Cooperative generation method for dynamic visual content based on cognitive logic chain

Abstract

The invention discloses a dynamic visual content collaborative generation method based on a cognitive logic chain, and belongs to the technical field of visual content generation, and the method comprises the following steps: S1, user intention analysis and data input; s2, dynamically constructing a cognitive logic chain; s3, intelligent scheduling of the multi-modal generation module; s4, cross-modal content collaborative generation is carried out; s5, collaborative editing and real-time feedback are carried out; s6, iterative optimization of logic chain driving; s7, multi-dimensional quality evaluation: constructing an evaluation matrix containing semantic consistency, visual attraction and user participation degree, predicting a content propagation effect in combination with a deep learning model, and generating a quantitative improvement suggestion report; and S8, updating the self-adaptive knowledge reversely marking the cognitive logic chain according to the finally adopted content version, extracting a new association rule, and injecting the new association rule into the rule base. Through deep semantic analysis and dynamic logic chain construction, the system accurately captures a core creation target of a user and converts the core creation target into an executable visual strategy.

Description

Technical Field

[0001] The present invention relates to the technical field of visual content generation, and in particular to a method for collaboratively generating dynamic visual content based on a cognitive logic chain. Background Art

[0002] Existing visual content generation technologies generally adopt static template-driven or human-led creation models, which have three core bottlenecks. First, at the level of multimodal content collaboration, text description image generation and video editing are often separated into independent processes, resulting in cross-modal semantic faults and the need for repeated manual verification. For example, the disconnection between scripts and storyboards in advertising creative production significantly increases communication costs. Secondly, the user intention conversion process relies on empirical interpretation. It is difficult for ordinary users to accurately convey abstract concepts through natural language, causing the final content to deviate from the original expectations. This is especially prone to knowledge transfer distortion in logic-intensive creations such as educational courseware. More critically, existing tools lack a dynamically evolving intelligent core and cannot optimize the generation strategy in a closed-loop based on user feedback, so the quality of creation remains at the upper limit of the initial algorithm capabilities.

[0003] Current technical solutions have not yet achieved a deep integration of cognitive logic and content generation. Although some systems have introduced knowledge graphs for simple semantic associations, they have not built a dynamically scalable intent node network, resulting in an inability to support the progressive expression of complex logical chains. Existing rule engines mostly use fixed threshold trigger mechanisms, which make it difficult to adaptively adjust the spatiotemporal constraints of multimodal content. For example, the rhythm misalignment problem between background music and key frames during automatic video synthesis has long existed, and the collaborative editing function is limited to surface modification records. Therefore, we propose a dynamic visual content collaborative generation method based on cognitive logic chains to solve this problem. Summary of the Invention

[0004] The purpose of the present invention is to provide a method for collaboratively generating dynamic visual content based on cognitive logic chains to solve the problems raised in the above background technology.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions: A method for collaboratively generating dynamic visual content based on cognitive logic chains includes the following steps: S1. User Intent Analysis and Data Input: Receive raw user input through a multi-channel interface, use natural language processing technology to extract core intent keywords, and verify the integrity and logical consistency of the input data, providing a structured foundation for building a cognitive logic chain. S2. Dynamic Construction of Cognitive Logic Chain: Create an extensible intent node network based on intent keywords, match semantically related logical reasoning paths from the association rule library, generate a cognitive logic chain framework containing causal relationships in real time, and mark key decision points; S3, intelligent scheduling of the multimodal generation module: automatically selects text, image, and video generation engines based on the attributes of logical chain nodes, allocates computing resources through a dynamic load balancing algorithm, and initializes multiple content generation threads in parallel; S4, Collaborative cross-modal content generation: Calling the text generation module to output the scene description script, driving the image generation engine to generate keyframe sequences, and synchronously triggering the video synthesis module to integrate timed visual elements to ensure the semantic coherence of multimodal content; S5. Collaborative editing and real-time feedback: Rendering the initial version of the content in a visual editing interface captures the operation trajectories and annotation data of multiple users, and using an operation conflict detection algorithm to automatically merge valid modifications, enabling real-time collaborative optimization among multiple users. S6. Logic chain-driven iterative optimization: Map user feedback to corresponding nodes in the cognitive logic chain, adjust content generation parameters using optimization strategies in the rule base, and perform A / B testing to compare the effectiveness of different versions of content. S7. Multi-dimensional quality assessment: Build an evaluation matrix that includes semantic consistency, visual appeal, and user engagement, combine it with a deep learning model to predict content dissemination effects, and generate a quantitative improvement suggestion report. S8. Adaptive knowledge base update: Reverse-annotate the cognitive logic chain based on the final adopted content version, extract new association rules and inject them into the rule base, update the user intent classification model parameters, and form a closed-loop learning mechanism.

[0006] Preferably, the S1 comprises the following steps: S101. Multi-source data acquisition: Receive structured and unstructured data through API interfaces, file uploads, and voice input, automatically convert heterogeneous formats such as JSON and XML into a unified data model, filter out noisy data, and supplement missing fields; S102, Deep Intent Mining: Use the BERT model to parse the implicit semantics of the input text, identify deep needs by linking entity relationships through the knowledge graph, and calibrate intent confidence based on historical user behavior data; S103. Dynamic data validation: Establish a domain-specific data logic rule base to detect boundary conditions and contradictions of input parameters, and trigger real-time interactive questionnaires to supplement key information.

[0007] Preferably, S2 comprises the following steps: S201, dynamic node generation: Instantiate the intent keyword into an operable node object, add metadata such as timestamp and weight coefficient, and establish a parent-child inheritance relationship between nodes; S202, rule intelligent matching: Use graph neural networks to traverse the association rule base, calculate the semantic similarity between rules and intent nodes, and dynamically load inference rule sets that meet threshold conditions; S203. Visual construction of logical chain: Automatically layout the node topology structure on the knowledge graph interface, adjust the rule application order through drag-and-drop interaction, and generate a logical flow diagram with weight identification.

[0008] Preferably, S3 includes the following steps: S301. Modal requirement analysis: Parse the content type identifiers in the logic chain nodes to identify the text descriptions, 3D models, or dynamic special effects requirements that need to be generated simultaneously; S302, Dynamic Resource Allocation: Monitor GPU memory and memory usage, allocate computing nodes based on task complexity, and prioritize critical path task resources. S303, initialization of the generation pipeline: configuring prompt templates for text generation, style parameters for image generation, and frame rate standards for video synthesis, and establishing a cross-module data exchange channel.

[0009] Preferably, the S4 comprises the following steps: S401, Text-driven visual generation: Input the scene description output by the logic chain into the Stable Diffusion model to generate the basic image, and call the ControlNet plug-in to add spatial constraints; S402, time sequence content arrangement: divide the video into segments according to the time nodes of the logic chain, automatically generate transition special effects scripts, and synchronize the background music rhythm with the screen switching points; S403, cross-modal alignment verification: Use the CLIP model to detect the semantic matching of images and text, verify the coherence of video actions through optical flow analysis, and repair logical conflicts of visual elements.

[0010] Preferably, the S5 comprises the following steps: S501, Real-time Collaborative Rendering: Use Operational Transformation technology to synchronize multi-user operations and update content modification effects in real time on the WebGL canvas; S502, intelligent conflict handling: Mark overlapping edit areas, intelligently recommend merging solutions based on user permission levels and edit history, and retain version traceability information; S503. Dynamic preview generation: Re-render the local area in real time according to the edited content, provide multi-version comparison views, and automatically save incremental modification records.

[0011] Preferably, the S6 comprises the following steps: S601, Feedback-Node Mapping: Establishing an association index between user comments and logic chain nodes, and quantifying the impact weight coefficient of negative feedback; S602, parameter adaptive adjustment: updating the hyperparameters of the generation module based on the reinforcement learning algorithm, such as adjusting the text creativity coefficient or image detail intensity; S603, Multi-version Experimental Verification: Generate three sets of optimization solutions in parallel, collect experience data through user focus group testing, and generate a decision heat map.

[0012] Preferably, the S7 includes the following steps: S701. Multi-dimensional indicator fusion: Calculate the cosine similarity between content semantics and original intent, analyze user gaze heat maps to assess visual appeal, and calculate interaction duration to quantify engagement. S702, Dissemination Effect Prediction: Use the LSTM model to simulate the dissemination path of content on social networks and predict the diffusion curves of key indicators such as likes and reposts; S703, Defect Location Report: Locate weak nodes in the logic chain through attribution analysis and generate an optimization roadmap including modification priority sorting.

[0013] Preferably, the S8 comprises the following steps: S801, Logical Chain Knowledge Extraction: Feed back the key decision points of the final version of the content to the intention node, expand the node attribute dimension, and strengthen the weight of the successful path; S802, incremental learning of rule base: Use federated learning technology to integrate distributed user feedback, automatically generate new association rules, and store them in the database after verification; S803. User portrait update: record user preferred content style and interaction mode, improve the personalized generation strategy library, and optimize the initial intent recognition accuracy.

[0014] The beneficial effects of the present invention are: 1. In the present invention, the method for collaboratively generating dynamic visual content based on cognitive logic chains, through deep semantic analysis and dynamic logic chain construction, accurately captures the user's core creative goals and transforms them into executable visualization strategies. The node-based expression of cognitive logic chains provides abstract thinking with editable operational paths. The cross-modal generation engine driven by the association rule library ensures that text descriptions, static images, and dynamic videos are deeply unified at the level of thematic expression, significantly reducing the distortion of intention and execution deviation in traditional creation. In particular, it maintains the integrity of original creativity in complex scene presentations, solving the industry pain point of the separation between creativity and execution. 2. The present invention describes a method for collaboratively generating dynamic visual content based on cognitive logic chains. Based on the cognitive logic chain's timed arrangement mechanism, it intelligently schedules text generation scripts as the foundational anchor for visual creation, synchronously drives image keyframe generation and video dynamic synthesis processes, and innovatively employs cross-modal alignment verification technology to automatically repair image-text conflicts and action faults, achieving spatiotemporal consistency across content elements in different modalities. This technical system revolutionizes the iterative revision model employed in traditional multi-tool collaboration, enabling a smoother production experience for scenarios such as educational courseware production and advertising creative development. 3. In the present invention, a method for collaboratively generating dynamic visual content based on a cognitive logic chain is described. In this method, user collaborative editing data is fed back to the logic chain nodes in real time to form an optimization closed loop. Federated learning technology continuously absorbs multi-party feedback to enhance the reasoning capabilities of the rule base. During the iteration process, the system autonomously expands the dimensions of intent nodes and the scale of association rules, gradually developing a creative knowledge system adapted to specific fields. This self-evolutionary feature significantly reduces the system's subsequent maintenance costs, while enabling the quality of content generation to continuously improve with frequency of use, providing sustainable intelligent support for long-term creative projects. 4. In the present invention, the method for collaboratively generating dynamic visual content based on cognitive logic chains uses a visual construction interface that makes the AI ​​decision-making process transparent. Users can directly intervene in the content generation logic by dragging and dropping nodes. A conflict merging algorithm ensures that effective innovations are retained when multiple people are editing. Dynamic preview technology presents the visual impact of parameter adjustments in real time. This collaborative model, where human creativity is the driving force and machine execution is optimized, retains the core decision-making power of professional creators while leveraging machine efficiency to solve the problem of repetitive labor, thereby reshaping the production relationship of digital art creation. 5. In the present invention, a method for collaboratively generating dynamic visual content based on a cognitive logic chain is described. A quantitative evaluation matrix is ​​used to comprehensively measure multi-dimensional indicators such as semantic accuracy and visual appeal. The communication prediction model is combined to judge the market value of the content in advance. The optimization direction is accurately confirmed through the logic chain defect positioning technology. An improvement roadmap including priority sorting is generated, which enables the creative team to break away from the empiricist decision-making model and obtain scientific quality prediction and optimization guidelines before the content is put into production, effectively reducing the risk of market trial and error and increasing the probability of producing high-quality content. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 This is a flowchart of a method for collaboratively generating dynamic visual content based on cognitive logic chains proposed by the present invention. DETAILED DESCRIPTION

[0016] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0017] Reference Figure 1 , a method for collaboratively generating dynamic visual content based on cognitive logic chain, comprising the following steps: S1. User Intent Analysis and Data Input: Receive raw user input through a multi-channel interface, use natural language processing technology to extract core intent keywords, and verify the integrity and logical consistency of the input data, providing a structured foundation for building a cognitive logic chain. S2. Dynamic Construction of Cognitive Logic Chain: Create an extensible intent node network based on intent keywords, match semantically related logical reasoning paths from the association rule library, generate a cognitive logic chain framework containing causal relationships in real time, and mark key decision points; S3, intelligent scheduling of the multimodal generation module: automatically selects text, image, and video generation engines based on the attributes of logical chain nodes, allocates computing resources through a dynamic load balancing algorithm, and initializes multiple content generation threads in parallel; S4, Collaborative cross-modal content generation: Calling the text generation module to output the scene description script, driving the image generation engine to generate keyframe sequences, and synchronously triggering the video synthesis module to integrate timed visual elements to ensure the semantic coherence of multimodal content; S5. Collaborative editing and real-time feedback: Rendering the initial version of the content in a visual editing interface captures the operation trajectories and annotation data of multiple users, and using an operation conflict detection algorithm to automatically merge valid modifications, enabling real-time collaborative optimization among multiple users. S6. Logic chain-driven iterative optimization: Map user feedback to corresponding nodes in the cognitive logic chain, adjust content generation parameters using optimization strategies in the rule base, and perform A / B testing to compare the effectiveness of different versions of content. S7. Multi-dimensional quality assessment: Build an evaluation matrix that includes semantic consistency, visual appeal, and user engagement, combine it with a deep learning model to predict content dissemination effects, and generate a quantitative improvement suggestion report. S8. Adaptive knowledge base update: Reverse-annotate the cognitive logic chain based on the final adopted content version, extract new association rules and inject them into the rule base, update the user intent classification model parameters, and form a closed-loop learning mechanism.

[0018] Preferably, the S1 comprises the following steps: S101. Multi-source data acquisition: Receive structured and unstructured data through API interfaces, file uploads, and voice input, automatically convert heterogeneous formats such as JSON and XML into a unified data model, filter out noisy data, and supplement missing fields; S102, Deep Intent Mining: Use the BERT model to parse the implicit semantics of the input text, identify deep needs by linking entity relationships through the knowledge graph, and calibrate intent confidence based on historical user behavior data; S103. Dynamic data validation: Establish a domain-specific data logic rule base to detect boundary conditions and contradictions of input parameters, and trigger real-time interactive questionnaires to supplement key information.

[0019] Preferably, S2 comprises the following steps: S201, dynamic node generation: Instantiate the intent keyword into an operable node object, add metadata such as timestamp and weight coefficient, and establish a parent-child inheritance relationship between nodes; S202, rule intelligent matching: Use graph neural networks to traverse the association rule base, calculate the semantic similarity between rules and intent nodes, and dynamically load inference rule sets that meet threshold conditions; S203. Visual construction of logical chain: Automatically layout the node topology structure on the knowledge graph interface, adjust the rule application order through drag-and-drop interaction, and generate a logical flow diagram with weight identification.

[0020] Preferably, S3 includes the following steps: S301. Modal requirement analysis: Parse the content type identifiers in the logic chain nodes to identify the text descriptions, 3D models, or dynamic special effects requirements that need to be generated simultaneously; S302, Dynamic Resource Allocation: Monitor GPU memory and memory usage, allocate computing nodes based on task complexity, and prioritize critical path task resources. S303, initialization of the generation pipeline: configuring prompt templates for text generation, style parameters for image generation, and frame rate standards for video synthesis, and establishing a cross-module data exchange channel.

[0021] Preferably, the S4 comprises the following steps: S401, Text-driven visual generation: Input the scene description output by the logic chain into the Stable Diffusion model to generate the basic image, and call the ControlNet plug-in to add spatial constraints; S402, time sequence content arrangement: divide the video into segments according to the time nodes of the logic chain, automatically generate transition special effects scripts, and synchronize the background music rhythm with the screen switching points; S403, cross-modal alignment verification: Use the CLIP model to detect the semantic matching of images and text, verify the coherence of video actions through optical flow analysis, and repair logical conflicts of visual elements.

[0022] Preferably, the S5 comprises the following steps: S501, Real-time Collaborative Rendering: Use Operational Transformation technology to synchronize multi-user operations and update content modification effects in real time on the WebGL canvas; S502, intelligent conflict handling: Mark overlapping edit areas, intelligently recommend merging solutions based on user permission levels and edit history, and retain version traceability information; S503. Dynamic preview generation: Re-render the local area in real time according to the edited content, provide multi-version comparison views, and automatically save incremental modification records.

[0023] Preferably, the S6 comprises the following steps: S601, Feedback-Node Mapping: Establishing an association index between user comments and logic chain nodes, and quantifying the impact weight coefficient of negative feedback; S602, parameter adaptive adjustment: updating the hyperparameters of the generation module based on the reinforcement learning algorithm, such as adjusting the text creativity coefficient or image detail intensity; S603, Multi-version Experimental Verification: Generate three sets of optimization solutions in parallel, collect experience data through user focus group testing, and generate a decision heat map.

[0024] Preferably, the S7 includes the following steps: S701. Multi-dimensional indicator fusion: Calculate the cosine similarity between content semantics and original intent, analyze user gaze heat maps to assess visual appeal, and calculate interaction duration to quantify engagement. S702, Dissemination Effect Prediction: Use the LSTM model to simulate the dissemination path of content on social networks and predict the diffusion curves of key indicators such as likes and reposts; S703, Defect Location Report: Locate weak nodes in the logic chain through attribution analysis and generate an optimization roadmap including modification priority sorting.

[0025] Preferably, the S8 comprises the following steps: S801, Logical Chain Knowledge Extraction: Feed back the key decision points of the final version of the content to the intention node, expand the node attribute dimension, and strengthen the weight of the successful path; S802, incremental learning of rule base: Use federated learning technology to integrate distributed user feedback, automatically generate new association rules, and store them in the database after verification; S803. User portrait update: record user preferred content style and interaction mode, improve the personalized generation strategy library, and optimize the initial intent recognition accuracy.

[0026] In this embodiment, when the system is deployed, the core module of the cognitive logic chain engine is first constructed, and the intent parsing service and rule base management service are deployed in the distributed computing cluster. The user submits the creation requirement text and material files through the multi-channel input interface of the web or mobile terminal. The system calls the pre-trained semantic parsing model to extract the intent keywords and instantiate them as node objects. The association rule library generates the initial logic chain framework through graph neural network matching. Then the scheduler activates the corresponding multimodal generation engine according to the attributes of the logic chain node. The text generation module uses the pre-trained large model to output the scene description script. The image generation engine receives the script and combines the style parameters to generate a key frame sequence. The video synthesis module automatically arranges the timing elements and adds transition effects. Finally, the initial version of the content is rendered in the collaborative editing interface for real-time operation by multiple users.

[0027] When users modify the visual logic chain nodes by dragging and dropping, the system automatically triggers the parameter adjustment service to update the content generation rules. The operation trajectory in the collaborative editing process is mapped to the corresponding logic node in real time. The conflict detection algorithm intelligently merges the edited content based on user permissions. The optimized feedback data drives the generation module to perform incremental iterations. The quality assessment service continuously monitors the semantic consistency and visual fluency indicators. The communication prediction model outputs the content market value assessment report. Finally, the system extracts the decision points in the optimization path and feeds them back to the rule base. The intention recognition model parameters are updated through federated learning technology to form a closed-loop workflow from creation to optimization. All process data is stored in the blockchain evidence system to ensure traceability.

[0028] The above is a detailed introduction to a method for collaboratively generating dynamic visual content based on a cognitive logic chain provided by the present invention. Specific embodiments are used herein to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. It should be pointed out that for ordinary technicians in this technical field, without departing from the principles of the present invention, the present invention can also be improved and modified, and these improvements and modifications also fall within the scope of protection of the claims of the present invention.

Claims

1. A method for collaboratively generating dynamic visual content based on cognitive logic chains, characterized in that: The following steps are involved: S1. User Intent Analysis and Data Input: Receive raw user input through a multi-channel interface, use natural language processing technology to extract core intent keywords, and verify the integrity and logical consistency of the input data, providing a structured foundation for building a cognitive logic chain. S2. Dynamic Construction of Cognitive Logic Chain: Create an extensible intent node network based on intent keywords, match semantically related logical reasoning paths from the association rule library, generate a cognitive logic chain framework containing causal relationships in real time, and mark key decision points; S3, intelligent scheduling of the multimodal generation module: automatically selects text, image, and video generation engines based on the attributes of logical chain nodes, allocates computing resources through a dynamic load balancing algorithm, and initializes multiple content generation threads in parallel; S4, Collaborative cross-modal content generation: Calling the text generation module to output the scene description script, driving the image generation engine to generate keyframe sequences, and synchronously triggering the video synthesis module to integrate timed visual elements to ensure the semantic coherence of multimodal content; S5. Collaborative editing and real-time feedback: Rendering the initial version of the content in a visual editing interface captures the operation trajectories and annotation data of multiple users, and using an operation conflict detection algorithm to automatically merge valid modifications, enabling real-time collaborative optimization among multiple users. S6. Logic chain-driven iterative optimization: Map user feedback to corresponding nodes in the cognitive logic chain, adjust content generation parameters using optimization strategies in the rule base, and perform A / B testing to compare the effectiveness of different versions of content. S7. Multi-dimensional quality assessment: Build an evaluation matrix that includes semantic consistency, visual appeal, and user engagement, combine it with a deep learning model to predict content dissemination effects, and generate a quantitative improvement suggestion report. S8. Adaptive knowledge base update: Reverse-annotate the cognitive logic chain based on the final adopted content version, extract new association rules and inject them into the rule base, update the user intent classification model parameters, and form a closed-loop learning mechanism.

2. The method for collaboratively generating dynamic visual content based on cognitive logic chains according to claim 1, characterized in that: The S1 comprises the following steps: S101. Multi-source data acquisition: Receive structured and unstructured data through API interfaces, file uploads, and voice input, automatically convert heterogeneous formats such as JSON and XML into a unified data model, filter out noisy data, and supplement missing fields; S102, Deep Intent Mining: Use the BERT model to parse the implicit semantics of the input text, identify deep needs by linking entity relationships through the knowledge graph, and calibrate intent confidence based on historical user behavior data; S103. Dynamic data validation: Establish a domain-specific data logic rule base to detect boundary conditions and contradictions of input parameters, and trigger real-time interactive questionnaires to supplement key information.

3. The method for collaboratively generating dynamic visual content based on cognitive logic chains according to claim 1, characterized in that: The S2 comprises the following steps: S201, dynamic node generation: Instantiate the intent keyword into an operable node object, add metadata such as timestamp and weight coefficient, and establish a parent-child inheritance relationship between nodes; S202, rule intelligent matching: Use graph neural networks to traverse the association rule base, calculate the semantic similarity between rules and intent nodes, and dynamically load inference rule sets that meet threshold conditions; S203. Visual construction of logical chain: Automatically layout the node topology structure on the knowledge graph interface, adjust the rule application order through drag-and-drop interaction, and generate a logical flow diagram with weight identification.

4. The method for collaboratively generating dynamic visual content based on cognitive logic chains according to claim 1, characterized in that: The S3 comprises the following steps: S301. Modal requirement analysis: Parse the content type identifiers in the logic chain nodes to identify the text descriptions, 3D models, or dynamic special effects requirements that need to be generated simultaneously; S302, Dynamic Resource Allocation: Monitor GPU memory and memory usage, allocate computing nodes based on task complexity, and prioritize critical path task resources. S303, initialization of the generation pipeline: configuring prompt templates for text generation, style parameters for image generation, and frame rate standards for video synthesis, and establishing a cross-module data exchange channel.

5. The method for collaboratively generating dynamic visual content based on cognitive logic chains according to claim 1, characterized in that: The S4 comprises the following steps: S401, Text-driven visual generation: Input the scene description output by the logic chain into the Stable Diffusion model to generate the basic image, and call the ControlNet plug-in to add spatial constraints; S402, time sequence content arrangement: divide the video into segments according to the time nodes of the logic chain, automatically generate transition special effects scripts, and synchronize the background music rhythm with the screen switching points; S403, cross-modal alignment verification: Use the CLIP model to detect the semantic matching of images and text, verify the coherence of video actions through optical flow analysis, and repair logical conflicts of visual elements.

6. The method for collaboratively generating dynamic visual content based on cognitive logic chains according to claim 1, characterized in that: The S5 comprises the following steps: S501, Real-time Collaborative Rendering: Use Operational Transformation technology to synchronize multi-user operations and update content modification effects in real time on the WebGL canvas; S502, intelligent conflict handling: Mark overlapping edit areas, intelligently recommend merging solutions based on user permission levels and edit history, and retain version traceability information; S503. Dynamic preview generation: Re-render the local area in real time according to the edited content, provide multi-version comparison views, and automatically save incremental modification records.

7. The method for collaboratively generating dynamic visual content based on cognitive logic chains according to claim 1, characterized in that: The S6 comprises the following steps: S601, Feedback-Node Mapping: Establishing an association index between user comments and logic chain nodes, and quantifying the impact weight coefficient of negative feedback; S602, parameter adaptive adjustment: updating the hyperparameters of the generation module based on the reinforcement learning algorithm, such as adjusting the text creativity coefficient or image detail intensity; S603, Multi-version Experimental Verification: Generate three sets of optimization solutions in parallel, collect experience data through user focus group testing, and generate a decision heat map.

8. The method for collaboratively generating dynamic visual content based on cognitive logic chains according to claim 1, characterized in that: The S7 comprises the following steps: S701. Multi-dimensional indicator fusion: Calculate the cosine similarity between content semantics and original intent, analyze user gaze heat maps to assess visual appeal, and calculate interaction duration to quantify engagement. S702, Dissemination Effect Prediction: Use the LSTM model to simulate the dissemination path of content on social networks and predict the diffusion curves of key indicators such as likes and reposts; S703, Defect Location Report: Locate weak nodes in the logic chain through attribution analysis and generate an optimization roadmap including modification priority sorting.

9. The method for collaboratively generating dynamic visual content based on cognitive logic chains according to claim 1, characterized in that: The S8 comprises the following steps: S801, Logical Chain Knowledge Extraction: Feed back the key decision points of the final version of the content to the intention node, expand the node attribute dimension, and strengthen the weight of the successful path; S802, incremental learning of rule base: Use federated learning technology to integrate distributed user feedback, automatically generate new association rules, and store them in the database after verification; S803. User portrait update: record user preferred content style and interaction mode, improve the personalized generation strategy library, and optimize the initial intent recognition accuracy.

Images