Methods and apparatus for processing meeting data, methods and apparatus for splitting media content, methods and apparatus for generating summaries, electronic devices, and computer-readable media

Abstract

Embodiments of this disclosure relate to a method and apparatus for processing conference data, a method and apparatus for segmenting media content, a method and apparatus for generating summaries, electronic equipment, and computer-readable media. The method for processing conference data includes acquiring conference data of a network conference, determining that the conference data includes audio data of the network conference, determining the conference type of the network conference based on the conference data, and segmenting the network conference based on the conference type and the conference content of the network conference to obtain conference segments of the network conference. The method for segmenting media content includes acquiring data of media content, determining that the media content includes content of at least two content dimensions, determining a target segmentation point of the media content based on the content dimensions and the media content data, and determining sub-content of the media content based on the target segmentation point. The summarization method includes obtaining content data for each sub-content included in the media content, where the sub-content is obtained by dividing the media content; determining a summary of each sub-content based on its content data; and merging the summaries of each sub-content based on their respective weights to obtain a summary of the media content, where the weights of the sub-contents indicate the importance of each sub-content within the media content.

Description

Technical Field

[0001] [Cross - reference to Related Applications] This application claims the priority of Chinese Patent Applications No. 202311569183.4, No. 202311569948.4, and No. 202311569394.8 filed on November 22, 2023, and incorporates by reference in its entirety the content disclosed in the above - mentioned Chinese patent applications as part of this application.

[0002] Embodiments of the present disclosure relate to a method and apparatus for processing conference data, a method and apparatus for splitting media content, a method and apparatus for generating summaries, an electronic device, and a computer - readable medium.

Background Art

[0003] Users need to quickly grasp the specific content of certain media content. For example, in the scenario of video playback, users may need to grasp the general content of the video to decide whether to continue watching. Also, for example, in the scenario of a meeting, after the meeting ends, users may need to check the meeting content to grasp the content discussed in this meeting.

[0004] Conventionally, users can quickly grasp the content from the summary of media content. However, it is difficult for the summary of media content to accurately summarize the main content of the media content, which affects users' understanding of the media content. A network meeting may also be called an online meeting, referring to a meeting realized via a network. Participants in a network meeting can hold and participate in the meeting via the network. Participants in a network meeting can generate meeting data of the network meeting by using the meeting content of the network meeting automatically recorded by a client that provides the network - meeting service. After the meeting ends, participants can use the meeting data to check the meeting content.

[0005] However, network meetings typically have a relatively large amount of meeting content, making it inconvenient for participants to review it. [Overview of the project] [Means for solving the problem]

[0006] Embodiments of the present disclosure provide a method for processing conference data, which includes obtaining conference data of a network conference, wherein the conference data includes audio data of the network conference; determining the conference type of the network conference based on the conference data; and dividing the network conference based on the conference type and the conference content of the network conference to obtain conference segments of the network conference.

[0007] Embodiments of the present disclosure provide a method for segmenting media content, which includes obtaining data of the media content, determining that the media content includes content of at least two content dimensions, determining a target segmentation point of the media content based on the content dimensions and the data of the media content, and determining sub-content of the media content based on the target segmentation point.

[0008] Embodiments of the present disclosure provide a summary generation method, which includes: obtaining content data for each sub-content included in media content, wherein the sub-content is obtained by dividing the media content; determining a summary of each sub-content based on the content data of each sub-content; and merging the summaries of each sub-content based on the weight of each sub-content to obtain a summary of the media content, wherein the weight of each sub-content indicates the importance of the sub-content in the media content.

[0009] Embodiments of the present disclosure provide a conference data processing device configured to acquire conference data of a network conference, the conference data comprising: a first acquisition unit configured to acquire conference data including audio data of the network conference; a first determination unit configured to determine the conference type of the network conference based on the conference data; and a first splitting unit configured to split the network conference based on the conference type and the conference content of the network conference, thereby acquiring conference segments of the network conference.

[0010] Embodiments of the present disclosure provide a media content splitting device configured to acquire data of media content, comprising: a second acquisition unit configured to acquire content of at least two content dimensions; a second determination unit configured to determine a target splitting point of the media content based on the content dimensions and the data of the media content; and a second splitting unit configured to determine sub-content of the media content based on the target splitting point.

[0011] Embodiments of the present disclosure provide a summary generation device comprising: a third acquisition unit configured to acquire content data of each sub-content contained in media content, wherein the sub-content is acquired by dividing the media content; a third determination unit configured to determine a summary of each sub-content based on the content data of each sub-content; and a generation unit configured to acquire a summary of the media content by fusing the summaries of each sub-content based on the weight of each sub-content, wherein the weight of each sub-content indicates the importance of the sub-content in the media content.

[0012] Embodiments of the present disclosure provide an electronic device comprising at least one processor and a storage device storing at least one program, wherein the at least one program, when executed by the at least one processor, causes the at least one processor to implement a method for processing meeting data according to any one embodiment of the present disclosure, a method for splitting media content according to any one embodiment of the present disclosure, or a method for generating a summary according to any one embodiment of the present disclosure.

[0013] Embodiments of the present disclosure provide a computer-readable medium on which a computer program is stored, and when executed by a processor, the program implements a method for processing meeting data according to any one embodiment of the present disclosure, a method for splitting media content according to any one embodiment of the present disclosure, or a method for generating a summary according to any one embodiment of the present disclosure. [Brief explanation of the drawing]

[0014] To further clarify the technical concepts of the embodiments of this disclosure, the drawings of the embodiments will be briefly described below. Clearly, the drawings described below relate to only some embodiments of this disclosure and do not limit this disclosure.

[0015] [Figure 1] Figure 1 is a flowchart of the method for processing meeting data according to an embodiment of this disclosure. [Figure 2] Figure 2 is a schematic diagram of the structure of a conference data processing device according to an embodiment of the present disclosure. [Figure 3] Figure 3 is a flowchart of a media content segmentation method according to an embodiment of the present disclosure. [Figure 4] Figure 4 is a schematic diagram of the structure of a media content splitting device according to an embodiment of the present disclosure. [Figure 5] Figure 5 is a flowchart of the summary generation method according to an embodiment of this disclosure. [Figure 6] Figure 6 is a schematic diagram of the structure of a summary generation device according to an embodiment of the present disclosure. [Figure 7]Figure 7 is a schematic diagram of the basic structure of an electronic device according to an embodiment of the present disclosure. [Modes for carrying out the invention]

[0016] The following describes the technical aspects of the embodiments of this disclosure clearly and completely, with reference to the drawings of the embodiments, in order to further clarify the purpose, technical proposal, and advantages of the embodiments of this disclosure. Clearly, the embodiments described are a part of, but not all, embodiments of this disclosure. All other embodiments that a person skilled in the art can obtain without requiring any creative work based on the embodiments of this disclosure described are all within the scope of this disclosure. After a network conference has ended, conference data can be generated to record the conference content. The conference data includes, for example, audio data of the network conference. Users authorized to review the conference content can review the conference data and understand the conference content. However, in the case of long conferences, the amount of conference content is relatively large. It is inconvenient for users to review a large amount of conference content.

[0017] Based on this, embodiments of the present disclosure provide a method, apparatus, equipment, and medium for processing conference data. In this method, conference data of a network conference is acquired. The conference data includes audio data of the network conference. Based on the conference data, the conference type of the network conference is determined. Furthermore, the network conference is divided based on the conference type and the conference content of the network conference to obtain conference segments of the network conference. By dividing the network conference into different segments based on the conference type and conference content, more accurate conference segments containing different conference content can be obtained, thereby achieving a better segmentation effect of the network conference. In this way, users can easily view conference content by checking the conference segments and understanding the conference content included in those segments, thereby improving the user experience.

[0018] The method for processing conference data according to an embodiment of the present disclosure can be applied to an electronic device having a conference data processing capability. The electronic device may be, for example, a server or a terminal. The terminal may include, but is not limited to, a smartphone, a tablet computer, a notebook computer, a personal digital assistant (PDA), or a smart wearable device. The server may be, for example, a central server in a central computing cluster or a cloud server such as an edge server in an edge computing cluster. Of course, the server may also be a server in a local data center. The local data center refers to a data center directly controlled by a user.

[0019] The electronic device acquires conference data of a network conference including audio data, and determines the conference type of the network conference by analyzing the conference data. The network conference is segmented based on the conference type to obtain conference segments of the network conference including different conference contents. The conference segments obtained by segmentation based on the conference type and conference content are more reasonable. In this way, the user can grasp the conference content included in the conference segment, improve the convenience for the user to check the conference content, and improve the user experience.

[0020] As can be understood by those skilled in the art, the above application scenario is only an example in which the embodiments of the present disclosure can be realized. The application scope of the embodiments of the present disclosure is not limited by the framework at all.

[0021] To facilitate understanding of the technical solution according to the embodiment of the present disclosure, the method for processing conference data according to the embodiment of the present disclosure will be described below with reference to the drawings.

[0022] Referring to FIG. 1. FIG. 1 is a flowchart of a method for processing conference data according to an embodiment of the present disclosure. As shown in FIG. 1, the method for processing conference data may include S101 to S103.

[0023] S101: Obtain the meeting data of the network meeting. The meeting data includes the audio data of the network meeting.

[0024] The meeting data of the network meeting is data for recording meeting content generated during the process of holding the network meeting. The embodiments of the present disclosure do not limit the source of the meeting data of the network meeting. In one possible implementation, the meeting data of the network meeting is generated by the participants of the network meeting triggering the recording function of the network meeting. In other possible implementations, the meeting data of the network meeting is data for realizing the interaction between participants generated during the progress of the network meeting.

[0025] The meeting data of the network meeting includes audio data. The audio data is generated based on the voices of the participants of the network meeting. In some possible implementations, the meeting data of the network meeting further includes other types of data such as video data, shared content display data, meeting reservation data, etc. The video data is generated based on the videos of the participants of the network meeting. The shared content display data is generated based on the content shared by the participants in the network meeting. The shared content is, for example, the screen of a device, a file, a video, a photo, etc. The meeting reservation data is the data of the network meeting reserved by the participants for participation. The meeting reservation data includes the relevant information of the network meeting input by the participants. The data types included in the meeting data of the network meeting are determined by the interaction method adopted by the participants in the network meeting.

[0026] S102: Determine the meeting type of the network meeting based on the meeting data.

[0027] The meeting type may be a pre-defined type according to the need for segmentation of the network meeting.

[0028] The embodiments of this disclosure do not limit the methods for classifying meeting types. For example, depending on the communication type of the different network meetings, network meetings may be classified into interview type, share type, multitopic type, and general type. An interview type network meeting may, for example, be a meeting with two participants, where one person asks questions and the other answers. A share type network meeting may, for example, be a meeting with at least two participants, where one participant gives a presentation. Specifically, a share type network meeting may further be a meeting with at least two participants, where one participant shares and explains content and the other participants ask questions. A multitopic type network meeting is a meeting that discusses multiple meeting topics. Multitopic type network meetings typically have multiple participants. A general-purpose type network meeting is a meeting that does not belong to the interview type, share type, or multitopic type. The above classification methods for meeting types are merely examples and do not limit the specific classification methods for network meeting types. As another example, network meetings can be classified according to the number of participants.

[0029] Network meetings of different conference types have different characteristics of conference content. The conference type of a network meeting is determined by analyzing the conference data of the network meeting. Of several possible implementations, embodiments of this disclosure provide two specific implementations for analyzing conference data to obtain the conference type of a network meeting, which will be described in detail later.

[0030] S103: Segment the network conference based on the conference type and the conference content of the network conference to obtain the conference segments of the network conference.

[0031] By segmenting network meetings based on meeting type and meeting content, it is possible to obtain network segments that contain different primary meeting content.

[0032] The embodiments of this disclosure are not limited to methods for segmenting network meetings based on meeting type and meeting content.

[0033] One possible implementation involves pre-setting segmentation rules corresponding to meeting types. These segmentation rules may be pre-set according to the meeting type and segmentation needs. For example, if the meeting type is an interview, the segmentation rule for an interview type is that one question and one answer constitute one meeting segment. Alternatively, the segmentation rule for an interview type is that questions and answers belonging to the same issue constitute one meeting segment. In other words, the first question, follow-up questions, and answers all belong to one meeting segment.

[0034] After determining the meeting type, the system first processes the network meeting data to obtain the network meeting content, and then segments the meeting content according to segmentation rules corresponding to the meeting type. For example, the network meeting content may be represented by meeting text data obtained by processing the network meeting audio data using speech recognition.

[0035] In other possible implementations, a segmentation model corresponding to the conference type is pre-trained. The segmentation model can recognize the conference content of a network conference based on the conference data and perform segmentation processing on the network conference based on the conference type. For example, the segmentation model corresponding to the conference type may be acquired by training using the conference data of that conference type and the segmentation results of the network conference. After determining the conference type of the network conference, the conference data of the network conference is processed using the segmentation model corresponding to the conference type, and the conference segments determined by the segmentation model are obtained. In the case of audio data from a general-purpose network conference, a general-purpose segmentation model can be used to divide the conference segments.

[0036] The embodiments of this disclosure do not limit the marking method for different conference segments. As an example, a segmentation point is determined in audio data of a network conference. The segmentation point is for marking audio data before the segmentation point and is audio data of a different conference segment than the audio data after the segmentation point.

[0037] As can be seen from the contents of S101 to S103 above, segmenting a network conference based on the conference type and conference content allows for the segmentation of network conferences containing different conference content based on the content communication characteristics of the conference type, resulting in a relatively high degree of relevance of the conference content in the acquired conference segments and more accurate segmentation of conference segments. Compared to segmenting network conferences according to a fixed segmentation method, the conference data processing method according to the embodiment of this disclosure has a superior network conference segmentation effect and satisfies the user's need to review conference content. By reviewing the conference content included in a conference segment, users can quickly grasp the content discussed in that segment, improving the efficiency of users reviewing conference content.

[0038] The embodiments of this disclosure provide two possible implementations for determining the conference type of a network conference based on conference data.

[0039] First implementation method: Analyze meeting data to obtain communication information of the network meeting. Network meeting communication information includes information related to the participants in the network meeting and information related to the process by which participants communicate in the network meeting. For example, network meeting communication information includes at least information indicating the number of participants, information indicating the communication format, and information indicating the communication content. Here, information indicating the number of participants is for indicating the number of participants, i.e., the number of participants participating in this network meeting. Information indicating the communication format is for indicating the communication format, and the communication format is the communication format between participants. For example, the communication format includes, for example, a question and answer format, an explanation format, a discussion format, etc. Information indicating the communication content indicates the communication content, and the communication content is the specific communication content between participants. Information indicating the communication format and information indicating the communication content may be determined based on participant behavior information. Participant behavior information includes the participant's speaking time interval and the participant's speaking content. The participant's speaking time interval is the time interval in which each participant speaks each time in the network meeting. The participant's speaking content is the speaking content that each participant speaks each time in the network meeting. Participant behavior information may be obtained by analyzing meeting data. In some possible implementations, meeting data further includes shared content display data for network meetings. Communication format and communication content may further be determined by the shared content display data. For example, based on the shared content display data, it may be determined that the communication information includes explanatory formatting. Based on the shared content display data, it may be determined that the communication content includes shared content.

[0040] Based on communication information, determine the type of network meeting.

[0041] In several possible implementations, the type of information network conference is determined based on one or more pieces of information, including the number of participants in the communication information, the communication format, and the communication content. As an example, the criteria for determining the conference type are predetermined based on the characteristics of each conference type. The criteria for determining the conference type that the communication information can satisfy are determined, and the conference type that satisfies the criteria is determined as the conference type of the network conference.

[0042] The criteria for determining the interview type may, for example, be that the number of participants is 2 and the communication format is question-and-answer. Based on the information indicating the number of participants and the communication format included in the communication information, it is possible to determine whether a network meeting is of the interview type. For example, a network meeting with 2 participants and a question-and-answer communication format is determined to be of the interview type.

[0043] As another example, the criteria for determining the interview type could include, for instance, that there are two participants, the communication format is such that the ratio of the speaking time each participant spends to the total duration of the network meeting is less than 20%, and the communication content is such that question content accounts for more than 80% of the content spoken by one of the two participants.

[0044] For example, the criteria for determining the share type may be, for instance, that the communication format includes an explanatory format. Based on information indicating the communication format included in the communication information, it is possible to determine whether a network meeting is of the share type. For example, a network meeting whose communication format includes an explanatory format may be determined to be of the interview type.

[0045] As another example, the criteria for determining the sharing type include, for instance, that there are two or more participants, the communication format is such that the ratio of the speaking time of the first participant to the total duration of the network meeting exceeds 60%, and the communication content is such that question content accounts for more than 80% of the second participant's spoken content, and that there is shared content included in the first participant's spoken content.

[0046] Furthermore, for example, the criteria for determining whether a meeting is multitopic may be, for instance, that the communication content includes multiple meeting topics. It is possible to determine whether a network meeting is multitopic based on information indicating the communication content included in the communication information. For example, a network meeting whose communication content includes multiple meeting topics may be determined to be an interview type.

[0047] As another example, the criteria for determining a multitopic type include having two or more participants, a communication format in which at least one participant speaks for more than 30% of the total meeting time, and communication content in which the relevance of participants' spoken content within different speaking time intervals is less than 20%.

[0048] Here, the participant's speaking time is the length of time included in the participant's speaking time interval. The proportion of question content in the participant's speaking content is the proportion of question sentences included in the speaking content.

[0049] In this way, the type of network meeting can be determined based on communication information.

[0050] As another example, an artificial intelligence model is pre-trained to determine the conference type of a network conference. The training data for the AI ​​model consists of, for example, training communication information and labels corresponding to the training communication information, obtained by analyzing training conference data from a training network conference. The training communication information includes one or more of the following: information indicating the number of participants, information indicating the communication format, and information indicating the communication content. The labels corresponding to the training communication information are the conference type of the training network conference. The communication information is input into the pre-trained AI model, and the conference type of the network conference output by the AI ​​model is obtained.

[0051] Second implementation method: The meeting data further includes meeting reservation data for network meetings. The meeting reservation data includes the meeting reservation type. The meeting reservation type may be determined by information about the reserved meeting that participants have entered in advance. For example, the meeting reservation type may be determined based on the title of the reserved meeting. For example, the title of the reserved meeting is "XX Share Meeting". Based on "XX Share Meeting", the meeting reservation type is determined to be a share type. Alternatively, for example, the meeting reservation type may be determined based on the meeting topic of the reserved meeting. For example, the meeting topic of the reserved meeting is "Discussion on Business A, Business B, and Business C". Based on "Discussion on Business A, Business B, and Business C", the meeting reservation type is determined to be a multitopic type. The embodiments of this disclosure do not limit the methods for determining the meeting reservation type. As an example, the meeting reservation type may be determined by analyzing the semantics of the meeting reservation data. As another example, the meeting reservation type may be determined by recognizing keywords contained in the text of the meeting reservation data.

[0052] The meeting reservation type may also serve as reference information for determining the type of network meeting. The type of network meeting is determined based on the communication information and the meeting reservation type.

[0053] One possible implementation involves determining whether the communication information can satisfy the meeting type determination criteria corresponding to the meeting reservation type. If the communication information can satisfy the meeting type determination criteria, the meeting type of the network meeting is determined as that meeting type, i.e., the meeting reservation type. If the communication information cannot satisfy the meeting type determination criteria, the determination criteria for other meeting types that the communication information satisfies are determined.

[0054] In another possible implementation, an artificial intelligence (AI) model is pre-trained to determine the conference type of a network conference. The training data for the AI ​​model includes, for example, training communication information, training conference booking type, and conference type label obtained by analyzing training conference data from a training network conference. The training communication information includes one or more of the following: information indicating the number of participants, information indicating the communication format, and information indicating the communication content. The conference type label is the conference type of the training network conference. The communication information and conference booking type are input into the trained AI model to obtain the conference type of the network conference output by the AI ​​model.

[0055] In some possible cases, a single network meeting may include several meetings belonging to different meeting types. Taking an interview scenario as an example, some meetings within a network meeting may be question-and-answer sessions between the interviewer and the interviewee, while others may be meetings in which the interviewee explains their answers to the interview questions.

[0056] Embodiments of this disclosure provide a possible implementation for segmenting a network conference based on the conference type and the conference content of the network conference, the implementation being: Depending on the type of meeting, the network meeting may be divided into multiple sub-meetings, This includes dividing sub-meetings based on the sub-meeting type and sub-meeting content.

[0057] For network meetings that belong to multiple meeting types, first, divide the network meeting into multiple sub-meetings based on the meeting type. A sub-meeting is a meeting segment of a network meeting that belongs to the same meeting type. For example, if the first 50% of a network meeting belongs to the interview type and the last 50% belongs to the share type, divide the first 50% and the last 50% of the network meeting into two sub-meetings.

[0058] Furthermore, by dividing sub-meetings based on the sub-meeting type and sub-meeting content, segmentation for network meetings can be achieved.

[0059] The segmentation method for sub-conferences is similar to the method for dividing network conferences in S103 above, so a detailed explanation is omitted here. The conference segments of each sub-conference obtained after segmentation may also be used as conference segments for network conferences.

[0060] In this way, more accurate segmentation of network meetings belonging to multiple meeting types can be achieved, resulting in more rational meeting segmentation of different acquired meeting content and improving the convenience for users to review meeting content.

[0061] In several possible scenarios, split meeting segments can be further combined. For example, if the number of meeting segments from a network meeting is relatively large, the split meeting segments can be combined. As an example, one can determine whether the number of meeting segments in a network meeting exceeds a quantity threshold. The quantity threshold is a pre-set maximum number of meeting segments for a network meeting. For example, the quantity threshold is 12 segments per hour. In other words, if the ratio of the number of meeting segments to the length of the network meeting exceeds 12, it indicates that the number of meeting segments exceeds the quantity threshold. For example, a network meeting with a total length of 2 hours has 30 meeting segments, which exceeds the quantity threshold of 24.

[0062] The method for processing meeting data according to the embodiments of this disclosure further includes the following steps:

[0063] Meeting segments are combined according to their meaning. The meaning is the meaning of the main content included in that meeting segment.

[0064] The embodiments of this disclosure are not limited to implementations for determining the semantics of each meeting segment. One possible implementation involves first extracting text data from the meeting data of each meeting segment, inputting the text data into a semantic extraction model, and obtaining the semantic of the meeting segment output from the semantic extraction model. Another possible implementation involves first extracting text data from the meeting data of each meeting segment, extracting keywords from the text data, and using the keywords to indicate the semantic of the meeting segment.

[0065] This process determines the semantic similarity of meeting segments. Meeting segments with adjacent time intervals in a network meeting and whose semantic similarity exceeds a similarity threshold are merged. This reduces the number of meeting segments while maintaining the similarity of the meeting content within each segment, improving the user's convenience in reviewing meeting content.

[0066] Based on the method for processing meeting data according to the above-described embodiment, the embodiments of this disclosure further provide a meeting data processing device. The meeting data processing device will be described below with reference to the drawings.

[0067] Refer to Figure 2. Figure 2 is a schematic diagram of the structure of a conference data processing device according to an embodiment of the present disclosure. As shown in Figure 2, the conference data processing device is The system is configured to acquire meeting data from a network conference, and the meeting data includes a first acquisition unit 201 which contains audio data from the network conference, A first decision unit 202 is configured to determine the type of network conference based on conference data, The system includes a first splitting unit 203 configured to split a network conference based on the conference type and the conference content of the network conference, and to obtain conference segments of the network conference.

[0068] In one possible implementation, the first decision unit 202 is configured to specifically acquire communication information for a network conference based on meeting data and to determine the conference type of the network conference based on the communication information.

[0069] In one possible implementation, communication information is, This includes one or more of the following: information indicating the number of participants, information indicating the communication format, and information indicating the communication content.

[0070] In one possible implementation, the meeting data further includes shared content display data for the network meeting, and the shared content display data is used to determine information indicating the communication format and information indicating the communication content.

[0071] In one possible implementation, the communication information includes information indicating the communication format, and the first decision unit 202 is configured to determine the conference type of the network conference based on the communication information. The first decision unit 202 is configured to determine the meeting type of a network meeting as a share type if the information indicating the communication format indicates that the communication format includes an explanatory format.

[0072] In one possible implementation, the communication information includes information indicating the number of participants and information indicating the communication format, and the first decision unit 202 is configured to determine the type of network meeting based on the communication information. The first decision unit 202 is configured to determine the type of network meeting as an interview type if the information indicating the number of participants indicates that there are two participants, and the information indicating the communication format indicates that the communication format is question and answer.

[0073] In one possible implementation, the communication information includes information indicating the communication content, and the first decision unit 202 is configured to determine the type of network conference based on the communication information. The first decision unit 202 is configured to determine the conference type of a network conference as a multitopic type if the information indicating the communication content indicates that the communication content includes multiple conference topics.

[0074] In one possible implementation, the meeting data further includes the meeting reservation type of the network meeting, and the first decision unit 202 specifically obtains the communication information and meeting reservation type of the network meeting based on the meeting data. It is configured to determine the type of network meeting based on communication information and meeting reservation type.

[0075] In one possible implementation, the first splitting unit 203 is configured to split the network conference based on the conference type and the conference content of the network conference. The first division unit 203 is configured to divide a network conference into multiple sub-conferences according to the conference type, where each sub-conference is a conference segment of the network conference belonging to the same conference type, and the sub-conferences are divided based on the conference type and conference content of each sub-conference.

[0076] In one possible implementation, the first splitting unit 203 is configured to split the network conference based on the conference type and the conference content of the network conference. The first segmentation unit 203 is configured to segment network conferences based on the conference content of the network conference using segmentation rules corresponding to the conference type.

[0077] In one possible implementation, the first splitting unit 203 is configured to split the network conference based on the conference type and the conference content of the network conference. The first segmentation unit 203 is configured to process network conference data using a segmentation model corresponding to the conference type, wherein the segmentation model is for segmenting the network conference.

[0078] In one possible implementation, the processing unit for the conference data is: The system further includes a joining unit configured to determine the meaning of each meeting segment and to join meeting segments whose time intervals in a network meeting are adjacent and whose semantic similarity is equal to or greater than a similarity threshold.

[0079] Media content is content that is presented through various distribution methods. Because media content can be distributed via the internet, it is convenient for users to access it online. Some media content contains a relatively large amount of other content. If a user is interested in a specific piece of media content, they cannot determine the location of that specific piece within the media content and thus access it without first reviewing the other content included. This leads to a diminished user experience when reviewing media content.

[0080] Based on this, embodiments of the present disclosure provide a method, apparatus, device, and medium for segmenting media content. In this method, data of media content is acquired, the media content includes data content of at least two content dimensions, a target segmentation point of the media content is determined based on the content dimensions and the media content data, and the sub-content of the media content is determined using the target segmentation point. By determining the target segmentation point of the media content based on multiple content dimensions, segmentation of media content based on the characteristics of different content dimensions can be achieved, and sub-content of the segmented media content can be obtained more accurately. In this way, it is made easier for the user to understand some of the content contained in the media content by checking the sub-content, improving the user experience.

[0081] The media content segmentation method according to the embodiments of this disclosure may be applied to electronic devices having data processing capabilities. These electronic devices may be, for example, servers or terminals. Terminals include, but are not limited to, smartphones, tablet computers, laptop computers, personal digital assistants (PDAs), or smart wearable devices. Servers may be cloud servers, such as a central server in a central computing cluster or an edge server in an edge computing cluster. Naturally, servers may also be servers in a local data center. A local data center refers to a data center directly controlled by the user.

[0082] The electronic device acquires data from media content, which includes data content from at least two content dimensions. Based on the content dimensions and the media content data, the device determines the target division point of the media content and uses the target division point to determine the sub-content of the media content.

[0083] As those skilled in the art will understand, the above application scenarios are merely examples of what embodiments of the present disclosure may achieve. The scope of application of embodiments of the present disclosure is not limited in any way by this framework.

[0084] To facilitate understanding of the technical proposals relating to the embodiments of this disclosure, the method for dividing media content relating to the embodiments of this disclosure will be explained below with reference to the drawings.

[0085] Refer to Figure 3. Figure 3 is a flowchart of a media content division method according to an embodiment of the present disclosure. As shown in Figure 3, the media content division method may include steps S301 to S303.

[0086] S301: Retrieve data for media content, where the media content includes content of at least two content dimensions.

[0087] Media content is content generated based on a distribution method. The embodiments of this disclosure do not limit the specific type of media content. For example, media content may be movie video content, or media content may be live video content, or media content may be conference content.

[0088] Media content data may include one or more types of data from video data, audio data, text data, and image data. The media content data is specifically determined based on the type of media content.

[0089] Media content includes content of multiple content dimensions. The embodiments of this disclosure do not limit the classification method of content dimensions. For example, different content dimensions may be classified according to different methods of generating the content. Another example is classifying different content dimensions according to different data types.

[0090] As an example, let's consider a case where media content is conference content. The content dimensions included in media content are: Includes one or more of the following dimensions: audio content dimension, shared screen content dimension, shared document content dimension, stage time content dimension, and stage content dimension.

[0091] A detailed explanation of these five content dimensions will follow later.

[0092] S302: Determine the target segmentation point for media content based on content dimensions and media content data.

[0093] Content of different content dimensions has different characteristics. Based on the characteristics of the multiple content dimensions contained in the media content, the data of the media content is processed using a partitioning method corresponding to each content dimension, and the division points for partitioning the media content are determined. The division points of the media content are used to mark the positions where the media content is to be divided. For example, if the media content data includes audio data or video data, the division point of the media content is the time on the media content's time axis when the media content needs to be divided. For example, if the media content data includes text data, the division point of the media content is the delimiter that divides the text in the text data.

[0094] Taking a scenario in which media content is meeting content as an example, the embodiments of this disclosure provide an audio content dimension, a shared screen content dimension, a shared document content dimension, a stage time content dimension, and a stage content dimension, respectively. Details of the method for determining the division points of media content using these five content dimensions may be found below.

[0095] There may be multiple points of division in media content, determined based on multiple content dimensions. From these multiple points of division, at least one target point is determined as the point of division from which to acquire media content and content.

[0096] The embodiments of this disclosure are not limited to possible implementation methods for determining the target division point.

[0097] In one possible implementation, each division point determined by multiple content dimensions is considered a target division point for media content.

[0098] In other possible implementations, the division points determined based on each content dimension are all considered candidate division points for media content. A target division point is then selected from these candidate division points for media content.

[0099] The embodiments of this disclosure are not limited to implementation methods for selecting a target division point from candidate division points.

[0100] As an example, multiple candidate split points that satisfy the joining conditions are joined together as a single target split point. Candidate split points that do not satisfy the joining conditions are designated as the target split point.

[0101] The joining condition is, for example, that the distance between the division points of the media content marked by candidate division points is less than a threshold. When the deviation of the division points for dividing the media content, indicated by multiple candidate division points that satisfy the joining condition, is relatively small, it indicates that the division accuracy at that division point is relatively high. For example, if the media content data includes audio data or video data, and the candidate division points of the media content are the times on the media content's time axis when the media content needs to be divided, the joining condition is, for example, that the time interval between candidate division points is less than 5 minutes. The embodiments of this disclosure do not limit the method of joining candidate division points. As an example, one of the candidate division points that satisfy the joining condition is selected as the target division point obtained by joining candidate division points that satisfy the joining condition. Alternatively, candidate division points that do not satisfy the joining condition may be used as target division points. In this way, the number of target division points can be reduced and the number of sub-contents can be reduced while ensuring the accuracy of media content division, and this can be easily verified by the user.

[0102] As another example, the segmentation confidence of each candidate split point is determined. Segmentation confidence evaluates the segmentation accuracy of the candidate split point and also indicates the reliability of splitting media content from that candidate split point.

[0103] The segmentation confidence level may be determined according to pre-defined segmentation confidence level setting rules.

[0104] For example, segmentation confidence is determined by the segmentation method of the candidate segment. For example, segmentation confidence is determined by the dimension type of the content dimension corresponding to the candidate segment. Using the five content dimensions mentioned above as examples, candidate segmentations determined based on the audio content dimension and the shared content dimension, i.e., the shared screen content dimension and the shared document content dimension, have relatively high segmentation confidence. Candidate segmentations determined based on the descriptive content dimension, i.e., the stage time content dimension and the stage content dimension, have relatively low segmentation confidence. Also, for example, segmentation confidence is determined based on the accuracy of the candidate segment determined by the segmentation method. For example, the accuracy of a candidate segment determined by the descriptive content dimension is determined by the similarity of the content before and after the candidate segment. If the similarity is high, the accuracy of the candidate segment is relatively low, and the segmentation confidence is low. If the similarity is low, the accuracy of the candidate segment is relatively high, and the segmentation confidence is high.

[0105] For example, segmentation confidence is determined by the segmentation granularity of the candidate segmentation point. Segmentation granularity refers to the granularity at which the media content of the candidate segmentation point is divided. Segmentation granularity may also be determined based on the type of content dimension. For example, using the five content dimensions mentioned above, candidate segmentation points determined based on the audio content dimension and the shared content dimension, i.e., the shared screen content dimension and the shared document content dimension, are fine-grained candidate segmentation points. Candidate segmentation points determined based on the descriptive content dimension, i.e., the stage time content dimension and the stage content dimension, are coarse-grained candidate segmentation points. Coarse-grained candidate segmentation points may have problems with inaccurate segmentation, and therefore have relatively low segmentation confidence. Fine-grained candidate segmentation points have relatively high segmentation accuracy and therefore relatively high segmentation confidence.

[0106] The two methods for determining segmentation confidence described above are merely examples of how to determine segmentation confidence, and the embodiments of this disclosure are not limited thereto. Furthermore, the multiple methods for determining segmentation confidence may be used independently or simultaneously. If a single candidate segmentation point has multiple segmentation confidence values, the weighted values ​​of each segmentation confidence value for that candidate segmentation point may be calculated as the segmentation confidence for that candidate segmentation point. The weight of each segmentation confidence value may be determined by the method for determining segmentation confidence. For example, the weight of a segmentation confidence value determined by a segmentation method is greater than the weight of a segmentation confidence value determined by segmentation granularity.

[0107] After determining the segmentation confidence of each candidate segmentation point, the segmentation confidence of each candidate segmentation point is sorted in descending order, and the candidate segmentation points that are sorted above a predetermined rank are selected as the target segmentation point. Alternatively, candidate segmentation points whose segmentation confidence exceeds the confidence threshold are designated as high-priority candidate segmentation points, and candidate segmentation points whose segmentation confidence is below the confidence threshold are designated as low-priority candidate segmentation points. The high-priority candidate segmentation points are selected as the target segmentation points.

[0108] S303: Determine the sub-content of media content based on the target segmentation point.

[0109] The media content is divided according to the target segmentation point, and the sub-content of the media content is obtained. For example, if the media content is meeting content, the sub-content would be meeting segment content.

[0110] As can be seen from the above S301-S303, the target division points of media content determined by the content dimension make it possible to divide media content based on the characteristics of the content dimension, the content relevance of the acquired sub-content is relatively high, and the division of media content is more accurate. In this way, users can understand the content of a sub-content by checking that sub-content, thereby enabling quick browsing of media content and improving the user experience by eliminating the need to carefully check the complete media content.

[0111] The following describes possible implementation methods for determining the division points of media content based on five content dimensions: audio content dimension, shared screen content dimension, shared document content dimension, stage time content dimension, and stage content dimension, as described in the embodiments of this disclosure.

[0112] In some possible implementations, the data of media content includes audio data. The media content includes audio content, and at least two content dimensions related to the media content include audio content dimensions.

[0113] The first division point for audio content dimensions is determined using the following method.

[0114] A1: Based on the audio data, retrieve the content of the participant's statements as the facilitator.

[0115] Audio data included in media content may be the audio data of people who participated in the creation of the media content. For example, the media content may be meeting content. The people are the participants in the meeting. The people may include the facilitator. The facilitator is responsible for guiding the communication process of the content. For example, the facilitator is the meeting moderator or organizer.

[0116] The embodiments of this disclosure are not limited to methods for identifying the person who is the facilitator.

[0117] One possible implementation is to determine the pre-assigned moderator based on the information set for each person in the media content. For example, the moderator could be determined based on the person's name or type. For instance, a person whose type is "host" could be designated as the moderator.

[0118] In other possible implementations, the facilitator's spoken content has the characteristic of guiding the process. The person who will act as facilitator is determined by analyzing the audio data contained in the media content.

[0119] As an example, the system analyzes the spoken content of individuals within media content. By detecting words related to the process guide or words with similar meanings to those related to the process guide within the spoken content, the system identifies the facilitator. The words related to the process guide may be pre-defined reference words. For example, reference words might include, "I will be leading today's meeting," or "First, I would like to discuss..." In some scenarios, the system analyzes the spoken content of individuals whose speaking order falls within a sequence threshold, based on the characteristic that the facilitator speaks earlier in the sequence. The sequence threshold may be determined based on the number of people included in the media content. For example, for media content containing 5 people, the sequence threshold might be 3. In this way, the scope for detecting the facilitator can be reduced, thereby lowering costs.

[0120] As another example, a recognition model capable of recognizing the identity of a person in media content is pre-trained. The recognition model can determine the identity of a person in media content using input audio data or text data obtained by recognizing the audio data. Audio data or text data obtained by recognizing the audio data is input into the recognition model, and the person who is the facilitator is obtained from the recognition model's output.

[0121] A2: Determine the first break point of the media content based on the different process stages indicated by the spoken content.

[0122] The facilitator's spoken content includes content that indicates different process stages of the media content. For example, semantic recognition of the facilitator's spoken content is performed to determine terms that indicate different process stages. These terms include, for example, process terms and summary terms. Based on the different process stages indicated by the facilitator's spoken content, the first dividing point of the media content for separating the different process stages is determined. For example, the time of the switch between different process stages is used as the first dividing point of the media content. The first dividing point is used to determine the target dividing point.

[0123] The methods for determining the first division point for the audio content dimensions described above are merely examples, and the embodiments of this disclosure are not limited thereto. For example, the first division point may be determined based on the timing of pauses in human speech in the audio data.

[0124] In some possible implementations, the data of media content includes shared data. The media content includes shared content, and at least two content dimensions related to the media content include shared content dimensions.

[0125] For example, shared data includes one or more of the following: shared screen data and shared document data. Shared screen data is, for example, video data generated by the screen of a shared device or the interface displayed on the screen of a shared device. Shared screen data can reflect the content of a screen shared and displayed by a person. Different screen content can reflect different content stages in media content. Shared document data is video data generated by a shared document. Shared document data can reflect the content of a shared document. Different document content displayed can reflect different content stages in media content.

[0126] If the shared data includes shared screen data, the shared content dimension includes the shared screen content dimension. If the shared data includes shared document data, the shared content dimension includes the shared document content dimension.

[0127] The second division point for the shared screen content dimension is determined using the following method.

[0128] B1: Determine the timing of content changes on the shared screen based on the shared screen content.

[0129] Switching the content of a shared screen can reflect changes in the content stages contained within the media content. The embodiments of this disclosure are not limited to the method for determining the switching of content on a shared screen. As an example, images displayed on the shared screen within different time intervals are obtained from the shared screen data. Then, the similarity of the images displayed on the screen in adjacent time intervals is determined to determine whether the content on the shared screen has changed. The time at which the similarity of the images is less than a similarity threshold is set as the content switching time on the shared screen. As another example, operations on the cursor on the shared screen are detected based on the shared screen data, and the content switching time on the shared screen is determined. For example, the time when the cursor is operated and a page switching button is clicked is set as the content switching time on the shared screen. As yet another example, the start time of screen sharing and the end time of screen sharing are set as the content switching times on the shared screen.

[0130] B2: Determine the second break point for media content based on the content switching time on the shared screen.

[0131] For example, the time of the content changeover on the determined shared screen is used as the second division point for splitting the media content. Alternatively, a portion of the time of the content changeover on the determined shared screen is used as the second division point for splitting the media content. The second division point is used to determine the target division point.

[0132] For shared document content, the third division point will be determined using the following method.

[0133] C1: Determine the document content switch time based on the shared document content.

[0134] The title of a shared document may reflect different content involved in the process of sharing the document. The title of a shared document may be determined based on processing the shared document data. In one possible implementation, the position of the user cursor in the document's display area is detected based on the shared document data. Based on the position of the user cursor in the document's display area, the content of the currently shared document is determined. For example, if it is detected that the user cursor is selecting a document title, or is within the display area of ​​the document title, or that the user cursor has moved to the display area of ​​content corresponding to a different document title, it is determined that the document content has changed, and that time is taken as the document content switch time. In another possible implementation, in a shared document, the document content being shared is displayed in a special display format. Based on the shared document data, the time of change in the title of the document to which the specially displayed document content belongs is taken as the document content switch time.

[0135] C2: Determine the third break point for media content based on the document content switch time.

[0136] For example, the transition time of the determined document content is designated as the third division point for media content. Alternatively, a portion of the transition times from the determined document content transition times are selected as the third division point for media content. The third division point is used to determine the target division point.

[0137] In some possible implementations, media content includes descriptive content. At least two content dimensions related to media content include descriptive content dimensions. Descriptive content is content that describes the media content. The data of media content includes descriptive data. Descriptive data is, for example, text data. For example, the media content is meeting content, and the descriptive content is meeting agenda content.

[0138] For example, descriptive data includes one or more of stage time data and stage content data. Stage time data includes time information for different content stages in media content. Stage content data includes information for the main content of different content stages in media content.

[0139] Taking the example where the descriptive data is meeting agenda data, the stage time data includes time information for different content stages in the meeting content, for example, the first 30 minutes of the meeting are dedicated to discussing topic a, then topic b. Or, for example, 4:00-5:00 is dedicated to communicating about issue x, and 5:00-5:30 is dedicated to communicating about issue y. The stage content data includes information for the main content of different content stages in the meeting content, for example, the meeting first discusses topic a, and secondly discusses topic b.

[0140] If the descriptive data includes stage-time data, the descriptive content dimension includes the stage-time content dimension. If the descriptive data includes stage-time data, the descriptive content dimension includes the stage-time content dimension.

[0141] The fourth division point for the stage time content dimension is determined using the following method.

[0142] Based on the time indicated by the chronological content, the fourth division point of the media content is determined.

[0143] In one possible implementation, each time point indicated by the time-staged content is designated as the fourth division point of the media content. In another possible implementation, the main content of the media content in the time interval preceding each time point indicated by the time-staged content is determined, as is the main content of the media content in the time interval following each time point. The length of the preceding and succeeding time intervals may be predetermined. If the content similarity between the main content of the media content in the time interval preceding a certain time point and the main content of the media content in the time interval succeeding to a certain time point is below a threshold, that time point is designated as the fourth division point of the media content. If the content similarity between the main content of the media content in the time interval preceding a time point and the main content of the media content in the time interval succeeding to a certain time point exceeds the threshold, that time point is dropped.

[0144] The fifth division point for the staged content dimension is determined using the following method.

[0145] Content clustering is performed on media content based on the staged content, and the fifth division point of the media content is determined based on the different clustered content.

[0146] The staged content data indicates the main content of each content stage contained within the media content. Based on the media content data, the media content is clustered according to the main content indicated by the staged content data. In one possible implementation, the media content data includes audio data, which is converted to text data. Clustering of the media content is achieved by performing semantic clustering on the words contained in the text data. The clustering categories are the categories of different main content indicated by the staged content data. After obtaining the clustered different content, a fifth division point of the media content is determined. The fifth division point is used to divide the clustered different content contained within the media content. The fifth division point is used to determine the target division point.

[0147] The above are possible implementations for determining division points based on the five content dimensions of the embodiments of this disclosure. All of the above implementations for determining division points are illustrative and not limited to determining division points based on content dimensions.

[0148] Based on the media content splitting method according to the above embodiment of the method, the embodiments of this disclosure further provide a media content splitting device, which will be described below with reference to the drawings.

[0149] Refer to Figure 4. Figure 4 is a schematic diagram of the structure of a media content splitting device according to an embodiment of the present disclosure. As shown in Figure 4, the media content splitting device is The system is configured to acquire data from media content, and the media content includes a second acquisition unit 401 which contains content of at least two content dimensions, A second decision unit 402 is configured to determine the target division point of media content based on content dimensions and media content data, It includes a second division unit 403 configured to determine sub-content of media content based on a target division point.

[0150] In one possible implementation, the second decision unit 402 is configured to determine candidate division points for media content in the content dimension based on the division scheme and media content data corresponding to the content dimension, and to determine target division points for media content based on the candidate division points.

[0151] In one possible implementation, the second decision unit 402 is configured to determine the target division point of the media content based on candidate division points. The second decision unit 402 is configured to determine the target segmentation point of media content based on the segmentation confidence of candidate segmentation points, wherein the segmentation confidence is used to evaluate the accuracy of segmentation by the candidate segmentation points.

[0152] In one possible implementation, segmentation confidence is determined based on the segmentation method that determines the candidate segmentation points.

[0153] In one possible implementation, segmentation confidence is determined based on the segmentation granularity of candidate segmentation points, which indicates the degree of subdivision of media content by the candidate segmentation points.

[0154] In one possible implementation, media content includes audio content, content dimension includes audio content dimension, and with respect to the audio content dimension, Based on the audio data, we retrieve the spoken content of the presenter included in the media content. The method used to determine the first dividing point of media content is to determine the first dividing point based on the different process stages indicated by the spoken content, and the first dividing point is used to determine the target dividing point.

[0155] In one possible implementation, media content includes shared content.

[0156] In one possible implementation, shared content includes shared screen content, content dimension includes shared screen content dimension, and with respect to the shared screen content dimension, Based on the shared screen content, the timing of content changes on the shared screen is determined. The second division point of media content is determined using a method that determines the second division point based on the content switching time of the shared screen. This second division point is used to determine the target division point.

[0157] In one possible implementation, shared content includes shared document content, content dimension includes shared document content dimension, and with respect to the shared document content dimension, Based on the shared document content, determine the document content switching time. The third division point of media content is determined using a method that determines the third division point based on the document content switching time, and this third division point is used to determine the target division point.

[0158] In one possible implementation, media content includes descriptive content.

[0159] In one possible implementation, the descriptive content includes the phase-time content, the phase-time content is the phase-time content dimension, and with respect to the phase-time content dimension, The fourth division point of the media content is determined using a method that determines the time indicated by the time-series content, and this fourth division point is used to determine the target division point.

[0160] In one possible implementation, the descriptive content includes the stage content, the stage content is the stage content dimension, and with respect to the stage content dimension, The method involves performing content clustering on media content based on tiered content, and then determining the fifth division point of the media content based on the different clustered content. The fifth division point is used to determine the target division point.

[0161] In one possible implementation, the media content is the conference content, and the sub-content of the media content is the conference segment content.

[0162] Media content refers to content presented through various distribution methods. Media content includes one or more of the following: video content, image content, audio content, and text content. Because media content can be distributed via the internet, it is convenient for users to access it online. Currently, there is a relatively large amount of media content available to users, and some media content contains relatively rich material. Therefore, to facilitate user access, summaries describing the main content of the media content can be provided. By reviewing these summaries, users can quickly grasp the specific content contained within the media content, making it easier for them to select and access the media content they need, or enhancing their understanding of the media content. However, traditional media content summaries often struggle to accurately describe the main content contained within the media content.

[0163] Based on this, embodiments of the present disclosure provide a summary generation method, apparatus, device, and medium. In the method, content data of each sub-content contained in media content is obtained, and a summary of each sub-content is obtained based on the content data of each sub-content. The sub-content summary can describe the main content of the sub-content. By extracting the sub-content summaries in advance, the difficulty of generating a summary of media content can be reduced. Based on the weight of each sub-content, the summaries of each sub-content are merged to obtain a summary of media content. Here, the weight of the sub-content can reflect the importance of the sub-content in the media content. In this way, by merging the summaries of the sub-content by referring to the importance of each sub-content in the media content, it is possible not only to reduce the omission of important content but also to avoid over-description of unimportant content, and the obtained summary of media content can more accurately describe the main content of the media content, making it easier for users to grasp the media content through the summary.

[0164] The summary generation method relating to the embodiments of this disclosure may be applied to electronic devices having data processing capabilities. The electronic devices may be, for example, servers or terminals. Terminals include, but are not limited to, smartphones, tablet computers, laptop computers, personal digital assistants (PDAs), or smart wearable devices. Servers may be cloud servers, such as a central server in a central computing cluster or an edge server in an edge computing cluster. Naturally, servers may also be servers in a local data center. A local data center refers to a data center directly controlled by the user.

[0165] The electronic device retrieves content data for each sub-content contained within the media content, obtains a summary of each sub-content based on its content data, and then merges these summaries based on their respective weights to obtain a summary of the media content. Here, the weight of the sub-content indicates the importance of the sub-content within the media content. By first extracting the summaries of the sub-content and then merging them to obtain a summary of the media content, the difficulty of generating the media content summary can be reduced, making summary generation easier. By merging the summaries of the sub-content based on the importance of each sub-content within the media content, it is possible to reduce the omission of important content and avoid over-description of unimportant content. The resulting media content summary can more accurately summarize the main content of the media content, meeting the needs of users who review media content summaries.

[0166] As those skilled in the art will understand, the above application scenarios are merely examples of what embodiments of the present disclosure may achieve. The scope of application of embodiments of the present disclosure is not limited in any way by this framework.

[0167] To facilitate understanding of the technical proposals relating to the embodiments of this disclosure, the summary generation method relating to the embodiments of this disclosure will be described below with reference to the drawings.

[0168] Refer to Figure 5. Figure 5 is a flowchart of the summary generation method according to an embodiment of the present disclosure. As shown in Figure 5, the summary generation method may include steps S501 to S503.

[0169] S501: Retrieves content data for each sub-content included in the media content.

[0170] Media content includes, but is not limited to, one or more of the following: video content, image content, audio content, and text content.

[0171] Sub-content is a portion of media content obtained by splitting it. Media content contains at least two sub-contents.

[0172] The embodiments of this disclosure do not limit the method of sub-content segmentation. In one possible implementation, media content is divided equally by time length to obtain each sub-content. For example, if the media content is video content, the video content is divided into segments of 20 minutes each to obtain multiple sub-contents. In another possible implementation, the content contained in the media content is analyzed, a clustering operation is performed on the content, and the media content is segmented based on similar types of content. In this way, a portion of the content with relatively high content relevance can be segmented into a single sub-content. The sub-content obtained in this way matches the actual content structure of the media content, allowing for the extraction of a more accurate summary of the sub-content and, furthermore, the generation of a more accurately summarized summary of the media content.

[0173] In one possible implementation, the media content is the content of the meeting. For example, the sub-content is the content of the sub-meetings, which are obtained by dividing the meeting.

[0174] The embodiments of this disclosure are not limited to the method of acquiring meeting data.

[0175] As an example, a meeting can be divided based on its meeting type and meeting content to obtain sub-meetings. Here, the meeting type may be determined based on the meeting's communication information. This communication information may include, for example, one or more of the following: information indicating the number of participants, information indicating the communication format, and information indicating the communication content. Sub-meetings can be obtained by dividing the meeting using a division model corresponding to the meeting type, or by dividing the meeting using division rules corresponding to the meeting type to determine the sub-meetings.

[0176] As another example, the content of a meeting includes content from at least two content dimensions. The meeting is divided based on the content dimensions to determine sub-meetings. In one possible implementation, a division scheme corresponding to the content dimensions is used to determine candidate division points of the meeting and the segmentation confidence of those candidate division points. A target division point is determined based on the segmentation confidence, and the meeting is divided using the target division point to obtain sub-meetings. In another possible implementation, the meeting is divided using a division scheme corresponding to the content dimensions to obtain meeting segments. Furthermore, meeting segments that meet certain criteria are combined to obtain sub-meetings. Meeting segments that meet the criteria are, for example, meeting segments whose time intervals in the meeting are adjacent and whose content similarity exceeds a similarity threshold.

[0177] Alternatively, the meeting is a regular scheduled meeting. The subcontent is the content of at least one scheduled meeting that is included in the regular scheduled meeting. For example, the meeting is a regular meeting held every Monday afternoon. The subcontent is the content of at least one Monday afternoon meeting that is included in the said regular meeting.

[0178] The content data of sub-content is data related to the sub-content. For example, if the sub-content is a sub-meeting, the content data of the sub-meeting is meeting data. The embodiments of this disclosure do not limit the specific type of content data of sub-content. For example, the content data of sub-content may be one or more of audio data, video data, image data, and text data.

[0179] S502: Determine the summary of each sub-content based on the content data of each sub-content.

[0180] A subcontent summary is intended to describe the main content of the subcontent. The embodiments of this disclosure are not limited to implementations that determine the summary of each subcontent based on the content data of each subcontent. One possible implementation determines the keywords of the subcontent by analyzing the content data of the subcontent. A summary of the subcontent is generated based on the keywords. Another possible implementation uses a second language processing model to process the content data of the subcontent and obtain a summary of the subcontent. The second language processing model has the capability to process natural language. The second language processing model can analyze the input content data and output a summary.

[0181] Because the amount of content data for each sub-content within media content is relatively small, it is easy to process the content data of the sub-content and obtain a summary of the sub-content. The cost of generating the sub-content summary is relatively low, and the accuracy of summarizing the sub-content is relatively high, which can improve the accuracy of the obtained media content summary.

[0182] S503: Based on the weight of each sub-content, the summaries of each sub-content are merged to obtain a summary of the media content.

[0183] Different sub-contents have different levels of importance within the media content. The importance of sub-content within the media content is indicated by its weight.

[0184] The embodiments of this disclosure do not limit the method for determining the weights of subcontent.

[0185] In one possible implementation, the weights of the sub-contents may be set, for example, by the creator of the media content. For example, if the media content is generated based on a network conference, the weight of each sub-content in the media content may be set, for example, by the organizer, moderator, or other person with conference management authority over the network conference.

[0186] In other possible implementations, the weight of sub-content is determined based on the sub-content's content information. For example, the sub-content's content information is determined by the sub-content's temporal information and content data within the media content. For instance, the sub-content's content information includes the sub-content's temporal information within the media content and relevant information about the specific content determined by the content data. The temporal information of the sub-content within the media content can reflect to some extent the importance of the sub-content within the media content. For example, sub-content at the beginning of media content is usually the introductory part and therefore has relatively low importance. Conversely, sub-content at the end of media content is usually the concluding part and therefore has relatively high importance. By analyzing the content data, the specific content of the sub-content can be obtained, and its importance and weight can be determined.

[0187] As an example, an artificial intelligence model is pre-trained to determine the weights of sub-content. For instance, the AI ​​model is trained using training data that includes training content information and labels for the training content information. The labels for the training content information are the weights of the training content information. The trained AI model can determine the weights of sub-content based on the content information of the input sub-content. As another example, the labels for the training content information are the importance values ​​of the training content information. The importance values ​​are used to evaluate the importance of the training content information. The trained AI model can determine the importance values ​​of sub-content based on the content information of the input sub-content. Furthermore, the weights of the sub-content are determined based on the weights corresponding to the importance values ​​of the sub-content.

[0188] As another example, the content information of the sub-content is analyzed to obtain sub-information of at least one dimension that can determine the weight of the sub-content. Embodiments of this disclosure do not limit the methods of partitioning the dimensions included in the content information. For example, the sub-information included in the content information may be one or more of the following: the duration of the sub-content, the number of people involved in the sub-content, and the position of the sub-content's time interval within the media content's time interval. Here, the duration of the sub-content and the position of the sub-content's time interval within the media content's time interval may be determined by analyzing the sub-content's time information in the media content. The number of people involved in the sub-content may be obtained by analyzing the sub-content's content data. In another possible implementation, in a scenario where the media content is the content of a meeting, the people involved in the sub-content are participants, and the sub-information further includes the participants' meeting status in the meeting. Meeting status may be determined by the participant's speaking style, speaking frequency, and other speaking characteristics included in the sub-content's content data. For example, if the speaking style is summarizing the event and the speaking frequency is relatively high, the participant may be determined to be a speaker. If a participant's speaking style is consistent with the progress of the process and they speak relatively frequently, that participant will be designated as the moderator.

[0189] Each sub-information has a corresponding sub-weight. The sub-weight of each sub-information may be determined, for example, according to a pre-defined rule for determining sub-weights. For example, sub-content at the beginning of media content may have a relatively low sub-weight, while sub-content at the end of media content may have a relatively high sub-weight.

[0190] When meeting information contains multiple sub-information items of different dimensions, the weight of the sub-content is determined based on the sub-weights of each sub-information item. For example, the statistical values ​​of the sub-weights of each sub-information item contained in the sub-content are calculated as the weight of the sub-content. These statistical values ​​are numerical values ​​obtained using data statistics methods, such as the mean, weighted mean, and median.

[0191] The weight of each sub-content can influence the proportion of the sub-content summaries included in the media content summary. The summaries of each sub-content are merged based on their respective weights to obtain the media content summary.

[0192] The embodiments of this disclosure are not limited to implementations that obtain a summary of media content by fusing the summaries of each subcontent based on the weight of each subcontent. In one possible implementation, a first language processing model is used to process the weights of each subcontent and the summaries of each subcontent to obtain a summary of media content. The first language processing model has the capability to process natural language. The first language processing model may be used to fuse the summaries based on the summaries and the weights of each summary to output a summary. In other possible implementations, rules for fusing summaries are predetermined. The rules for fusing summaries define a method for fusing summaries with different weights. The summaries of each subcontent are processed according to the rules for fusing summaries to obtain a summary of media content.

[0193] As can be seen from the above S501-S503, by first extracting summaries of sub-content and then fusing these summaries to obtain a summary of media content, the difficulty of generating a summary of media content can be reduced, making summary generation easier. By fusing the summaries of sub-content while referring to the importance of each sub-content in the media content, it is possible not only to reduce the omission of important content but also to avoid over-description of unimportant content. The resulting summary of media content can more accurately describe the main content of the media content, meet the needs of users who are reviewing the summary of media content, and improve the user experience.

[0194] Based on the summary generation method according to the above embodiment, the embodiments of this disclosure further provide a summary generation device, which will be described below with reference to the drawings.

[0195] Refer to Figure 6. Figure 6 is a schematic diagram of the structure of a summary generation device according to an embodiment of the present disclosure. As shown in Figure 6, the summary generation device is A third acquisition unit 601 is configured to acquire content data for each sub-content contained in media content, and the sub-content is acquired by dividing the media content, A third decision unit 602 is configured to determine a summary of each sub-content based on the content data of each sub-content, The system includes a generation unit 603 configured to obtain a summary of media content by fusing summaries of each subcontent based on the weight of each subcontent, wherein the weight of the subcontent indicates the importance of the subcontent in the media content.

[0196] In one possible implementation, the weight of the subcontent is determined by the content information of the subcontent, which is determined by the time information of the subcontent and the content data of the subcontent in the media content.

[0197] In one possible implementation, the weights of subcontent are determined based on the subweights corresponding to the subinformation contained within the subcontent's content information.

[0198] In one possible implementation, content information includes one or more sub-information items, such as the duration of the sub-content, the number of people involved in the sub-content, and the position of the sub-content's time segment within the media content's time segment.

[0199] In one possible implementation, the weights of the sub-content are determined based on an artificial intelligence model, which outputs weights based on the input content information.

[0200] In one possible implementation, the generation unit 603 is configured to specifically process the weights of each subcontent and the summaries of each subcontent based on a first language processing model to obtain a summary of the media content.

[0201] In one possible implementation, the content data is text data, and the third decision unit 602 is configured to process the content data of each sub-content based on a second language processing model to obtain a summary of each sub-content.

[0202] In one possible implementation, the media content is the content of the meeting, and the sub-content is the content of a sub-meeting obtained by splitting the meeting, or the meeting is a regular scheduled meeting, and the sub-content is the content of at least one regular meeting included in the regular scheduled meeting.

[0203] One possible implementation is to split the meeting and retrieve the sub-meetings by dividing the meeting based on the meeting type and meeting content of the meeting.

[0204] In one possible implementation, sub-meetings are obtained by splitting the meeting, which is done by dividing the meeting based on at least two content dimensions of the meeting.

[0205] Based on the meeting data processing method, media content segmentation method, or summary generation method according to the above embodiment, the present disclosure further provides an electronic device comprising one or more processors and a storage device storing one or more programs, wherein when one or more programs are executed by one or more processors, the one or more processors implement the meeting data processing method, media content segmentation method, or summary generation method according to any one of the above embodiments.

[0206] Refer to Figure 7 below. Figure 7 shows a schematic diagram of the structure of an electronic device 700 suitable for realizing an embodiment of the present disclosure. The terminal devices in the embodiment of the present disclosure may include, but are not limited to, mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (portable Android devices, tablet computers), PMPs (Portable Media Players), and in-vehicle terminals (e.g., in-vehicle navigation terminals), as well as fixed terminals such as digital TVs (televisions) and desktop computers. The electronic device shown in Figure 7 is merely an example and does not limit the functions and scope of use of the embodiment of the present disclosure in any way.

[0207] As shown in Figure 7, the electronic device 700 includes a processing unit (e.g., a central processing unit, a graphics processing unit, etc.) 701 which may perform various appropriate operations and processes based on a program stored in a read-only memory (ROM) 702 or a program loaded from a storage device 708 into a random access memory (RAM) 703. The RAM 703 further stores various programs and data necessary for the operation of the electronic device 700. The processing unit 701, ROM 702, and RAM 703 are connected to each other via a bus 704. An input / output (I / O) interface 705 is also connected to the bus 704.

[0208] Typically, input devices 708, including, for example, touch panels, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc., output devices 707, including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc., storage devices 708, including, for example, magnetic tape, hard disks, etc., and communication devices 709 may be connected to the I / O interface 705. The communication devices 709 may also allow the electronic device 700 to exchange data with other devices via wireless or wired communication. Figure 7 shows an electronic device 700 equipped with various devices, but it should be understood that it is not necessary to implement or include all the devices shown. Instead, more or fewer devices may be implemented or included.

[0209] In particular, according to embodiments of the present disclosure, the process described above with reference to flowcharts (e.g., Figures 1, 3, and 5) may be implemented as a computer software program. For example, embodiments of the present disclosure include a computer program product which includes a computer program contained in a non-temporary computer-readable medium, the computer program including program code for performing the method shown in the flowchart. In such embodiments, the computer program may be downloaded and installed from a network via a communication device 709, or installed from a storage device 708, or installed from a ROM 702. When executed by a processing device 701, the computer program performs the functions limited to those described above in the conference data processing method, media content segmentation method, or summary generation method according to embodiments of the present disclosure.

[0210] The electronic devices according to the embodiments of this disclosure belong to the same inventive concept as the method for processing meeting data, dividing media content, or generating summaries according to the above embodiments, and for technical details not described in detail in these embodiments, refer to the above embodiments, and these embodiments have the same beneficial effects as the above embodiments.

[0211] Based on the meeting data processing method, media content segmentation method, or summary generation method according to the above-described embodiment, the embodiment of the present disclosure provides a computer storage medium, a computer program is stored in the computer storage medium, and when the program is executed by a processor, the meeting data processing method, media content segmentation method, or summary generation method according to any one of the above embodiments is realized.

[0212] It should be noted that the computer-readable medium in this disclosure may be a computer-readable signal medium, a computer-readable storage medium, or any combination of both. The computer-readable storage medium may be, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of computer-readable storage medium may include, but are not limited to, an electrical connection having one or more wires, a portable computer magnetic disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination of the above. In this disclosure, the computer-readable storage medium may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, apparatus, or device. In this disclosure, the computer-readable signal medium includes a data signal propagated in the baseband or as part of a carrier wave, which may include computer-readable program code. The data signals propagated in this manner may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. The computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium, and may transmit, propagate, or transmit programs for use by or in combination with instruction execution systems, apparatus, or devices. The program code contained in the computer-readable medium may be transmitted by any suitable medium, including but not limited to electric wires, optical cables, RF (radio frequency), or any suitable combination thereof.

[0213] In some embodiments, clients and servers may communicate using any known or future-developed network protocols, such as HTTP (Hypertext Transfer Protocol), and interconnect with digital data communications (e.g., communication networks) of any form or medium. Examples of communication networks include local area networks ("LANs"), wide area networks ("WANs"), internetworks (e.g., the Internet), end-to-end networks (e.g., ad hoc end-to-end networks), and any known or future-developed networks.

[0214] The computer-readable medium described above may be included in the electronic device described above, or it may exist separately without being incorporated into the electronic device.

[0215] The computer-readable medium described above includes one or more programs, and when the one or more programs are executed by the electronic device, the electronic device is instructed to execute the conference data processing method, media content segmentation method, or summary generation method according to any one of the embodiments described above.

[0216] Computer program code for performing the operations disclosed herein may be written in one or more programming languages ​​or a combination thereof. Such programming languages ​​may include, but are not limited to, object-oriented programming languages ​​such as Java, Smalltalk, and C++, and may further include conventional procedural programming languages ​​such as the C language or similar programming languages. The program code may run entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, it may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or it may be connected to an external computer (for example, via the Internet using an Internet service provider).

[0217] The flowcharts and block diagrams in the drawings illustrate the implementable architectures, functions, and operations of the systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in a flowchart or block diagram may represent a module, program segment, or part of code, which includes one or more executable instructions for implementing a defined logical function. It should also be noted that in some alternative implementations, the functions shown within a block may be executed in an order different from that shown. For example, two blocks shown in sequence may actually be executed substantially in parallel, or in reverse order depending on the related functions. It should also be noted that each block in a block diagram and / or flowchart, and combinations of blocks in a block diagram and / or flowchart, may be implemented by a dedicated hardware-based system that performs the defined function or operation, or by a combination of dedicated hardware and computer instructions.

[0218] The units described in the embodiments of this disclosure may be implemented in software or in hardware. The names of the units / modules are not limited in any case to the units / modules themselves; for example, an audio data acquisition module may be described as a "data acquisition module." The units relating to the embodiments of this disclosure may be implemented in software or in hardware. The names of the units / modules may not be limited to the units themselves; for example, an audio data acquisition module may be further described as a "data acquisition module."

[0219] The functions described above in this specification may be performed, at least in part, by one or more hardware logic components. Typical types of hardware logic components available may include, but are not limited to, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), on-chip systems (SOCs), complex programmable logic devices (CPLDs), and the like.

[0220] In the context of this disclosure, a machine-readable medium may be a tangible medium containing or storing a program used in or in combination with an instruction execution system, device, or apparatus. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or apparatus, or any appropriate combination of the above. More specific examples of machine-readable storage media may include one or more wire-based electrical connections, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any appropriate combination of the above.

[0221] Each embodiment in this specification is described progressively, with emphasis on the differences between each embodiment and the others, and any similar or identical parts between embodiments should be referenced to one another. The systems or apparatus disclosed in the embodiments are briefly described to correspond to the methods for processing meeting data, segmenting media content, or generating summaries disclosed in the embodiments, and relevant sections should be referred to in the description of the methods.

[0222] In this disclosure, “at least one” means one or more, and “multiple” means two or more. “And / or” is used to describe the relationship between related objects and indicates that there may be three relationships. For example, “A and / or B” may indicate three situations: “only A exists,” “only B exists,” and “A and B exist simultaneously,” where A and B may be singular or plural. The letter “ / ” generally indicates that the related objects before and after it are in an “or” relationship. “At least one of the following” or similar expressions means any combination of these objects, including any combination of singular or plural objects. For example, at least one of a, b, or c may mean a, b, c, “a and b,” “a and c,” “b and c,” or “a, b and c,” where a, b, and c may be singular or plural.

[0223] Furthermore, it should be explained that, in this specification, relational terms such as “first” and “second” are merely used to distinguish one entity or operation from another, and do not necessarily require or imply that such an actual relationship or order exists between these entities or operations. Also, the terms “include,” “incorporate,” or any variation thereof are intended to encompass non-exclusive inclusion. Thus, a process, method, article, or apparatus that includes a set of elements includes not only those elements, but also other elements not explicitly enumerated, or elements specific to such process, method, article, or apparatus. Unless specifically limited, an element limited by the phrase “includes one…” does not preclude the existence of other identical elements in a process, method, article, or apparatus that includes such element.

[0224] Based on the above description of the disclosed embodiments, those skilled in the art can implement or use the present disclosure. Various modifications to these embodiments will be obvious 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 the present disclosure. Accordingly, the present disclosure is not limited to these embodiments shown herein and should adhere to the broadest scope consistent with the principles and novel features disclosed herein.

Claims

1. The conference data of the network conference is acquired, and the conference data includes the audio data of the network conference. Based on the aforementioned meeting data, the type of meeting for the network meeting is determined, A method for processing meeting data, comprising: dividing the network meeting based on the meeting type and the meeting content of the network meeting to obtain meeting segments of the network meeting.

2. Determining the type of meeting for the network meeting based on the aforementioned meeting data is: Based on the aforementioned meeting data, communication information of the network meeting is obtained, A method for processing meeting data according to claim 1, comprising determining the type of meeting of the network meeting based on the communication information.

3. The aforementioned communication information is A method for processing meeting data according to claim 2, comprising one or more of the following: information indicating the number of participants, information indicating the communication format, and information indicating the communication content.

4. The method for processing meeting data according to claim 3, wherein the meeting data further includes shared content display data of the network meeting, and the shared content display data is for determining information indicating the communication format and information indicating the communication content.

5. The aforementioned communication information includes information indicating the format of communication, Determining the type of meeting for the network conference based on the aforementioned communication information is: The method for processing meeting data according to claim 3, further comprising determining the meeting type of the network meeting as a share type when the information indicating the communication format indicates that the communication format includes an explanatory format.

6. The aforementioned communication information includes information indicating the number of participants and information indicating the communication format. Determining the type of meeting for the network conference based on the aforementioned communication information is: A method for processing meeting data according to claim 3, comprising determining the type of the network meeting as an interview type when the information indicating the number of participants indicates that the number of participants is two, and the information indicating the communication format indicates that the communication format is a question-and-answer format.

7. The aforementioned communication information includes information indicating the communication content, Determining the type of meeting for the network conference based on the aforementioned communication information is: The method for processing meeting data according to claim 3, further comprising determining the meeting type of the network meeting as a multitopic type when the information indicating the communication content indicates that the communication content includes multiple meeting topics.

8. The aforementioned meeting data further includes the meeting reservation type of the network meeting, Determining the type of meeting for the network meeting based on the aforementioned meeting data is: Based on the aforementioned meeting data, the communication information of the network meeting and the meeting reservation type are obtained, A method for processing meeting data according to claim 1, comprising determining the type of network meeting based on the communication information and the meeting reservation type.

9. Dividing the network conference based on the conference type and the conference content of the network conference is: Depending on the conference type, the network conference is divided into multiple sub-conferences, and each sub-conference is a conference segment of the network conference belonging to the same conference type. A method for processing meeting data according to any one of claims 1 to 8, comprising dividing the sub-meeting based on the meeting type of the sub-meeting and the meeting content of the sub-meeting.

10. Dividing the network conference based on the conference type and the conference content of the network conference is: A method for processing meeting data according to any one of claims 1 to 8, comprising dividing the network meeting based on a segmentation rule corresponding to the meeting type and the meeting content of the network meeting.

11. Dividing the network conference based on the conference type and the conference content of the network conference is: A method for processing meeting data according to any one of claims 1 to 8, comprising processing meeting data of a network meeting using a segmentation model corresponding to the meeting type, wherein the segmentation model is for segmenting the network meeting.

12. To determine the meaning of each of the aforementioned meeting segments, A method for processing conference data according to any one of claims 1 to 8, further comprising combining conference segments in the aforementioned network conference whose time intervals are adjacent and whose semantic similarity is equal to or greater than a similarity threshold.

13. The media content data is obtained, and the media content includes content of at least two content dimensions. Based on the content dimensions and the media content data, the target division point of the media content is determined. A method for dividing media content, comprising determining sub-content of the media content based on the aforementioned target division point.

14. Determining the target division point of the media content based on the content dimensions and the media content data is: Based on the division method corresponding to the content dimension and the data of the media content, candidate division points for the media content in the content dimension are determined. A method for dividing media content according to claim 13, comprising determining a target division point of the media content based on the candidate division points.

15. Determining the target division point of the media content based on the candidate division points is: A method for segmenting media content according to claim 14, comprising determining a target segmentation point of the media content based on the segmentation confidence of the candidate segmentation points, wherein the segmentation confidence is for evaluating the accuracy of segmentation by the candidate segmentation points.

16. The method for segmenting media content according to claim 15, wherein the segmentation confidence is determined based on the division method of the candidate division points.

17. The segmentation confidence is determined by the segmentation granularity of the candidate segmentation points. The method for dividing media content according to claim 15, wherein the segmentation granularity indicates the degree of subdivision of the media content by the candidate division points.

18. The media content includes audio content, and the content dimension includes the audio content dimension. Regarding the aforementioned audio content dimension, Based on the audio data, the spoken content of the presenter included in the media content is obtained. A method for dividing media content according to claim 13, wherein a first division point is determined using a method in which a first division point of the media content is determined based on different process stages indicated by the aforementioned statement content, and the first division point is for determining the target division point.

19. The method according to claim 13, wherein the media content includes shared content.

20. The shared content includes shared screen content, and the content dimension includes the shared screen content dimension. Regarding the shared screen content dimension, Based on the shared screen content, the time for switching the content on the shared screen is determined. A method for dividing media content according to claim 19, wherein a second division point is determined using a method in which a second division point of the media content is determined based on the switching time of the content of the shared screen, and the second division point is for determining the target division point.

21. The shared content includes shared document content, and the content dimension includes the shared document content dimension. Regarding the shared document content dimension, Based on the shared document content, the document content switching time is determined. A method for dividing media content according to claim 19, wherein a third division point is determined using a method in which a third division point of the media content is determined based on the switching time of the document content, and the third division point is for determining the target division point.

22. The method for dividing media content according to claim 13, wherein the media content includes descriptive content.

23. The description content includes stage time content, and the content dimension includes stage time content dimension. Regarding the aforementioned time-based content dimension, A method for dividing media content according to claim 22, wherein a fourth division point is determined using a method in which a fourth division point of the media content is determined based on the time indicated by the stage time content, and the fourth division point is for determining the target division point.

24. The description content includes stage content, and the content dimension includes stage content dimension, Regarding the aforementioned stage content dimensions, A method for dividing media content according to claim 22, wherein a fifth division point is determined by performing content clustering on the media content based on the aforementioned stage content, and determining a fifth division point of the media content based on the clustered different content, and the fifth division point is for determining the target division point.

25. The method for dividing media content according to any one of claims 13 to 24, wherein the media content is conference content, and the sub-content of the media content is conference segment content.

26. The content data of each sub-content included in the media content is obtained, and the sub-content is obtained by dividing the media content. Based on the content data of each sub-content, a summary of the sub-content is determined. A summary generation method comprising: fusing summaries of each sub-content based on the weight of each sub-content to obtain a summary of the media content, wherein the weight of the sub-content is intended to indicate the importance of the sub-content in the media content.

27. The weight of the sub-content is determined by the content information of the sub-content. The summary generation method according to claim 26, wherein the content information is determined based on the time information of the sub-content and the content data of the sub-content in the media content.

28. The summary generation method according to claim 27, wherein the weight of the sub-content is determined based on the sub-weights corresponding to the sub-information contained in the content information of the sub-content.

29. The aforementioned content information is, The summary generation method according to claim 28, comprising one or more sub-informations, including the duration of the sub-content, the number of people involved in the sub-content, and the position of the sub-content's time interval within the media content's time interval.

30. The summarization method according to claim 27, wherein the weights of the sub-contents are determined based on an artificial intelligence model, and the artificial intelligence model is for outputting weights based on input content information.

31. To obtain a summary of the media content by fusing summaries of each sub-content based on the weight of each sub-content, A method for generating a summary according to claim 26, comprising processing the weights of each sub-content and the summaries of each sub-content based on a first language processing model to obtain a summary of the media content.

32. The content data is text data, and the summary of each sub-content is determined based on the content data of each sub-content. The summary generation method according to claim 26, comprising processing the content data of each sub-content based on a second language processing model to obtain a summary of each sub-content.

33. The media content is the content of the meeting, and the sub-content is the content of a sub-meeting obtained by dividing the meeting, or The summary generation method according to any one of claims 26 to 32, wherein the meeting is a regular scheduled meeting, and the sub-content is the content of at least one regular meeting included in the regular scheduled meeting.

34. The aforementioned sub-meeting was The summary generation method according to claim 33, wherein the meeting is divided and obtained based on the meeting type and meeting content of the meeting.

35. The aforementioned sub-meeting was The summary generation method according to claim 33, wherein the meeting is divided and obtained based on at least two content dimensions of the meeting.

36. A first acquisition unit is configured to acquire meeting data of a network conference, wherein the meeting data includes audio data of the network conference. A first decision unit configured to determine the conference type of the network conference based on the conference data, A conference data processing device comprising: a first splitting unit configured to split the network conference based on the conference type and the conference content of the network conference, and to obtain conference segments of the network conference.

37. A second acquisition unit is configured to acquire data from media content, wherein the media content includes content of at least two content dimensions. A second decision unit configured to determine the target division point of the media content based on the content dimension and the media content data, A media content splitting device comprising: a second splitting unit configured to determine sub-content of the media content based on the aforementioned target splitting point.

38. A third acquisition unit is configured to acquire content data for each sub-content contained in media content, wherein the sub-content is acquired by dividing the media content, A third decision unit configured to determine a summary of each sub-content based on the content data of each sub-content, A summary generation device comprising: a generation unit configured to obtain a summary of media content by fusing summaries of each subcontent based on the weight of each subcontent, wherein the weight of the subcontent is intended to indicate the importance of the subcontent in the media content.

39. At least one processor, A memory device that stores at least one program, An electronic device wherein, when the at least one program is executed by the at least one processor, it causes the at least one processor to implement the method for processing conference data according to any one of claims 1 to 12, the method for dividing media content according to any one of claims 13 to 25, or the method for generating a summary according to any one of claims 26 to 35.

40. A computer-readable medium on which computer programs are stored, The program, when executed by a processor, realizes a computer-readable medium that implements a method for processing conference data according to any one of claims 1 to 12, a method for dividing media content according to any one of claims 13 to 25, or a method for generating a summary according to any one of claims 26 to 35.

Images