level is divided according to the completion parameters of the labor education goals. Feedback data that are at the first and last tier of target completion level are removed to obtain the second-best tier data.

4. The AI-based integrated media labor education content dynamic generation and distribution system according to claim 3, characterized in that, The completion parameters are data collection indicators used to quantify the efficiency of labor education dissemination; For educational objectives, the data collection metrics should include at least the number of valid views and the percentage of time spent on the content. For practical objectives, the data collection indicators should include at least the number of participants and the number of records of group participation.

5. The AI-based dynamic generation and distribution system for converged media labor education content according to claim 3, characterized in that, The characteristics of the potential audience also include student group interaction characteristics, which include at least: student group gathering scene identifiers, including at least dormitories, classes or clubs; synchronous access time of the group; and group-related interaction trajectories, including at least records of multiple people in the same group clicking and forwarding in succession. The adjustment module adjusts the dissemination combination based on the characteristics of student group interaction behavior. Specifically, it adds group collaboration guidance when generating labor education content, pushes content in a targeted manner according to the time period when the group gathers, and opens a group sharing portal on the dissemination platform.

6. A method for dynamically generating and distributing converged media labor education content based on AI, applied to the AI-based converged media labor education content dynamic generation and distribution system as described in any one of claims 1 to 5, characterized in that, The method includes: In response to the input labor education goals, a dissemination combination is output, which includes labor education content and a dissemination platform; wherein: the labor education content is based on the labor education goals and is obtained through AI content generation technology, and its media form includes text and / or video; and the labor education goals are set with corresponding completion parameters, which are used to quantify the achievement of the labor education goals; Feedback data corresponding to different communication combinations is collected, and this feedback data is used to characterize the achievement of labor education goals. Based on a preset threshold range, suboptimal data is obtained from the feedback data, and then further filtered to obtain suboptimal feedback data, which is used to map potential audiences. The filtering of the suboptimal feedback data includes: Input the completion parameters corresponding to the labor education goals into the preset large language model, and construct the first prompt word through the large language model. The first prompt word is used to clarify the type of feedback data collection. Based on the first prompt word, feedback data generated by different propagation combinations on the corresponding propagation platforms is collected. The collected feedback data is correlated with the completion parameters to obtain the completion status of each propagation combination. Input the completion status into the large language model, and construct a second cue word through the large language model. The second cue word is used to define the audience feature dimensions corresponding to different completion statuses. Based on the audience characteristic dimensions defined by the second prompt word, cluster analysis was performed on the feedback data to obtain multiple feedback data clusters corresponding to different audience groups; The feedback data clusters within each suboptimal level are sorted independently according to their completion status. The data clusters at the beginning and end of the sorting within each suboptimal level are removed, and the feedback data corresponding to the remaining data clusters are defined as suboptimal feedback data. Based on suboptimal feedback data, a potential audience group is identified, and the communication mix is ​​adjusted based on this potential audience group to optimize the dynamic generation and distribution of labor education content; wherein, adjusting the communication mix includes adjusting the labor education content and / or the communication platform; the operation of the adjustment module includes: Receive suboptimal feedback data and extract the potential audience characteristics associated with the data. The potential audience characteristics include at least the audience's preference for media forms, focus of content interest, and interactive behavior habits. The content of labor education is adjusted based on the characteristics of the potential audience, without adding any new labor education objectives; Simultaneously adjust the communication platform based on the characteristics of the potential audience, including at least switching to a communication platform where the potential audience is more active; The combined and adjusted labor education content and dissemination platform will output new dissemination combinations, optimize the dynamic generation and distribution of labor education content, and await the next round of feedback data collection and iteration.

7. A computer device for dynamically generating and distributing converged media labor education content based on AI, characterized in that, Includes at least one processor, at least one memory, and a data bus; The processor and the memory communicate with each other via the data bus; The memory stores program instructions that can be executed by the processor, and the processor invokes the program instructions to execute the AI-based method for dynamically generating and distributing labor education content based on converged media as described in claim 6.

8. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the AI-based method for dynamically generating and distributing converged media labor education content as described in claim 6.

Images