Crisis public relations model based on offline priority double-line closed loop

Abstract

The application relates to the technical field of information processing and data analysis, and discloses a crisis public relations model based on offline priority and double-line closed loop, which comprises a basic support module, an offline processing module, an online response module, a double-line cooperation module, a trust reconstruction module and a prevention reinforcement module; through the construction of the whole-process system of "offline priority and double-line closed loop", the offline core contradiction is preferentially solved, online and offline information real-time cooperation and closed loop verification are simultaneously realized, the problem of repeated public opinions caused by the traditional mode of "emphasizing online and ignoring offline" is effectively avoided, the transparency and credibility of crisis disposal are significantly improved through the introduction of third-party supervision and multidimensional evaluation, and the public trust is systematically repaired from the initial stage of the crisis.

Description

Technical Field

[0001] This invention relates to the field of information processing and data analysis technology, and more specifically discloses a crisis public relations model based on offline priority and dual-line closed loop. Background Technology

[0002] Crisis public relations refers to management activities that use a series of strategies and actions to resolve conflicts, repair image, and rebuild trust when facing a crisis. With the rapid development of social media and network technology, information dissemination has broken through the limitations of time and space, and the speed and influence of crisis events have increased exponentially, posing a severe challenge to the survival and development of organizations such as government and enterprises.

[0003] Current crisis communication technologies overemphasize online public opinion management while neglecting offline conflict resolution, leading to repeated public opinion crises. In addition, the disconnect and lack of coordination between online and offline information can easily trigger a crisis of trust. Furthermore, trust reconstruction is slow and the methods are limited, making it difficult to restore public trust. The neglect of grassroots capacity building makes it difficult to resolve potential risks in a timely manner. Finally, the response efficiency is low and the assessment dimensions are limited to online, failing to form a complete closed loop and making it difficult to adapt to complex crisis environments. Summary of the Invention

[0004] The main technical problem solved by this invention is to provide a crisis public relations model based on offline priority and dual-line closed loop, which can solve the problems mentioned in the background.

[0005] To address the aforementioned technical problems, according to one aspect of the present invention, more specifically, a crisis public relations model based on an offline-priority dual-line closed loop, comprising:

[0006] The basic support module integrates grassroots public opinion clues, historical case data, and access control to provide basic data and security for crisis response, addressing the problem of delayed response caused by information dispersion and chaotic access in traditional crisis management. The offline processing module quickly establishes specialized response teams, conducts on-site investigations and problem-solving, and preserves evidence to ensure the root causes of offline conflicts are resolved and to prevent the crisis from escalating due to delays. The online response module monitors online public opinion dynamics in real time, generates tiered response content, and releases it through multiple channels to achieve precise public opinion guidance, addressing the problems of delayed and insufficiently targeted online responses in traditional systems. The dual-line collaboration module achieves two-way synchronization between offline response progress and online public opinion feedback, verifies information consistency, collects public suggestions, constructs a complete response loop, and improves information transparency and public participation. The trust reconstruction module introduces third-party supervision, publicizes long-term improvement mechanisms, and showcases trust restoration results to rebuild public trust in the organization, addressing the problems of slow and unconvincing trust restoration after a crisis. The prevention enhancement module conducts risk warning training, assesses grassroots response capabilities, and dynamically updates the case database to reduce the probability of crises from the source, addressing the problem of emphasizing response over prevention in traditional crisis management.

[0007] Furthermore, the basic support modules include: a basic public opinion monitoring module, a case management module, and a permissions and security module;

[0008] Basic public opinion monitoring module: Through reports from grassroots personnel and basic monitoring tools, it collects offline dispute clues, public demands and potential risk signals to form a risk warning list, providing a basis for crisis prediction;

[0009] Case Management Module: Stores historical crisis case handling plans, public opinion evolution paths, and lessons learned. It supports searching by crisis type and industry characteristics and intelligent matching with similar cases, providing a reference for current crisis response.

[0010] Access Control and Security Module: Set information access permissions for different roles, encrypt and store sensitive data, ensure that crisis information flows within a controllable range, and prevent information leakage or misuse.

[0011] Furthermore, the offline processing module includes: a processing scheduling module, an on-site processing module, and an evidence management module;

[0012] The dispatch module ensures that a special response team is formed and tasks are assigned within one hour of a crisis being triggered, and that the timeframes for response (1 hour), investigation plan development (2 hours), and on-site intervention (4 hours) are clearly defined to ensure efficient initiation of the response.

[0013] On-site handling module: Conduct in-depth on-site investigations at crisis sites, verify the essence of the problem, communicate and negotiate with the parties involved, formulate and implement solutions based on laws, regulations and actual circumstances, and promote the substantive resolution of offline conflicts;

[0014] Evidence Management Module: Collects, organizes, and stores key evidence from the offline handling process, including investigation records, recordings of communications between the parties, settlement agreements, and rectification certificates, forming a complete chain of evidence to provide a basis for subsequent online responses and tracing.

[0015] Furthermore, the online response module includes: a public opinion analysis module, a response generation module, and a multi-channel publishing module;

[0016] Public opinion analysis module: It uses web crawler technology to capture relevant information from social media, news websites, and forum platforms, uses natural language processing technology to filter keywords, analyze public sentiment, and determine the development trend of public opinion based on the scope of dissemination;

[0017] Response generation module: Based on the stage of public opinion, it calls a preset hierarchical script library and generates response content after review by public relations experts and legal counsel;

[0018] Multi-channel release module: Synchronizes approved response content to multiple channel platforms to promptly release the progress of crisis management, measures taken, and results achieved to the public.

[0019] Furthermore, the dual-line collaboration module includes: an information synchronization module, a progress and verification module, and a public feedback module;

[0020] Information synchronization module: Through the information sharing platform, the progress of offline handling and the dynamics of online public opinion are synchronized in two ways, ensuring that the offline investigation results, solutions and online content are consistent with each other;

[0021] Progress and Verification Module: Generates a time-stamped progress chart for the handling of cases, publishes offline evidence online, invites media and netizens to witness the handling results, forms a closed-loop verification, and improves information transparency;

[0022] Public feedback module: Collect public suggestions on the handling plan through the official WeChat account's message area and online questionnaires, screen frequently asked questions and send them to the offline handling team to optimize the solution and enhance public participation.

[0023] Furthermore, the trust reconstruction module includes: a third-party collaboration module, a long-term mechanism module, and a display module;

[0024] Third-party collaboration module: Invite authoritative third-party organizations to participate in the investigation process, supervise the compliance of the solution, and issue independent evaluation reports to address the lack of credibility of organizations proving their innocence;

[0025] Long-term mechanism module: Publicize and implement long-term improvement mechanisms, clarify responsible departments, implementation processes and assessment standards, and regularly update implementation data through the official website to address the problem of recurrence after a crisis;

[0026] Display Module: A "Trust Reconstruction Zone" has been set up on the official website to display third-party evaluation reports, the results of the implementation of long-term mechanisms, and the results of public satisfaction surveys, presenting the progress of public trust restoration in a visual way.

[0027] Furthermore, the prevention enhancement module includes: a risk warning training module, a grassroots assessment module, and an update module;

[0028] Risk warning training module: Based on historical case data, a risk prediction model is trained to provide training on dispute identification and response for grassroots staff, thereby improving their ability to detect risks early.

[0029] Grassroots assessment module: The indicators of "offline dispute resolution rate within 48 hours" and "repeated complaint rate" will be included in the performance assessment of grassroots staff, with a higher weight than "online public opinion deletion volume". Those who perform well will be rewarded, and those who fail to handle the situation will be trained and supervised to guide the grassroots to pay attention to solving practical problems.

[0030] Update module: Collect new crisis cases and handling experience every quarter, update the case library, training materials and risk prediction model parameters to ensure that the prevention system is adapted to new scenarios.

[0031] The beneficial effects of this invention, based on an offline-priority, dual-line closed-loop crisis public relations model, are as follows:

[0032] By constructing a full-process system that prioritizes offline operations and features a dual-loop system, the system prioritizes resolving core offline conflicts while simultaneously achieving real-time online and offline information collaboration and closed-loop verification. This effectively avoids the recurring public opinion issues caused by the traditional model's emphasis on online over offline operations. Furthermore, by introducing third-party supervision and multi-dimensional evaluation, the system significantly enhances the transparency and credibility of crisis management, ensuring that public trust is systematically restored from the early stages of the crisis.

[0033] By strengthening grassroots capacity building and risk early warning mechanisms, crisis prevention is embedded in daily management. Combined with a dynamically updated case library and standardized handling procedures, the efficiency of crisis response and the ability to resolve conflicts at the grassroots level have been greatly improved. This forms a complete closed loop from prevention and handling to trust reconstruction, completely changing the shortcomings of the traditional model of delayed response and one-sided assessment, and providing a scientific and efficient paradigm for crisis response in complex environments. Attached Figure Description

[0034] The present invention will now be described in further detail with reference to the accompanying drawings and specific implementation methods.

[0035] Figure 1 This is a schematic diagram of the system principle;

[0036] Figure 2 This is a flowchart illustrating the steps. Detailed Implementation

[0037] The present invention will be described in detail below with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in the present application can be combined with each other.

[0038] According to one aspect of the invention, such as Figures 1-2 As shown, a crisis public relations model based on a dual-loop system prioritizing offline operations is provided, including:

[0039] The basic support module integrates grassroots public opinion clues, historical case data, and access control, providing fundamental data and security for crisis response and solving the problem of delayed response caused by information fragmentation and chaotic access in traditional crisis response. This module includes:

[0040] Basic public opinion monitoring module: Through reports from grassroots personnel and basic monitoring tools, it collects offline dispute clues, public demands and potential risk signals (specifically including multi-source data such as community feedback forms, customer service complaint records, and grassroots inspection logs. After collection, the data is processed in a standardized manner (in a unified JSON format, including timestamps, data types and core demand fields) to form a risk warning list, providing a basis for crisis prediction.

[0041] Data standardization is achieved through preset field templates. For example, "product quality issues" are uniformly labeled as "quality_issue" and "service attitude complaints" are labeled as "service_complaint", ensuring that data from different sources are in a consistent format.

[0042] In addition, the collected data is initially classified according to "urgency" (e.g., mass incidents are marked as "high" and individual demands are marked as "medium / low"), and the classification rules are set based on the probability of risk escalation in historical cases.

[0043] Case Management Module: Stores the handling plans, public opinion evolution paths, and lessons learned from historical crisis cases (specifically including the types of disputes involved in the case, the timeline of handling, key decision-making nodes, channels of public opinion dissemination, and evaluation of the effectiveness of the solution (such as public satisfaction and the time it takes for public opinion to subside)). It supports searching by crisis type and industry characteristics and intelligent matching with similar cases to provide a reference for current crisis response.

[0044] The search function is implemented through keyword indexing; for example, entering "food quality" can quickly locate similar cases.

[0045] Similar case matching is based on the calculation of case feature vectors, extracting key features such as "dispute parties, scope of impact, and core demands" from the case, and calculating the degree of matching with the current crisis using the cosine similarity algorithm. The formula is as follows:

[0046]

[0047] Where A is the feature vector of the current crisis, and B is the feature vector of historical cases. i and B i These are the values ​​of the i-th dimension of the vector. For example, cases with a matching degree of ≥80% are automatically pushed to the handling group to shorten the decision-making time.

[0048] The permissions and security module sets information access permissions for different roles (specifically divided into three levels: administrator level (access to all data and operation logs), special task force level (access to cases and handling data only related to the current task), and frontline staff level (access to upload clues and view public warning information only)). Sensitive handling data is encrypted and stored to ensure that crisis information flows within a controllable range and prevent information leakage or misuse.

[0049] Sensitive data encryption uses the AES-256 symmetric encryption algorithm. The encrypted objects include information of the parties involved, details of the settlement agreement, etc. The key is automatically generated by the system and rotated regularly (e.g., updated every 7 days).

[0050] This setup controls information access boundaries through hierarchical permission settings, ensures data transmission and storage security through encryption technology, and, combined with operation logs (including access time, operator, and operation content), achieves full-process traceability, thus solving the problems of information misuse and leakage in traditional models.

[0051] The offline handling module enables the rapid formation of specialized response teams, the execution of on-site investigations and problem-solving, and the preservation of evidence to ensure that the root causes of offline conflicts are resolved and to prevent crises from escalating due to delays. This module includes:

[0052] The dispatch module ensures that a special response team is formed and tasks are assigned within one hour of a crisis being triggered, and that the timeframes for response (1 hour), investigation plan development (2 hours), and on-site intervention (4 hours) are clearly defined to ensure efficient initiation of the response.

[0053] The special task force was established through an expert database tag matching system. The system pre-sets expert tags in fields such as law, technology, and public relations (e.g., "food testing" and "consumer rights law"). When a crisis type (e.g., "product quality") is identified, the system automatically retrieves and pushes the list of experts with the highest matching degree. By using keyword mapping (e.g., associating "food safety crisis" with "food testing technology experts" and "food and drug administration regulations experts"), the system reduces the time spent on manual screening and ensures that the team members are professionally matched.

[0054] Task allocation uses a visual Gantt chart tool to clearly define the responsibilities of each member (such as technical personnel being responsible for tracing the source of problems, and legal personnel being responsible for reviewing the compliance of solutions) and delivery time. The system automatically pushes task reminders to members' terminals, and tasks that are not completed on time will be marked in red as a warning.

[0055] On-site handling module: Conduct in-depth on-site investigations at crisis sites, verify the essence of the problem, communicate and negotiate with the parties involved, formulate and implement solutions based on laws, regulations and actual circumstances, and promote the substantive resolution of offline conflicts;

[0056] The communication and negotiation process is recorded through electronic transcripts, which support real-time speech-to-text conversion (using an offline speech recognition engine to ensure usability in environments without network access), automatically extract and mark key information such as "demands" and "rectification requirements" to avoid omissions in manual recording, and require electronic signature confirmation from the parties involved after the transcript is generated to enhance its legal effect.

[0057] When formulating solutions, the system works in real time with the case management module via an API interface. It automatically retrieves historical solutions for similar crises (such as "compensation standards for similar product complaints"), dynamically adjusts them in conjunction with the current demands of the parties involved and relevant laws and regulations, and generates a draft solution with legal citations to improve the rationality and efficiency of the solution.

[0058] Evidence Management Module: Collects, organizes, and stores key evidence from the offline handling process, including investigation records, recordings of communications between the parties, settlement agreements, and rectification certificates, forming a complete chain of evidence to provide a basis for subsequent online responses and tracing.

[0059] The evidence storage uses blockchain technology to generate a unique hash value for each evidence document and associate it with a timestamp. This hash value is then uploaded to a consortium blockchain node (such as a node jointly maintained by an enterprise, law firm, or third-party organization) to ensure that the evidence is tamper-proof and to support tracing the original document at any time through the hash value.

[0060] The above measures have enabled the standardization of the entire process from personnel dispatch and on-site handling to evidence management, ensuring that offline conflicts are substantially resolved within the golden time frame and laying the foundation for a closed-loop system.

[0061] The online response module monitors online public opinion dynamics in real time, generates tiered response content, and releases it through multiple channels to achieve precise guidance of public opinion, solving the problems of delayed and insufficient targeting in traditional online responses. This module includes:

[0062] Public opinion analysis module: It uses web crawler technology to capture relevant information from social media, news websites, and forum platforms, uses natural language processing technology to filter keywords, analyze public sentiment, and determine the development trend of public opinion based on the scope of dissemination;

[0063] The web crawler technology adopts a distributed crawling architecture, which can simultaneously crawl targeted information from platforms such as Weibo, Douyin, and Zhihu. The crawling frequency is dynamically adjusted according to the popularity of public opinion (e.g., crawling once every 5 minutes during an outbreak) to ensure that no key information is missed.

[0064] Sentiment analysis is achieved through a pre-set sentiment dictionary (containing negative words such as "anger" and "questioning" and positive words such as "understanding" and "support"), which scores the text on sentiment (e.g., -100 to 100 points, with scores below 0 marked as negative information) to help quickly identify public opinion focus;

[0065] In addition, the scope of dissemination is determined by combining three-dimensional indicators: number of reposts, number of comments, and number of media reprints (for example, when the number of reposts exceeds 100,000 times in a single hour or more than 5 mainstream media outlets reprint it), it is automatically determined to be a period of public opinion outbreak and triggers a high-level response mechanism.

[0066] Response generation module: Based on the stage of public opinion, it calls a preset hierarchical script library and generates response content after review by public relations experts and legal counsel;

[0067] The public opinion situation is divided into three stages: the attention period, the fermentation period, and the outbreak period. The pre-set graded script library stores the scripts according to the attention period, fermentation period, and outbreak period. The scripts for the attention period focus on reassuring statements such as "an investigation has been launched". The fermentation period adds "preliminary progress + resolution timeframe" and the outbreak period includes the core elements of "apology + third-party supervision" to ensure that the scripts match the public opinion situation stage.

[0068] The review process is conducted through an online collaboration platform, where public relations experts review the communication effect and legal counsel reviews the legal risks. Once both reviews are passed, a publishable version is generated, thus shortening the review time.

[0069] In addition, quick reply templates are set up for frequently asked questions (such as "compensation standards" and "rectification period"), which can be directly embedded with the response content to improve response efficiency.

[0070] Multi-channel release module: Synchronizes approved response content to multiple channel platforms to promptly release the progress of crisis management, measures taken, and results achieved to the public;

[0071] These multiple channels and platforms include the official website, Weibo, Douyin, and live press conferences;

[0072] In addition, keyword alerts (such as "perfunctory" or "lying") are set for negative comments in the comment section to automatically remind operators to respond first and prevent the spread of negative emotions.

[0073] The above modules work together to achieve precision across the entire process from public opinion monitoring and content generation to multi-channel release, ensuring that online responses and offline actions are synchronized and effectively guiding public opinion.

[0074] The dual-track collaboration module enables two-way synchronization between offline response progress and online public opinion feedback, verifies information consistency, collects public suggestions, constructs a complete response loop, and improves information transparency and public participation. This module includes:

[0075] Information synchronization module: Through the information sharing platform, the progress of offline handling and the dynamics of online public opinion are synchronized in two ways, ensuring that the offline investigation results, solutions and online content are consistent with each other;

[0076] The system sets verification rules for bidirectional synchronized data. For example, the offline "solution execution completed" status must match the online "execution result" description. If they are inconsistent, the system will issue an alert to prompt relevant personnel to check and correct the data.

[0077] In addition, sensitive information (such as the privacy of the parties involved) is automatically anonymized, and fields such as ID number and contact information are hidden during synchronization, retaining only the necessary public information.

[0078] Progress and Verification Module: Generates a time-stamped progress chart for the handling of cases, publishes offline evidence online, invites media and netizens to witness the handling results, forms a closed-loop verification, and improves information transparency;

[0079] The progress tracking uses a timestamp chain, with each node (such as "investigation launched" or "plan announced") generating a unique timestamp that cannot be tampered with and is linked to offline operation logs to ensure that the progress is real and traceable.

[0080] The evidence displayed online is linked to a blockchain-based notarization link, allowing the public to view the original evidence files with hash values ​​and verify that the evidence has not been tampered with.

[0081] Public feedback module: Collect public suggestions on the handling plan through the official WeChat account's message area and online questionnaires, screen frequently asked questions and send them to the offline handling team to optimize the solution and enhance public participation;

[0082] The online questionnaire channel incorporates AI semantic analysis tools to automatically identify core demands such as "compensation standards" and "rectification time limits" in public suggestions, and summarizes them by topic.

[0083] When filtering frequently asked questions, they are sorted by frequency of mention, and the top ten questions are automatically generated into a list of questions to be responded to, which is then pushed to the offline handling team through the information synchronization module.

[0084] At the same time, the adoption of public suggestions (such as the optimization points of the plan) will be publicized.

[0085] This achieves a deep integration of offline handling and online response, ensuring both the timeliness and accuracy of information synchronization and enhancing the credibility of crisis management through public participation, thus forming a complete response loop.

[0086] The trust rebuilding module introduces third-party oversight, publicizes long-term improvement mechanisms, and showcases trust restoration results to rebuild public trust in the organization and address the issues of slow and unconvincing trust restoration after a crisis. This module includes:

[0087] Third-party collaboration module: Invite authoritative third-party organizations to participate in the investigation process, supervise the compliance of the solution, and issue independent evaluation reports to address the lack of credibility of organizations proving their innocence;

[0088] The selection of third-party organizations is carried out through random selection from a publicly available list, which includes qualified industry associations, testing institutions, law firms, etc. The entire selection process is recorded and archived to ensure impartiality.

[0089] The supervision process adopts a "node confirmation system," whereby third parties are required to sign and confirm key nodes such as investigation initiation, plan formulation, and implementation acceptance. These opinions, along with organizational handling records, are uploaded to the blockchain and cannot be tampered with.

[0090] Long-term mechanism module: Publicize and implement long-term improvement mechanisms, clarify responsible departments, implementation processes and assessment standards, and regularly update implementation data through the official website to address the problem of recurrence after a crisis;

[0091] The assessment criteria are directly linked to the performance of grassroots staff (for example, department heads whose "dispute resolution rate within 7 days" does not reach 90% need to receive training), and the assessment data is synchronized to the display module in real time.

[0092] Display module: A "Trust Reconstruction Zone" has been set up on the official website to display third-party evaluation reports, the results of the implementation of long-term mechanisms, and the results of public satisfaction surveys, so as to intuitively present the progress of public trust restoration;

[0093] The Trust Rebuilding Zone uses dynamic data visualization, such as using trend charts to present the process of "public satisfaction," and hovering the mouse over it to view the specific events of each data node (such as "satisfaction improved after the release of the third-party report").

[0094] This approach transforms the abstract concept of "repairing public trust" into quantifiable and traceable concrete actions, gradually rebuilding public trust in the organization through third-party endorsement and transparent presentation.

[0095] The prevention enhancement module conducts risk warning training, assesses grassroots response capabilities, and dynamically updates the case database to reduce the probability of crises occurring at the source, addressing the problem of traditional crisis management that emphasizes response over prevention. This module includes:

[0096] Risk warning training module: Based on historical case data, a risk prediction model is trained to provide training on dispute identification and response for grassroots staff, thereby improving their ability to detect risks early.

[0097] The risk prediction model uses a decision tree algorithm. It takes the characteristics of disputes reported by the grassroots level (such as "complaint type" and "number of people involved") as input, and then outputs the risk level (such as low / medium / high) and early warning suggestions (such as "high risk must be reported within 2 hours"). The model is iterated and optimized monthly based on new case data.

[0098] The training uses VR simulation technology to recreate high-frequency scenarios such as "shopping mall return and exchange conflicts" and "community property disputes," allowing grassroots staff to practice communication skills and handling procedures through immersive operation.

[0099] Grassroots assessment module: The indicators of "offline dispute resolution rate within 48 hours" and "repeated complaint rate" will be included in the performance assessment of grassroots staff, with a higher weight than "online public opinion deletion volume". Those who perform well will be rewarded, and those who fail to handle the situation will be trained and supervised to guide the grassroots to pay attention to solving practical problems.

[0100] The offline dispute resolution rate within 48 hours is:

[0101]

[0102] The data will be updated and visualized in real time.

[0103] Update module: Collect new crisis cases and handling experience every quarter, update the case library, training materials and risk prediction model parameters to ensure that the prevention system is adapted to new scenarios.

[0104] The above design enables the prevention enhancement module to shift from "post-event response" to "pre-event prevention," continuously improving the organization's risk resistance capabilities through data-driven risk warning, scenario-based training, and dynamic optimization mechanisms.

[0105] Of course, the above description is not a limitation of the present invention, and the present invention is not limited to the examples given above. Any changes, modifications, additions or substitutions made by those skilled in the art within the scope of the present invention are also within the protection scope of the present invention.

Claims

1. A crisis public relations model based on offline priority and dual-line closed loop, characterized in that, include: The basic support module is used to integrate grassroots public opinion clues, historical case data, and access control, providing basic data and security for crisis response and solving the problem of delayed response caused by information dispersion and chaotic access in traditional crisis response; the offline processing module is used to quickly form special handling teams, carry out on-site investigations and problem solving, and retain handling evidence to ensure that the root causes of offline conflicts are resolved and to prevent the crisis from worsening due to delays. The online response module is used to monitor online public opinion dynamics in real time, generate tiered response content, and release it through multiple channels to achieve precise guidance of public opinion, solving the problems of delayed and insufficient targeting in traditional online responses. The dual-line collaboration module is used to achieve two-way synchronization between offline handling progress and online public opinion feedback, verify information consistency, collect public suggestions, build a complete response loop, and improve information transparency and public participation. The trust reconstruction module is used to introduce third-party supervision, publicize long-term improvement mechanisms, and showcase trust restoration results to rebuild public trust in the organization, solving the problems of slow and unconvincing trust restoration after a crisis. The prevention enhancement module is used to conduct risk warning training, assess grassroots handling capabilities, and dynamically update the case library to reduce the probability of crisis occurrence from the source, solving the problem of emphasizing response over prevention in traditional crisis management.

2. The crisis public relations model based on offline priority dual-line closed loop as described in claim 1, characterized in that: The basic support modules include: a basic public opinion monitoring module, a case management module, and a permissions and security module; Basic public opinion monitoring module: Through reports from grassroots personnel and basic monitoring tools, it collects offline dispute clues, public demands and potential risk signals to form a risk warning list, providing a basis for crisis prediction; Case Management Module: Stores historical crisis case handling plans, public opinion evolution paths, and lessons learned. It supports searching by crisis type and industry characteristics and intelligent matching with similar cases, providing a reference for current crisis response. Access Control and Security Module: Set information access permissions for different roles, encrypt and store sensitive data, ensure that crisis information flows within a controllable range, and prevent information leakage or misuse.

3. The crisis public relations model based on offline priority dual-line closed loop as described in claim 1, characterized in that: The offline processing module includes: a processing scheduling module, an on-site processing module, and an evidence management module; The dispatch module ensures that a special response team is formed and tasks are assigned within one hour of a crisis being triggered, and that the timeframes for response (1 hour), investigation plan development (2 hours), and on-site intervention (4 hours) are clearly defined to ensure efficient initiation of the response. On-site handling module: Conduct in-depth on-site investigations at crisis sites, verify the essence of the problem, communicate and negotiate with the parties involved, formulate and implement solutions based on laws, regulations and actual circumstances, and promote the substantive resolution of offline conflicts; Evidence Management Module: Collects, organizes, and stores key evidence from the offline handling process, including investigation records, recordings of communications between the parties, settlement agreements, and rectification certificates, forming a complete chain of evidence to provide a basis for subsequent online responses and tracing.

4. The crisis public relations model based on offline priority dual-line closed loop as described in claim 1, characterized in that: The online response module includes: a public opinion analysis module, a response generation module, and a multi-channel publishing module; Public opinion analysis module: It uses web crawler technology to capture relevant information from social media, news websites, and forum platforms, uses natural language processing technology to filter keywords, analyze public sentiment, and determine the development trend of public opinion based on the scope of dissemination; Response generation module: Based on the stage of public opinion, it calls a preset hierarchical script library and generates response content after review by public relations experts and legal counsel; Multi-channel release module: Synchronizes approved response content to multiple channel platforms to promptly release the progress of crisis management, measures taken, and results achieved to the public.

5. The crisis public relations model based on offline priority dual-line closed loop as described in claim 1, characterized in that: The dual-line collaboration module includes: an information synchronization module, a progress and verification module, and a public feedback module; Information synchronization module: Through the information sharing platform, the progress of offline handling and the dynamics of online public opinion are synchronized in two ways, ensuring that the offline investigation results, solutions and online content are consistent with each other; Progress and Verification Module: Generates a time-stamped progress chart for the handling of cases, publishes offline evidence online, invites media and netizens to witness the handling results, forms a closed-loop verification, and improves information transparency; Public feedback module: Collect public suggestions on the handling plan through the official WeChat account's message area and online questionnaires, screen frequently asked questions and send them to the offline handling team to optimize the solution and enhance public participation.

6. The crisis public relations model based on offline priority dual-line closed loop as described in claim 1, characterized in that: The trust reconstruction module includes: a third-party collaboration module, a long-term mechanism module, and a display module; Third-party collaboration module: Invite authoritative third-party organizations to participate in the investigation process, supervise the compliance of the solution, and issue independent evaluation reports to address the lack of credibility of organizations proving their innocence; Long-term mechanism module: Publicize and implement long-term improvement mechanisms, clarify responsible departments, implementation processes and assessment standards, and regularly update implementation data through the official website to address the problem of recurrence after a crisis; Display module: A "Trust Reconstruction Zone" is set up on the official website to display third-party evaluation reports, the results of the implementation of long-term mechanisms, and the results of public satisfaction surveys, so as to intuitively present the progress of public trust restoration.

7. The crisis public relations model based on offline priority dual-line closed loop as described in claim 1, characterized in that: The prevention enhancement module includes: a risk warning training module, a grassroots assessment module, and an update module; Risk warning training module: Based on historical case data, a risk prediction model is trained to provide training on dispute identification and response for grassroots staff, thereby improving their ability to detect risks early. Grassroots assessment module: The indicators of "offline dispute resolution rate within 48 hours" and "repeated complaint rate" will be included in the performance assessment of grassroots staff, with a higher weight than "online public opinion deletion volume". Those who perform well will be rewarded, and those who fail to handle the situation will be trained and supervised to guide the grassroots to pay attention to solving practical problems. Update module: Collect new crisis cases and handling experience every quarter, update the case library, training materials and risk prediction model parameters to ensure that the prevention system is adapted to new scenarios.

Images