A method and system for private group chats based on keyword interception and message visibility management

Abstract

This application provides a method and system for private group chat based on keyword interception and message visibility management, relating to the field of communications. The method includes: receiving a first message sent by a first member in a target private group and identifying the first member's permissions; determining a keyword set based on the attributes of the target private group and the first member's permissions; intercepting the first message if its content includes any keyword from the keyword set; dynamically generating target visibility configuration information for each second member of the first message, combining a context-aware algorithm and personalized visibility configuration results, if the first message does not include all keywords from the keyword set; and selectively broadcasting the first message to all members or designated members in the target private group based on the target visibility configuration information for each second member. The technical solution provided in this application improves the flexibility of keyword interception and message visibility management.

Description

Technical Field

[0001] This application relates to the field of communication technology, and in particular to a private group chat method and system based on keyword interception and message visibility management. Background Technology

[0002] In today's instant messaging and social networking environment, private group chat functionality is widely used in scenarios such as project management, team communication, and file sharing. To protect user privacy and ensure message security, developers have been exploring ways to more effectively manage and control the scope and visibility of messages. Traditional private group chat systems typically use a broadcast messaging approach, meaning that once a message is sent, all members in the group can see it. While simple and direct, this method lacks fine-grained control over message propagation, especially when sensitive content is involved, easily leading to information leaks or inappropriate information exposure.

[0003] Existing private group chat systems primarily manage and control the scope and visibility of messages through static access control, fixed keyword blocking, and rule-based visibility configuration. Static access control is simple to implement; for example, administrators can set different speaking permissions for different members, making it suitable for situations with relatively fixed permission structures. Fixed keyword blocking effectively prevents the spread of obviously inappropriate messages. Rule-based visibility configuration allows manual configuration of message visibility for different members, such as setting certain messages to be visible only to specific teams or roles.

[0004] However, static access control lacks flexibility, fixed keyword blocking has false positives and false negatives when the scenario changes, and rule-based visibility configuration is inefficient. Therefore, existing private group chat systems suffer from low effectiveness of keyword blocking, low efficiency of message visibility management, and poor flexibility in complex language environments and changing communication scenarios. Summary of the Invention

[0005] This application provides a method and system for private group chat based on keyword interception and message visibility management, in order to solve the problems of low effectiveness of keyword interception, low efficiency and poor flexibility of message visibility management in the prior art.

[0006] In a first aspect, embodiments of this application provide a method for private group chat based on keyword interception and message visibility management, including:

[0007] Receive the first message sent by the first member in the target private group and identify the first member's permissions;

[0008] Based on the attributes of the target private group and the permissions of the first member, a set of keywords is determined. When the content of the first message includes any keyword from the set of keywords, the first message is intercepted, making the first message invisible to all members of the target private group.

[0009] If the first message does not include all the keywords in the keyword set, the target visibility configuration information of the first message for each second member is dynamically generated by combining the context-aware algorithm and the personalized visibility configuration result. The second member is the other member in the target private group other than the first member.

[0010] Based on the target visibility configuration information for each second member in the first message, the first message is selectively broadcast to all members or designated members in the target private group.

[0011] Optionally, the step of dynamically generating the target visibility configuration information of the first message for each second member by combining the context-aware algorithm and the personalized visibility configuration results includes:

[0012] Obtain the initial visibility configuration information of the first message for each second member;

[0013] When the group configuration information of the target private group indicates that there are context-aware requirements and personalized visibility configuration requirements, the features of the first message are extracted using a multimodal parsing algorithm. Based on the features of the first message, a preset classification algorithm is applied to classify the media type of the first message to obtain the media type of the first message. The media type is text, image, audio, video, file, business card, location message, email or chat history.

[0014] Based on the media type of the first message, combined with the role and permissions of each second member, a security assessment algorithm is applied to assess the security of the first message for each second member, and the security assessment results of the first message for each second member are obtained.

[0015] Based on the security assessment results of the first message for each second member, the visibility configuration information is adjusted to obtain the intermediate visibility configuration information of the first message for each second member.

[0016] By combining the context-aware algorithm and the results of personalized visibility configuration, the intermediate visibility configuration information of the first message for each second member is optimized to obtain the target visibility configuration information of the first message for each second member.

[0017] Optionally, obtaining the initial visibility configuration information of the first message for each second member includes:

[0018] A multi-level knowledge graph corresponding to the target private group is constructed. The multi-level knowledge graph includes multi-level nodes and edges. The first-level nodes in the multi-level nodes are used to represent members in the target private group, the second-level nodes are used to represent the teams to which the members belong, and the third-level nodes are used to represent the projects that the teams participate in. Different edges are used to reflect the relationships between different nodes, and each node has corresponding node attributes.

[0019] Based on the content of the first message and the multi-level knowledge graph, the purpose of the group chat in the first message is determined. Based on the purpose of the group chat in the first message, the initial visibility configuration information of the first message for each second member is determined. The content of the first message includes at least one of the following: project name, team name, and member name.

[0020] Optionally, the step of combining the context-aware algorithm and the personalized visibility configuration results to optimize the intermediate visibility configuration information of the first message for each second member, thereby obtaining the target visibility configuration information of the first message for each second member, includes:

[0021] Based on the node attributes in the multi-level knowledge graph, the context perception result of each second member is determined using a context perception algorithm. The node attributes include context features, which include time, geographical location, and activity status. The context perception result of the second member is used to indicate whether the second member is in a working period, in an office area, or in a project discussion phase.

[0022] Based on the node attributes in the multi-level knowledge graph, combined with user profiling technology and collaborative filtering algorithm, the personalized visibility configuration results of each second member are constructed. The node attributes include the second member's role, permissions, and historical messages.

[0023] The context awareness results and personalized visibility configuration results of each second member are fused to obtain a fusion result. Based on the fusion result, the intermediate visibility configuration information of the first message for each second member is optimized to obtain the target visibility configuration information of the first message for each second member.

[0024] Optionally, the process of constructing personalized visibility configuration results for each second member based on node attributes in a multi-level knowledge graph, combined with user profiling technology and collaborative filtering algorithms, includes:

[0025] Based on the node attributes in a multi-level knowledge graph, user profile information of the second member is constructed using user profiling technology.

[0026] Based on the user profile information of the second member, collaborative filtering algorithms are used to perform collaborative filtering between members and between messages to obtain the collaborative filtering results of the second member and the filtering results of the first message; during the collaborative filtering process, cosine similarity is used to calculate the similarity between the second members.

[0027] By integrating the user profile information of the second member, the collaborative filtering results of the second member, and the filtering results of the first message, the personalized visibility configuration results of each second member are obtained.

[0028] Optionally, the step of applying a security assessment algorithm based on the media type of the first message, combined with the role and permissions of each second member, to assess the security of the first message for each second member, and obtaining the security assessment result of the first message for each second member, includes:

[0029] Obtain the mapping relationship between the role and permissions of each second member and the media type of the first message, as well as the context information of the first message;

[0030] Based on the mapping relationship, a security assessment algorithm is applied to evaluate the security of the first message for each second member to obtain an initial security score. The security of the context information for each second member is then evaluated to obtain a context security score.

[0031] Based on the correlation between context information and the first message, the influence of the context security score is determined, and the product of the influence of the context security score and the context security score is determined as the adjustment value.

[0032] The adjusted value and the initial security score are added together to obtain the security assessment result of the first message for each second member.

[0033] Optionally, the step of selectively broadcasting the first message to all members or designated members in the target private group based on the target visibility configuration information of the first message for each second member includes:

[0034] The first message is parsed to obtain the target visibility configuration information for each second member, resulting in the visibility level, time window, and geographical restrictions.

[0035] The second member is classified into multiple groups based on visibility level, time window, and geographic restrictions;

[0036] Execute the message broadcasting policy corresponding to each group, and selectively broadcast the first message to all members or designated members in the target private group according to the message broadcasting policy.

[0037] Secondly, embodiments of this application provide a private group chat system based on keyword interception and message visibility management, including:

[0038] The receiving and identification module is used to receive the first message sent by the first member in the target private group and identify the permissions of the first member;

[0039] The interception module is used to determine a set of keywords based on the attributes of the target private group and the permissions of the first member. When the content of the first message includes any keyword in the set of keywords, the first message is intercepted, so that the first message is not visible to all members of the target private group.

[0040] The generation module is used to dynamically generate target visibility configuration information for each second member of the first message, combining context-aware algorithms and personalized visibility configuration results, when the first message does not include all keywords in the keyword set. The second members are other members in the target private group other than the first member.

[0041] The broadcast module invokes and selectively broadcasts the first message to all members or designated members in the target private group, based on the target visibility configuration information for each second member according to the first message.

[0042] Thirdly, embodiments of this application provide a computing device, including a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are to be invoked and executed by the processing component to implement a private group chat method based on keyword interception and message visibility management as described in any of the first aspects.

[0043] Fourthly, embodiments of this application provide a computer storage medium storing a computer program, which, when executed by a computer, implements a private group chat method based on keyword interception and message visibility management as described in any of the first aspects.

[0044] This application provides a method for private group chat based on keyword interception and message visibility management, including: receiving a first message sent by a first member in a target private group and identifying the first member's permissions; determining a keyword set based on the attributes of the target private group and the first member's permissions; intercepting the first message when its content includes any keyword from the keyword set, making the first message invisible to all members in the target private group; dynamically generating target visibility configuration information for each second member of the first message, where the second members are members in the target private group other than the first member, by combining a context-aware algorithm and personalized visibility configuration results when the first message does not include all keywords from the keyword set; and selectively broadcasting the first message to all members or designated members in the target private group based on the target visibility configuration information for each second member of the first message.

[0045] This application's embodiments comprehensively consider the attributes of the target private group and the permissions of the first member, enabling more accurate determination of the corresponding keyword set and ensuring the effectiveness and targeting of keyword interception. By combining context-aware algorithms and personalized visibility configuration results, the target visibility configuration information for each second member in the first message is precisely managed. This allows the target visibility configuration information to be flexibly adjusted according to actual conditions, improving the efficiency and flexibility of message visibility management. This, in turn, enhances the security and user experience of the private group chat system, adapting to complex language environments and changing communication scenarios. Specifically, by constructing a multi-level knowledge graph and using multimodal parsing algorithms to extract message features, the content and background of messages can be understood, automatically determining the initial visibility configuration and improving the intelligence level of message management. Considering the context-aware needs and personalized visibility configuration requirements of the target private group, the final target visibility configuration information is dynamically generated. This flexibility allows the private group chat system to adapt to diverse usage scenarios and provide a more personalized service experience.

[0046] These or other aspects of this application will become more apparent in the following description of the embodiments. Attached Figure Description

[0047] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0048] Figure 1 A flowchart illustrating a private group chat method based on keyword interception and message visibility management, provided for an embodiment of this application;

[0049] Figure 2 A schematic diagram of the structure of a private group chat system based on keyword interception and message visibility management, provided for an embodiment of this application;

[0050] Figure 3 This is a schematic diagram of the structure of a computing device provided in an embodiment of this application. Detailed Implementation

[0051] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0052] Figure 1 A flowchart illustrating a private group chat method based on keyword interception and message visibility management, provided as an embodiment of this application, is shown below. Figure 1 As shown, the method includes:

[0053] S11. Receive the first message sent by the first member in the target private group and identify the permissions of the first member.

[0054] It should be understood that the private group chat method based on keyword interception and message visibility management can be executed by a private group chat system that can manage multiple private groups. The target private chat group is any one of the private groups managed by the private group chat platform. The first message can be any multimedia message type, such as text, images, audio, video, files, business cards, location messages, emails, chat logs, etc. The first member is the member who sent the first message.

[0055] It should also be understood that in the target private group, different members have different permissions. For example, senior administrators have senior management permissions, ordinary administrators have ordinary management permissions, and ordinary members have general permissions.

[0056] Specifically, senior administrators with advanced management privileges have the following functions: (1.1) Manage all settings of the entire private group, including adding / removing members, assigning permissions, etc.; (1.2) View all messages blocked by keyword sets and decide whether to allow them or process them further; (1.3) Adjust the message visibility configuration of any member and set personalized visibility rules; (1.4) Develop and update security policies such as keyword sets and context-aware rules. For example, senior administrator A can add a new member B to the development team's private group chat and grant him ordinary administrator privileges. A can view and approve a message containing sensitive information that was originally automatically blocked by the system because it contained the keyword "confidential". A can dynamically adjust the message visibility configuration of certain members according to the project progress, for example, allowing some members of the marketing team to view part of the development progress during key stages of the project.

[0057] A regular administrator with general management privileges has the following functions: (2.1) They can add or remove regular members, but cannot adjust the privileges of other administrators; (2.2) They can view and manage messages sent by regular members, but cannot view private messages of senior administrators; (2.3) They can set basic visibility configurations for regular members, but cannot modify complex rules set by senior administrators; (2.4) They can adjust keyword sets and context-aware rules within a certain range, but require approval from senior administrators. For example, a regular administrator B can add a newly hired developer C to the development team and grant them regular member privileges. B can view and process messages sent by regular members, ensuring that the message content meets security standards. B can set simple visibility rules for regular members, such as restricting certain sensitive information to be visible only during working hours.

[0058] Ordinary members with general permissions have the following functions: (3.1) They can send and receive messages within the group, but are limited by the keyword set and corresponding visibility configuration; (3.2) Based on the visibility configuration, they can only view message content that they have permission to view; (3.3) They can report potentially inappropriate messages to the administrator, but do not have the right to handle them directly. For example, ordinary member C can send a message about code issues in the target private group, but if the message contains the keyword "password," it will be automatically blocked by the private group chat system. C can only see message content that the administrator has approved and deemed that C has permission to view, such as project progress and task assignments. If C finds a potentially inappropriate message, C can report it to the administrator for further processing.

[0059] The aforementioned permission identification ensures the legitimacy of the message source; only members with the corresponding permissions can send messages, thus enhancing the security of the private group chat system.

[0060] S12. Based on the attributes of the target private group and the permissions of the first member, determine the keyword set. When the content of the first message includes any keyword in the keyword set, intercept the first message so that the first message is not visible to any member in the target private group.

[0061] It should be understood that a keyword set is a group of sensitive or prohibited keywords, usually set by the administrator, used to filter message content. Interception is the act of preventing the first message from being seen by other members. Embodiments of this application can customize keyword sets for different types of private groups, such as industry-specific terms in finance, healthcare, and education.

[0062] It should also be understood that the attributes of the target private group include business characteristics of the industry, group type (such as the entire company team, project team, department discussion group, temporary work group, etc.), and security level. Different security levels affect the strictness of the keyword set. Specifically, high-security-level keyword sets are more stringent, containing more sensitive words such as "confidential," "password," and "source code." Private group chat systems may employ more complex filtering rules and context-aware mechanisms to ensure all messages meet the highest security standards. Medium-security-level keyword sets are more moderate, containing some common but important sensitive words such as "contract" and "quotation." Private group chat systems employ a balanced security strategy, ensuring information security without impacting work efficiency. Low-security-level keyword sets are relatively lenient, mainly containing a small number of basic sensitive words such as "urgent" and "important." Private group chat systems may allow more messages to pass through, suitable for open or low-sensitivity communication scenarios.

[0063] Analysis of the above technical solutions reveals that: keyword sets effectively prevent the leakage of sensitive information; and by dynamically adjusting the keyword set, they can adapt to different business needs and security requirements. The real-time interception mechanism improves the response speed of the private group chat system, preventing the spread of inappropriate messages at the first instance. Combining the attributes of the target private group and the permissions of the first member, the private group chat system can more accurately manage and control the visibility and security of messages within the private group chat, ensuring that the transmission of the first message is both efficient and secure. This layered and customized security management mechanism not only improves the flexibility of the private group chat system but also better adapts to complex and ever-changing business needs.

[0064] S13. If the first message does not include all keywords in the keyword set, combine the context-aware algorithm and the personalized visibility configuration result to dynamically generate the target visibility configuration information of the first message for each second member. The second member is the other member in the target private group other than the first member.

[0065] It should be understood that context-aware algorithms can utilize factors such as time, geographic location, and activity status to assess the impact of the current context on message visibility. Personalized visibility configuration results can refer to message visibility configurations customized based on each member's role, permissions, historical behavior, and other characteristics. Target visibility configuration information refers to the final determined message visibility settings, including visibility level, time window, and geographic restrictions.

[0066] For example, all members of the development team in the target private group can view all the content in the first message during working hours and in the office area, while some members of the marketing team (such as the marketing team leader) can view some of the content in the first message outside of working hours and in the office area.

[0067] In step S13, context awareness and personalized configuration improve the flexibility and accuracy of message management, ensuring that message delivery meets both security requirements and personalized needs. The dynamic generation mechanism can adaptively adjust and continuously optimize message management strategies, improving user experience.

[0068] S14. Based on the target visibility configuration information of the first message for each second member, selectively broadcast the first message to all members or designated members in the target private group.

[0069] Specifically, based on the target visibility configuration information for each second member in the first message, perform one or more of the following operations:

[0070] a) Visible to all: If the first message's target visibility configuration information for all second members indicates that all content is publicly available to all members, then the first message will be broadcast directly to all members within the entire target private group.

[0071] b) Partially visible modes, including at least one of the following:

[0072] If the first message indicates that the target visibility configuration information for some second members is to disclose all content to these members, then the first message will be broadcast to these members. Therefore, in this embodiment of the application, the first message is only visible to certain specific members, ensuring that the first message will not be leaked to irrelevant personnel.

[0073] If the first message indicates that some content is to be made public to some members based on the target visibility configuration information of some members, then the other non-public parts of the first message will be obfuscated or made private in order to broadcast the part of the content to be made public to some members.

[0074] c) Visible after review mode: For first messages that require additional review, it can be set to submit them to the administrator or a member with review authority for approval first, and then determine the final visibility range of the first message based on the review results.

[0075] d) Private message mode: Allows the first member to select one or more specific members as recipients to form a one-to-one or one-to-many private conversation. In this case, the message is only visible to the selected recipients.

[0076] Therefore, selective broadcasting refers to deciding which members or all members to broadcast a message to based on target visibility configuration information. Selective broadcasting ensures message security and relevance, avoids unnecessary information exposure, and protects sensitive data. This refined management improves communication efficiency, ensuring messages are only delivered to those who need them and reducing interference from irrelevant personnel.

[0077] Analysis of the above technical solutions reveals that by executing steps S11-S14, this embodiment of the application can effectively prevent messages containing sensitive content from spreading within the target private group using keyword sets and a real-time interception mechanism. The strictness of the keyword set is determined by the group's security level, ensuring that security requirements under different business scenarios are met. The real-time interception mechanism ensures that sensitive information is not unintentionally or maliciously spread to other members, enhancing the security of the private group chat system. Combining context-aware algorithms and personalized visibility configuration results, target visibility configuration information for each member is dynamically generated, ensuring that message delivery meets both security requirements and personalized needs.

[0078] Furthermore, the visibility configuration in this application can be continuously optimized based on actual conditions and member feedback to maintain optimal security strategies and communication effectiveness. The selective broadcast mechanism ensures that messages are only delivered to those who need them, reducing interference from irrelevant personnel, improving communication efficiency, and ensuring the targeted and secure delivery of the first message, thereby enhancing the user experience.

[0079] By employing context awareness and personalized configuration, the private group chat system can more accurately assess the security and visibility of the first message, reducing unnecessary false positives and false negatives. Through hierarchical permission management and customized keyword sets, the private group chat system can more finely control the visibility and security of the first message. Therefore, the embodiments of this application not only accurately manage and control the visibility and security of messages in private group chats but also ensure that the security and personalization needs of the first message are fully considered. This method overcomes the shortcomings of existing technologies and provides a more flexible, accurate, and efficient private group chat management solution.

[0080] In one possible embodiment, after step S14, the present application embodiment may also perform steps 15-16.

[0081] S15. Based on the completed message processing, record and save the results of each message processing step to generate a detailed message processing log. Specifically, the message processing results include sender information, receiver information, message content summary (de-sensitized), keyword matching results, visibility configuration details, broadcast range, etc. The message processing results are stored in a structured manner in a secure log system to ensure data integrity and security. This step clarifies the processing procedure for each message, facilitating accountability and ensuring system transparency and traceability. In case of abnormal situations, this embodiment of the application can quickly locate problems through logs, improving troubleshooting efficiency.

[0082] S16. Based on the processing status of the first message, send a second message to the first member to provide immediate feedback, informing them whether the first message was successfully sent or intercepted, and providing details of its visibility configuration. The second message can be a pop-up, SMS, email, or displayed directly within the group, ensuring the first member receives prompt, immediate feedback. If the first message is intercepted or not sent as expected, the first member can adjust the message content or resend it based on the immediate feedback. For first messages containing sensitive information, immediate feedback helps the first member recognize potential security risks and take necessary preventative measures.

[0083] As described above, this embodiment provides detailed logging and saving functions, ensuring the traceability and security of message processing, and providing strong support for auditing, troubleshooting, and security monitoring. Through an instant feedback mechanism, the first member can quickly understand the processing status and visibility settings details of the first message, enhancing the transparency and accuracy of communication, while helping users correct errors promptly and promoting team collaboration. This embodiment is not only suitable for complex communication scenarios within enterprises but can also be widely applied to various industries.

[0084] To accurately and dynamically generate the target visibility configuration information of the first message for each second member, embodiments of this application can establish a dynamic generation process. For example, in one possible embodiment, combining a context-aware algorithm and personalized visibility configuration results, dynamically generating the target visibility configuration information of the first message for each second member includes:

[0085] Step 131: Obtain the initial visibility configuration information for each second member of the first message. It should be understood that the initial visibility configuration information refers to the message visibility rules set by the private group chat system for each second member at the initial stage of message processing, based on the multi-level knowledge graph and the content of the first message. These rules determine whether the second member can see, partially see, or not see the first message at all.

[0086] As one possible implementation, step 131, obtaining the initial visibility configuration information of the first message for each second member, includes:

[0087] Step a1: Construct a multi-level knowledge graph corresponding to the target private group. The multi-level knowledge graph includes multi-level nodes and edges. First-level nodes represent members of the target private group, second-level nodes represent the teams to which members belong, and third-level nodes represent the projects the teams participate in. Different edges reflect the relationships between different nodes, and each node has corresponding node attributes. Edges represent relationships between different nodes, such as a member belonging to a team or a team participating in a project. The multi-level knowledge graph provides a clear organizational structure view, facilitating the understanding and management of complex relationship networks. It covers multi-level information from individual members to the entire project, ensuring the comprehensiveness and integrity of information management. The multi-level knowledge graph can be dynamically updated according to actual conditions, ensuring the real-time nature and accuracy of information. Based on the rich node attributes in the multi-level knowledge graph, more refined message visibility configuration can be achieved to meet personalized needs in different scenarios.

[0088] Step a2: Based on the content of the first message and the multi-level knowledge graph, determine the group chat purpose of the first message. Based on this purpose, determine the initial visibility configuration information for each second member. The content of the first message includes at least one of the following: project name, team name, and member name. Specifically, analyze the content of the first message and extract key information such as the project name, team name, or member name. Combined with the multi-level knowledge graph, determine the group chat purpose of the first message, such as whether it concerns project progress, task allocation, or daily communication. Based on the group chat purpose, generate initial visibility configuration information for each second member, ensuring that message delivery meets both security requirements and personalized needs.

[0089] Analysis of the above technical solutions reveals that by executing steps a1-a2, this embodiment provides a clear organizational structure view, facilitating the understanding and management of complex relationship networks. Furthermore, by clearly defining the purpose of the group chat, this embodiment enables the private group chat system to more accurately configure the visibility of the first message, ensuring the effectiveness and relevance of the first message delivery. Based on the group chat purpose and member attributes, personalized visibility configurations are generated, improving the flexibility and adaptability of the first message delivery. A clear group chat purpose helps members better understand the message context, promoting effective communication and collaboration, ultimately making the private group chat system more intelligent, efficient, and secure, meeting complex and ever-changing business needs.

[0090] Step 132: Given that the target private group's group configuration information indicates a need for context-aware and personalized visibility configuration, a multimodal parsing algorithm is used to extract features from the first message. Based on these features, a pre-defined classification algorithm is applied to classify the first message by media type, resulting in the media type of the first message. The media type can be text, image, audio, video, file, business card, location message, email, or chat history. It should be understood that a multimodal parsing algorithm is an algorithm used to process multiple data types, capable of simultaneously parsing message content in various formats such as text, images, audio, and video, and extracting their features. Feature extraction refers to extracting key information from the message that is helpful for classification and evaluation, such as keywords in text, visual features in images, and voiceprint features in audio. The pre-defined classification algorithm is a pre-set machine learning or deep learning algorithm used to classify the first message into a specific media type.

[0091] It should be noted that when the media type of the first message is text, the embodiments of this application can apply natural language processing technology to identify the language structure type of the first message, and generate visibility configuration information of the first message for each second member based on the media type of the first message and the language structure type of the first message.

[0092] Step 133: Based on the media type of the first message, combined with the role and permissions of each second member, apply a security assessment algorithm to evaluate the security of the first message for each second member, and obtain the security assessment result of the first message for each second member. Step 133 enables personalized security assessment, ensuring the security and effectiveness of the first message transmission, reducing false positives and false negatives, and improving user experience and the reliability of the private group chat system.

[0093] For example, as one possible implementation, step 133 involves applying a security assessment algorithm based on the media type of the first message, combined with the role and permissions of each second member, to assess the security of the first message for each second member, and obtaining the security assessment results of the first message for each second member, including:

[0094] Step b1: Obtain the mapping relationship between each second member's role, permissions, and the media type of the first message, as well as the context information of the first message. The context information may refer to the historical messages within the target private group prior to the first message.

[0095] Step b2: Based on the mapping relationship, apply the security assessment algorithm to evaluate the security of the first message for each second member and obtain the initial security score. Then evaluate the security of the context information for each second member and obtain the context security score.

[0096] Step b3: Based on the correlation between context information and the first message, determine the influence of the context security score, and determine the product of the influence of the context security score and the context security score as the adjustment value.

[0097] Step b4: Add the adjusted value and the initial security score value to obtain the security assessment result of the first message for each second member.

[0098] For example, the security assessment results of the first message for each second member are as follows:

[0099] F i =w S ·S i +w C ·(C i ×D i );

[0100] Among them, F i The first message represents the security assessment result for the i-th second member in the target private group, where i starts from 1 and goes up to N, where N is the total number of second members in the target private group. i Let C be the initial security score for the i-th second member in the target private group as defined in the first message. i D represents the contextual security score for the i-th second member in the target private group. i To determine the impact of the contextual security score on the i-th second member in the target private group, C i ×D i Used to indicate adjustment values; w S and w C These are the adjustment factors for the corresponding terms, expressed as w. S and w C With the introduction of this feature, the embodiments of this application can dynamically adjust the importance of the initial security scoring item and the context scoring item, making the embodiments of this application more flexible and adaptable to different application scenarios.

[0101] By executing steps b1-b4, this embodiment combines roles, permissions, and context information to enable the private group chat system to more flexibly adjust security policies and adapt to different communication scenarios. This embodiment ensures that each second member's permissions match the types of messages they can receive, and better understands the context of the current message through historical message context information. This embodiment generates personalized initial and contextual security scores by independently evaluating the first message and contextual information, improving the accuracy and relevance of the evaluation. This embodiment ensures the rationality of adjusted values ​​by evaluating the correlation between contextual information and the first message, achieving more refined security management. The final security evaluation result comprehensively considers the security scores of the first message and contextual information, ensuring the comprehensiveness and accuracy of the evaluation results. This embodiment, through precise security evaluation, ensures that the transmission of the first message meets both security requirements and personalized needs, enhancing user trust and satisfaction, while improving communication efficiency.

[0102] Step 134: Based on the security assessment results of the first message for each second member, adjust the visibility configuration information to obtain the intermediate visibility configuration information of the first message for each second member.

[0103] For example, before the adjustment, a certain initial visibility configuration was visible to all content of member A1 in the marketing team; after the adjustment, the corresponding intermediate visibility configuration is visible to some content of member A1 in the marketing team. As another example, before the adjustment, a certain initial visibility configuration was invisible to member B1 in the development team; after the adjustment, the corresponding intermediate visibility configuration is visible to some content of member B1 in the development team.

[0104] Step 135: Combining the context-aware algorithm and the personalized visibility configuration results, optimize the intermediate visibility configuration information of the first message for each second member to obtain the target visibility configuration information of the first message for each second member.

[0105] As one possible implementation, step 135 involves combining the context-aware algorithm and the personalized visibility configuration results to optimize the intermediate visibility configuration information of the first message for each second member, thereby obtaining the target visibility configuration information of the first message for each second member, including:

[0106] Step c1: Based on the node attributes in the multi-level knowledge graph, a context-aware algorithm is used to determine the context-awareness result of each second member. Node attributes include context features, which include time, geographical location, and activity status. The context-awareness result of the second member indicates whether the second member is in a working time period, in an office area, or in a project discussion phase. It should be understood that a context-aware algorithm is an algorithm that can dynamically adjust the system behavior based on context features (such as time, location, and activity status) to evaluate the impact of the current context on message visibility.

[0107] Step c2: Based on the node attributes in the multi-level knowledge graph, combined with user profiling technology and collaborative filtering algorithm, construct the personalized visibility configuration results of each second member. The node attributes include the second member's role, permissions, and historical messages.

[0108] Specifically, step c2, based on the node attributes in the multi-level knowledge graph, combined with user profiling technology and collaborative filtering algorithms, constructs the personalized visibility configuration results for each second member, including:

[0109] Step c21: Based on the node attributes in the multi-level knowledge graph, construct the user profile information of the second member using user profiling technology. Node attributes in the multi-level knowledge graph include roles, permissions, departments, historical behaviors, etc. The user profile information of the second member is used to reflect the second member's interests, preferences, and behavioral patterns. Through detailed user profiles, the private group chat system can more comprehensively understand the needs and behavioral patterns of each member, providing technical support for subsequent personalized configuration.

[0110] Step c22: Based on the user profile information of the second member, perform collaborative filtering algorithms for both inter-member and inter-message collaborative filtering to obtain the collaborative filtering results for the second member and the filtering results for the first message. During the collaborative filtering process, cosine similarity is used to calculate the similarity between the second members. It should be understood that inter-member collaborative filtering can refer to using user profile information to calculate the similarity between second members (e.g., using cosine similarity) to find members with similar interests and behavioral patterns. Inter-message collaborative filtering can refer to evaluating the relevance between different messages to determine which messages might be more valuable to a specific member.

[0111] Step c23: Integrate the user profile information of the second member, the collaborative filtering results of the second member, and the filtering results of the first message to obtain the personalized visibility configuration results for each second member. It should be understood that integration can be interpreted as information fusion. Through collaborative filtering among members, the system can intelligently recommend that members who may be interested view specific messages, improving the efficiency and accuracy of the first message delivery. Through collaborative filtering between messages, the system can discover potential message relationships, ensuring that relevant information is not missed and enhancing the completeness of the information.

[0112] By executing steps c21-c23, the embodiments of this application can achieve precise visibility configuration, ensuring that the transmission of the first message meets both security requirements and personalization needs, enhancing user trust and satisfaction, and improving communication efficiency.

[0113] Step c3: Integrate the context awareness results and personalized visibility configuration results of each second member to obtain the fusion result. Based on the fusion result, optimize the intermediate visibility configuration information of the first message for each second member to obtain the target visibility configuration information of the first message for each second member.

[0114] The fusion result combines context-aware results and personalized visibility configuration results to form a comprehensive evaluation result, used to ensure the security and relevance of the first message delivery. For example, the fusion result includes the visibility time window and geographical restrictions of the first message for the second member; for instance, text may be displayed for 10 days, and images for 3 hours; it may be visible within office areas and outside office areas.

[0115] By executing steps c1-c3, this embodiment of the application can utilize the contextual features in a multi-level knowledge graph to determine the real-time contextual awareness result of the second member through a context-aware algorithm, judging whether they are in a working time period, office area, or project discussion stage. Then, combining user profiling technology and collaborative filtering algorithms, a personalized visibility configuration result is constructed based on the second member's role, permissions, and historical messages.

[0116] Finally, by integrating the context-aware results and personalized visibility configuration results, the intermediate visibility configuration information of the first message for each second member is optimized to generate the final target visibility configuration information. Through the combination of context-aware algorithms and user profiling technology, the private group chat system can dynamically adapt to the real-time situation of different members, accurately control the visibility of the first message, reduce unnecessary information exposure, and improve communication efficiency and user experience. Simultaneously, personalized configuration based on collaborative filtering algorithms makes the delivery of the first message more targeted, promoting effective team collaboration. In summary, the embodiments of this application not only enhance the flexibility and accuracy of message management but also significantly improve the reliability and user satisfaction of the private group chat system.

[0117] By executing steps 131-135, this embodiment of the application can ensure that the features of different types of messages can be effectively extracted through multimodal parsing, thereby improving the accuracy of classification. Context-aware requirements and personalized visibility configuration requirements enable the system to flexibly adjust strategies according to actual conditions, adapting to complex and ever-changing communication scenarios. Automated feature extraction and classification reduce manual intervention and improve processing speed and efficiency. The private group chat system can handle various types of multimedia messages, ensuring that different forms of information can be effectively managed and controlled.

[0118] In one possible embodiment, step S14, selectively broadcasting the first message to all members or designated members in the target private group based on the target visibility configuration information for each second member in the first message, includes:

[0119] Step 141: Parse the target visibility configuration information of the first message for each second member to obtain the visibility level, time window, and geographical restrictions. The visibility level includes full visibility of all content, partial visibility of all content, and complete invisibility of all content.

[0120] Step 142: Classify the second member according to visibility level, time window and geographical restrictions to obtain multiple groups.

[0121] For example, Group 1 includes 5 second members. All content in the first message is fully visible to these 5 second members for 6 hours, and all 5 second members are in an office area. Group 2 includes 3 second members. Partial content in the first message is visible to these 3 second members for 6 hours, and these 3 second members may be in an office area or a non-office area.

[0122] Step 143: Execute the message broadcasting strategy corresponding to each group. According to the message broadcasting strategy, selectively broadcast the first message to all members or designated members in the target private group.

[0123] For example, for group 1, all the contents of the first message are broadcast to all members of group 1; for group 2, part of the contents of the first message are broadcast to all members of group 2.

[0124] Optionally, group 1 has a higher priority than group 2, so that in the event of high network pressure, the first message is received by the second member in group 1 first.

[0125] By executing steps 141-143, this embodiment of the application can more precisely control the visibility and propagation scope of the first message through detailed parsing of visibility configuration information, ensuring the security and effectiveness of the first message transmission. This embodiment of the application simplifies the subsequent message broadcasting process by classifying second members, improving processing speed and efficiency, while ensuring the personalization and accuracy of the first message transmission. Based on group classification and broadcasting strategies, this embodiment of the application enables the private group chat system to accurately deliver the first message to those who need to know it, reducing interference from irrelevant personnel and improving communication efficiency. Second members can clearly know which content in the first message is visible to them and which content is restricted, enhancing user trust and satisfaction.

[0126] It should be understood that, in order to implement the private group chat method based on keyword interception and message visibility management, embodiments of this application can establish a private group chat platform that supports keyword interception and message visibility management. This private group chat platform, also referred to as a private group chat system based on keyword interception and message visibility management, can also support multimedia types, such as text, images, audio, video, files, business cards, location messages, emails, chat logs, etc. For example, Figure 2 This application provides a schematic diagram of the structure of a private group chat system based on keyword interception and message visibility management, as shown in the embodiments of this application. Figure 2 As shown, the system includes:

[0127] The receiving and identification module 21 is used to receive the first message sent by the first member in the target private group and identify the permissions of the first member.

[0128] The interception module 22 is used to determine the keyword set based on the attributes of the target private group and the permissions of the first member. When the content of the first message includes any keyword in the keyword set, the first message is intercepted, making the first message invisible to all members of the target private group.

[0129] The generation module 23 is used to dynamically generate target visibility configuration information for each second member of the first message, combining context-aware algorithms and personalized visibility configuration results, when the first message does not include all keywords in the keyword set. The second members are other members in the target private group besides the first member.

[0130] Broadcast module 24 calls and selectively broadcasts the first message to all members or specified members in the target private group based on the target visibility configuration information for each second member in the first message.

[0131] Figure 2 The aforementioned private group chat system based on keyword interception and message visibility management can execute... Figure 1The implementation principle and technical effects of the private group chat method based on keyword interception and message visibility management described in the illustrated embodiment will not be repeated here. The specific methods by which each module and unit of the private group chat system based on keyword interception and message visibility management in the above embodiments are described in detail in the embodiments related to this method, and will not be elaborated upon here.

[0132] In one possible design, Figure 2 The private group chat system based on keyword interception and message visibility management in the illustrated embodiment can be implemented as a computing device, such as... Figure 3 As shown, the computing device may include a storage component 31 and a processing component 32.

[0133] The storage component 31 stores one or more computer instructions, wherein the one or more computer instructions are invoked and executed by the processing component 32.

[0134] The processing component 32 is configured to: receive a first message sent by a first member in a target private group and identify the permissions of the first member; determine a set of keywords based on the attributes of the target private group and the permissions of the first member; intercept the first message when the content of the first message includes any keyword in the set of keywords, making the first message invisible to all members in the target private group; dynamically generate target visibility configuration information for each second member of the first message, where the second members are members in the target private group other than the first member, by combining the context-aware algorithm and the personalized visibility configuration results, when the first message does not include all keywords in the set of keywords; and selectively broadcast the first message to all members or designated members in the target private group based on the target visibility configuration information for each second member of the first message.

[0135] The processing component 32 may include one or more processors to execute computer instructions to complete all or part of the steps in the above method.

[0136] Storage component 31 is configured to store various types of data to support operations on the terminal. The storage component can be implemented by any type of volatile or non-volatile storage device or a combination thereof.

[0137] This application also provides a computer storage medium storing a computer program, which, when executed by a computer, can perform the above-described functions. Figure 1 The illustrated embodiment is a private group chat method based on keyword interception and message visibility management.

[0138] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the system and unit described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0139] The system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0140] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method for private group chat based on keyword interception and message visibility management, characterized in that, include: Receive the first message sent by the first member in the target private group and identify the first member's permissions; Based on the attributes of the target private group and the permissions of the first member, a set of keywords is determined. When the content of the first message includes any keyword from the set of keywords, the first message is intercepted, making the first message invisible to all members of the target private group. If the first message does not include all the keywords in the keyword set, the target visibility configuration information of the first message for each second member is dynamically generated by combining the context-aware algorithm and the personalized visibility configuration result. The second member is the other member in the target private group other than the first member. Based on the target visibility configuration information of the first message for each second member, the first message is selectively broadcast to all members or designated members in the target private group; The dynamic generation of target visibility configuration information for each second member of the first message, combining context-aware algorithms and personalized visibility configuration results, includes: Obtain the initial visibility configuration information of the first message for each second member; When the group configuration information of the target private group indicates that there are context-aware requirements and personalized visibility configuration requirements, the features of the first message are extracted using a multimodal parsing algorithm. Based on the features of the first message, a preset classification algorithm is applied to classify the media type of the first message to obtain the media type of the first message. The media type is text, image, audio, video, file, business card, location message, email or chat history. Based on the media type of the first message, combined with the role and permissions of each second member, a security assessment algorithm is applied to assess the security of the first message for each second member, and the security assessment results of the first message for each second member are obtained. Based on the security assessment results of the first message for each second member, the visibility configuration information is adjusted to obtain the intermediate visibility configuration information of the first message for each second member. By combining the context-aware algorithm and the personalized visibility configuration results, the intermediate visibility configuration information of the first message for each second member is optimized to obtain the target visibility configuration information of the first message for each second member. Based on the media type of the first message, combined with the role and permissions of each second member, a security assessment algorithm is applied to evaluate the security of the first message for each second member, resulting in a security assessment result for the first message for each second member, including: Obtain the mapping relationship between the role and permissions of each second member and the media type of the first message, as well as the context information of the first message; Based on the mapping relationship, a security assessment algorithm is applied to evaluate the security of the first message for each second member to obtain an initial security score. The security of the context information for each second member is then evaluated to obtain a context security score. Based on the correlation between context information and the first message, the influence of the context security score is determined, and the product of the influence of the context security score and the context security score is determined as the adjustment value. The adjusted value and the initial security score are added together to obtain the security assessment result of the first message for each second member; The security assessment results for each second member from First Message are as follows: ； in, The first message is targeted at the private group of the target. The security assessment result of the second member, i starts from 1 and goes up to N, where N is the total number of second members in the target private group. The first message is targeted at the private group of the target. The initial security score of the second member, For contextual information targeting the first private group The context security score of the second member, To target the private group of the first For the second member, the impact of the contextual security score value Used to indicate adjustment values; and These are the adjustment coefficients for the corresponding items.

2. The method according to claim 1, characterized in that, The step of obtaining the initial visibility configuration information of the first message for each second member includes: A multi-level knowledge graph corresponding to the target private group is constructed. The multi-level knowledge graph includes multi-level nodes and edges. The first-level nodes in the multi-level nodes are used to represent members in the target private group, the second-level nodes are used to represent the teams to which the members belong, and the third-level nodes are used to represent the projects that the teams participate in. Different edges are used to reflect the relationships between different nodes, and each node has corresponding node attributes. Based on the content of the first message and the multi-level knowledge graph, the purpose of the group chat in the first message is determined. Based on the purpose of the group chat in the first message, the initial visibility configuration information of the first message for each second member is determined. The content of the first message includes at least one of the following: project name, team name, and member name.

3. The method according to claim 1, characterized in that, The process of combining context-aware algorithms and personalized visibility configuration results to optimize the intermediate visibility configuration information of the first message for each second member yields the target visibility configuration information of the first message for each second member, including: Based on the node attributes in the multi-level knowledge graph, the context perception result of each second member is determined using a context perception algorithm. The node attributes include context features, which include time, geographical location, and activity status. The context perception result of the second member is used to indicate whether the second member is in a working period, in an office area, or in a project discussion phase. Based on the node attributes in the multi-level knowledge graph, combined with user profiling technology and collaborative filtering algorithm, the personalized visibility configuration results of each second member are constructed. The node attributes include the second member's role, permissions, and historical messages. The context awareness results and personalized visibility configuration results of each second member are fused to obtain a fusion result. Based on the fusion result, the intermediate visibility configuration information of the first message for each second member is optimized to obtain the target visibility configuration information of the first message for each second member.

4. The method according to claim 3, characterized in that, The personalized visibility configuration results for each second member are constructed based on node attributes in a multi-level knowledge graph, combined with user profiling technology and collaborative filtering algorithms, including: Based on the node attributes in a multi-level knowledge graph, user profile information of the second member is constructed using user profiling technology. Based on the user profile information of the second member, collaborative filtering algorithms are used to perform collaborative filtering between members and between messages to obtain the collaborative filtering results of the second member and the filtering results of the first message; during the collaborative filtering process, cosine similarity is used to calculate the similarity between the second members. By integrating the user profile information of the second member, the collaborative filtering results of the second member, and the filtering results of the first message, the personalized visibility configuration results of each second member are obtained.

5. The method according to claim 1, characterized in that, The step of selectively broadcasting the first message to all members or designated members in the target private group based on the target visibility configuration information for each second member in the first message includes: The first message is parsed to obtain the target visibility configuration information for each second member, resulting in the visibility level, time window, and geographical restrictions. The second member is classified into multiple groups based on visibility level, time window, and geographic restrictions; Execute the message broadcasting policy corresponding to each group, and selectively broadcast the first message to all members or designated members in the target private group according to the message broadcasting policy.

6. A private group chat system based on keyword interception and message visibility management, characterized in that, include: The receiving and identification module is used to receive the first message sent by the first member in the target private group and identify the permissions of the first member; The interception module is used to determine a set of keywords based on the attributes of the target private group and the permissions of the first member. When the content of the first message includes any keyword in the set of keywords, the first message is intercepted, so that the first message is not visible to all members of the target private group. The generation module is used to dynamically generate target visibility configuration information for each second member of the first message, combining context-aware algorithms and personalized visibility configuration results, when the first message does not include all keywords in the keyword set. The second members are other members in the target private group other than the first member. The broadcast module invokes and selectively broadcasts the first message to all members or designated members in the target private group, based on the target visibility configuration information for each second member according to the first message. The dynamic generation of target visibility configuration information for each second member of the first message, combining context-aware algorithms and personalized visibility configuration results, includes: Obtain the initial visibility configuration information of the first message for each second member; When the group configuration information of the target private group indicates that there are context-aware requirements and personalized visibility configuration requirements, the features of the first message are extracted using a multimodal parsing algorithm. Based on the features of the first message, a preset classification algorithm is applied to classify the media type of the first message to obtain the media type of the first message. The media type is text, image, audio, video, file, business card, location message, email or chat history. Based on the media type of the first message, combined with the role and permissions of each second member, a security assessment algorithm is applied to assess the security of the first message for each second member, and the security assessment results of the first message for each second member are obtained. Based on the security assessment results of the first message for each second member, the visibility configuration information is adjusted to obtain the intermediate visibility configuration information of the first message for each second member. By combining the context-aware algorithm and the personalized visibility configuration results, the intermediate visibility configuration information of the first message for each second member is optimized to obtain the target visibility configuration information of the first message for each second member. Based on the media type of the first message, combined with the role and permissions of each second member, a security assessment algorithm is applied to evaluate the security of the first message for each second member, resulting in a security assessment result for the first message for each second member, including: Obtain the mapping relationship between the role and permissions of each second member and the media type of the first message, as well as the context information of the first message; Based on the mapping relationship, a security assessment algorithm is applied to evaluate the security of the first message for each second member to obtain an initial security score. The security of the context information for each second member is then evaluated to obtain a context security score. Based on the correlation between context information and the first message, the influence of the context security score is determined, and the product of the influence of the context security score and the context security score is determined as the adjustment value. The adjusted value and the initial security score are added together to obtain the security assessment result of the first message for each second member; The security assessment results for each second member from First Message are as follows: ； in, The first message is targeted at the private group of the target. The security assessment result of the second member, i starts from 1 and goes up to N, where N is the total number of second members in the target private group. The first message is targeted at the private group of the target. The initial security score of the second member, For contextual information targeting the first private group The context security score of the second member, To target the private group of the first For the second member, the impact of the contextual security score value Used to indicate adjustment values; and These are the adjustment coefficients for the corresponding items.

7. A computing device, characterized in that, It includes a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are invoked and executed by the processing component to implement a private group chat method based on keyword interception and message visibility management as described in any one of claims 1-5.

8. A computer storage medium, characterized in that, The system stores a computer program, which, when executed by a computer, implements a private group chat method based on keyword interception and message visibility management as described in any one of claims 1-5.

Images