A conference information generation method and device, a storage medium and an electronic device

Abstract

The application discloses a conference information generation method and device, a storage medium and an electronic device, relates to the technical field of data processing, and comprises the following steps: acquiring audio and video data in a target conference; determining target topic information of the target conference based on the audio and video data; determining historical topic information of a historical conference associated with the target topic information, and constructing a conference decision graph corresponding to the target conference according to the target topic information and the historical topic information, wherein the conference decision graph is used to identify a decision association relationship between the historical conference and the target conference; and generating decision conflict evaluation information of the target conference according to the conference decision graph. Compared with the prior art, the application can trace historical conference information for associated decision making based on the conference decision graph in the case that there is conference information conflict between the target conference and the historical conference, so that the possible conflict situation of the conference can be accurately identified, and the user experience is improved.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and in particular to a method, apparatus, storage medium and electronic device for generating conference information. Background Technology

[0002] Metaverse Conference is an immersive conference system built in a virtual three-dimensional digital space: participants can use digital humans to enter the virtual venue, move freely, have face-to-face conversations, participate in group roundtables, give presentations, and use interactive whiteboards. Combined with voice / real-time video projection, virtual props, and spatial sound effects, it can replicate the real experience of offline conferences, forums, and seminars.

[0003] Currently, the meeting information of the Metaverse Meeting is stored separately. When there is a conflict between the meeting information of the current meeting and the meeting information of the past meeting, the Metaverse Meeting cannot trace the information of the past meeting to make related decisions, which leads to the inability to accurately identify decision conflicts and affects the user experience. Summary of the Invention

[0004] In view of this, this application provides a meeting information generation method, apparatus, storage medium, and electronic device. The main purpose is to improve the technical problem that in the current metaverse meeting, meeting information is stored separately. When there is a conflict between the current meeting and the meeting information of the previous meeting, the metaverse meeting cannot trace the historical meeting information to make related decisions, which leads to the inability to accurately identify decision conflicts and affects the user experience.

[0005] Firstly, this application provides a method for generating meeting information, including: Acquire audio and video data from the target meeting; Based on the audio and video data, the target topic information of the target meeting is determined; Determine the historical topic information of the historical meetings associated with the target topic information, and construct a meeting decision graph corresponding to the target meeting based on the target topic information and the historical topic information. The meeting decision graph is used to identify the decision association relationship between the historical meetings and the target meeting. Based on the meeting decision graph, generate decision conflict evaluation information for the target meeting.

[0006] Secondly, this application provides a meeting information generation device, comprising: The acquisition module is configured to acquire audio and video data from the target meeting. The determination module is configured to determine the target topic information of the target meeting based on the audio and video data; The construction module is configured to determine the historical topic information of the historical meetings associated with the target topic information, and construct a meeting decision graph corresponding to the target meeting based on the target topic information and the historical topic information. The meeting decision graph is used to identify the decision association relationship between the historical meetings and the target meeting. The generation module is configured to generate decision conflict evaluation information for the target meeting based on the meeting decision graph.

[0007] Thirdly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the meeting information generation method described in the first aspect.

[0008] Fourthly, this application provides an electronic device, including a storage medium, a processor, and a computer program stored on the storage medium and executable on the processor, wherein the processor executes the computer program to implement the conference information generation method described in the first aspect.

[0009] Fifthly, this application provides a computer program product, which includes a computer program that, when executed by a processor, implements the meeting information generation method described in the first aspect.

[0010] By means of the above technical solutions, this application provides a method, apparatus, storage medium, and electronic device for generating meeting information, including: acquiring audio and video data from a target meeting; determining target topic information of the target meeting based on the audio and video data; determining historical topic information of historical meetings associated with the target topic information, and constructing a meeting decision graph corresponding to the target meeting based on the target topic information and the historical topic information, wherein the meeting decision graph is used to identify the decision-making relationship between historical meetings and the target meeting; and generating decision conflict evaluation information of the target meeting based on the meeting decision graph. Compared with existing technologies, this application obtains audio and video data from the target meeting; based on the audio and video data, it determines the target topic information of the target meeting, thus accurately identifying the meeting topic; further, by determining the historical topic information of historical meetings associated with the target topic information, and constructing a meeting decision graph corresponding to the target meeting based on the target topic information and the historical topic information, the meeting decision graph is used to identify the decision-making relationship between the historical meetings and the target meeting, and can associate the meeting topics and meeting decisions of the historical meetings and the target meeting to obtain the meeting decision graph. Based on the meeting decision graph, decision conflict evaluation information of the target meeting is generated. This allows the application to trace historical meeting information and make associated decisions based on the meeting decision graph when there is a conflict between the target meeting and historical meetings, thereby accurately identifying potential conflicts in the meeting and improving the user experience. Attached Figure Description

[0011] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0012] 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, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 A flowchart illustrating a meeting information generation method provided in an embodiment of this application is shown; Figure 2 This paper illustrates a schematic diagram of the system architecture of a conference information generation system provided in an embodiment of this application. Figure 3 This illustration shows a data flow diagram of a meeting information generation system provided in an embodiment of this application; Figure 4 A flowchart illustrating an example of audio and video processing provided in an embodiment of this application is shown; Figure 5 A flowchart illustrating a meeting information generation method provided in an embodiment of this application is shown; Figure 6 A flowchart illustrating an example of generating user minutes information provided in an embodiment of this application is shown. Figure 7 A flowchart illustrating an example of generating a user knowledge model according to an embodiment of this application is shown; Figure 8 A flowchart illustrating an example of generating a user log report according to an embodiment of this application is shown; Figure 9 The illustration shows a flowchart of an example of a map binding interactive hotspot provided in an embodiment of this application; Figure 10 This illustration shows a schematic diagram of an example of user interaction with a knowledge base provided in an embodiment of this application; Figure 11 A flowchart illustrating a multimodal interaction example provided in an embodiment of this application is shown; Figure 12 This paper shows a schematic diagram of the structure of a conference information generation device provided in an embodiment of this application; Figure 13 A schematic diagram of the structure of an electronic device provided in an embodiment of this application is shown. Detailed Implementation

[0014] The embodiments of this application will now be described in more detail with reference to the accompanying drawings. It should be noted that, unless otherwise specified, the embodiments and features described herein can be combined with each other.

[0015] To address the technical problem of existing Metaverse Meeting technologies where meeting information is stored separately, leading to conflicts between current and historical meeting information, Metaverse Meeting cannot trace historical meeting information for correlation decision-making, resulting in inaccurate identification of decision conflicts and impacting user experience, this embodiment provides a meeting information generation method, such as... Figure 1 As shown, the method includes: Step 101: Obtain audio and video data from the target meeting.

[0016] In this application embodiment, the target meeting can be the Metaverse Meeting; the Metaverse Meeting is an immersive meeting system built in a virtual three-dimensional digital space: participants can use digital humans to enter the virtual venue, and can move freely, talk face-to-face, participate in group roundtables, give speeches on stage, and use interactive whiteboards. Combined with voice / real-time video projection, virtual props, and spatial sound effects, it can replicate the real experience of offline conferences, forums, and seminars.

[0017] In some examples, embodiments of this application can capture multimodal data in the metaverse conference in real time, namely the audio and video data in the embodiments of this application, including multi-microphone audio streams (48kHz sampling), multi-view camera video streams (1080p, 30fps), and 6-DOF motion sensor data of virtual reality devices, etc.

[0018] Step 102: Based on the audio and video data, determine the target topic information for the target meeting.

[0019] In this embodiment of the application, the meeting topic information can be the main discussion topics of the target meeting; for example, environmental issues, development issues, project research issues, etc., and so on, without further examples here.

[0020] For example, such as Figure 2 The diagram shows the system architecture of the meeting information generation system. Data is input into three execution paths: the first generates a personal knowledge graph; the second identifies the decision content and constructs a decision tree; and the third identifies decision conflicts in real time and handles differentiated role-playing interactions and conflict stalemate resolution. Figure 3 The image shown can be... Figure 2 A schematic diagram illustrating an example of data flow within the architecture.

[0021] For example, embodiments of this application can use the conference audio and video stream parsing module in the system to obtain and process audio and video data; such as Figure 4As shown, this can be a flowchart of audio and video processing. After acquiring audio and video data, voiceprint separation technology can be used to distinguish the voice segments of different speakers. At the same time, gesture recognition algorithm is used to parse the action semantics in the video stream, and finally, a structured data packet including metadata is output.

[0022] For example, after obtaining a structured data packet including metadata, the structured data packet including metadata can be input into... Figure 2 In the digital human avatar engine module, which serves as the central hub for user personification and data routing, the module receives structured audio and video data packets and first completes user identity binding through a voiceprint matching algorithm (activated when matching degree > 0.92). It then generates a user behavior feature vector (e.g., [speaking frequency: 0.32, interruption count: 2]) and drives synchronized digital human body movements. Most importantly, it performs data routing and distribution: directing individual speaking data to a dedicated minutes generator, while simultaneously sending global topic data (topic decisions, topic summaries, etc.), i.e., the target topic information in this embodiment, to... Figure 2 The decision graph builder module forms a dual-channel data processing mechanism.

[0023] Step 103: Determine the historical topic information of the historical meetings associated with the target topic information, and construct the meeting decision map corresponding to the target meeting based on the target topic information and the historical topic information.

[0024] Among them, the meeting decision graph is used to identify the decision-making relationships between historical meetings and target meetings.

[0025] In this embodiment of the application, the historical meeting can be a historical meeting associated with the target topic information; for example, if there are multiple historical meetings in the historical meeting library, they can include meetings with various topic information, such as environmental topics, development topics, project research topics, etc. If the target meeting is a project research topic, the historical meeting associated with the project research topic (i.e., the target topic information in this embodiment of the application) can be determined from the multiple historical meetings in the historical meeting library.

[0026] In some examples, historical issue information can be global issue data of historical issues, such as issue decisions and issue summaries. Based on the target issue information and historical issue information, a meeting decision graph corresponding to the target meeting can be constructed. Thus, the decision conflicts between historical meetings and the target meeting can be identified through the meeting decision graph corresponding to the target meeting.

[0027] Step 104: Generate decision conflict evaluation information for the target meeting based on the meeting decision map.

[0028] In this embodiment, based on the correlation between the global issue data of historical issues and the target issue information in the meeting decision graph, it is possible to identify whether the target meeting has a decision conflict with the global issue data of historical issues, thereby generating decision conflict evaluation information for the target meeting.

[0029] Compared with existing technologies, this embodiment acquires audio and video data from the target meeting; based on the audio and video data, it determines the target topic information of the target meeting, thus accurately identifying the meeting topic; further, it determines the historical topic information of historical meetings associated with the target topic information, and constructs a meeting decision graph corresponding to the target meeting based on the target topic information and the historical topic information. The meeting decision graph is used to identify the decision-making relationship between the historical meeting and the target meeting, and can associate the meeting topics and meeting decisions of the historical meeting and the target meeting to obtain the meeting decision graph. Based on the meeting decision graph, decision conflict evaluation information of the target meeting is generated. This embodiment enables the tracing of historical meeting information to make associated decisions based on the meeting decision graph when there is a conflict between the target meeting and the historical meeting, thereby accurately identifying possible conflicts in the meeting and improving the user experience.

[0030] As a refinement and extension of the above embodiments, when performing the task of "determining the historical topic information of historical meetings associated with the target topic information, and constructing a meeting decision graph corresponding to the target meeting based on the target topic information and the historical topic information," the following methods can be used, but are not limited to them: Figure 2 As shown, the method includes: Step 201: Based on the semantic information of the target topic information, determine the historical meetings associated with the target topic information.

[0031] In this embodiment of the application, semantic recognition can be performed on the target topic, thereby determining the historical meetings related to the semantics of the target topic based on the identified semantics of the target topic, that is, the historical meetings associated with the target topic information in this embodiment of the application.

[0032] For example, after identifying the semantics of the target issue, embodiments of this application may also use... Figure 2 The dedicated minutes generator module generates personalized minutes information for each participating user in the meeting, for example, such as... Figure 6The diagram illustrates an example of generating user minutes information. The personalized minutes generator module focuses on personalized knowledge extraction. After receiving the user-specific speech data stream distributed by the digital human engine, it uses technologies such as speech-to-text (ASR), natural language processing (NLP) entity extraction, commitment detection, and sentiment analysis to transform the raw speech into structured task items. For example, "Submit the report before this Friday" can be transformed into {task: "Submit the report", owner: "U123", deadline: "This Friday"}, ultimately generating a personalized log object containing speaking percentage statistics and a personal to-do list, which is stored in the personal knowledge base, resulting in personalized minutes information for each participating user in the meeting.

[0033] Furthermore, after generating personalized minutes for each participant in the meeting, the personalized minutes for each participant can be input... Figure 2 The personal knowledge base module serves as a private knowledge storage center, such as... Figure 7 The diagram illustrates an example of generating a user knowledge model. It receives structured logs from a dedicated minutes generator and performs encrypted storage. When a user queries, a permission filtering mechanism returns only relevant decision graph nodes. Its core operation is knowledge fusion: when new meeting logs are added, the old and new task lists are automatically merged, and the decision preference profile is updated based on the user's attitude, forming a dynamically evolving personal knowledge model.

[0034] As an optional approach, it can be used after a dynamically evolving personal knowledge model is formed. Figure 2 The customized meeting log module generates individual reports for participating users; the customized meeting log module serves as a personalized report generation engine, such as... Figure 8 The diagram illustrates an example of generating user log reports. This module integrates basic data from a personal knowledge base with insights from decision trees. It generates multi-format reports via a dynamic template engine, including contribution statistics, to-do lists, and decision conflict alerts (such as "Your reasons for opposing the budget conflict with the decision"). A special intelligent conflict detection function is integrated, automatically highlighting contradictions between the current decision and historical logic.

[0035] Step 202: Determine the historical topics of the historical conferences from the historical conference knowledge base.

[0036] In this embodiment of the application, the historical meeting knowledge base can be a knowledge base that stores global meeting data of historical meetings; through the historical meeting knowledge base, the historical topic information of the historical meeting associated with the target meeting can be determined from the meeting topic information of the meeting including a variety of topic information.

[0037] Step 203: Incrementally update the historical topic information based on the target topic information to construct the meeting decision map corresponding to the target meeting.

[0038] Optionally, when performing the "incremental update of historical topic information based on target topic information to construct the meeting decision graph corresponding to the target meeting", the following methods can be used, but are not limited to: determining the historical meeting decision graph corresponding to the historical topic information, and the target decision information corresponding to the target meeting, wherein the historical meeting decision graph includes the historical decision information of the historical meetings; in the historical meeting decision graph, incrementally updating the historical decision information based on the target decision information to construct the meeting decision graph corresponding to the target meeting.

[0039] For example, in embodiments of this application, target issue information and historical issue information can be input... Figure 2 The decision graph builder module, acting as a cross-temporal decision association engine, receives global issue data and retrieves historical knowledge bases. It calculates the semantic similarity between the current issue and historical nodes using a spatiotemporal association algorithm, and then overlays the time decay weight calculated using Formula 1. The construction process employs an incremental update mechanism. For example, it establishes an "opposition reason inheritance" relationship edge between the "a1 time budget" node and the "a2 time procurement" node, outputting an incremental update package for the decision graph. Since time a2 is earlier than time a1, this provides a relational data foundation for visualization. Formula 1 is shown below: W=1 / (1+log(days_passed)) (Formula 1) In Formula 1, W represents the time decay weight, and days_passed represents the decay time.

[0040] For example, after constructing the meeting decision graph, the constructed meeting decision graph can be input... Figure 2 The cross-meeting decision tree module in the middle; such as Figure 9 The diagram illustrates the process of binding interactive hotspots to a meeting decision graph. This module enables visualized knowledge graph management, converting raw relational data output by the decision graph builder into a 3D visualization model. It determines node positions using a spatial layout algorithm, generates curved relational paths, and binds interactive hotspots to each node. When a user is detected staring at a specific node, it automatically triggers highlighting and playback of associated historical segments, transforming data relationships into an interactive experience.

[0041] As an optional approach, in this embodiment of the application, after automatically triggering highlighting and playback of associated historical segments, the user can enter an interactive decision graph module; this 3D interactive decision tracing interface receives the visualized data model of the decision tree and converts the node / edge data into 3D objects through a real-time rendering engine. For example... Figure 10The diagram illustrates an example of user interaction with a knowledge base, establishing an interactive feedback loop: when a user clicks a node, the knowledge base is queried to retrieve historical segments, while the user's annotations on the video generate new knowledge nodes that are then returned to storage. This achieves a shift from passive viewing to active knowledge creation.

[0042] It should be noted that the implementation of this application uses an incremental update mechanism to establish a relationship network between decision nodes in the decision graph builder. The core operation is to establish an edge for the inheritance of objections between the current meeting node and historical nodes, forming a graph update package with a timestamp. The interactive decision graph module transforms data into a three-dimensional node model through a spatial layout algorithm, binding an interactive hotspot to each node. When a user gazes at a specific node, it automatically triggers a highlight and plays related historical meeting clips. The personal knowledge base synchronously integrates newly added to-do items and user attitude profiles. In the three-dimensional metaverse interface, eye-tracking technology realizes millisecond-level gaze detection mapping instructions, forming a natural and intuitive operation paradigm where gazing at the digital human arm triggers the timeline and looking at topic tags plays historical clips.

[0043] Optionally, when performing the "Generate decision conflict evaluation information for the target meeting based on the meeting decision graph" step, the following methods may be used, but are not limited to: based on the meeting decision graph, identify the first conflict information between historical meetings and the target meeting in the semantic conflict dimension, the second conflict information in the temporal conflict dimension, the third conflict information in the object conflict dimension, and the fourth conflict information in the scheme conflict dimension; generate decision conflict evaluation information for the target meeting based on the first, second, third, and fourth conflict information; and, if the decision conflict evaluation information meets the conflict conditions corresponding to the target meeting, display the meeting conflict content between the target meeting and the historical meetings in the target meeting.

[0044] In this embodiment of the application, the meeting decision graph can also be used. Figure 2 The decision conflict identifier module in the application receives data such as current issue decision data and issue topic, i.e., target issue information in this embodiment. It uses a decision conflict algorithm to identify whether the meeting decision conflicts with similar meeting decisions in the past. The algorithm may include, but is not limited to, a semantic comparison model to calculate semantic similarity, and a conflict dimension evaluation model to analyze conflict information in semantic conflict dimension, time conflict dimension, object conflict dimension and scheme conflict dimension, so as to obtain the first conflict information, second conflict information, third conflict information and fourth conflict information in this embodiment.

[0045] For example, the decision conflict identification algorithm mainly includes a semantic comparison model and a conflict dimension evaluation model. The semantic comparison model is used to evaluate the similarity between the current decision and the historical decision discourse. Let the semantic vector of the current meeting decision be Vcurr and the semantic vector of the historical decision be Vhist. Then the cosine similarity between the two can be expressed by Formula 2, which is shown below: (Formula 2) In Formula 2, Vcurr·Vhist represents the vector dot product. These represent the magnitudes of the vectors.

[0046] In some examples, the conflict dimension assessment model is used to evaluate the conflict between the current decision and historical decisions. The defined conflict assessment function, Conflictscore, can comprehensively consider semantic conflict, temporal conflict, object conflict, and solution conflict dimensions. The semantic conflict dimension can be based on semantic similarity (Sim), where... ( When the semantic similarity threshold is reached, a semantic conflict assessment is triggered.

[0047] For example, the time conflict dimension can be set as Tcurr for the current decision time and Thist for the historical decision time. The time conflict coefficient Ctime is defined by Formula 3, which is shown below: (Formula 3) For example, the object conflict dimension can be set as Ocurr for the current decision pair and Ohist for the historical decision objects. The object conflict coefficient Cobj is defined by Formula 4, which is shown below: (Formula 4) For example, the conflict dimension can be defined as Scurr for the current decision and Shist for the historical decisions. The conflict coefficient Csol is defined by Formula 5, which is shown below: (Formula 5) Based on the two models mentioned above, a comprehensive conflict score calculation formula can be designed for conflict calculation. That is, based on the first conflict information, the second conflict information, the third conflict information, and the fourth conflict information, the decision conflict evaluation information of the target meeting is generated. The comprehensive conflict score calculation formula is Formula Six, which is shown below:

[0048] (Formula 6) In Formula Six, Here are the weight coefficients for each dimension, where... Need to meet Furthermore, weight values ​​are set according to actual needs.

[0049] After calculating the decision conflict score using Formula Six above, it can be determined whether the threshold of the current meeting scenario has been reached. That is, it is necessary to determine whether the decision conflict evaluation information in this embodiment meets the conflict conditions corresponding to the target meeting. If the threshold is reached, it means that the current decision conflicts with the content of a decision in a previous meeting, and the content of the current decision conflict needs to be displayed in the metaverse meeting space. For example, when the comprehensive conflict score reaches the threshold of the current meeting scenario, that is, when the decision conflict evaluation information in this embodiment meets the conflict conditions corresponding to the target meeting, the data of this decision, the topic and summary, historical conflict scenarios, meeting role identifiers, etc., will be transmitted to... Figure 2 The decision conflict display module in the program.

[0050] Optionally, when performing the action of "displaying the conflicting content between the target meeting and historical meetings in the target meeting", the following methods may be used, but are not limited to: determining the digital human image corresponding to at least one participant based on the user type information of at least one participant in the target meeting; determining the content dashboard corresponding to at least one participant based on the at least one digital human image; and displaying the conflicting content to the participant in at least one content dashboard.

[0051] In this embodiment of the application, if the decision conflict identifyer detects a conflict in meeting decisions, it will initiate the following... Figure 2 The Decision Conflict Display Module includes features such as role recognition and Kanban binding, optimized Kanban display for metaverse meeting scenarios, and the ability to manage stalemates in conflict discussions. It can also transmit the data and interactive logic to be displayed to the 3D metaverse interface for execution.

[0052] In some examples, the role recognition and dashboard binding function receives data transmitted by the decision conflict recognizer (current decision data, topic and summary, historical conflict scenarios, meeting role identifiers, etc.), namely the historical topic information and target topic information in this application embodiment. It uses digital human identity tags to define and bind VR device IDs, digital human clothing badges to identify and detect armbands or legal code floating icons, and dynamically and manually assigns roles to achieve role matching. That is, based on the user type information of at least one participant in the target meeting, it determines the digital human image corresponding to at least one participant. Different roles display key content and interaction methods on the dashboards differently. After determining the role, the system calls the role-specific dashboard library to display a 3D strategic dashboard, API compatibility conflict points, or legal clause warning boxes. That is, based on at least one digital human image, it determines the content dashboard corresponding to at least one participant, thereby displaying the meeting conflict content in at least one content dashboard.

[0053] As an alternative approach, embodiments of this application can define the role of the attendee through a digital human identity tag, such as binding the VR device ID worn by the attendee to the corresponding role; for example, identification can be performed through digital human clothing / badges, real-time detection of clothing features, identification of badge / accessory metadata to bind the role, and setting specific clothing or badge accessories for special roles, such as identifying executives by wearing virtual corporate badges (such as gold armbands), legal personnel by wearing standard suits with virtual legal code floating icons, and engineers by wearing tech T-shirts and AR glasses models. In addition to the above examples, there can be more role settings.

[0054] As an optional approach, this application embodiment can also dynamically and manually assign roles. For example, before the meeting starts, the host can assign roles (such as "This meeting: Zhang San is the technical reviewer, and Li Si is the risk controller"), and the system will automatically adjust the dashboard permissions according to the roles (such as "The risk controller can view all historical risk data").

[0055] It should be noted that after determining the roles of the participants according to the above method, the embodiments of this application can select a specific Kanban library from the Kanban library according to the role and display it in combination with conflicting content. The following are examples of Kanban boards for different roles, and it also supports adding different roles and corresponding Kanban boards.

[0056] As an optional approach, senior management's core needs focus on strategic conflict assessment. Key aspects include the impact of conflict on strategic goals such as revenue and market share, the success rate of similar historical decisions (e.g., two out of three similar decisions resulted in project delays), and the historical voting tendencies of key opponents (e.g., senior management's conservative stance on 70% of budget decisions). The presentation uses a floating 3D strategic dashboard to dynamically display core performance metrics such as return on investment and risk index, along with a scrollable timeline to compare the results of similar decisions over the years. Interactive optimizations support voice shortcuts (e.g., "Show most relevant historical cases") for automatic projection filtering, as well as pinch-to-zoom gestures to expand detailed data and wave gestures to close non-critical information.

[0057] As an optional approach, technical experts (architects / engineers) can focus more on the technical details of the conflicts, highlighting conflicts in technical parameters such as API compatibility and performance overhead, outlining subsequent issues caused by historical technical decisions (e.g., adopting Solution A resulted in three online failures), and providing links to relevant technical documents such as test reports. The presentation uses interactive 3D code / architecture diagrams to highlight conflict points (e.g., flashing red interface modules), and a floating sidebar document library comparing the advantages and disadvantages of different technical solutions. Interactive optimizations support a virtual whiteboard annotation function, allowing users to select code segments with gestures and add comments (e.g., "A performance bottleneck exists here"), and also enables automatic data location via voice search.

[0058] As an optional approach, legal or compliance personnel focus on legal risk management, highlighting key aspects such as clause conflicts (e.g., contract breach risks), citing historical litigation / arbitration cases (e.g., lawsuits arising from similar clauses), and providing compliance recommendations (e.g., adding user data encryption clauses). The presentation uses red warning boxes paired with holographic projections of legal provisions, automatically highlighting conflicting clauses, and supplemented by a timeline risk graph showing the timeline and outcomes of similar legal disputes. Interactive optimizations support AI voice interpretation of relevant legal explanations after clicking on a clause, and the automatic generation of compliance amendment drafts based on gesture selection of conflicting paragraphs.

[0059] As an optional approach, the Product Manager (PM) focuses on user value, with key content including user impact analysis (e.g., canceling a feature would cause a 5% drop in Net Promoter Score), competitor comparisons (e.g., both competitors A and B retain this feature), and reference to historical user feedback (e.g., research shows 70% of users need this feature). The presentation uses a 3D projection of user profiles to visualize the age and geographic distribution of the affected group, and displays A / B test results comparison data in a floating display. Interaction optimizations support gesture-based "User Priority Sorting" for quick real-time voting, and voice commands to "View Related Requirements" to display related items in the requirement pool.

[0060] As an optional approach, financial / financial analysts focus on conflicting economic indicators, with key content including cost conflicts such as costs exceeding budget by 30%, analyzing historical financial decision returns (e.g., an average investment return cycle of 18 months), and assessing cash flow impact (e.g., approval leading to Q3 cash reserves falling below the safety line). The presentation uses a dynamic financial model with 3D bar charts comparing costs and benefits, and a risk heatmap highlighting financial risk areas in red. Interactive optimizations include hypothetical analysis tools; voice input such as "If we cut the budget by 20%", real-time updates to financial forecasts, and drag-and-drop functionality for quickly exporting financial charts to a personal virtual workspace.

[0061] In the context of a metaverse conference, to address the issue of information acquisition when participating digital androids rotate or move, this application embodiment employs an intelligent dashboard perspective-following and multi-user collaborative perspective mechanism. The perspective-following function cleverly integrates user head posture data (collected in real-time via the VR device's gyroscope), enabling the dashboard to dynamically rotate while maintaining a readable angle. Simultaneously, it incorporates an anti-occlusion algorithm; when a collision object detects other digital androids obscuring the dashboard, the system intelligently makes the dashboard transparent or moves it to the nearest available area, ensuring comprehensive information display.

[0062] It's important to note that when multiple people simultaneously focus on the same controversial decision, the system will trigger a multi-user collaborative perspective mode. First, it uses head / eye tracking technology (VR devices continuously detect the user's gaze area, such as focusing on a conflicting section for more than 3 seconds) and voice hotspot detection (automatically marking topics discussed by multiple people as "high-controversy points") to identify the focus. When it detects two or more people operating the same Kanban board simultaneously, the system automatically splits it into multiple instances for each user to operate independently. The user's annotations on their personal copy will be displayed in real-time as "Item Created (Pending)". Other users can merge their annotations into the main Kanban board by gesturing "Accept". If no consensus is reached within 5 minutes, the moderator will be prompted to make a mandatory decision. Furthermore, the system will visualize the level of controversy through a floating controversy index (intuitively displaying attention data such as "7 people participated in the discussion, 4 people opposed") and an emotion aggregation chart (real-time statistics of participant emotion distribution such as "60% anxious, 20% support"), providing an intuitive reference for group decision-making based on emotion and participation.

[0063] It should be noted that the embodiments of this application can first detect conflicts between meeting decisions and historical decisions in real time through a decision conflict algorithm. This algorithm includes a semantic comparison model to calculate semantic similarity, and a conflict dimension evaluation model to analyze semantic conflict dimensions, temporal conflict dimensions, object conflict dimensions, and scheme conflict dimensions. When the comprehensive conflict score reaches the threshold of the current meeting scenario, a role-differentiated response mechanism is automatically triggered. Role matching is achieved by defining and binding VR device IDs using digital human identity tags, recognizing armbands or floating icons of legal codes using digital human clothing badges, and dynamically assigning roles manually. Subsequently, a role-specific dashboard library is called to display a 3D strategic dashboard, API compatibility conflict points, or legal clause warning boxes. The dashboard view-following function combines VR device gyroscope head posture data to dynamically rotate the display surface, and the anti-occlusion algorithm performs transparency or empty area displacement when a collision object detects occlusion, ensuring unobstructed presentation of conflict information.

[0064] Optionally, this embodiment may also employ the following methods, but not limited to these: obtaining evaluation information of at least one participating user on the content of the meeting conflict; when the information difference between the evaluation information of at least two target participating users meets the difference condition, exchanging the target digital human images corresponding to at least two target participating users, and generating updated comment information of at least two target participating users after the exchange, so as to update the information difference.

[0065] In this embodiment of the application, the evaluation information of at least one participating user on the content of the meeting conflict can be the discussion information of at least one user on the content of the meeting conflict; then, if the information difference of the evaluation information of at least two target participating users meets the difference condition, if there is a dispute in the discussion information of at least two target participating users, the target digital human images corresponding to at least two target participating users can be exchanged to generate updated comment information of at least two target participating users after the exchange, so as to update the information difference, wherein the target participating users can be the participating users with the difference in evaluation information.

[0066] For example, in a metaverse meeting scenario, when conflicts are presented, the system can use digital human behavior analysis to help the facilitator understand the participants' attitudes and facilitate conflict resolution. For instance, it can first construct personality profiles based on historical voting data. For example, conservative decision-makers (who historically voted against change 80%) will be prompted to "provide more data to persuade," while radical decision-makers (who often support high-risk solutions) will be marked as "the potential risks need to be emphasized." Simultaneously, the system captures the digital human's body language in real time—resistance signals such as crossed arms and furrowed brows will trigger a push notification indicating "User D may strongly oppose," while supportive signals such as nodding and leaning forward will indicate "User E tends to agree." Furthermore, the facilitator's dedicated hidden panel will display real-time participant emotion change curves, historical decision-making tendency radar charts, reference data from the company's knowledge base, and even recommended scripts based on conflict type (such as "cite similar cases from 2023 to reduce controversy"), providing multi-dimensional references for controlling the meeting's pace.

[0067] As an optional approach, when a discussion reaches a stalemate (lasting 5 minutes and with an emotion index > 0.7), i.e., when the information differences in the evaluation information of at least two target participants meet the difference criteria, the system will activate a dual-progression mechanism. On one hand, dynamic authoritative data intervention will suspend the original text of the clauses in the center of the disputing parties' field of vision, with key items highlighted by pulsed red light (e.g., projecting the latest architecture specifications and playing a holographic video of expert explanations during technical route disputes). Simultaneously, it will retrieve a 12-month impact report on similar conflicts from the enterprise knowledge base, displaying changes in key indicators after the decision in a timeline format. On the other hand, the role perception reconstruction function will forcibly swap the digital human models of the stalemate parties, exchanging the target digital human images corresponding to at least two target participants. For example, the technical director and product manager must use each other's role templates when speaking, with the system providing a real-time prompt: "You now represent the technical team and must prioritize the stability of the application programming interface." Simultaneously, a unified cost-benefit heatmap will be displayed on both parties' interfaces, eliminating information gaps. The reconstructed viewpoints will automatically generate a comparison report, intuitively presenting the points of cognitive difference, helping participants to think from each other's perspectives and break the deadlock, thereby changing the discussion conflict situation of at least two target participants.

[0068] It should be noted that when a meeting lasts for a period of time (e.g., five minutes) and the emotional index exceeds the threshold, leading to a stalemate, this embodiment of the application can use a digital human behavior analysis module to identify conservative or radical decision-makers through personality profiling, and a body language capture system to detect resistance signals such as crossed arms and furrowed brows. The system pushes a hidden panel displaying emotional change curves and recommended dialogue, while simultaneously activating a dual-debunking engine: dynamic authoritative data intervenes by suspending the original ISO clause in the center of the field of vision and highlighting key items with pulsed red light; cognitive role reconstruction forces the exchange of digital human templates between the stalemate parties and generates a cost-benefit heatmap. The multi-person collaborative perspective splitting system identifies highly controversial points based on head eye tracking. When it detects two or more people operating the same Kanban board, it automatically splits the labeled copy of the Kanban board, and users accept the merged annotations by clicking with gestures. A floating controversy index, such as four people opposing a discussion out of seven participants, is used in conjunction with an emotional aggregation chart to quantify the controversy.

[0069] For example, embodiments of this application can be implemented through... Figure 2 The 3D metaverse interface module interacts with the user, such as... Figure 11 The diagram illustrates a multimodal interaction example. The 3D Metaverse interface module serves as the multimodal interaction terminal, integrating all front-end outputs and processing user physiological signals. It utilizes eye-tracking technology to detect gaze (300ms response time), mapping 2D screen coordinates to 3D spatial positions. Gesture recognition analyzes VR actions, converting them into graphical operation commands (such as swipe → rotate). Its core principle is mapping physical actions to digital interactions: gazing at a digital human arm triggers a timeline, looking at a topic label plays historical clips, forming a natural and intuitive operation paradigm.

[0070] It should be noted that, among the related technologies, metaverse meetings are mostly used for static data association, lacking the ability to model dynamic decision flows in real time, and cannot handle the temporal and controversial characteristics of meeting scenarios; they focus on visual presentation and are not deeply bound to personal knowledge bases and decision-making behaviors, thus failing to provide personalized decision support; traditional meeting minutes are homogenized, meaning everyone receives the same content, making it impossible to quickly locate an individual's contribution to the meeting (e.g., "my speech accounted for 30% but was buried in the full text"); traditional meetings suffer from decision-making gaps, meaning that single meeting records are stored in isolation, making it impossible to trace historical decisions and identify decision conflicts in real time, leading to post-meeting work progress blockages and shirking of responsibility; traditional meetings suffer from information overload: written minutes require manual secondary processing and cannot intuitively present the division of responsibilities and timeline.

[0071] In some embodiments, this application can automatically trigger differentiated responses through multi-dimensional conflict analysis, enabling senior management to intuitively grasp the historical cases related to strategic risks, technical experts to quickly locate core conflict points such as API compatibility, and legal personnel to instantly identify the risk of breach of legal clauses. Dynamic dashboard tracking and anti-obstruction mechanisms ensure the continuous visibility of key information in complex spatial environments, significantly reducing decision-making delays caused by role perception biases. This application can also construct a decision logic chain based on the inheritance relationship of objections, combined with historical fragment backtracking triggered by gaze, forming a coherent evidence loop from previously scattered decision-making evidence. Users can directly perceive the evolution of the decision-making process through three-dimensional spatial interaction, and the real-time fusion capability of personal knowledge bases enables meeting logs to proactively warn of historical logical conflicts, effectively avoiding repetitive disputes and strengthening decision continuity. When the technical team and business departments are embroiled in disputes, this application can also force the exchange of digital human role templates between the two parties, requiring participants to express their views based on the other party's perspective; simultaneously, it automatically projects unified data views, such as cost-benefit heatmaps and highlighted compliance clauses, eliminating information asymmetry. The split display and merging function for collaborative annotation by multiple people, along with real-time dispute index prompts, breaks down complex disputes into actionable processes, promoting consensus.

[0072] Compared with existing technologies, this embodiment acquires audio and video data from the target meeting; based on the audio and video data, it determines the target topic information of the target meeting, thus accurately identifying the meeting topic; further, it determines the historical topic information of historical meetings associated with the target topic information, and constructs a meeting decision graph corresponding to the target meeting based on the target topic information and the historical topic information. The meeting decision graph is used to identify the decision-making relationship between the historical meeting and the target meeting, and can associate the meeting topics and meeting decisions of the historical meeting and the target meeting to obtain the meeting decision graph. Based on the meeting decision graph, decision conflict evaluation information of the target meeting is generated. This embodiment enables the tracing of historical meeting information to make associated decisions based on the meeting decision graph when there is a conflict between the target meeting and the historical meeting, thereby accurately identifying possible conflicts in the meeting and improving the user experience.

[0073] Furthermore, as Figure 1 and Figure 5 The specific implementation of the method shown in this embodiment provides a meeting information generation device, such as... Figure 12 As shown, the device includes: an acquisition module 31, a determination module 32, a construction module 33, and a generation module 34.

[0074] The acquisition module 31 is configured to acquire audio and video data from the target meeting; The determining module 32 is configured to determine the target topic information of the target meeting based on the audio and video data; The construction module 33 is configured to determine the historical topic information of the historical meetings associated with the target topic information, and construct a meeting decision graph corresponding to the target meeting based on the target topic information and the historical topic information. The meeting decision graph is used to identify the decision association relationship between the historical meetings and the target meeting. The generation module 34 is configured to generate decision conflict evaluation information for the target meeting based on the meeting decision graph.

[0075] In some examples of this embodiment, the construction module 33 is specifically configured to determine the historical meetings associated with the target topic information based on the semantic information of the target topic information; determine the historical topic information of the historical meetings from the historical meeting knowledge base; incrementally update the historical topic information based on the target topic information; and construct the meeting decision graph corresponding to the target meeting.

[0076] In some examples of this embodiment, the construction module 33 is further configured to determine the historical meeting decision graph corresponding to the historical topic information and the target decision information corresponding to the target meeting. The historical meeting decision graph includes the historical decision information of the historical meeting. In the historical meeting decision graph, the historical decision information is incrementally updated based on the target decision information to construct the meeting decision graph corresponding to the target meeting.

[0077] In some examples of this embodiment, the generation module 34 is specifically configured to identify, based on the meeting decision graph, the first conflict information between the historical meeting and the target meeting in the semantic conflict dimension, the second conflict information in the time conflict dimension, the third conflict information in the object conflict dimension, and the fourth conflict information in the scheme conflict dimension; generate the decision conflict evaluation information of the target meeting based on the first conflict information, the second conflict information, the third conflict information, and the fourth conflict information; and, if the decision conflict evaluation information meets the conflict conditions corresponding to the target meeting, display the meeting conflict content between the target meeting and the historical meeting in the target meeting.

[0078] In some examples of this embodiment, the generation module 34 is further configured to determine at least one digital human image corresponding to each of the at least one participating user based on the user type information of at least one participating user of the target meeting; determine at least one content dashboard corresponding to each of the at least one participating user based on the at least one digital human image; and display the meeting conflict content to the participating user in at least one of the content dashboards.

[0079] In some examples of this embodiment, the generation module 34 is further configured to obtain evaluation information of at least one of the participating users on the content of the meeting conflict; when the information difference of the evaluation information of at least two target participating users meets the difference condition, the target digital human images corresponding to the at least two target participating users are exchanged, and updated comment information of the at least two target participating users after the exchange is generated to update the information difference.

[0080] It should be noted that other corresponding descriptions of the functional units involved in the meeting information generation device provided in this embodiment can be found in [reference]. Figure 1 and Figure 5 The corresponding descriptions in [the document] will not be repeated here.

[0081] Based on the above, Figure 1 and Figure 5 Accordingly, this embodiment also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described method. Figure 1 and Figure 5 The method shown.

[0082] Based on this understanding, the technical solution of this application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (such as CD-ROM, USB flash drive, mobile hard drive, etc.) and includes several instructions to cause a computer device (such as personal computer, server, or network device, etc.) to execute the methods of various implementation scenarios of this application.

[0083] like Figure 13 The diagram shown is a hardware structure schematic of an electronic device according to the present invention, comprising: At least one processor 401; and, A memory 402 is communicatively connected to at least one of the processors 401; wherein, The memory 402 stores instructions that can be executed by at least one of the processors to enable the at least one of the processors to perform the conference information generation method as described above.

[0084] Figure 13 Take a processor 401 as an example.

[0085] The electronic device may also include an input device 403 and a display device 404.

[0086] The processor 401, memory 402, input device 403, and display device 404 can be connected via a bus or other means. Figure 13 Taking the example of a connection between China and Israel via a bus.

[0087] Memory 402, as a non-volatile computer-readable storage medium, can be used to store non-volatile software programs, non-volatile computer-executable programs, and modules, such as the program instructions / modules corresponding to the meeting information generation method in the embodiments of this application, for example, Figure 1 and Figure 5 The method flow is shown. The processor 401 executes various functional applications and data processing by running non-volatile software programs, instructions, and modules stored in the memory 402, thereby realizing the meeting information generation method in the above embodiments.

[0088] The memory 402 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created according to the use of the conference information generation method. Furthermore, the memory 402 may include high-speed random access memory and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state storage device. In some embodiments, the memory 402 may optionally include memory remotely located relative to the processor 401, and these remote memories may be connected via a network to the apparatus performing the conference information generation method. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0089] Input device 403 can receive user clicks and generate signal inputs related to user settings and function control of the meeting information generation method. Display device 404 may include display devices such as a display screen.

[0090] When one or more modules are stored in the memory 402, and are run by one or more processors 401, the meeting information generation method in any of the above method embodiments is executed.

[0091] Optionally, the aforementioned physical devices may also include a user interface, a network interface, a camera, radio frequency (RF) circuitry, sensors, audio circuitry, a Wi-Fi module, etc. The user interface may include a display screen, input units such as a keyboard, etc., and optional user interfaces may also include USB interfaces, card reader interfaces, etc. The network interface may optionally include standard wired interfaces, wireless interfaces (such as Wi-Fi interfaces), etc.

[0092] Those skilled in the art will understand that the physical device structure provided in this embodiment does not constitute a limitation on the physical device, and may include more or fewer components, or combine certain components, or have different component arrangements.

[0093] The storage medium may also include an operating system and a network communication module. The operating system is a program that manages the hardware and software resources of the aforementioned physical device, supporting the operation of information processing programs and other software and / or programs. The network communication module is used to enable communication between the various components within the storage medium, as well as communication with other hardware and software in the information processing physical device.

[0094] Through the above description of the embodiments, those skilled in the art can clearly understand that this application can be implemented by means of software plus necessary general-purpose hardware platforms, or it can be implemented by hardware. By applying the solution of this embodiment, compared with the existing technology, this embodiment can accurately determine the meeting topics of the target meeting by acquiring audio and video data in the target meeting; determining the target topic information of the target meeting based on the audio and video data; further determining the historical topic information of historical meetings associated with the target topic information, and constructing a meeting decision graph corresponding to the target meeting based on the target topic information and the historical topic information. The meeting decision graph is used to identify the decision association relationship between the historical meeting and the target meeting, and can associate the meeting topics and meeting decisions of the historical meeting and the target meeting to obtain the meeting decision graph. Based on the meeting decision graph, decision conflict evaluation information of the target meeting is generated. This embodiment can trace the historical meeting information based on the meeting decision graph to make associated decisions when there is a conflict between the target meeting and the historical meeting, thereby accurately identifying the possible conflict situation of the meeting and improving the user experience.

[0095] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0096] The above description is merely a specific embodiment of this application, enabling those skilled in the art to understand or implement this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments described herein, but is to be accorded the widest scope consistent with the principles and novel features claimed herein.

Claims

1. A method for generating meeting information, characterized in that, include: Acquire audio and video data from the target meeting; Based on the audio and video data, the target topic information of the target meeting is determined; Determine the historical topic information of the historical meetings associated with the target topic information, and construct a meeting decision graph corresponding to the target meeting based on the target topic information and the historical topic information. The meeting decision graph is used to identify the decision association relationship between the historical meetings and the target meeting. Based on the meeting decision graph, generate decision conflict evaluation information for the target meeting.

2. The method according to claim 1, characterized in that, The step of determining the historical topic information of the historical meetings associated with the target topic information, and constructing a meeting decision graph corresponding to the target meeting based on the target topic information and the historical topic information, includes: Based on the semantic information of the target topic information, determine the historical meetings associated with the target topic information; Determine the historical topics of the aforementioned historical conferences from the historical conference knowledge base; Based on the target topic information, the historical topic information is incrementally updated to construct the meeting decision graph corresponding to the target meeting.

3. The method according to claim 2, characterized in that, The step of incrementally updating the historical topic information based on the target topic information to construct the meeting decision graph corresponding to the target meeting includes: Determine the historical meeting decision map corresponding to the historical topic information, and the target decision information corresponding to the target meeting, wherein the historical meeting decision map includes the historical decision information of the historical meetings; In the historical meeting decision graph, the historical decision information is incrementally updated based on the target decision information to construct the meeting decision graph corresponding to the target meeting.

4. The method according to claim 1, characterized in that, Based on the meeting decision graph, generate decision conflict evaluation information for the target meeting, including: Based on the meeting decision graph, identify the first conflict information between the historical meeting and the target meeting in the semantic conflict dimension, the second conflict information in the time conflict dimension, the third conflict information in the object conflict dimension, and the fourth conflict information in the scheme conflict dimension. Based on the first conflict information, the second conflict information, the third conflict information, and the fourth conflict information, the decision conflict evaluation information of the target meeting is generated; If the decision conflict evaluation information meets the conflict conditions corresponding to the target meeting, the conflict content between the target meeting and the historical meeting will be displayed in the target meeting.

5. The method according to claim 4, characterized in that, The step of displaying the conflicting content between the target meeting and the historical meetings in the target meeting includes: Based on the user type information of at least one participant in the target meeting, determine at least one digital human image corresponding to each participant; Based on at least one of the digital human figures, determine at least one content dashboard corresponding to the participating user; The meeting conflict content is displayed to the participating users in at least one of the aforementioned content dashboards.

6. The method according to claim 5, characterized in that, The method further includes: Obtain evaluation information from at least one of the participating users regarding the content of the meeting conflicts; If the information difference between the evaluation information of at least two target participants meets the difference condition, the target digital human images corresponding to the at least two target participants are exchanged, and updated comment information of the at least two target participants after the exchange is generated to update the information difference.

7. A meeting information generation device, characterized in that, include: The acquisition module is configured to acquire audio and video data from the target meeting. The determination module is configured to determine the target topic information of the target meeting based on the audio and video data; The construction module is configured to determine the historical topic information of the historical meetings associated with the target topic information, and construct a meeting decision graph corresponding to the target meeting based on the target topic information and the historical topic information. The meeting decision graph is used to identify the decision association relationship between the historical meetings and the target meeting. The generation module is configured to generate decision conflict evaluation information for the target meeting based on the meeting decision graph.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1 to 6.

9. An electronic device, comprising a storage medium, a processor, and a computer program stored on the storage medium and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method of any one of claims 1 to 6.

10. A computer program product, the computer program product comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 6.

Images