Real-time conference summary generation method and device, equipment, medium and computer program product

Abstract

The invention discloses a real-time conference summary generation method, device and equipment, a medium and a computer program product, relates to the technical field of natural language processing, and can be applied to the field of financial science and technology. Acquiring real-time accumulated conference content of the target conference; performing text truncation on the real-time accumulated conference content to obtain fixed conference content of at least one round of conference, each round of conference including an interval conference of at least one time interval; and summarizing the fixed conference content and the interval conference content of the interval conference to obtain a real-time conference summary of the target conference. By adopting the technical scheme, the real-time conference content abstract is generated by performing text truncation on the conference content, the key content of the conference is accurately reflected, and the real-time conference abstract can be accurately generated in various conference scenes.

Description

Technical Field

[0001] The present invention relates to the field of natural language processing technology, which can be applied to the field of financial technology, and in particular to a real-time conference summary generation method, device, equipment, medium and computer program product. Background Art

[0002] The development of summary generation technology involves progress in many fields. There is little research on the generation technology of conference summaries. Unlike general document summaries, it needs to comprehensively consider the conversation context information. Conference summaries have high real-time requirements, so the technical difficulty and detail requirements will be more stringent.

[0003] Existing summary generation often relies on rule-based methods, which extract key sentences from text content to generate summaries through predefined rules and keyword matching; or identify important sentences to generate summaries by analyzing the relationship between word frequency and word co-occurrence in the text; or learn how to generate summaries from a large amount of annotated data through training models.

[0004] However, for texts with complex sentence structures or involving professional terms, it may not be possible to correctly understand and generate summaries, and key information may be omitted during the summary generation process, resulting in incomplete summaries or loss of important details. For models, they perform well in specific fields, but their generalization ability between different fields is limited. The performance of the model is highly dependent on the quality and diversity of the training data. If the training data is insufficient or biased towards a specific field, the generated conference summary may be biased. Summary of the invention

[0005] The present invention provides a real-time conference summary generation method, device, equipment, storage medium and computer program product. By truncating the text of the conference content and summarizing the fixed conference content and the interval conference content of the interval meeting, the key content of the meeting can be accurately reflected, ensuring that the real-time conference summary can be accurately generated in various conference scenarios.

[0006] According to one aspect of the present invention, a method for generating a real-time conference summary is provided, comprising:

[0007] Obtain the real-time cumulative meeting content of the target meeting;

[0008] Performing text truncation on the real-time accumulated meeting content to obtain fixed meeting content of at least one round of meetings, each round of meetings including at least one interval meeting of a time interval;

[0009] It is used to summarize the fixed meeting content and the interval meeting content of the interval meeting to obtain a real-time meeting summary of the target meeting.

[0010] According to another aspect of the present invention, there is provided a real-time conference summary generating device, comprising:

[0011] A conference content acquisition module is used to acquire the real-time accumulated conference content of the target conference;

[0012] A fixed meeting content acquisition module, used to perform text truncation on the real-time accumulated meeting content to obtain fixed meeting content of at least one round of meetings, each round of meetings including at least one interval meeting of a time interval;

[0013] The real-time conference summary determination module is used to summarize the fixed conference content and the interval conference content of the interval conference to obtain the real-time conference summary of the target conference.

[0014] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0015] at least one processor; and

[0016] a memory communicatively connected to the at least one processor; wherein,

[0017] The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute any of the methods for generating a real-time conference summary of the present invention.

[0018] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement any method for generating a real-time conference summary according to the present invention when executed.

[0019] According to another aspect of the present invention, a computer program product is provided. The computer program product includes a computer program. When the computer program is executed by a processor, the computer program implements any method for generating a real-time conference summary according to the present invention.

[0020] The technical solution of the embodiment of the present invention solves the problem of real-time generation of meeting summaries in various meeting scenarios by truncating the text of the target meeting content, summarizing the content based on the fixed meeting content and the interval meeting content, and generating a real-time meeting content summary, thereby facilitating users to review and grasp the previous key points in a timely manner and ensuring that they can keep up with the pace of the meeting when they return to the meeting after leaving.

[0021] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0023] Figure 1 is a flow chart of a real-time conference summary generation method provided according to an embodiment of the present invention;

[0024] Figure 2 It is a logical design diagram for generating summary of each interval meeting provided in accordance with an embodiment of the present invention;

[0025] Figure 3 is a flow chart of a real-time conference summary generation method provided according to an embodiment of the present invention;

[0026] Figure 4 is a structural block diagram of a real-time conference summary generating device provided according to an embodiment of the present invention;

[0027] Figure 5 It is a structural block diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0028] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0029] The large model is a large neural network model for deep learning with large-scale parameters and high-performance computing capabilities. It can utilize the powerful understanding ability of the large model to avoid technical problems such as poor generalization of small models, difficulty in collecting training data, and high training costs.

[0030] Figure 1 The flowchart of a real-time conference summary generation method provided by an embodiment of the present invention is applicable to the case of generating conference summaries in real time. The method can be executed by a real-time conference summary generation device provided by an embodiment of the present invention. The real-time conference summary generation device can be implemented in the form of hardware and / or software, and can be integrated into an electronic device that carries the conference summary generation function, such as a server. Figure 1 As shown, the method specifically comprises the following steps:

[0031] S110: Acquire the real-time accumulated meeting content of the target meeting.

[0032] The target meeting is a meeting that is currently in progress and for which a meeting summary needs to be generated; the real-time accumulated meeting content is the text content that is converted into text and accumulated in chronological order when the target meeting is in progress.

[0033] Specifically, after the target meeting starts, the meeting speech is converted into text content in real time through natural language processing technology, and the text content is accumulated and stored in chronological order. Preferably, the natural language processing technology can be an automatic speech recognition (ASR) engine that converts speech into text. It should be noted that relevant technicians can use other natural language processing technologies for speech conversion, and the embodiments of the present invention do not specifically limit this.

[0034] S120: Perform text truncation on the real-time accumulated conference content to obtain fixed conference content of at least one round of meetings, where each round of meetings includes at least one interval meeting of a time interval.

[0035] Among them, text truncation can be to split the accumulated meeting content within a certain time interval or a certain word range; then a certain time interval or a certain word range is a round of meetings, and the corresponding meeting content is the fixed meeting content of this round of meetings; the time interval is pre-set; a round of meetings can be composed of at least one time interval.

[0036] Specifically, the real-time accumulated meeting content is accumulated and acquired according to the preset time interval. When the accumulated meeting content reaches the text truncation condition, the accumulated meeting content is truncated, and the current accumulated meeting content is used as the fixed meeting content of a round of meetings and stored in the database, and continues to accumulate from the next time interval. It can be understood that by setting the text truncation condition to truncate the accumulated meeting content, the input limit of the current 10 billion large model is adapted to avoid affecting the output quality of the model when the input meeting content exceeds the maximum tolerance limit.

[0037] Optionally, the text cutoff condition is the word limit of the real-time accumulated meeting content, which can be 2,000 words, 3,000 words, 4,000 words or 5,000 words, etc. Preferably, in the embodiment of the present invention, the text cutoff condition is that when the real-time accumulated meeting content reaches 4,000 words, the current accumulated meeting content is stored as the fixed meeting content of this round. It can be understood that when the meeting content of about 4,000 words is fed into the large model to output the meeting summary, the adoption rate and usability index of the summary generated are optimal, and the topics conveyed by the meeting content are relatively stable at this time.

[0038] It should be noted that the time interval is pre-set by relevant technical personnel based on experience according to the specific meeting scenario, the speaker's speaking habits, speaking frequency and the number of words in the speech. A time interval can be set for one minute, two minutes or five minutes. The present invention does not limit the specific time interval setting.

[0039] For example, at the beginning of a meeting, after obtaining the meeting content within the first time interval, the meeting content of the next time interval is obtained and accumulated as real-time accumulated meeting content. When the real-time accumulated meeting content reaches 4,000 words, the real-time accumulated meeting content from the first time interval to 4,000 words is used as the fixed meeting content of the first round of meetings, and the meeting content of the next time interval is accumulated, and so on, until the target meeting ends. It can be seen that the time required for each round of fixed meeting content speech varies according to different meeting scenarios and the speaking frequency of the speaker, so a round of fixed meetings can have at least one interval meeting of time.

[0040] S130: Summarize the fixed meeting content and the interval meeting content of the interval meeting to obtain a real-time meeting summary of the target meeting.

[0041] Each interval meeting corresponds to a meeting summary; each round of meetings includes at least one meeting summary.

[0042] Specifically, the real-time accumulated meeting content is obtained, and the interval meeting content of a time interval is input into the large model to obtain the interval meeting summary. When the number of words in the current interval meeting content meets the text truncation condition, the 4,000-word text content before the text truncation condition is used as a round of fixed meeting content, and the round of fixed meeting content is input into the large model to obtain the fixed meeting summary of the round of fixed meeting content and save it; the interval meeting content of the next time interval is further obtained, and the interval meeting content is input into the large model to output the interval meeting summary corresponding to the current interval meeting content; and so on, after N minutes, the real-time accumulated meeting content reaches 4,000 words again, and then another round of fixed meeting summary is composed of the previous round of fixed summary content and the interval meeting content corresponding to the current N minutes. Wherein, N is a natural number greater than 0.

[0043] In an optional implementation of the embodiment of the present invention, before generating the interval meeting summary, it is necessary to determine whether the interval meeting content within the current time interval meets the summary generation condition:

[0044] Determine whether the number of words in the content of the interval meeting of the current time interval meets the summary generation condition;

[0045] If it is satisfied, the content of the interval meeting of the current time interval is input into the large model to generate the meeting summary;

[0046] If not satisfied, a prompt message is returned.

[0047] Among them, the condition for summary generation is that the number of words in the interval meeting content is greater than 200 words; the prompt message is "There are currently few transcribed text records, and there is no intelligent summary text yet."

[0048] like Figure 2 As shown, the content of the interval meeting of the current time interval is obtained through the ASR engine, and it is determined whether the number of words in the interval meeting content is greater than 200 words. If the number of words in the interval meeting content is greater than 200 words, it is input into the large model, and the interval meeting summary corresponding to the interval meeting content is output; if the number of words in the interval meeting content is less than 200 words, it directly returns "the current transcription text record is small, and there is no intelligent summary text yet."

[0049] It is understandable that by judging the number of words in the content of the interval meetings within a time interval, it is possible to avoid inaccurate summary of the meeting summary when the number of words is too small, thereby ensuring the completeness and richness of the data required for generating the meeting summary.

[0050] Optionally, the fixed meeting content and the interval meeting content of the interval meeting are summarized to obtain a real-time meeting summary of the target meeting, including:

[0051] For each round of meetings:

[0052] For each interval meeting in the round of meetings, determining the interval meeting content corresponding to the interval meeting;

[0053] Based on the big model, the interval meeting content and fixed meeting content are summarized and summarized to determine the real-time meeting summary of the target meeting.

[0054] The interval meeting content is the meeting content corresponding to the start and end of the current time interval in the round of meetings; the real-time meeting summary may include the interval meeting summary and the fixed meeting summary.

[0055] Specifically, for each round of meetings, each interval meeting is obtained, and the interval meeting content corresponding to each interval meeting is obtained, the interval meeting content is input into the big model to obtain the corresponding interval meeting summary; the fixed meeting content is input into the big model, and the corresponding fixed meeting summary is output; based on the interval meeting summary and the fixed meeting summary, the real-time meeting summary of the target meeting is summarized.

[0056] It can be understood that the large model is used to output the meeting summary corresponding to the meeting content, ensuring the accuracy and efficiency of the target meeting summary generation.

[0057] Optionally, the interval meeting content and the fixed meeting content are summarized based on the big model to determine the real-time meeting summary of the target meeting, including:

[0058] Based on the big model, the fixed meeting contents of the previous round of meetings are summarized to obtain the fixed meeting summary of the previous round of meetings;

[0059] Based on the big model, the fixed meeting summary of the interval meeting content and the previous round of meetings is summarized to determine the real-time meeting summary of the target meeting.

[0060] Specifically, when the meeting content reaches the first four thousand words, the fixed meeting content of this round is input into the big model for summary, and the fixed meeting summary of this round of meetings is output; continue to obtain the interval meeting content of the next time interval, and after the interval time ends, input the interval meeting content and the fixed meeting summary of the previous round of meetings into the big model, and output the interval meeting summary corresponding to the current interval meeting; the generation logic of each subsequent meeting summary is similar.

[0061] It can be understood that by taking the current interval meeting content and the previous round of fixed meeting summary as the input of the large model, it is ensured that the output meeting summary fully learns the context, achieves in-depth understanding and accurate expression of the meeting content, and thus accurately reflects the key content of the meeting.

[0062] The embodiment of the present invention obtains the real-time cumulative meeting content of the target meeting; performs text truncation on the real-time cumulative meeting content to obtain the fixed meeting content of at least one round of meetings, each round of meetings includes at least one interval meeting at a time interval; summarizes the fixed meeting content and the interval meeting content of the interval meeting to obtain the real-time meeting summary of the target meeting. The above technical scheme solves the problem of real-time generation of meeting summaries in various meeting scenarios, ensuring accurate generation of real-time meeting summaries, and by performing text truncation on the real-time meeting content, generating real-time meeting summaries based on the fixed meeting content and the interval meeting content, it can effectively understand long texts or complex contextual relationships, ensure the overall coherence of the meeting summary, and further ensure that the intent and semantics of the original text can be accurately reflected in meeting scenarios in various fields.

[0063] Optionally, after summarizing the fixed meeting content and the interval meeting content of the interval meeting to obtain the real-time meeting summary of the target meeting, the following is further included:

[0064] The meeting summaries of two adjacent rounds of meetings are merged based on the big model to optimize the real-time meeting summary of the target meeting.

[0065] The meeting summaries of two adjacent rounds of meetings include the last two interval meeting summaries of the previous round of meetings and the first two interval meeting summaries of the next round of meetings.

[0066] Specifically, based on the aforementioned text truncation processing of the content of a round of meetings, and based on the preset time interval to obtain the interval meeting content of a single time interval, after the target meeting is over, the summaries of two adjacent rounds of fixed meetings are merged, including: the last two topics of the first round of summaries and the first two topics of the second round of summaries, the last two topics of the second round of summaries and the first two rounds of summaries of the third round of summaries... are input into the large model for summary merging, until all rounds of meetings of the target meeting are merged. It should be noted that in the large model, different prompt words can be used to merge multiple summaries, or each summary can be vectorized, or each summary can be calculated for similarity; the present invention does not specifically limit the way of merging summaries.

[0067] It is understandable that since each round of meetings is truncated to 4,000 words, the same topic will appear in different rounds of meetings. Therefore, by merging the summaries of each round of the target meeting, redundant topics can be optimized, the number of topics can be simplified, and the readability and accuracy of the meeting summaries can be further increased.

[0068] Figure 3 This is a flowchart of a method for generating a real-time conference summary according to an embodiment of the present invention. Based on the above embodiment, the embodiment of the present invention summarizes the content of interval conferences and fixed conferences based on a large model, and optimizes the specific determination method of the real-time conference summary of the target conference. It should be noted that for the part not described in detail in the embodiment of the present invention, please refer to the relevant description of other embodiments. Figure 3 As shown, the method includes:

[0069] S210: Obtain the real-time accumulated meeting content of the target meeting.

[0070] S220: Perform text truncation on the real-time accumulated conference content to obtain fixed conference content of at least one round of meetings, where each round of meetings includes at least one interval meeting of a time interval.

[0071] S230: For each round of meetings and each interval meeting in the round of meetings, determine the interval meeting content corresponding to the interval meeting.

[0072] S240: Determine whether the word count of the interval meeting content of the interval meeting meets the word count threshold.

[0073] The word count threshold is pre-set and is the word count limit for each round of meetings, preferably 4,000 words. It should be noted that relevant technicians can set the corresponding word count threshold according to actual needs, and the embodiment of the present invention does not specifically limit this.

[0074] Specifically, when a round of meetings is finished, the meeting content of the next time interval is obtained, and the meeting content is recalculated to determine whether the number of words in the meeting content after the next time interval reaches 4,000 words.

[0075] Exemplarily, after the first round of meetings ends, the content of the interval meetings at the first time interval is obtained and recalculated, and it is determined whether the content of the interval meetings at the first time interval reaches four thousand words, and the corresponding interval meeting summary is completed; the content of the interval meetings at the second time interval is continued to be obtained, and accumulated with the content of the interval meetings at the first time interval to determine the number of words in the cumulative interval meeting content, and it is determined whether the number of words in the cumulative interval meeting content reaches the word count threshold of four thousand words, and so on... until the cumulative interval meeting content reaches four thousand words and is recalculated again.

[0076] S250: If not, input the interval meeting content of the interval meeting into the large model to obtain the interval meeting summary, and determine the real-time meeting summary of the target meeting based on the interval meeting summary and the fixed meeting summary of the previous round of meetings.

[0077] If the number of words in the cumulative interval meeting content corresponding to the current interval meeting is less than four thousand words, the interval meeting content of the current interval meeting is input into the large model, and the interval meeting summary of the current interval meeting is output. Based on the interval meeting summary of the current interval meeting and the fixed meeting summary of the previous round of meetings before the current round of meetings, a summary is re-obtained as the interval meeting summary of the current interval meeting content.

[0078] It can be understood that generating a new meeting summary by combining the current interval meeting summary and the fixed meeting summary of the previous round of meetings ensures that the input content expression style and form of the large model are at the same latitude, ensuring that the large model can accurately learn and fully follow the style of the summary.

[0079] S260: If the conditions are met, the fixed meeting content of the current round of meetings is input into the big model to obtain the fixed meeting summary of the current round of meetings, and the real-time meeting summary of the target meeting is determined based on the fixed meeting summary of the current round of meetings and the fixed meeting summary of the previous round of meetings.

[0080] If the word count of the accumulated interval meeting content corresponding to the current interval meeting reaches four thousand words, the fixed meeting content of this round of meetings is input into the big model to obtain the fixed meeting summary of this round of meetings. The fixed meeting summary of this round of meetings and the fixed meeting summary of the previous round of meetings are further used as inputs of the big model to further determine the fixed meeting summary of this round of meetings.

[0081] It is understandable that the fixed meeting summary of the current round of meetings is further determined by the fixed meeting summary of two rounds of meetings, so as to avoid the same topic content being divided into the fixed meeting content of two rounds of meetings due to text truncation, which in turn leads to repeated summaries, and further summarize based on the fixed meeting summary of the previous round, so as to improve the adoption rate and usability of the meeting summary output by the big model, and further ensure that the big model can understand long texts and contextual relationships, and ensure the overall coherence of the meeting summary.

[0082] The embodiment of the present invention judges the number of words in the content of interval meetings, and in one round of meetings, summarizes the interval summary of the current time interval and the fixed meeting summary of the previous round; at the end of one round of meetings, summarizes the fixed meeting summaries of two adjacent rounds; ensures that the expression style and form of the input content of the large model are at the same latitude, accurately learns and follows the summary style, avoids missing the topic content, and further ensures that when the content of two rounds of fixed meetings is the same topic, the context can be accurately understood to summarize, and ensures that real-time meeting summaries can be accurately generated in various meeting scenarios.

[0083] Figure 4 : is a structural block diagram of a real-time conference summary generation device provided according to an embodiment of the present invention. The present invention is applicable to the case of generating conference summaries in real time. The method can be executed by the real-time conference summary generation device provided by an embodiment of the present invention. The real-time conference summary generation device can be implemented in the form of hardware and / or software, and can be integrated into an electronic device that carries the summary generation function, such as a server. Figure 4 As shown, the real-time conference summary generating apparatus 300 includes a conference content acquiring module 310 , a fixed conference content acquiring module 320 , and a real-time conference summary determining module 330 .

[0084] The conference content acquisition module 310 is used to acquire the real-time accumulated conference content of the target conference;

[0085] The fixed meeting content acquisition module 320 is used to perform text truncation on the real-time accumulated meeting content to obtain the fixed meeting content of at least one round of meetings, each round of meetings including at least one interval meeting of a time interval;

[0086] The real-time conference summary determination module 330 is used to summarize the fixed conference content and the interval conference content of the interval conference to obtain the real-time conference summary of the target conference.

[0087] The embodiment of the present invention obtains the real-time cumulative meeting content of the target meeting; performs text truncation on the real-time cumulative meeting content to obtain the fixed meeting content of at least one round of meetings, each round of meetings includes at least one interval meeting at a time interval; summarizes the fixed meeting content and the interval meeting content of the interval meeting to obtain the real-time meeting summary of the target meeting. The above technical scheme solves the problem of real-time generation of meeting summaries in various meeting scenarios, ensuring accurate generation of real-time meeting summaries, and by performing text truncation on the real-time meeting content, generating real-time meeting summaries based on the fixed meeting content and the interval meeting content, it can effectively understand long texts or complex contextual relationships, ensure the overall coherence of the meeting summary, and further ensure that the intent and semantics of the original text can be accurately reflected in meeting scenarios in various fields.

[0088] Optionally, the real-time meeting summary determination module 330 is specifically used for: for each round of meetings, for each interval meeting in the round of meetings, determining the interval meeting content corresponding to the interval meeting; summarizing the interval meeting content and the fixed meeting content based on the big model to determine the real-time meeting summary of the target meeting.

[0089] Optionally, the real-time meeting summary determination module 330 is specifically used to determine whether the word count of the interval meeting content of the interval meeting meets the word count threshold; if not, the interval meeting content of the interval meeting is input into the big model to obtain the interval meeting summary, and the real-time meeting summary of the target meeting is determined based on the interval meeting summary and the fixed meeting summary of the previous round of meetings of this round of meetings; if satisfied, the fixed meeting content of this round of meetings is input into the big model to obtain the fixed meeting summary of this round of meetings, and the real-time meeting summary of the target meeting is determined based on the fixed meeting summary of this round of meetings and the fixed meeting summary of the previous round of meetings of this round of meetings.

[0090] Optionally, the real-time meeting summary module 330 is specifically used to summarize the fixed meeting content of the previous round of meetings of this round of meetings based on the big model to obtain the fixed meeting summary of the previous round of meetings; based on the big model, summarize the interval meeting content and the fixed meeting summary of the previous round of meetings to determine the real-time meeting summary of the target meeting.

[0091] Each interval meeting corresponds to a meeting summary; each round of meetings includes at least one meeting summary.

[0092] Optionally, the device also includes a conference summary merging module for merging the conference summaries of two adjacent rounds of conferences based on the large model to optimize the real-time conference summary of the target conference.

[0093] The automated test case processing device provided in the embodiment of the present invention can execute the automated test case processing method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0094] According to an embodiment of the present invention, the present invention also provides an electronic device, a readable storage medium and a computer program product.

[0095] Figure 5 A schematic diagram of the structure of an electronic device 10 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or claimed herein.

[0096] like Figure 5 As shown, the electronic device 10 includes at least one processor 11, and a memory connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 to the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0097] A number of components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0098] The processor 11 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as a real-time conference summary generation method.

[0099] In some embodiments, the real-time conference summary generation method may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the automated test case processing method described above may be performed. Alternatively, in other embodiments, the processor 11 may be configured to execute the real-time conference summary generation method in any other appropriate manner (e.g., by means of firmware).

[0100] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0101] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0102] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0103] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0104] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0105] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0106] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0107] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A real-time conference summary generation method, characterized in that: include: Obtain the real-time cumulative meeting content of the target meeting; Performing text truncation on the real-time accumulated meeting content to obtain fixed meeting content of at least one round of meetings, each round of meetings including at least one interval meeting of a time interval; The fixed meeting content and the interval meeting content of the interval meeting are summarized to obtain a real-time meeting summary of the target meeting.

2. The method according to claim 1, characterized in that Summarizing the fixed meeting content and the interval meeting content of the interval meeting to obtain a real-time meeting summary of the target meeting, including: For each round of meetings: For each interval meeting in the round of meetings, determining the interval meeting content corresponding to the interval meeting; The interval meeting content and the fixed meeting content are summarized based on the big model to determine a real-time meeting summary of the target meeting.

3. The method according to claim 2, characterized in that Summarizing the interval meeting content and the fixed meeting content based on the big model to determine a real-time meeting summary of the target meeting, including: Determine whether the word count of the content of the interval meeting of the interval meeting meets the word count threshold; If not, the interval meeting content of the interval meeting is input into the large model to obtain the interval meeting summary, and the real-time meeting summary of the target meeting is determined according to the interval meeting summary and the fixed meeting summary of the previous round of meetings of this round of meetings; If it is satisfied, the fixed meeting content of this round of meetings is input into the big model to obtain the fixed meeting summary of this round of meetings, and the real-time meeting summary of the target meeting is determined based on the fixed meeting summary of this round of meetings and the fixed meeting summary of the previous round of meetings.

4. The method according to claim 2, characterized in that: Summarizing the interval meeting content and the fixed meeting content based on the big model to determine a real-time meeting summary of the target meeting, including: Based on the big model, the fixed meeting contents of the previous round of meetings are summarized to obtain the fixed meeting summary of the previous round of meetings; Based on the large model, the fixed meeting summary of the interval meeting content and the previous round of meetings is summarized to determine the real-time meeting summary of the target meeting.

5. The method according to claim 1, characterized in that Each interval meeting corresponds to a meeting summary; each round of meetings includes at least one meeting summary.

6. The method according to claim 5, characterized in that After summarizing the fixed meeting content and the interval meeting content of the interval meeting to obtain the real-time meeting summary of the target meeting, the method further includes: The meeting summaries of two adjacent rounds of meetings are merged based on the big model, and the real-time meeting summary of the target meeting has been optimized.

7. A real-time conference summary generation device, characterized in that: include: A conference content acquisition module is used to acquire the real-time accumulated conference content of the target conference; A fixed meeting content acquisition module, used to perform text truncation on the real-time accumulated meeting content to obtain fixed meeting content of at least one round of meetings, each round of meetings including at least one interval meeting of a time interval; The real-time conference summary determination module is used to summarize the fixed conference content and the interval conference content of the interval conference to obtain the real-time conference summary of the target conference.

8. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the real-time conference summary generating method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the real-time conference summary generation method according to any one of claims 1 to 6 when executed.

10. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the computer program implements the real-time conference summary generating method according to any one of claims 1 to 6.

Images