Video conference grouping method and system, electronic equipment and storage medium

Abstract

The embodiment of the invention provides a video conference grouping method and system, and the method comprises the steps: obtaining the user information of a plurality of to-be-grouped participant users; extracting feature data of the plurality of participant users according to the user information; grouping the plurality of participant users by using the feature data to obtain a plurality of conference groups; and displaying the plurality of conference groups. According to the embodiment of the invention, the problems of low efficiency and insufficient rationality in traditional manual grouping are effectively solved, the organization efficiency of the conference is improved, and the conference effect is optimized.

Description

Technical Field

[0001] The present invention relates to the technical field of video conferencing, and in particular to a video conferencing grouping method, a video conferencing grouping system, an electronic device and a computer-readable storage medium. Background Art

[0002] With the popularity of remote work and online collaboration, video conferencing systems have become an indispensable tool for enterprises and organizations. However, in actual applications, existing video conferencing systems usually rely on manual grouping of participants. This manual grouping method is not only inefficient, but may also lead to unreasonable grouping due to human factors, thus affecting the overall effect of the meeting. Summary of the invention

[0003] In view of the above problems, embodiments of the present invention are proposed to provide a video conference grouping method, a video conference grouping system, an electronic device, and a computer-readable storage medium that overcome the above problems or at least partially solve the above problems.

[0004] In order to solve the above problem, an embodiment of the present invention discloses a video conference grouping method, the method comprising:

[0005] Obtain user information of multiple conference participants to be grouped;

[0006] Extracting characteristic data of a plurality of the conference-participating users according to the user information;

[0007] Using the characteristic data, grouping the plurality of conference participants to obtain a plurality of conference groups;

[0008] A plurality of said conference groups are displayed.

[0009] Optionally, the using the characteristic data to group the plurality of conference participants to obtain a plurality of conference groups includes:

[0010] Based on the feature data, a clustering algorithm or a graph theory algorithm is used to group the multiple conference participants to obtain the multiple conference groups.

[0011] Optionally, the step of grouping the plurality of conference participants using a clustering algorithm based on the feature data to obtain the plurality of conference groups includes:

[0012] Selecting a plurality of initial conference-party users from the plurality of conference-party users, and using the characteristic data of the plurality of initial conference-party users as characteristic vectors of a plurality of initial cluster centers;

[0013] The multiple conference party users are grouped according to the feature data of each of the conference party users and the feature vectors of the multiple initial cluster centers to obtain the multiple conference groups.

[0014] Optionally, the selecting a plurality of initial conference participant users from the plurality of conference participant users comprises:

[0015] Set the number of groups according to the meeting information of the video conference;

[0016] The initial participating party users of the group number are randomly selected from the multiple participating party users.

[0017] Optionally, grouping the plurality of conference party users according to the feature data of each of the conference party users and the feature vectors of the plurality of initial cluster centers to obtain the plurality of conference groups includes:

[0018] Calculate the Euclidean distance from each of the participating users to each of the initial cluster centers according to the feature data of each of the participating users and the feature vector of each of the initial cluster centers;

[0019] Each of the conference-participating users is assigned to the cluster where the initial cluster center with the closest Euclidean distance to the user is located to obtain a plurality of conference groups.

[0020] Optionally, after allocating each of the participating users to the cluster where the initial cluster center having the closest Euclidean distance to the participating user is located to obtain the plurality of conference groups, the method further includes:

[0021] Recalculating a feature vector of a new cluster center of each conference group according to the feature data of the conference participant users in each conference group;

[0022] When the maximum value of the change between the feature vector of the new cluster center of each conference group and the feature vector of the initial cluster center is less than the change threshold, or when the number of iterative operations of allocating the participating users to clusters and recalculating the feature vector of the new cluster center is equal to the number threshold, the current multiple conference groups are taken as multiple final conference groups.

[0023] Optionally, recalculating the feature vector of the new cluster center of each conference group according to the feature data of the participant users in each conference group includes:

[0024] Calculating the average of the characteristic data of all the participating users in each of the conference groups;

[0025] The mean is used as the feature vector of the new cluster center.

[0026] The embodiment of the present invention further discloses a video conference grouping system, the system comprising:

[0027] A user information acquisition module, used to acquire user information of multiple conference participants to be grouped;

[0028] A feature data extraction module, used to extract feature data of a plurality of conference-participating users according to the user information;

[0029] A user grouping determination module, configured to group the plurality of conference participants using the characteristic data to obtain a plurality of conference groups;

[0030] The conference group display module is used to display multiple conference groups.

[0031] Optionally, the user group determination module is used to group the plurality of conference participants based on the feature data by using a clustering algorithm or a graph theory algorithm to obtain the plurality of conference groups.

[0032] Optionally, the user group determination module includes:

[0033] An initial cluster center determination module, used to select a plurality of initial conference participants from the plurality of conference participants, and use the feature data of the plurality of initial conference participants as feature vectors of a plurality of initial cluster centers;

[0034] The conference participant user grouping module is used to group the multiple conference participant users according to the feature data of each of the conference participant users and the feature vectors of the multiple initial cluster centers to obtain the multiple conference groups.

[0035] Optionally, the initial cluster center determination module includes:

[0036] A group quantity setting module is used to set the group quantity according to the conference information of the video conference;

[0037] The initial participant user selection module is used to randomly select the grouped number of initial participant users from the multiple participant users.

[0038] Optionally, the participant user grouping module includes:

[0039] A Euclidean distance calculation module, used to calculate the Euclidean distance from each of the participating users to each of the initial cluster centers according to the feature data of each of the participating users and the feature vector of each of the initial cluster centers;

[0040] The conference participant user allocation module is used to allocate each of the conference participant users to the cluster where the initial cluster center with the closest Euclidean distance to the user is located to obtain multiple conference groups.

[0041] Optionally, the system further comprises:

[0042] A new cluster center calculation module, configured to recalculate a feature vector of a new cluster center of each conference group according to the feature data of the conference user in each conference group after the conference user allocation module allocates each conference user to the cluster where the initial cluster center with the closest Euclidean distance to the conference user is located to obtain a plurality of conference groups;

[0043] A convergence module is used to take the current multiple conference groups as multiple final conference groups when the maximum value of the change between the feature vector of the new cluster center of each conference group and the feature vector of the initial cluster center is less than a change threshold, or when the number of iterative operations of allocating the participating users to clusters and recalculating the feature vector of the new cluster center is equal to a number threshold.

[0044] Optionally, the new cluster center calculation module includes:

[0045] A feature mean calculation module, used to calculate the mean of the feature data of all the participating users in each conference group;

[0046] A new cluster center determination module is used to use the mean as a feature vector of the new cluster center.

[0047] An embodiment of the present invention also discloses an electronic device, comprising: one or more processors; and one or more machine-readable media storing instructions thereon, which, when executed by the one or more processors, enables the electronic device to execute the video conferencing grouping method as described above.

[0048] The embodiment of the present invention further discloses a computer-readable storage medium, wherein a computer program stored in the storage medium enables a processor to execute the above-mentioned video conference grouping method.

[0049] The embodiments of the present invention include the following advantages:

[0050] The video conference grouping scheme provided by the embodiment of the present invention obtains user information of multiple participating users to be grouped; extracts feature data of the multiple participating users based on the user information; groups the multiple participating users using the feature data to obtain multiple conference groups; and displays the multiple conference groups.

[0051] Compared with the background technology, the embodiments of the present invention have the following beneficial effects:

[0052] 1. Improve grouping efficiency

[0053] Through a systematic process, feature data is directly extracted from the user information of the participants and grouped automatically, which greatly reduces the manual participation and significantly shortens the time required for grouping. Especially in large-scale meetings, the automatic grouping solution can achieve rapid response and avoid the tedious steps of traditional manual grouping.

[0054] 2. Enhance the rationality of grouping

[0055] Feature data is extracted based on user information, such as professional background, areas of interest, job role or relevance to the conference topic, and grouping is performed using algorithms or rules to ensure that members of each conference group have high complementarity or common interests, thereby improving the scientific nature and applicability of the grouping.

[0056] 3. Improve meeting discussion results

[0057] The automatic grouping solution can reasonably allocate members according to the matching degree of the characteristic data of the participants, avoiding inefficient or dull discussions caused by mismatched interests or abilities, thereby improving the depth and quality of group discussions and promoting the efficient achievement of meeting goals.

[0058] 4. Reduce human subjective bias

[0059] It replaces manual intervention, avoids grouping errors caused by subjective factors or incomplete information, evaluates the characteristics of participating users more objectively and comprehensively, and generates optimized conference groups.

[0060] In summary, the embodiments of the present invention effectively solve the problems of low efficiency and insufficient rationality in traditional manual grouping, which not only improves the organizational efficiency of the meeting, but also optimizes the meeting effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 It is a flowchart of the steps of a video conference grouping method according to an embodiment of the present invention;

[0062] Figure 2 This is a flowchart of a step of an AI-based automatic grouping solution for video conferencing according to an embodiment of the present invention;

[0063] Figure 3 The present invention is a structural block diagram of a video conferencing grouping system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0064] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0065] The video conference grouping scheme proposed in the embodiment of the present invention realizes efficient and scientific grouping of participants through intelligent and automated means. First, the user information of multiple participants to be grouped is obtained, and feature data is extracted therefrom. Subsequently, the participant users are analyzed and matched using algorithms or rules using these feature data, and are grouped into multiple conference groups to ensure that the members of each conference group are highly complementary or consistent in terms of knowledge, skills or interests. Finally, the grouping results, i.e., multiple conference groups, are intuitively displayed to the conference organizer and / or the participants, thereby supporting the smooth conduct of the conference discussion. The embodiment of the present invention not only improves the grouping efficiency, but also significantly improves the rationality of the grouping and the effect of the conference discussion, which helps to improve the overall quality and output of the conference.

[0066] Reference Figure 1 , shows a flowchart of the steps of a video conference grouping method according to an embodiment of the present invention. The video conference grouping method can be applied to a video conference grouping system or a video conference system, referred to as a system. The video conference grouping method can specifically include the following steps:

[0067] Step 101: Obtain user information of multiple conference participants to be grouped.

[0068] User information can usually be obtained in a variety of ways, including the personal information filled out by the user when registering for the video conferencing system, historical meeting records, or instant messages provided by the user before the meeting. This user information usually includes but is not limited to name, position, department, professional field, hobbies, and roles related to the meeting (such as speaker, audience, technical support, etc.). In addition, more dimensional user information can be supplemented from external sources (such as internal enterprise systems or third-party platforms authorized by users), such as skill tags, past projects participated in, and current work tasks.

[0069] Step 102: extracting characteristic data of multiple conference participants based on user information.

[0070] The core of this step is to classify and process user information so that the user information can accurately reflect the key attributes of the participants in the form of feature data. The type of feature data can be flexibly defined according to the needs of the meeting, but usually includes field expertise, interest direction, job role, team division of labor, etc. For example, for a technical seminar, the feature data may focus on the participant's technology stack, research direction and related experience, while for a marketing meeting, it may focus more on the market experience, customer relationships and innovation capabilities of the participants.

[0071] The feature data extraction process can utilize a variety of technical methods, including natural language processing, keyword matching, and machine learning models. For example, the user's key skills or interests can be extracted through text analysis technology and automatically classified into predefined feature labels. In addition, for structured data (such as forms or options), options can be directly mapped to specific feature values.

[0072] Step 103: grouping multiple conference participants using the feature data to obtain multiple conference groups.

[0073] The key to this step is to ensure the rationality and effectiveness of grouping through algorithms or rules. Common grouping methods include clustering algorithms based on feature similarity, target-based optimization grouping algorithms, and rule-guided artificial intelligence decision trees. For example, in a technology sharing conference, a clustering algorithm can be used to group participants with similar technical backgrounds or common interests into a group, thereby ensuring that group members have a good basis for discussion.

[0074] In addition, for cross-functional team collaboration meetings, you can adopt a complementary grouping strategy, such as assigning employees from different departments to the same group based on the principle of complementary skills to promote cross-departmental collaboration. During the grouping process, you can define multiple constraints based on specific meeting requirements, such as the range of the number of members in each group, the balance of role distribution within the group, etc.

[0075] Step 104: Display multiple conference groups.

[0076] The display method needs to take into account both user experience and comprehensiveness of information. Common forms include charts, lists, tree structures, or visual interactive interfaces. For example, for smaller meetings, the members and related information of each group can be presented directly in the form of a list, while for larger meetings, dynamic charts or group diagrams can be used to help users quickly understand the group structure and member distribution.

[0077] In order to improve the transparency of the meeting, you can also provide the logic behind the grouping, such as showing the characteristic labels of each participant or the basis for grouping. This transparent design not only enhances the user's sense of trust, but also provides a feedback entry when necessary, allowing users to make modification suggestions or further optimize the grouping results.

[0078] The video conference grouping scheme provided by the embodiment of the present invention obtains user information of multiple participating users to be grouped; extracts feature data of the multiple participating users based on the user information; groups the multiple participating users using the feature data to obtain multiple conference groups; and displays the multiple conference groups.

[0079] Compared with the background technology, the embodiments of the present invention have the following beneficial effects:

[0080] 1. Improve grouping efficiency

[0081] Through a systematic process, feature data is directly extracted from the user information of the participants and grouped automatically, which greatly reduces the manual participation and significantly shortens the time required for grouping. Especially in large-scale meetings, the automatic grouping solution can achieve rapid response and avoid the tedious steps of traditional manual grouping.

[0082] 2. Enhance the rationality of grouping

[0083] Feature data is extracted based on user information, such as professional background, areas of interest, job role or relevance to the conference topic, and grouping is performed using algorithms or rules to ensure that members of each conference group have high complementarity or common interests, thereby improving the scientific nature and applicability of the grouping.

[0084] 3. Improve meeting discussion results

[0085] The automatic grouping solution can reasonably allocate members according to the matching degree of the characteristic data of the participants, avoiding inefficient or dull discussions caused by mismatched interests or abilities, thereby improving the depth and quality of group discussions and promoting the efficient achievement of meeting goals.

[0086] 4. Reduce human subjective bias

[0087] It replaces manual intervention, avoids grouping errors caused by subjective factors or incomplete information, evaluates the characteristics of participating users more objectively and comprehensively, and generates optimized conference groups.

[0088] In summary, the embodiments of the present invention effectively solve the problems of low efficiency and insufficient rationality in traditional manual grouping, which not only improves the organizational efficiency of the meeting, but also optimizes the meeting effect.

[0089] In an exemplary embodiment of the present invention, an implementation method of grouping multiple conference participants using feature data to obtain multiple conference groups is: based on the feature data, a clustering algorithm or a graph theory algorithm is used to group multiple conference participants to obtain multiple conference groups.

[0090] Specifically, the clustering algorithm is an unsupervised learning method that analyzes the similarity of the feature data of the participants and divides the participants with similar features into the same group. For example, in the K-means clustering algorithm, the feature data of the participants are first distributed in the feature space, and the optimal grouping center (i.e., cluster center) is found through iterative calculation, and each participant is assigned to the group closest to the cluster center. This method can effectively bring together participants with common interests or similar backgrounds, which helps to improve the discussion efficiency and collaboration quality within the group.

[0091] Graph theory algorithms are suitable for grouping scenarios that require complex constraints. Graph theory algorithms treat the participating users as nodes in the graph and represent the degree of feature correlation between the two participating users (such as overlapping interests and complementary skills) as the weight of the edge. By constructing a weighted graph, algorithms such as maximum weight matching, graph segmentation, or community detection can be used to achieve intelligent grouping based on relationship networks. For example, in a team collaboration meeting, participants from different departments with complementary skills can be assigned to the same group through a graph partitioning algorithm to maximize the effect of cross-domain collaboration. In addition, graph theory algorithms can also flexibly introduce other grouping conditions, such as a limit on the number of people in each group or the allocation requirements of specific roles, thereby optimizing the grouping strategy while meeting various practical needs.

[0092] This implementation method automatically processes feature data through relevant algorithms, greatly improving the efficiency of grouping, and can quickly respond to large-scale meetings or complex scenarios with many participants. The relevant algorithms are based on mathematical models and data-driven principles, avoiding possible subjective biases in manual grouping, making the grouping results more scientific and reasonable. For example, clustering algorithms can ensure the similarity of features within groups, and graph theory algorithms can achieve multi-dimensional optimized grouping according to specific task requirements.

[0093] In an exemplary embodiment of the present invention, based on feature data, a clustering algorithm is used to group multiple participating users to obtain multiple conference groups. One implementation method is: multiple initial participating users are selected from multiple participating users, and the feature data of the multiple initial participating users are used as feature vectors of multiple initial clustering centers; multiple participating users are grouped according to the feature data of each participating user and the feature vectors of the multiple initial clustering centers to obtain multiple conference groups.

[0094] Specifically, we first select a number of initial participants from the participants, and determine the feature vectors of the initial cluster centers based on their feature data (such as skill tags, areas of interest, or work background). These initial cluster centers can represent each initial grouping in the algorithm, providing a directional reference for subsequent grouping. Next, we calculate the feature similarity between other participants and the initial cluster centers, and assign the participants to the group with the closest features.

[0095] The key to this implementation lies in the selection of the initial clustering center. Reasonable selection of the initial participant users helps optimize the grouping effect. For example, if the initial participant users have obvious professional directions (such as data science, marketing, or human resource management), the generated initial clustering center can well reflect the gathering points of different interests or professional backgrounds, ensuring that each conference group has a high degree of similarity or consistency.

[0096] This implementation method can quickly group users by using the initial cluster center as a guide, without having to analyze the relationship between each participant and other participants one by one. The grouping method based on feature similarity avoids the bias of human subjective judgment and ensures the objectivity of the grouping result.

[0097] In an exemplary embodiment of the present invention, an implementation method of selecting multiple initial participating users from multiple participating users is: setting the number of groups according to conference information of the video conference; and randomly selecting the number of initial participating users from the multiple participating users.

[0098] Specifically, the number of groups is usually set based on the theme of the meeting, the number of participants, and the goal of the group discussion. For example, in a technical seminar, the number of groups may be consistent with the number of technical fields that need to be discussed; while in a team building activity, the number of groups may be determined by the total number of participants and the expected group size. After the number of groups is set, a number of users are selected from all the participants as the initial participants through random sampling, and the feature vector of the initial cluster center is generated based on their feature data.

[0099] The random selection of initial participants has certain fairness and diversity. Especially when there is no obvious distribution pattern of feature data, random sampling can effectively avoid subjective bias or prior assumptions that may be introduced by human selection. It ensures that the grouping process has high generalization. Even when the feature data of the participants are relatively scattered, different parts of the feature space can be covered by random initial clustering centers. For example, in a cross-departmental corporate meeting, the randomly selected initial participants may come from the technical department, marketing department, and human resources department, thus laying the foundation for diversified grouping at the initial stage.

[0100] This implementation method can quickly start the grouping process by randomly selecting the initial participants, reducing complex preliminary preparations, and is particularly efficient when the meeting is large and time is tight. Randomness introduces the diversity of the initial participants and avoids grouping bias or local optimality problems that may be caused by fixed rules. Through random selection, all participants have the opportunity to participate in the formation of the initial cluster center, reflecting an unbiased grouping principle.

[0101] In an exemplary embodiment of the present invention, a plurality of participating users are grouped according to feature data of each participating user and feature vectors of a plurality of initial clustering centers to obtain a plurality of conference groups. An implementation method is as follows: based on the feature data of each participating user and the feature vector of each initial clustering center, the Euclidean distance of each participating user to each initial clustering center is calculated respectively; and each participating user is assigned to a cluster where the initial clustering center with the closest Euclidean distance to the participating user is located to obtain a plurality of conference groups.

[0102] Specifically, during the grouping process, the position of each participant's feature data (such as interests, skills, or professional background) in the feature space is first calculated, and the feature vectors of multiple initial cluster centers are recorded. These feature vectors can be regarded as the coordinate points of the initial cluster centers in the feature space. Next, the distance value between each participant and all initial cluster centers is calculated using the Euclidean distance formula (i.e., the straight-line distance between two points), and the participant is classified into the cluster with the closest initial cluster center. By repeating this process, all participant users will be assigned to their most suitable conference groups.

[0103] The core role of Euclidean distance in this implementation is that it provides a method to quantify the similarity between the participant user and the initial cluster center. The shorter the Euclidean distance, the closer the characteristics of the participant user are to the characteristics of the initial cluster center. For example, if the initial cluster centers of a technology sharing conference are "artificial intelligence", "blockchain" and "big data", then through Euclidean distance calculation, the participant users who are interested in these topics can be clustered into corresponding conference groups.

[0104] The Euclidean distance of this implementation is a widely used mathematical tool that can effectively process feature data and map the complex feature data similarity into a simple distance value in a quantitative manner, which significantly improves the accuracy and operability of grouping. By assigning the participants to the cluster where the nearest initial cluster center is located, it is ensured that there is a high consistency or correlation between the features of each group of participants. This grouping strategy is particularly suitable for conference scenarios that require in-depth discussion or collaboration, such as academic seminars or technical exchanges. In addition, the calculation of the Euclidean distance is highly efficient, can quickly process large-scale data, and has good adaptability to the grouping needs of large conferences.

[0105] In an exemplary embodiment of the present invention, after each participating user is assigned to the cluster where the initial cluster center with the closest Euclidean distance to the participating user is located to obtain multiple conference groups, an implementation method is: based on the characteristic data of the participating users in each conference group, the characteristic vector of the new cluster center of each conference group is recalculated; when the maximum value of the change between the characteristic vector of the new cluster center of each conference group and the characteristic vector of the initial cluster center is less than the change threshold, or when the number of iterative operations of assigning participating users to clusters and recalculating the characteristic vector of the new cluster center is equal to the number threshold, the current multiple conference groups are used as multiple final conference groups.

[0106] Specifically, based on the initial grouping, the feature vector of the new cluster center of each group is recalculated according to the feature data of the participants in each conference group. This process is an update of the initial cluster center. By considering the actual distribution of the feature data of the participants assigned to each group, a new cluster center that better represents the current grouping characteristics is generated. This recalculation method can dynamically adjust the position of the cluster center to make it closer to the actual user distribution, thereby improving the rationality of the grouping.

[0107] Next, evaluate the change between the eigenvector of the new cluster center and the eigenvector of the previous cluster center. If the maximum value of all changes is lower than the preset change threshold, it means that the position of the cluster center has stabilized and the current grouping can be regarded as the final result. At the same time, in order to prevent the algorithm from falling into endless iterations, a number threshold is also set. When the number of iterations reaches this number threshold, even if the change of the cluster center has not fully converged, the current conference group will be used as the final conference group. This dual condition design (change threshold and number threshold) not only ensures the quality of grouping, but also effectively controls the computational cost, avoiding the impact of excessive computing time on system efficiency.

[0108] Through iterative optimization, the grouping results of this implementation can fully consider the actual feature distribution of the participating users, making the feature center of each conference group more representative. Setting the dual conditions of the change threshold and the number threshold not only improves the convergence of the algorithm, but also ensures the efficiency of the grouping process. It is particularly suitable for large-scale conferences with a large number of participating users and diverse features. It can save time and resources while ensuring the grouping effect. In addition, this iterative optimization process can accommodate a certain degree of feature data uncertainty or distribution anomalies, and dynamically adjust the clustering center to smooth the noise or deviation of the data, thereby improving the robustness of the final grouping.

[0109] In an exemplary embodiment of the present invention, an implementation method of recalculating the feature vector of the new cluster center of each conference group based on the feature data of the participating users in each conference group is: calculating the mean of the feature data of all participating users in each conference group; and using the mean as the feature vector of the new cluster center.

[0110] Specifically, after the initial grouping is completed, each conference group contains a group of participants with specific feature data. These feature data may include attributes such as interest areas, professional skills, and functional backgrounds. By averaging these feature data, a new feature vector is generated, which can be regarded as a comprehensive expression or representative value of the characteristics of users in the group. The feature vector of the new cluster center replaces the feature vector of the initial cluster center and is used for grouping adjustments in the next round of iterations.

[0111] As a commonly used central tendency indicator in statistics, the mean can effectively balance the differences in characteristics of users from different parties in a group. For example, in a technical exchange meeting, if most users in a group are interested in "artificial intelligence" and "blockchain", while a small number of users are interested in "big data", then the process of calculating the mean will adjust the cluster center according to the weight of the number of users, making it more inclined to "artificial intelligence" and "blockchain", while taking into account a certain degree of "big data" characteristics. This adjustment ensures the integrity and representativeness of the characteristics within the group, while also appropriately accommodating the existence of a few characteristics, enhancing the diversity and inclusiveness of the discussion within the group.

[0112] This implementation adjusts the position of the cluster center through mean calculation, which can more realistically reflect the actual feature distribution of the users of the participating parties in the group, thereby optimizing the scientificity and rationality of the grouping. It is particularly effective for processing scenarios with uneven distribution of feature data. For example, in cross-domain meetings with diverse backgrounds of participants, the extreme differences in features can be smoothed through mean calculation to avoid imbalance in grouping results. The mean calculation method is simple and efficient, with low computational complexity. It is particularly suitable for large-scale conference scenarios, and can quickly complete the update of cluster centers and improve grouping efficiency. In addition, the use of the mean as the feature vector of the new cluster center has good stability. Even if there are anomalies or noise in the feature data of individual users, the impact on the overall result is relatively small, thereby improving the robustness and reliability of the grouping results.

[0113] Based on the above description of an embodiment of a grouping method for a video conference, an automatic grouping solution for a video conference based on artificial intelligence (AI) is introduced below.

[0114] This AI-based video conferencing automatic grouping solution mainly includes the following components:

[0115] Data collection module: used to collect basic information of participants, including but not limited to name, position, department, interests, historical meeting participation, etc.

[0116] Feature extraction module: processes the collected basic information and extracts key features for grouping, such as the professional fields, interest preferences, historical cooperation status, etc. of the participants.

[0117] Grouping algorithm module: Based on the extracted feature data, clustering algorithms (such as K-means, hierarchical clustering, etc.) or graph theory algorithms (such as community discovery algorithms) are used to group the participants. The goal of grouping is to make the participants within the group have a high degree of similarity in terms of attributes, interests, roles, etc., while the participants between groups have a low degree of similarity.

[0118] Optimization module: further optimize the grouping results, consider factors such as meeting objectives and agenda arrangements, and adjust the groups to ensure that each group can effectively participate in the discussion.

[0119] Result output module: displays the grouping results to the conference organizer or participants in an intuitive manner, such as displaying the grouping status through the interface of the conference system.

[0120] Reference Figure 2 , shows a flowchart of the steps of an AI-based automatic grouping solution for video conferencing according to an embodiment of the present invention. The AI-based automatic grouping solution for video conferencing may specifically include the following steps:

[0121] Step 201: Collect user information of conference participants.

[0122] Before the video conference begins, the basic information of the participants is collected through the conference system or enterprise information system. It can also be connected to the conference system to obtain the registration information of the participants, such as name, position, department, etc. It can also obtain detailed information of the participants through the internal human resources system or employee management system of the enterprise, such as educational background, work experience, historical meeting participation, etc. The above basic information, registration information and detailed information can all be used as user information.

[0123] Step 202: extracting characteristic data of the conference participant user from the user information.

[0124] Natural language processing and machine learning technologies are used to process the collected user information and extract characteristic data such as the professional fields, interest preferences, etc. of the participants.

[0125] Specifically, the text information of the participant's job description, educational background, etc. is pre-processed by natural language processing technology such as word segmentation, stop word removal, and stem extraction to obtain text features. Numerical information such as the participant's age and years of work experience can also be directly standardized. The participant's historical meeting participation can be quantitatively scored, such as assigning different weights according to the number of speeches and questions to obtain numerical features.

[0126] Finally, the text features and numerical features are combined to form the feature data of the participating users.

[0127] Step 203: Based on the feature data, a clustering algorithm is used to group the conference participants to obtain multiple conference groups.

[0128] Specifically, the K-means clustering algorithm is used to group the participants, the number of groups is set to K, and the Euclidean distance from each participant to each cluster center is calculated, and the participant is assigned to the group with the closest cluster center. The following is a detailed description:

[0129] Step 1: Set the number of groups K, which can be preset according to factors such as the meeting scale and agenda requirements, and randomly select the feature data of K participating users as the feature vector of the initial cluster center.

[0130] Step 2: For each participant user, calculate the distance from the participant to all cluster centers (usually using the Euclidean distance), and assign each participant user to the cluster with the nearest cluster center.

[0131] For multi-dimensional feature data, the calculation formula of Euclidean distance is:

[0132] d(x,c)=∑i=1n(xi-ci)2

[0133] Among them, x is the feature data of the participating users, c is the feature vector of the cluster center, n is the dimension of the feature data, and i is the number of participating users and cluster centers.

[0134] Step 3: For each cluster, recalculate its cluster center. The new cluster center is usually the mean (i.e., centroid) of the feature data of all participating users in the cluster.

[0135] For the kth cluster, the calculation formula for its new cluster center ck is:

[0136] ck=|Ck|1∑x∈Ckx

[0137] Among them, x is the characteristic data of the participating user, Ck is the set of all participating users in the kth cluster, and |Ck| is the number of participating users in the cluster.

[0138] Step 4: Repeat steps 2 and 3 until a convergence condition is reached. The convergence condition may be that the cluster center no longer changes (ie, the new cluster center is the same as the cluster center of the previous iteration), or a preset number of iterations is reached.

[0139] A threshold (such as the maximum value of the cluster center change is less than a certain value) or a preset number of iterations can be set as a convergence condition. When the convergence condition is met, the algorithm stops iterating and outputs the grouping result.

[0140] Step 5: Finally, each participant user is assigned to a cluster, i.e., a group. The grouping results can be displayed in a list or graphic form on the interface of the conference system for the conference organizer or the participant user to view.

[0141] It should be noted that, for the sake of simplicity, the method embodiments are described as a series of action combinations, but those skilled in the art should be aware that the embodiments of the present invention are not limited by the order of the actions described, because according to the embodiments of the present invention, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions involved are not necessarily required by the embodiments of the present invention.

[0142] Reference Figure 3 , shows a structural block diagram of a video conference grouping system according to an embodiment of the present invention. The video conference grouping system may specifically include the following modules.

[0143] A user information acquisition module 31 is used to acquire user information of multiple conference participants to be grouped;

[0144] A feature data extraction module 32 is used to extract feature data of a plurality of the conference participants based on the user information;

[0145] A user group determination module 33, configured to group the plurality of conference participants using the characteristic data to obtain a plurality of conference groups;

[0146] The conference group display module 34 is used to display a plurality of the conference groups.

[0147] In an exemplary embodiment of the present invention, the user group determination module 33 is used to group the plurality of conference participants based on the feature data using a clustering algorithm or a graph theory algorithm to obtain the plurality of conference groups.

[0148] In an exemplary embodiment of the present invention, the user group determination module 33 includes:

[0149] An initial cluster center determination module, used to select a plurality of initial conference participants from the plurality of conference participants, and use the feature data of the plurality of initial conference participants as feature vectors of a plurality of initial cluster centers;

[0150] The conference participant user grouping module is used to group the multiple conference participant users according to the feature data of each of the conference participant users and the feature vectors of the multiple initial cluster centers to obtain the multiple conference groups.

[0151] In an exemplary embodiment of the present invention, the initial cluster center determination module includes:

[0152] A group quantity setting module is used to set the group quantity according to the conference information of the video conference;

[0153] The initial participant user selection module is used to randomly select the grouped number of initial participant users from the multiple participant users.

[0154] In an exemplary embodiment of the present invention, the participant user grouping module includes:

[0155] A Euclidean distance calculation module, used to calculate the Euclidean distance from each of the participating users to each of the initial cluster centers according to the feature data of each of the participating users and the feature vector of each of the initial cluster centers;

[0156] The conference participant user allocation module is used to allocate each of the conference participant users to the cluster where the initial cluster center with the closest Euclidean distance to the user is located to obtain multiple conference groups.

[0157] In an exemplary embodiment of the present invention, the system further comprises:

[0158] A new cluster center calculation module, configured to recalculate a feature vector of a new cluster center of each conference group according to the feature data of the conference user in each conference group after the conference user allocation module allocates each conference user to the cluster where the initial cluster center with the closest Euclidean distance to the conference user is located to obtain a plurality of conference groups;

[0159] A convergence module is used to take the current multiple conference groups as multiple final conference groups when the maximum value of the change between the feature vector of the new cluster center of each conference group and the feature vector of the initial cluster center is less than a change threshold, or when the number of iterative operations of allocating the participating users to clusters and recalculating the feature vector of the new cluster center is equal to a number threshold.

[0160] In an exemplary embodiment of the present invention, the new cluster center calculation module includes:

[0161] A feature mean calculation module, used to calculate the mean of the feature data of all the participating users in each conference group;

[0162] A new cluster center determination module is used to use the mean as a feature vector of the new cluster center.

[0163] As for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0164] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0165] It will be appreciated by those skilled in the art that the embodiments of the present invention may be provided as methods, devices, or computer program products. Therefore, the embodiments of the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the embodiments of the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0166] The embodiments of the present invention are described with reference to the flowcharts and / or block diagrams of the methods, terminal devices (systems), and computer program products according to the embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing terminal device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing terminal device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0167] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing terminal device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0168] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal device so that a series of operating steps are executed on the computer or other programmable terminal device to produce a computer-implemented process, thereby providing instructions for executing on the computer or other programmable terminal device to implement the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0169] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the embodiments of the present invention.

[0170] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or terminal device including the elements.

[0171] The above is a detailed introduction to a video conference grouping method and a video conference grouping system provided by the present invention. Specific examples are used in this article to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for general technical personnel in this field, according to the idea of ​​the present invention, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A video conference grouping method, characterized in that: The method comprises: Obtain user information of multiple conference participants to be grouped; Extracting characteristic data of a plurality of the conference-participating users according to the user information; Using the characteristic data, grouping the plurality of conference participants to obtain a plurality of conference groups; A plurality of said conference groups are displayed.

2. The method according to claim 1, characterized in that The step of grouping the plurality of conference participants using the characteristic data to obtain a plurality of conference groups includes: Based on the feature data, a clustering algorithm or a graph theory algorithm is used to group the multiple conference participants to obtain the multiple conference groups.

3. The method according to claim 2, characterized in that The step of grouping the plurality of conference participants based on the characteristic data using a clustering algorithm to obtain the plurality of conference groups includes: Selecting a plurality of initial conference-party users from the plurality of conference-party users, and using the characteristic data of the plurality of initial conference-party users as characteristic vectors of a plurality of initial cluster centers; The multiple conference party users are grouped according to the feature data of each of the conference party users and the feature vectors of the multiple initial cluster centers to obtain the multiple conference groups.

4. The method according to claim 3, characterized in that The step of selecting a plurality of initial conference participants from the plurality of conference participants includes: Set the number of groups according to the meeting information of the video conference; The initial participating party users of the group number are randomly selected from the multiple participating party users.

5. The method according to claim 3, characterized in that: The step of grouping the plurality of conference participants according to the characteristic data of each of the conference participants and the characteristic vectors of the plurality of initial cluster centers to obtain the plurality of conference groups includes: Calculate the Euclidean distance from each of the participating users to each of the initial cluster centers according to the feature data of each of the participating users and the feature vector of each of the initial cluster centers; Each of the conference-participating users is assigned to the cluster where the initial cluster center with the closest Euclidean distance to the user is located to obtain a plurality of conference groups.

6. The method according to claim 5, characterized in that After allocating each of the conference participants to the cluster where the initial cluster center with the closest Euclidean distance to the user is located to obtain the plurality of conference groups, the method further includes: Recalculating a feature vector of a new cluster center of each conference group according to the feature data of the conference participant users in each conference group; When the maximum value of the change between the feature vector of the new cluster center of each conference group and the feature vector of the initial cluster center is less than the change threshold, or when the number of iterative operations of allocating the participating users to clusters and recalculating the feature vector of the new cluster center is equal to the number threshold, the current multiple conference groups are taken as multiple final conference groups.

7. The method according to claim 6, characterized in that The step of recalculating the feature vector of the new cluster center of each conference group according to the feature data of the participant users in each conference group includes: Calculating the average of the characteristic data of all the participating users in each of the conference groups; The mean is used as the feature vector of the new cluster center.

8. A video conference grouping system, characterized in that: The system comprises: A user information acquisition module, used to acquire user information of multiple conference participants to be grouped; A feature data extraction module, used to extract feature data of a plurality of conference-participating users according to the user information; A user grouping determination module, configured to group the plurality of conference participants using the characteristic data to obtain a plurality of conference groups; The conference group display module is used to display multiple conference groups.

9. An electronic device, characterized in that: include: one or more processors; and One or more machine-readable media having instructions stored thereon, when executed by the one or more processors, enable the electronic device to execute the video conference grouping method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer program stored therein enables the processor to execute the video conference grouping method as described in any one of claims 1 to 7.

Images