An international public opinion event information dissemination and control system based on artificial intelligence

Abstract

The application relates to an international public opinion event information dissemination and control system based on artificial intelligence, and relates to the field of social communication science. The system can monitor, identify, analyze, track and effectively control the information dissemination in international public opinion events in real time. The system uses natural language processing (NLP), machine learning (ML), deep learning (DL) and other artificial intelligence technologies to deeply mine and understand massive cross-language and cross-platform data, thereby realizing accurate identification of public opinion events, intelligent analysis of dissemination paths, early warning and intervention of negative information, and guidance and amplification of positive information. The system aims to improve the control efficiency and accuracy of international public opinion events, maintain social stability, and provide decision support and action basis for relevant departments.

Description

Technical Field

[0001] This invention relates to the field of social communication studies, specifically to an artificial intelligence-based system for the dissemination and control of international public opinion events. Background Technology

[0002] With the deepening of globalization and the rapid development of internet technology, the spread of international public opinion events is characterized by speed, wide scope, significant impact, and diverse forms. Traditional information monitoring and control methods are no longer sufficient to meet current needs, especially when facing complex cross-cultural and cross-linguistic international public opinion. The rapid spread of negative information can seriously damage economic development, social stability, and even international relations. Therefore, developing a system capable of intelligently and efficiently monitoring and controlling the spread of international public opinion events is of paramount importance.

[0003] Problems with existing technology:

[0004] 1. Information overload and difficulty in filtering: The explosive growth of information on the Internet makes it difficult to effectively filter and analyze the massive amount of information manually.

[0005] 2. Cross-language and cross-cultural barriers: Language and cultural differences in different regions increase the difficulty of information understanding and analysis.

[0006] 3. Complex and dynamic dissemination mechanisms: The paths of public opinion dissemination are varied, and the nodes of dissemination are difficult to predict and track.

[0007] 4. Delayed early warning and low efficiency in control of negative information: Once negative information spreads, the difficulty and cost of control increase sharply.

[0008] 5. Diverse sources of information and difficulty in verifying their authenticity: Information sources such as social media, news reports, and forum discussions are numerous and complex, making it difficult to determine their authenticity.

[0009] 6. Lack of precise intervention methods: Traditional control methods are often crude and ineffective, lacking specificity. Summary of the Invention

[0010] Purpose of the invention: To provide an artificial intelligence-based system for the dissemination and control of international public opinion events, in order to solve the aforementioned problems existing in the prior art.

[0011] Technical solution: An artificial intelligence-based international public opinion event information dissemination and control system, comprising six parts: intelligent information collection module, data preprocessing module, intelligent identification and analysis module for public opinion events, intelligent analysis and tracking module for dissemination paths, public opinion event control and intervention module, and user interaction and decision support module.

[0012] The intelligent information acquisition module is used to collect data in various formats such as text, images, and videos in real time from multi-source heterogeneous information channels around the world.

[0013] The data preprocessing module is used to perform preprocessing operations such as cross-language translation, language recognition, data cleaning, and deduplication on the collected data;

[0014] The intelligent identification and analysis module for public opinion events utilizes natural language processing, machine learning, and deep learning technologies to perform operations such as event discovery, sentiment analysis, entity recognition, relationship extraction, and information influence assessment on preprocessed data.

[0015] The intelligent analysis and tracking module for propagation paths is used to construct social network graphs, identify key propagation nodes, track information propagation paths, and identify and trace the source of rumors or false information.

[0016] The public opinion event management and intervention module performs risk warnings and classifications of public opinion events based on the analysis results, and generates or recommends corresponding management and intervention strategies.

[0017] The user interaction and decision support module provides functions such as visual monitoring dashboards, customized report generation, and early warning notifications to support user decision-making.

[0018] In a further embodiment, the intelligent information collection module collects data through web crawlers, API calls, third-party data services, and other means.

[0019] In a further embodiment, the data preprocessing module integrates a high-quality machine translation engine capable of real-time translation of multilingual text.

[0020] In a further embodiment, the intelligent identification and analysis module for public opinion events employs topic models, clustering algorithms, sentiment analysis models (such as BERT-based models), named entity recognition models (such as Bi-LSTM-CRF), and relation extraction models.

[0021] In a further embodiment, the propagation path intelligent analysis and tracking module uses a graph database to store social network graphs and employs graph theory algorithms such as PageRank and Betweenness Centrality to identify key nodes.

[0022] In a further embodiment, the public opinion event management and intervention module can classify public opinion events into risk levels based on the event's sensitivity, spread speed, emotional tendency, and scope of influence, and generate intervention strategies such as suppressing negative information, guiding positive information, guiding public opinion, or countering the generation of information.

[0023] In a further embodiment, the process of generating and recommending the intervention strategy can be dynamically optimized using a reinforcement learning algorithm.

[0024] In a further embodiment, the user interaction and decision support module can provide visualization and simulation of event propagation paths.

[0025] In a further embodiment, the dissemination and control method of the artificial intelligence-based international public opinion event information dissemination and control system includes the following steps:

[0026] S1. Data is collected in real time from multi-source heterogeneous information channels around the world through the intelligent information collection module;

[0027] S2. Through the data preprocessing module, perform preprocessing operations such as cross-language translation, language recognition, data cleaning, and deduplication on the collected data;

[0028] S3. Utilizing natural language processing, machine learning, and deep learning technologies, the intelligent public opinion event identification and analysis module performs operations such as event discovery, sentiment analysis, entity recognition, relationship extraction, and information influence assessment on the pre-processed data.

[0029] S4. Through the intelligent analysis and tracking module of the propagation path, construct a social network graph, identify key propagation nodes, track the information propagation path, and identify and trace the source of rumors or false information;

[0030] S5. Based on the analysis results, conduct risk warnings and classifications of public opinion events through the public opinion event management and intervention module, and generate or recommend corresponding management and intervention strategies.

[0031] S6. Through the user interaction and decision support module, it provides users with functions such as visual monitoring dashboard, customized report generation, and early warning notification, thus providing decision support for users.

[0032] In a further embodiment, when conducting risk warnings and classifications, the sensitivity, speed of spread, emotional sentiment, and scope of impact of the event are taken into consideration.

[0033] The resulting control and intervention strategies include suppressing negative information, guiding positive information, guiding public opinion, or countering the generation of information.

[0034] Beneficial Effects: This invention relates to an artificial intelligence-based system for the dissemination and control of international public opinion events, which has the following beneficial effects:

[0035] 1. Intelligentization and Automation: Utilize AI technology to achieve automated collection, analysis and control of public opinion information, significantly improving work efficiency.

[0036] 2. High precision and deep understanding: It can deeply understand complex public opinion information across languages ​​and cultures, and identify subtle emotions and intentions.

[0037] 3. Real-time monitoring and early warning capabilities: Enables real-time monitoring and early warning of public opinion events, allowing valuable time for timely intervention.

[0038] 4. Refined Management and Effective Intervention: Provide targeted intervention strategy suggestions to achieve precise management of public opinion events.

[0039] 5. Scalability and adaptability: The modular design facilitates system upgrades and functional expansion, enabling it to adapt to the ever-changing public opinion environment.

[0040] 6. Decision Support and Risk Mitigation: Provide important decision support information to relevant departments and enterprises, and help mitigate public opinion risks. Attached Figure Description

[0041] Figure 1 This is a schematic diagram of the overall system framework of the present invention. Detailed Implementation

[0042] In the following description, numerous specific details are set forth in order to provide a more thorough understanding of the invention. However, it will be apparent to those skilled in the art that the invention can be practiced without one or more of these details. In other instances, certain technical features well-known in the art have not been described in order to avoid obscuring the invention.

[0043] The international public opinion event information dissemination and control system based on artificial intelligence involved in this invention mainly includes six parts: intelligent information collection module, data preprocessing module, intelligent identification and analysis module for public opinion events, intelligent analysis and tracking module for dissemination paths, public opinion event control and intervention module, and user interaction and decision support module.

[0044] The intelligent information acquisition module is used to collect data in various formats such as text, images, and videos in real time from multi-source heterogeneous information channels around the world.

[0045] The data preprocessing module is used to perform preprocessing operations such as cross-language translation, language recognition, data cleaning, and deduplication on the collected data;

[0046] The intelligent identification and analysis module for public opinion events utilizes natural language processing, machine learning, and deep learning technologies to perform operations such as event discovery, sentiment analysis, entity recognition, relationship extraction, and information influence assessment on preprocessed data.

[0047] The intelligent analysis and tracking module for propagation paths is used to construct social network graphs, identify key propagation nodes, track information propagation paths, and identify and trace the source of rumors or false information.

[0048] The public opinion event management and intervention module performs risk warnings and classifications of public opinion events based on the analysis results, and generates or recommends corresponding management and intervention strategies.

[0049] The user interaction and decision support module provides functions such as visual monitoring dashboards, customized report generation, and early warning notifications to support user decision-making.

[0050] The intelligent information collection module collects data through web crawlers, API calls, and third-party data services.

[0051] The data preprocessing module integrates a high-quality machine translation engine, enabling real-time translation of multilingual text.

[0052] The intelligent identification and analysis module for public opinion events employs topic models, clustering algorithms, sentiment analysis models (such as BERT-based models), named entity recognition models (such as Bi-LSTM-CRF), and relation extraction models.

[0053] The propagation path intelligent analysis and tracking module uses a graph database to store social network graphs and employs graph theory algorithms such as PageRank and Betweenness Centrality to identify key nodes.

[0054] The public opinion event management and intervention module can classify public opinion events into risk levels based on their sensitivity, speed of dissemination, emotional tendency, and scope of influence, and generate intervention strategies such as suppressing negative information, guiding positive information, guiding public opinion, or countering the generation of counter-information.

[0055] The generation and recommendation process of the intervention strategy can be dynamically optimized using reinforcement learning algorithms.

[0056] The user interaction and decision support module can provide visualization and simulation of event propagation paths.

[0057] The dissemination and control methods of the AI-based international public opinion event information dissemination and control system include the following steps:

[0058] S1. Data is collected in real time from multi-source heterogeneous information channels around the world through the intelligent information collection module;

[0059] S2. Through the data preprocessing module, perform preprocessing operations such as cross-language translation, language recognition, data cleaning, and deduplication on the collected data;

[0060] S3. Utilizing natural language processing, machine learning, and deep learning technologies, the intelligent public opinion event identification and analysis module performs operations such as event discovery, sentiment analysis, entity recognition, relationship extraction, and information influence assessment on the pre-processed data.

[0061] S4. Through the intelligent analysis and tracking module of the propagation path, construct a social network graph, identify key propagation nodes, track the information propagation path, and identify and trace the source of rumors or false information;

[0062] S5. Based on the analysis results, conduct risk warnings and classifications of public opinion events through the public opinion event management and intervention module, and generate or recommend corresponding management and intervention strategies.

[0063] S6. Through the user interaction and decision support module, it provides users with functions such as visual monitoring dashboard, customized report generation, and early warning notification, thus providing decision support for users.

[0064] When conducting risk warnings and classifications, the sensitivity, speed of dissemination, emotional sentiment, and scope of impact of the event should be considered.

[0065] The resulting control and intervention strategies include suppressing negative information, guiding positive information, guiding public opinion, or countering the generation of information.

[0066] Furthermore, this invention provides an artificial intelligence-based system for the dissemination and control of international public opinion events, the core components and functions of which include:

[0067] 1. Intelligent Information Acquisition and Preprocessing Module:

[0068] Multi-source heterogeneous data collector: Integrates multiple methods such as web crawlers, API calls, and third-party data services to collect data in various formats, including text, images, and videos, in real time from global social media platforms (such as Twitter, Facebook, Instagram, Weibo, VK, etc.), news websites, blogs, forums, instant messaging tools (authorization required), and official news release platforms.

[0069] Cross-language processing module:

[0070] Machine translation engine: Integrates high-quality cross-language translation models (such as the Transformer architecture) to achieve real-time and accurate translation of collected multilingual text.

[0071] Language recognition module: Automatically identifies the language type of text.

[0072] Cultural Context Analysis Module: Adapts to different language and cultural backgrounds in terms of expression, idioms, and slang, improving the accuracy of comprehension.

[0073] Data cleaning and deduplication module: Denoises the collected raw data (e.g., remove HTML tags and advertisements), standardizes the format, corrects errors, and performs efficient deduplication to avoid redundant analysis.

[0074] Data annotation and augmentation module: Manual or semi-automatic annotation of some key data for training and optimization of models; at the same time, data augmentation techniques (such as back translation and synonym replacement) can be used to increase the diversity of the dataset.

[0075] 2. Intelligent identification and analysis module for public opinion events:

[0076] Event detection and clustering algorithms: Using topic models (such as LDA), keyword extraction, and clustering algorithms (such as K-means, DBSCAN), potential public opinion events are identified and related scattered information is aggregated into a unified event.

[0077] Sentiment Analysis and Propensity Identification:

[0078] Fine-grained sentiment analysis: It can not only distinguish between positive, negative, and neutral emotions, but also identify specific emotional intensities (such as extreme anger or slight dissatisfaction) and implicit emotions.

[0079] Goal-oriented sentiment analysis: Identifies the specific object (person, event, organization, etc.) to which the sentiment is directed.

[0080] Emotion recognition: By combining information such as facial expressions and body language in text and images / videos, deeper emotional states can be identified.

[0081] Entity recognition and relation extraction: Identify named entities such as people's names, place names, organization names, time, and events in the text, and analyze the relationships between entities (such as who did what, and who is related to whom).

[0082] Keyword and Hot Topic Extraction: Extract keywords and hot topics that reflect the core content and focus of public opinion events in real time.

[0083] Anomaly detection and mutation point identification: Monitor indicators such as the frequency of information release and the rate of change in sentiment to identify the suddenness or abnormal propagation patterns of public opinion events.

[0084] Information impact assessment: This involves evaluating the dissemination potential and impact of information by combining factors such as the number of reads, shares, comments, likes, and the social network levels through which the information is disseminated.

[0085] 3. Intelligent analysis and tracking module for propagation paths:

[0086] Social network graph construction: Based on data such as user relationships, information forwarding / commenting interactions, etc., construct a social network graph that reflects the spread of information.

[0087] Key node identification: Graph theory algorithms (such as PageRank, Betweenness Centrality) are used to identify nodes that play a key role in the information dissemination process (such as opinion leaders and KOLs).

[0088] Visualization and simulation of dissemination paths: Visualizes the path, speed, and scope of information dissemination, and can simulate the dissemination effects under different intervention measures.

[0089] Identification and tracing of rumors and misinformation:

[0090] Rumor detection model: By combining multiple factors such as text features, dissemination patterns, and the credibility of information sources, the model is trained to identify potential rumors.

[0091] Information tracing algorithm: Tracing the original source and dissemination chain of information to identify behind-the-scenes manipulators or malicious disseminators.

[0092] 4. Public Opinion Incident Management and Intervention Module:

[0093] Risk warning and classification: Based on factors such as the sensitivity, speed of dissemination, emotional tendency, and scope of impact of an event, the risk level of public opinion events is classified and corresponding warning mechanisms are triggered.

[0094] Intervention strategy generation and recommendation:

[0095] Negative information suppression: Identify and suggest actions such as fact-clarification, debunking, and content removal for negative information, or reduce its spread priority through algorithms.

[0096] Positive information guidance and amplification: Identify and recommend positive information and facts to be published through official channels, credible media, or opinion leaders, and amplify their reach through algorithms.

[0097] Public opinion guidance strategies: Based on the analysis of audience psychology and communication patterns, we recommend public opinion guidance strategies such as topic setting, content format, and release timing.

[0098] Countermeasure Information Generation: When necessary, the system can assist in generating persuasive and policy-aligned debunking or countermeasure information.

[0099] Intervention effectiveness evaluation and optimization: Monitor the dissemination effect of intervention measures in real time, and adjust and optimize the intervention strategy based on feedback data.

[0100] 5. User Interaction and Decision Support Module:

[0101] Visualized monitoring dashboard: Provides an intuitive and easy-to-use user interface to display information such as global public opinion trends, key events, hot topics, sentiment distribution, and dissemination paths.

[0102] Customized report generation: Generate public opinion analysis reports with different dimensions and depths based on user needs.

[0103] Early warning and alarm system: Sends early warning information and event updates to relevant personnel in a timely manner via email, SMS, and App push notifications.

[0104] Strategy Simulation and Deduction: Provides simulation tools that allow users to test the potential effects of different intervention strategies in a safe environment.

[0105] Knowledge Base and Case Study Library: Store historical public opinion event analysis data, intervention measures and effects, and build a knowledge base and case study library for future reference and learning.

[0106] Furthermore, this system adopts a modular design, with each module exchanging data and collaborating through standardized interfaces, as detailed below:

[0107] Data acquisition layer: It is mainly implemented using scripting languages ​​such as Python combined with distributed crawling frameworks (such as Scrapy), and uses message queues such as Kafka for data buffering and transmission.

[0108] Data preprocessing layer:

[0109] NLP technology: Primarily utilizes NLP libraries such as spaCy, NLTK, and Hugging Face Transformers to achieve text segmentation, part-of-speech tagging, named entity recognition, syntactic analysis, and sentiment analysis.

[0110] Machine translation: Integrate Google Translate API, DeepL API, or other self-built Transformer-based translation models.

[0111] Image / Video Analysis: Using frameworks such as OpenCV, TensorFlow / PyTorch, perform image feature extraction, face recognition, object recognition, and preliminary analysis of video content.

[0112] Event recognition and analysis layer:

[0113] Machine learning and deep learning models:

[0114] Event detection: LDA, NMF topic models, text clustering algorithms (K-means, HDBSCAN).

[0115] Sentiment analysis: Fine-tuning of pre-trained models such as BERT and RoBERTa, and deep learning models such as CNN and LSTM.

[0116] Entity recognition: Bi-LSTM-CRF model, BERT-based NER model.

[0117] Relation extraction: Graph Neural Networks (GNNs), Sequence labeling-based models.

[0118] Rumor detection: A classification model based on graph convolutional network (GCN) analysis of propagation graphs and combined with text features.

[0119] Anomaly detection: Algorithms such as ARIMA, Isolation Forest, and One-Class SVM.

[0120] Propagation path analysis layer:

[0121] Graph databases: Neo4j, ArangoDB, etc. are used to store and query social network graphs.

[0122] Network analysis algorithms: PageRank, Betweenness Centrality, Community Detection (such as Louvain).

[0123] Control and Intervention Layer:

[0124] Recommendation system algorithms: content-based recommendation and collaborative filtering, used for recommendation intervention strategies.

[0125] Reinforcement learning: It can be used to dynamically optimize intervention strategies and adjust actions based on real-time feedback.

[0126] User interaction layer: The front end uses frameworks such as React and Vue.js to build a visual dashboard, and the back end uses frameworks such as Django and Flask to provide API services.

[0127] Furthermore, the detailed descriptions of each module of the AI-based international public opinion event information dissemination and control system are as follows:

[0128] Intelligent information acquisition and preprocessing module:

[0129] Multi-source heterogeneous data acquisition device:

[0130] This system integrates multiple methods, including web crawling, API calls, and third-party data services, to collect data in real-time from various formats such as text, images, and videos from global social media platforms (e.g., Twitter, Facebook, Instagram, Weibo, VK), news websites, blogs, forums, instant messaging tools (authorization required), and official news release platforms. Web crawlers automatically capture publicly available information from the internet, API calls obtain structured data from some platforms, and third-party data services provide public opinion data in specialized fields. This combination of collection methods ensures comprehensive and timely acquisition of international public opinion information.

[0131] Cross-language processing module:

[0132] Machine translation engine: Integrates high-quality cross-language translation models (such as the Transformer architecture) to achieve real-time and accurate translation of collected multilingual text. The Transformer architecture translation model has powerful language understanding and generation capabilities, and can handle complex language structures and semantic information, improving the accuracy and efficiency of translation.

[0133] Language recognition module: Automatically identifies the language type of text. Through feature analysis and machine learning algorithms, it can quickly and accurately determine the language of the text, providing a foundation for subsequent translation and processing.

[0134] The Cultural Context Analysis module adapts to different language and cultural backgrounds, including expressions, idioms, and slang, to improve comprehension accuracy. This module combines a cultural knowledge base and machine learning models to identify and analyze cultural elements in text, avoiding misunderstandings caused by cultural differences.

[0135] Data cleaning and deduplication module:

[0136] The collected raw data undergoes noise reduction (e.g., removing HTML tags and advertisements), format standardization, error correction, and efficient deduplication to avoid redundant analysis. Data cleaning improves data quality and usability by removing useless information and interfering factors; deduplication reduces data redundancy and improves system processing efficiency.

[0137] Data annotation and enhancement module:

[0138] Manual or semi-automatic annotation of key data is used for model training and optimization. Simultaneously, data augmentation techniques (such as back-translation and synonym replacement) can be used to increase the diversity of the dataset. Data annotation is a crucial step in machine learning model training, providing supervised learning samples. Data augmentation techniques can expand the scale of the dataset and improve the model's generalization ability.

[0139] Intelligent identification and analysis module for public opinion events:

[0140] Event detection and clustering algorithms:

[0141] By utilizing topic models (such as LDA), keyword extraction, and clustering algorithms (such as K-means and DBSCAN), potential public opinion events can be identified, and related scattered information can be aggregated into unified events. Topic models can discover latent topics in text, keyword extraction can extract keywords that reflect the core content of the text, and clustering algorithms can group similar information into one category, thereby achieving preliminary identification and classification of public opinion events.

[0142] Sentiment Analysis and Propensity Identification:

[0143] Fine-grained sentiment analysis: It not only distinguishes between positive, negative, and neutral emotions, but also identifies specific emotional intensities (such as extreme anger or slight dissatisfaction) and implicit emotions. By combining semantic analysis of text with a sentiment lexicon, it can more accurately determine the sentiment tendency and intensity of the text.

[0144] Goal-oriented sentiment analysis identifies the specific objects (people, events, organizations, etc.) to which sentiment is directed. This analysis can help understand the public's emotional attitudes towards different objects, providing more targeted information for public opinion management.

[0145] Emotion recognition: By combining information such as facial expressions and body language from text and images / videos, deeper emotional states can be identified. Through the combination of computer vision and natural language processing technologies, a more comprehensive understanding of public emotional responses can be achieved.

[0146] Entity recognition and relation extraction:

[0147] It identifies named entities such as people's names, places, organizations, times, and events in text and analyzes the relationships between entities (e.g., who did what, who is related to whom). Entity identification and relationship extraction can help build a knowledge graph of public opinion events, more clearly showing the subjects and relationships of public opinion events, and providing support for subsequent analysis and decision-making.

[0148] Keyword and hot topic extraction:

[0149] It extracts keywords and trending terms that reflect the core content and focus of public opinion events in real time. Through text frequency statistics, TF-IDF algorithms, and other technologies, it can quickly and accurately extract keywords and trending terms, helping staff to quickly understand the core content of public opinion events.

[0150] Anomaly detection and mutation point identification:

[0151] By monitoring indicators such as the frequency of information dissemination and the rate of sentiment shift, we can identify sudden or abnormal propagation patterns of public opinion events. Through the analysis of historical data and the application of machine learning algorithms, we can establish a normal public opinion propagation model and issue timely warnings when anomalies occur.

[0152] Information impact assessment:

[0153] By combining factors such as the number of reads, shares, comments, likes, and the social network levels through which the information is disseminated, the potential for information dissemination and its impact can be assessed. Information impact assessment can help staff determine which public opinion information requires focused attention and handling, thereby improving the efficiency of public opinion management.

[0154] Intelligent analysis and tracking module for propagation paths

[0155] Social network graph construction

[0156] Based on user relationships, information forwarding / commenting interactions, and other data, a social network graph reflecting information dissemination is constructed. This graph can visually display the dissemination path and nodes of information, helping staff understand the patterns and scope of public opinion information dissemination.

[0157] Key node identification:

[0158] Graph theory algorithms (such as PageRank and Betweenness Centrality) can be used to identify key nodes (such as opinion leaders and key opinion leaders) that play a crucial role in the information dissemination process. Key nodes have significant influence on public opinion dissemination, and identifying these nodes can provide important targets for public opinion intervention.

[0159] Propagation path visualization and simulation:

[0160] It visualizes the path, speed, and scope of information dissemination and can simulate the dissemination effects under different intervention measures. Through visualization, staff can gain a more intuitive understanding of public opinion dissemination; simulating the dissemination effects under different intervention measures can help staff select the optimal intervention strategy.

[0161] Identification and tracing of rumors and misinformation:

[0162] Rumor detection models combine multiple factors such as text features, propagation patterns, and the credibility of information sources to train models that can identify potential rumors. These models can be trained using machine learning and deep learning algorithms to improve the accuracy and efficiency of rumor identification.

[0163] Information tracing algorithms: These algorithms trace the original source and dissemination chain of information, identifying behind-the-scenes manipulators or malicious disseminators. Through analysis of information dissemination paths and data mining techniques, information tracing algorithms can pinpoint the source of information and key nodes in the dissemination process.

[0164] Public opinion event management and intervention module:

[0165] Risk warning and classification:

[0166] Based on factors such as the sensitivity, speed of dissemination, emotional sentiment, and scope of impact of an event, risk levels are categorized, and corresponding early warning mechanisms are triggered. Risk warnings and classifications can help staff quickly assess the severity of public opinion events and take appropriate measures.

[0167] Intervention strategy generation and recommendation:

[0168] Negative information suppression: Identifying and recommending actions such as fact-clarification, debunking, and content removal for negative information, or using algorithms to lower its dissemination priority. Suppressing negative information can reduce the impact of negative public opinion and maintain social stability and image.

[0169] Positive information guidance and amplification: This involves identifying and recommending positive information and factual facts to be published through official channels, credible media, or opinion leaders, and amplifying their reach through algorithms. Positive information guidance and amplification can influence public opinion and enhance public confidence in society.

[0170] Public opinion guidance strategies: Based on the analysis of audience psychology and communication patterns, we recommend strategies for guiding public opinion, including topic selection, content format, and timing of release. These strategies can help staff better guide public opinion and achieve the goal of public opinion management.

[0171] Countermeasure Information Generation: When necessary, the system can assist in generating persuasive and policy-aligned refutation or countermeasure information. This generation can help staff respond promptly to negative public opinion and reduce its negative impact.

[0172] Intervention effectiveness evaluation and optimization:

[0173] Real-time monitoring of the dissemination effects of intervention measures, and adjustment and optimization of intervention strategies based on feedback data. Effectiveness evaluation and optimization ensure the effectiveness of intervention measures and improve the efficiency and accuracy of public opinion management.

[0174] User interaction and decision support module:

[0175] Visual monitoring dashboard:

[0176] It provides an intuitive and user-friendly interface, displaying information such as global public opinion trends, key events, trending topics, sentiment distribution, and dissemination paths. The visual monitoring dashboard helps staff quickly understand the overall public opinion situation, improving work efficiency.

[0177] Customized report generation:

[0178] Based on user needs, we generate public opinion analysis reports with different dimensions and depths. Customized report generation can meet the needs of different users and provide them with personalized decision support.

[0179] Early warning notification and alarm system:

[0180] Early warning information and event updates are promptly sent to relevant personnel via email, SMS, and app push notifications. The early warning notification and alert system ensures that relevant personnel are aware of changes in public opinion and can take appropriate measures.

[0181] Strategy simulation and deduction:

[0182] Simulation tools are provided, allowing users to test the potential effects of different intervention strategies in a safe environment. Strategy simulation and deduction can help staff select the optimal intervention strategy, improving the scientific rigor and accuracy of decision-making.

[0183] Knowledge base and case library:

[0184] This system stores analytical data, intervention measures, and their effects from historical public opinion events, building a knowledge base and case study library for future reference and learning. The knowledge base and case study library can provide staff with valuable experience, improving their ability and level of public opinion management.

[0185] As described above, although the invention has been shown and described with reference to specific preferred embodiments, it should not be construed as limiting the invention itself. Various changes in form and detail may be made without departing from the spirit and scope of the invention as defined in the appended claims.

Claims

1. An artificial intelligence-based system for the dissemination and control of international public opinion events, characterized by: include: The intelligent information acquisition module is used to collect data in various formats, including text, images, and videos, from multiple heterogeneous information sources in real time. The data preprocessing module is used to perform cross-language translation, language recognition, data cleaning, and deduplication preprocessing operations on the collected data. The intelligent identification and analysis module for public opinion events utilizes natural language processing, machine learning, and deep learning technologies to perform event discovery, sentiment analysis, entity recognition, relationship extraction, and information impact assessment on pre-processed data. The intelligent analysis and tracking module for propagation paths is used to construct social network graphs, identify key propagation nodes, track information propagation paths, and identify and trace the source of rumors or false information. The public opinion event management and intervention module provides risk warnings and classifications for public opinion events based on the analysis results, and generates or recommends corresponding management and intervention strategies. The user interaction and decision support module provides users with visual monitoring dashboards, customized report generation, and early warning notification functions to support their decision-making.

2. The international public opinion event information dissemination and control system based on artificial intelligence according to claim 1, characterized in that: The intelligent information collection module collects data through web crawlers, API calls, and third-party data services.

3. The international public opinion event information dissemination and control system based on artificial intelligence according to claim 1, characterized in that: The data preprocessing module integrates a high-quality machine translation engine, enabling real-time translation of multilingual text.

4. The international public opinion event information dissemination and control system based on artificial intelligence according to claim 1, characterized in that: The intelligent identification and analysis module for public opinion events employs topic models, clustering algorithms, sentiment analysis models, named entity recognition models, and relation extraction models.

5. The international public opinion event information dissemination and control system based on artificial intelligence according to claim 1, characterized in that: The propagation path intelligent analysis and tracking module uses a graph database to store social network graphs and employs PageRank and Betweenness Centrality graph theory algorithms to identify key nodes.

6. The international public opinion event information dissemination and control system based on artificial intelligence according to claim 1, characterized in that: The public opinion event management and intervention module can classify public opinion events into risk levels based on their sensitivity, speed of dissemination, emotional tendency, and scope of influence, and generate intervention strategies such as suppressing negative information, guiding positive information, guiding public opinion, or countering the generation of counter-information.

7. The international public opinion event information dissemination and control system based on artificial intelligence according to claim 1, characterized in that: The process of generating or recommending the intervention strategy can be dynamically optimized using reinforcement learning algorithms.

8. The international public opinion event information dissemination and control system based on artificial intelligence according to claim 1, characterized in that: The user interaction and decision support module can provide visualization and simulation of event propagation paths.

9. The international public opinion event information dissemination and control system based on artificial intelligence according to claim 1, characterized in that: The dissemination and control methods of the AI-based international public opinion event information dissemination and control system include the following steps: S1. Data is collected in real time from multi-source heterogeneous information channels around the world through the intelligent information collection module; S2. Through the data preprocessing module, perform preprocessing operations such as cross-language translation, language recognition, data cleaning, and deduplication on the collected data; S3. Utilizing natural language processing, machine learning, and deep learning technologies, the intelligent public opinion event identification and analysis module performs operations such as event discovery, sentiment analysis, entity recognition, relationship extraction, and information influence assessment on the pre-processed data. S4. Through the intelligent analysis and tracking module of the propagation path, construct a social network graph, identify key propagation nodes, track the information propagation path, and identify and trace the source of rumors or false information; S5. Based on the analysis results, conduct risk warnings and classifications of public opinion events through the public opinion event management and intervention module, and generate or recommend corresponding management and intervention strategies. S6. Through the user interaction and decision support module, it provides users with functions such as visual monitoring dashboard, customized report generation, and early warning notification, thus providing decision support for users.

10. The international public opinion event information dissemination and control system based on artificial intelligence according to claim 9, characterized in that: When conducting risk warnings and classifications, the sensitivity, speed of dissemination, emotional sentiment, and scope of impact of the event should be considered. The resulting control and intervention strategies include suppressing negative information, guiding positive information, guiding public opinion, or countering the generation of information.

Images