Social network advertisement propagation and user behavior analysis integrated system

Abstract

The invention aims to provide a social network advertisement propagation and user behavior analysis integrated system, which belongs to the technical field of marketing data analysis and comprises a data acquisition module, a user portrait module, a propagation path module, a prediction engine module and an optimization regulation module. The data acquisition module acquires social data, advertisement interaction data and user behavior data. The user portrait module constructs a social influence model and generates a user feature tag and an interest map. The propagation path module identifies key propagation nodes, constructs a social network propagation model, and generates network diffusion vectors. The prediction engine module establishes a prediction model, calculates an advertisement interaction probability index, generates a resource allocation and optimal push strategy, generates a performance evaluation signal, and is bidirectionally connected with the optimization regulation and control module. The optimization regulation and control module monitors and predicts the operation state of the engine module in real time and sends out an optimization instruction according to the performance evaluation signal. The system can accurately analyze the user behavior, predict the advertisement propagation path, optimize the advertisement putting effect and improve the advertisement interaction possibility.

Description

Technical Field

[0001] The present invention relates to the technical field of marketing data analysis, and in particular to an integrated system for social network advertising dissemination and user behavior analysis. Background Art

[0002] With the rapid development and popularity of social networks, social media platforms have become a crucial advertising channel for businesses. Traditional social network advertising methods rely primarily on demographic characteristics and interest tags provided by the platforms for targeting, lacking in-depth analysis of user social behavior and content dissemination patterns, making it difficult to accurately evaluate and optimize advertising effectiveness.

[0003] Therefore, there is an urgent need for an integrated advertising dissemination and user behavior analysis system that can integrate social network data, deeply analyze user behavior characteristics, accurately predict advertising dissemination paths, and provide real-time optimization suggestions. Summary of the Invention

[0004] The purpose of the present invention is to overcome the shortcomings of the existing technology and provide an integrated system for social network advertising dissemination and user behavior analysis. By constructing a social influence model, a social network dissemination model and an advertising interaction prediction model, it can accurately analyze user behavior, predict the advertising dissemination path, and optimize the advertising delivery strategy in real time. It effectively solves the problems of inaccurate user portraits, unclear dissemination paths, and difficult to evaluate interaction probabilities in traditional advertising delivery, and significantly improves the advertising interaction rate and delivery effect.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] An integrated system for social network advertising dissemination and user behavior analysis, including a data collection module, a user portrait module, a dissemination path module, a prediction engine module, and an optimization and control module;

[0007] Data collection module: used to obtain first social data, first advertising interaction data and first user behavior data, and send them to the user portrait module;

[0008] User portrait module: used to build a social influence model, generate the first user feature label and the first user interest map, and send them to the communication path module;

[0009] Propagation path module: used to identify key propagation nodes based on the first user feature label and the first user interest map, build a social network propagation model, generate a first network diffusion vector, and send it to the prediction engine module;

[0010] Prediction engine module: used to establish a prediction model, calculate the first advertising interaction probability index, generate resource allocation strategy and optimization push strategy, generate performance evaluation signal, and send it to the optimization control module;

[0011] Optimization and control module: used to monitor the operating status of the prediction engine module in real time, receive performance evaluation signals, and issue corresponding optimization instructions based on the performance evaluation signals.

[0012] Furthermore, the workflow of the data collection module is as follows: collect social data, advertising interaction data and user behavior data for cleaning, denoising, and standardization, and perform feature extraction to generate first social data, first advertising interaction data and first user behavior data.

[0013] Furthermore, the workflow of the user portrait module is as follows:

[0014] Step S101: building a social influence model based on first social data, first advertising interaction data, and first user behavior data;

[0015] Step S102: The label generation unit analyzes user characteristics according to the social influence model to generate a first user characteristic label;

[0016] Step S103: The interest graph unit constructs the first user interest graph based on the first user feature label and the social influence model;

[0017] Step S104: Send the first user feature tag and the first user interest graph to the propagation path module.

[0018] Furthermore, the social influence model establishment process is as follows:

[0019] Step S201: vectorize the first social data, the first advertising interaction data, and the first user behavior data to generate a feature vector set P: P = (p1, p2, p3, ..., p n ); where set P includes user interaction frequency, content sharing volume, and social connection strength data;

[0020] Step S202: Calculate the social influence function Γ(P):

[0021]

[0022] Among them, m represents the number of samples; n represents the feature dimension; p ij represents the jth eigenvalue of the i-th sample; α ij represents the feature weight coefficient; β j Represents the characteristic impact factor.

[0023] Furthermore, the workflow of the propagation path module is as follows:

[0024] Step S301: identifying key communication nodes in the social network based on the first user feature tag and the first user interest graph, including opinion leader nodes, information bridge nodes, and high-activity nodes;

[0025] Step S302: Based on the identified key communication nodes, a social network communication model is constructed to analyze the communication path and diffusion characteristics of information in the social network;

[0026] Step S303: Using the position and role of key propagation nodes in the propagation model, a first network diffusion vector is generated to describe the propagation characteristics of the advertising information among different user groups;

[0027] Step S304: Send the first network diffusion vector to the prediction engine module.

[0028] Furthermore, the social network propagation model is expressed as:

[0029]

[0030] Where t represents the time variable; k represents the number of influencing factors; l represents the number of node types; λ represents the global propagation coefficient; ω i represents the weight of the i-th influencing factor; μ i represents the decay rate of the i-th influencing factor; γ j represents the influence weight of the j-th type of node; N j (t) represents the activity function of the j-th type node at time t.

[0031] Furthermore, the prediction engine module workflow is as follows:

[0032] Step S401: Establish an advertisement interaction prediction model based on the first network diffusion vector, and calculate a first advertisement interaction probability index η:

[0033]

[0034] Among them, r represents the number of propagation factors; s represents the number of user factors; D i represents the i-th communication-related factor, including quantitative indicators of information coverage, communication speed, and interaction depth; E j Represents the jth user-related factors, including quantitative indicators of user interest matching and historical interaction probability; δ i represents the weight of the i-th propagation factor; θ j represents the jth user factor weight; κ represents the adjustment parameter;

[0035] Step S402: When the first advertisement interaction probability index η is greater than or equal to a preset threshold, an optimization push strategy is generated; when the first advertisement interaction probability index η is less than the preset threshold, a resource allocation strategy is generated;

[0036] Step S403: Generate a performance evaluation signal according to the running status of the prediction model and send it to the optimization and control module.

[0037] Furthermore, the optimization and control module workflow is as follows:

[0038] Monitor the running status of the prediction engine module in real time, receive performance evaluation signals, and issue corresponding optimization instructions based on the performance evaluation signals;

[0039] When the first-level performance evaluation signal is received, an instruction to adjust the user portrait parameters is issued; when the second-level performance evaluation signal is received, an instruction to optimize the propagation path parameters is issued; when the third-level performance evaluation signal is received, an instruction to reconstruct the prediction model is issued.

[0040] Furthermore, the first user feature label includes demographic characteristics, behavioral preference characteristics and social activity characteristics; the first user interest graph is a multi-dimensional network structure, in which nodes represent interest points, edges represent interest relevance, and weights represent interest intensity.

[0041] Furthermore, the first network diffusion vector includes a propagation speed component, a coverage component, an interaction depth component and a conversion efficiency component; the value range of the first advertising interaction probability index η is 0 to 1, and a higher value indicates a greater possibility of advertising interaction.

[0042] Compared with the prior art, the present invention has the following beneficial effects:

[0043] (1) Through social influence models and multi-dimensional data analysis, including interaction frequency, content sharing volume, and social connection strength, user tags containing demographic characteristics, behavioral preference characteristics, and social activity characteristics, as well as multi-dimensional interest maps, are generated to significantly improve the accuracy and comprehensiveness of user portraits and build accurate user portraits.

[0044] (2) Based on user feature tags and interest graphs, identify key communication nodes in social networks and build a dynamic communication model, focusing on the attenuation rate and activity function, accurately predict the advertising diffusion path, and optimize the efficiency of information dissemination to achieve key communication node identification and path optimization.

[0045] (3) Using propagation vectors and user factors, a prediction model is established to calculate the advertising interaction probability index η, quantitatively evaluate the advertising effect, and generate resource allocation or optimal push strategies through threshold judgment, thereby improving the targeting and success rate of advertising delivery and achieving a scientific prediction of advertising interaction probability.

[0046] (4) Real-time dynamic optimization and control, achieving closed-loop feedback through performance evaluation signals, real-time adjustment of user portrait parameters, propagation path parameters or reconstruction of prediction models, ensuring that the system continuously adapts to data changes and improving the real-time and adaptability of advertising delivery.

[0047] (5) Provide data-driven strategy generation, integrate social data, advertising interaction data and user behavior data, generate data-supported resource allocation strategies and push strategies through standardized processing and feature extraction, reduce subjective bias, and improve the scientific nature and operability of advertising delivery.

[0048] (6) Through precise prediction and dynamic optimization, the system makes the η value approach 1, significantly improving the interactive possibility and conversion efficiency of advertising, reducing the cost of ineffective delivery, and maximizing the advertising marketing effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments of the present invention and, together with the description, serve to explain the principles of the present invention.

[0050] The present invention can be more clearly understood from the following detailed description with reference to the accompanying drawings, in which:

[0051] Figure 1 This is the overall flow chart of the social network advertising dissemination and user behavior analysis integrated system of the present invention;

[0052] Figure 2 This is a task decomposition flow chart of the data acquisition module of the social network advertising dissemination and user behavior analysis integrated system provided by an embodiment of the present invention; it includes data collection, cleaning, feature extraction, and generating first social data, advertising interaction data and user behavior data.

[0053] Figure 3 This is a task decomposition flowchart of the user portrait module of the integrated system for social network advertising dissemination and user behavior analysis provided by an embodiment of the present invention, including building a social influence model, generating user feature tags and interest maps;

[0054] Figure 4 This is a task decomposition flowchart of the propagation path module of the integrated system for social network advertising propagation and user behavior analysis provided by an embodiment of the present invention, including identifying key propagation nodes, building a social network propagation model, and generating a network diffusion vector;

[0055] Figure 5 This is a task decomposition flowchart of the prediction engine module of the integrated system for social network advertising dissemination and user behavior analysis provided by an embodiment of the present invention, including establishing an advertising interaction prediction model, calculating an advertising interaction probability index, and generating a push strategy or resource allocation strategy;

[0056] Figure 6 This is a task decomposition flowchart of the propagation path module of the social network advertising propagation and user behavior analysis integrated system provided by an embodiment of the present invention, including monitoring the running status of the prediction engine module, receiving performance evaluation signals, and issuing optimization instructions. DETAILED DESCRIPTION

[0057] The technical solution of the present invention is described in detail below through the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present application and the specific features in the embodiments are detailed descriptions of the technical solution of the present application, rather than limitations of the technical solution of the present application. In the absence of conflict, the technical features in the embodiments of the present application and the embodiments can be combined with each other. The following embodiments are only used to more clearly illustrate the technical solution of the present invention and are not intended to limit the scope of protection of the present invention.

[0058] The term "and / or" in this document simply describes a relationship between related objects, indicating that three possible relationships exist. For example, "A and / or B" can mean: A exists alone, A and B exist simultaneously, or B exists alone. Additionally, the character " / " in this document generally indicates an "or" relationship between the related objects.

[0059] See Figure 1 As shown, the integrated system for social network advertising dissemination and user behavior analysis of this embodiment includes a data collection module, a user portrait module, a dissemination path module, a prediction engine module and an optimization and control module;

[0060] Data collection module: used to obtain first social data, first advertising interaction data, and first user behavior data, and send them to the user profiling module; wherein the first social data includes the amount of content published by the user, the frequency of interaction, and the number of social network connections; the first advertising interaction data includes the browsing time, click behavior, and the number of forwarding and sharing; the first user behavior data includes interest preferences, active time period, and content consumption habits;

[0061] The user portrait module includes a label generation unit and an interest graph unit; the data acquisition module and the user portrait module are connected in a one-way signal, the user portrait module and the propagation path module are connected in a one-way signal, the propagation path module and the prediction engine module are connected in a one-way signal, and the prediction engine module and the optimization and control module are connected in a two-way signal;

[0062] User portrait module: used to build a social influence model, generate the first user feature label and the first user interest map, and send them to the communication path module;

[0063] Propagation path module: used to identify key propagation nodes based on the first user feature label and the first user interest map, build a social network propagation model, generate a first network diffusion vector, and send it to the prediction engine module;

[0064] Prediction engine module: used to establish a prediction model, calculate the first advertising interaction probability index, generate resource allocation strategy and optimization push strategy, generate performance evaluation signal, and send it to the optimization and control module; Optimization and control module: used to monitor the operation status of the prediction engine module in real time, receive performance evaluation signal, and issue corresponding optimization instructions based on the performance evaluation signal.

[0065] Furthermore, the workflow of the data collection module is as follows: collect social data, advertising interaction data and user behavior data for cleaning, denoising, standardization, and feature extraction to generate first social data, first advertising interaction data and first user behavior data; wherein, the cleaning process includes removing duplicate data, correcting outliers and filling missing values, standardization converts data of different dimensions into a unified numerical range, and feature extraction extracts key features of the data through methods such as principal component analysis and factor analysis.

[0066] Specifically, the workflow of the user portrait module is as follows:

[0067] Step S101: building a social influence model based on first social data, first advertising interaction data, and first user behavior data;

[0068] Step S102: The label generation unit analyzes user characteristics according to the social influence model to generate a first user characteristic label;

[0069] Step S103: The interest graph unit constructs the first user interest graph based on the first user feature label and the social influence model;

[0070] Step S104: Send the first user feature tag and the first user interest graph to the propagation path module.

[0071] This embodiment organically integrates a data acquisition module, a user profile module, a communication path module, a prediction engine module, and an optimization and control module to construct a complete closed-loop system for social network advertising analysis and optimization. The system can extract valuable information from raw data, construct dynamic user profiles, identify key communication nodes, predict advertising interaction effects, and optimize delivery strategies in real time, significantly improving the delivery accuracy and dissemination effectiveness of social network advertising. By cleaning, denoising, standardizing, and extracting features from social data, advertising interaction data, and user behavior data, the system significantly improves data quality and availability, providing a high-quality data foundation for subsequent analysis, reducing analytical bias caused by data inconsistencies or noise, and improving the overall analytical accuracy of the system. By constructing a social influence model, the system can deeply analyze the role and influence of users in social networks, generate multi-dimensional user feature tags and dynamic interest maps, breaking through the limitations of traditional static user profiles and more comprehensively and accurately describing user characteristics and interest changes, providing a reliable basis for targeted advertising delivery.

[0072] Example 2

[0073] This example provides a process for establishing a social influence model:

[0074] Step S201, vectorize the first social data, the first advertising interaction data and the first user behavior data to generate a feature vector set P: P = (p1, p2, p3, ..., pn); wherein the set P includes user interaction frequency, content sharing volume and social connection strength data; wherein p1 represents the average daily number of published content, p2 represents the average content interaction volume, p3 represents the number of social network connections, p4 represents the fan interaction rate, p5 represents the content dissemination breadth, p6 represents the content residence time, p7 to p8 represent the content sharing frequency, p9 represent the content sharing frequency, p10 represents the content sharing frequency, p110 represents the content sharing frequency, p120 represents the content sharing frequency, p130 represents the content sharing frequency, p140 represents the content sharing frequency, p150 represents the content sharing frequency, p160 represents the content sharing frequency, p170 represents the content sharing frequency, p180 represents the content sharing frequency, p190 represents the content sharing frequency, p20 ... n Indicates other social behavior indicators;

[0075] Step S202: Calculate the social influence function Γ(P):

[0076]

[0077] Among them, m represents the number of samples; n represents the feature dimension; p ij represents the jth eigenvalue of the i-th sample; α ij Represents the feature weight coefficient, which is learned from historical data through machine learning algorithms; β j It represents the feature influence factor, reflecting the relative importance of the feature to social influence.

[0078] Through feature vector processing and calculation of social influence functions, the system can quantitatively evaluate the size and type of user influence in social networks, identify user groups with high communication value, provide a scientific basis for selecting target groups for advertising, and improve the efficiency and quality of advertising coverage.

[0079] Example 3

[0080] This example provides the workflow of the propagation path module:

[0081] Step S301: identifying key communication nodes in the social network based on the first user feature tag and the first user interest graph, including opinion leader nodes, information bridge nodes, and high-activity nodes;

[0082] Step S302: Based on the identified key communication nodes, a social network communication model is constructed to analyze the communication path and diffusion characteristics of information in the social network;

[0083] Step S303: Using the position and role of key propagation nodes in the propagation model, a first network diffusion vector is generated to describe the propagation characteristics of the advertising information among different user groups;

[0084] Step S304: Send the first network diffusion vector to the prediction engine module.

[0085] By identifying key communication nodes and constructing a social network communication model, the system can predict and simulate the communication path and diffusion characteristics of advertising information in social networks, generate network diffusion vectors, provide guidance at the communication path level for the formulation of advertising delivery strategies, optimize information dissemination efficiency, and expand the scope of advertising influence.

[0086] Furthermore, the social network propagation model is expressed as:

[0087]

[0088] Where t represents the time variable; k represents the number of influencing factors; l represents the number of node types. Specifically, the node types include opinion leader nodes, information bridging nodes, high-activity nodes, and ordinary nodes. Different types of nodes play different roles in the information dissemination process; λ represents the global propagation coefficient; ω i represents the weight of the i-th influencing factor; μ i represents the decay rate of the i-th influencing factor; γ j represents the influence weight of the j-th type node; N j(t) represents the activity function of the j-th type of node at time t. By comprehensively considering multi-dimensional factors such as time variables, influencing factors, node types, etc., a mathematical communication dynamics model is established, which can accurately describe the dissemination rules and evolution process of information in social networks, provide a theoretical basis for predicting the effect of advertising communication, and improve the accuracy and interpretability of communication predictions.

[0089] As an embodiment, the workflow of the prediction engine module is as follows: Step S401: Establish an advertising interaction prediction model based on the first network diffusion vector, and calculate the first advertising interaction probability index η:

[0090]

[0091] Among them, r represents the number of propagation factors; s represents the number of user factors; D i represents the i-th communication-related factor, including quantitative indicators of information coverage, communication speed, and interaction depth; E j Represents the jth user-related factors, including quantitative indicators of user interest matching and historical interaction probability; δ i represents the weight of the i-th propagation factor; θ j represents the weight of the jth user factor; κ represents the adjustment parameter; the combined impact of communication factors and user factors on advertising interaction is integrated through weighted summation, and after normalization, a probability index ranging from 0 to 1 is obtained;

[0092] Step S402: When the first advertisement interaction probability index η is greater than or equal to a preset threshold, an optimization push strategy is generated; when the first advertisement interaction probability index η is less than the preset threshold, a resource allocation strategy is generated;

[0093] Step S403: Generate a performance evaluation signal according to the running status of the prediction model and send it to the optimization and control module.

[0094] By establishing a prediction model that comprehensively considers the network diffusion characteristics and calculating the advertising interaction probability index, the system can scientifically evaluate the potential effects of different advertising strategies, and intelligently generate resource allocation strategies and optimized push strategies based on the prediction results, thereby improving the accuracy of advertising delivery and resource utilization efficiency, and reducing invalid exposure and resource waste.

[0095] Furthermore, the workflow of the optimization control module is as follows:

[0096] Monitor the running status of the prediction engine module in real time, receive performance evaluation signals, and issue corresponding optimization instructions based on the performance evaluation signals;

[0097] When the first-level performance evaluation signal is received, an instruction to adjust the user portrait parameters is issued; when the second-level performance evaluation signal is received, an instruction to optimize the propagation path parameters is issued; when the third-level performance evaluation signal is received, an instruction to reconstruct the prediction model is issued; among them, the first-level performance evaluation signal indicates a slight decrease in prediction accuracy, the second-level performance evaluation signal indicates a significant increase in prediction deviation, and the third-level performance evaluation signal indicates that the prediction model fails.

[0098] By real-time monitoring of the operating status of the prediction engine module and receiving performance evaluation signals, the system can issue targeted optimization instructions based on performance signals at different levels, achieving adaptive optimization and closed-loop management of advertising delivery strategies, ensuring the system's continued efficient operation in a complex and changing social network environment, and improving the system's robustness and adaptability.

[0099] Specifically, the first user feature label includes demographic characteristics, behavioral preference characteristics and social activity characteristics; the first user interest map is a multi-dimensional network structure, in which nodes represent interest points, edges represent interest relevance, and weights represent interest intensity, wherein demographic characteristics include information such as age group, gender, region and occupation; behavioral preference characteristics include information such as content consumption type, interaction habits and purchasing tendencies; social activity characteristics include information such as login frequency, interaction frequency and participation; and interest points include specific topics, product categories or content themes. The present invention constructs an interest map with a multi-dimensional network structure by integrating demographic characteristics, behavioral preference characteristics and social activity characteristics. The system can comprehensively and deeply characterize the user's interest preferences and changing trends, realize fine-grained expression and dynamic tracking of user interests, provide accurate basis for personalized advertising push, and significantly improve user response rate and satisfaction.

[0100] Furthermore, the first network diffusion vector includes a propagation speed component, a coverage component, an interaction depth component, and a conversion efficiency component; the first advertising interaction probability index η ranges from 0 to 1, with higher values ​​indicating a greater likelihood of advertising interaction. The propagation speed component represents the number of nodes through which information is propagated per unit time; the coverage component represents the ratio of the number of users accessible to the information to the total number of target users; the interaction depth component represents the depth of user interaction with the information; and the conversion efficiency component represents the ratio of interaction to subsequent behavior. By integrating key indicators such as propagation speed, coverage, interaction depth, and conversion efficiency into a network diffusion vector and calculating a standardized advertising interaction probability index, the present invention enables the system to quantitatively evaluate and predict advertising communication effectiveness, providing advertisers with intuitive and comparable effectiveness evaluation indicators, facilitating the optimization and adjustment of advertising strategies and the evaluation of return on investment, thereby improving the scientific nature and transparency of advertising placement decisions.

[0101] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A social network advertising dissemination and user behavior analysis integrated system, characterized by: It includes data collection module, user portrait module, propagation path module, prediction engine module and optimization and control module; Data collection module: used to obtain first social data, first advertising interaction data and first user behavior data, and send them to the user portrait module; User portrait module: used to build a social influence model, generate the first user feature label and the first user interest map, and send them to the communication path module; Propagation path module: used to identify key propagation nodes based on the first user feature label and the first user interest map, build a social network propagation model, generate a first network diffusion vector, and send it to the prediction engine module; Prediction engine module: used to establish a prediction model, calculate the first advertising interaction probability index, generate resource allocation strategy and optimization push strategy, generate performance evaluation signal, and send it to the optimization control module; Optimization and control module: used to monitor the operating status of the prediction engine module in real time, receive performance evaluation signals, and issue corresponding optimization instructions based on the performance evaluation signals.

2. The integrated system for social network advertising dissemination and user behavior analysis according to claim 1, characterized in that: The data collection module has the following workflow: collects social data, advertising interaction data and user behavior data, performs cleaning, denoising and standardization processing, and performs feature extraction to generate first social data, first advertising interaction data and first user behavior data.

3. The integrated system for social network advertising dissemination and user behavior analysis according to claim 1, characterized in that: The user portrait module includes a tag generation unit and an interest graph unit; the user portrait module workflow is as follows: Step S101: building a social influence model based on first social data, first advertising interaction data, and first user behavior data; Step S102: The label generation unit analyzes user characteristics according to the social influence model to generate a first user characteristic label; Step S103: The interest graph unit constructs the first user interest graph based on the first user feature label and the social influence model; Step S104: Send the first user feature tag and the first user interest graph to the propagation path module.

4. The integrated system for social network advertising dissemination and user behavior analysis according to claim 1, characterized in that: The social influence model establishment process is as follows: Step S201: vectorize the first social data, the first advertising interaction data, and the first user behavior data to generate a feature vector set P: P = (p1, p2, p3, ..., p n ); where set P includes user interaction frequency, content sharing volume, and social connection strength data; Step S202: Calculate the social influence function Γ(P): Among them, m represents the number of samples; n represents the feature dimension; p ij represents the jth eigenvalue of the i-th sample; α ij represents the feature weight coefficient; β j Represents the characteristic impact factor.

5. The integrated system for social network advertising dissemination and user behavior analysis according to claim 1, characterized in that: The workflow of the propagation path module is as follows: Step S301: identifying key communication nodes in the social network based on the first user feature tag and the first user interest graph, including opinion leader nodes, information bridge nodes, and high-activity nodes; Step S302: Based on the identified key communication nodes, a social network communication model is constructed to analyze the communication path and diffusion characteristics of information in the social network; Step S303: Using the position and role of key propagation nodes in the propagation model, a first network diffusion vector is generated to describe the propagation characteristics of the advertising information among different user groups; Step S304: Send the first network diffusion vector to the prediction engine module.

6. The integrated system for social network advertising dissemination and user behavior analysis according to claim 5, characterized in that: The social network propagation model is expressed as: Where t represents the time variable; k represents the number of influencing factors; l represents the number of node types; λ represents the global propagation coefficient; ω i represents the weight of the i-th influencing factor; μ i represents the decay rate of the i-th influencing factor; γ j represents the influence weight of the j-th type of node; N j (t) represents the activity function of the j-th type node at time t.

7. The integrated system for social network advertising dissemination and user behavior analysis according to claim 1, characterized in that: The prediction engine module workflow is as follows: Step S401: Establish an advertisement interaction prediction model based on the first network diffusion vector, and calculate a first advertisement interaction probability index η: Among them, r represents the number of propagation factors; s represents the number of user factors; D i represents the i-th communication-related factor, including quantitative indicators of information coverage, communication speed, and interaction depth; E j Represents the jth user-related factors, including quantitative indicators of user interest matching and historical interaction probability; δ i represents the weight of the i-th propagation factor; θ j represents the jth user factor weight; κ represents the adjustment parameter; Step S402: When the first advertisement interaction probability index η is greater than or equal to a preset threshold, an optimization push strategy is generated; when the first advertisement interaction probability index η is less than the preset threshold, a resource allocation strategy is generated; Step S403: Generate a performance evaluation signal according to the running status of the prediction model and send it to the optimization and control module.

8. The integrated system for social network advertising dissemination and user behavior analysis according to claim 1, characterized in that: The optimization and control module workflow is as follows: Monitor the running status of the prediction engine module in real time, receive performance evaluation signals, and issue corresponding optimization instructions based on the performance evaluation signals; When the first-level performance evaluation signal is received, an instruction to adjust the user portrait parameters is issued; when the second-level performance evaluation signal is received, an instruction to optimize the propagation path parameters is issued; when the third-level performance evaluation signal is received, an instruction to reconstruct the prediction model is issued.

9. The integrated system for social network advertising dissemination and user behavior analysis according to claim 1, characterized in that: The first user feature label includes demographic characteristics, behavioral preference characteristics and social activity characteristics; the first user interest graph is a multi-dimensional network structure, where nodes represent interest points, edges represent interest relevance, and weights represent interest intensity.

10. The integrated system for social network advertising dissemination and user behavior analysis according to claim 1, characterized in that: The first network diffusion vector includes a propagation speed component, a coverage component, an interaction depth component and a conversion efficiency component; the value range of the first advertising interaction probability index η is 0 to 1, and a higher value indicates a greater possibility of advertising interaction.

Images