WO2019095586A1

WO2019095586A1 - Meeting minutes generation method, application server, and computer readable storage medium

Info

Publication number: WO2019095586A1
Application number: PCT/CN2018/077628
Authority: WO
Inventors: 王健宗; 黄章成; 程宁; 肖京
Original assignee: 平安科技（深圳）有限公司
Priority date: 2017-11-17
Filing date: 2018-02-28
Publication date: 2019-05-23
Also published as: CN108022583A

Abstract

Disclosed in the present application is a meeting minutes generation method. The method comprises: obtaining audio record information of a meeting, and extracting speech content of each speaker from the audio record information according to sound characteristics of each speaker; extracting a keyword from the speech content of each speaker; and generating meeting minutes corresponding to the meeting according to the extracted keywords. The present application also provides an application server and a computer readable storage medium. By means of the meeting minutes generation method, the application server and the computer readable storage medium provided in the present application, meeting minutes can be automatically summarized and generated according to meeting content records, thereby reducing costs of human resources.

Description

Meeting minutes generation method, application server, and computer readable storage medium

This application claims the priority of the Chinese patent application filed on November 17, 2017, the Chinese Patent Office, the application number is 201711141751.5, and the invention name is "meeting minutes generation method, application server and computer readable storage medium", the entire contents of which are The citation is incorporated in the application.

Technical field

The present application relates to the field of voice processing technologies, and in particular, to a conference minutes generation method, an application server, and a computer readable storage medium.

Background technique

In the process of government and company work, almost every working day may face various meetings, from the important decision-making level to the meeting, to the discussion of an event or the function of the group. The meeting "This form is done. In the process of participation, the participants generally focus on following up the content and process of the meeting. After the meeting, the meeting minutes often need to rely on the special staff to collect and organize according to the participation process, which leads to the process of organizing the meeting minutes. Input of labor costs. For some small group meetings, often due to time and manpower, no dedicated staff to organize the meeting minutes will not be conducive to promoting team building and growth.

Summary of the invention

In view of this, the present application provides a method for generating a meeting minutes, an application server, and a computer readable storage medium, which can automatically summarize and generate meeting minutes according to meeting content records, thereby saving human resource costs.

First, in order to achieve the above object, the present application provides an application server, where the application server includes a memory, a processor, and a memory meeting generation system that can be run on the processor, where the meeting minutes are generated. When the system is executed by the processor, the following steps are performed: acquiring audio record information of a conference, and extracting, from the audio record information, the content of each speaker according to the voice feature of each speaker; The content of the speech of the speaker is subjected to keyword extraction; and the meeting minutes corresponding to the meeting are generated according to the extracted keywords.

In addition, to achieve the above object, the present application further provides a method for generating a meeting minutes, which is applied to an application server, the method comprising: acquiring audio record information of a conference, and recording the audio record according to the voice feature of each speaker. Extracting the content of each of the speakers of the information; performing keyword extraction on the content of the speech of each of the speakers; and generating a meeting minutes corresponding to the meeting according to the extracted keywords.

Further, in order to achieve the above object, the present application further provides a computer readable storage medium storing a meeting minutes generating system, the meeting minutes generating system being executable by at least one processor, so that The at least one processor performs the steps of the method of generating a meeting minutes as described above.

Compared with the prior art, the conference minutes generating method, the application server, and the computer readable storage medium proposed by the present application first acquire audio recording information of a conference, and from the audio recording according to the voice characteristics of each speaker. The content of each speaker of the speaker is extracted from the information; secondly, keyword extraction is performed on the content of the speech of each of the speakers; and finally, the meeting minutes corresponding to the meeting are generated according to the extracted keywords. In this way, it is possible to automatically summarize and generate meeting minutes according to the meeting content records, so that the participants can review the content of the meeting. The participants in the meeting can focus more on the content and process of the meeting. After the meeting, the meeting summary is streamlined and accurate. It can also be used for reference and reference by other people in need. Compared with traditional manual recording, this solution is more efficient and accurate, and saves human resource costs.

DRAWINGS

1 is a schematic diagram of an optional application environment of each embodiment of the present application;

2 is a schematic diagram of an optional hardware architecture of an application server of the present application;

3 is a schematic diagram of a program module of a first embodiment of a meeting minutes generation system of the present application;

4 is a schematic diagram of a program module of a second embodiment of the meeting minutes generating system of the present application;

5 is a schematic flowchart of an implementation process of a first embodiment of a method for generating a meeting minutes of the present application;

FIG. 6 is a schematic diagram of an implementation process of a second embodiment of a method for generating a meeting minutes of the present application.

The implementation, functional features and advantages of the present application will be further described with reference to the accompanying drawings.

Detailed ways

In order to make the objects, technical solutions, and advantages of the present application more comprehensible, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It is understood that the specific embodiments described herein are merely illustrative of the application and are not intended to be limiting. All other embodiments obtained by a person of ordinary skill in the art based on the embodiments of the present application without departing from the inventive scope are the scope of the present application.

It should be noted that the descriptions of "first", "second" and the like in the present application are for the purpose of description only, and are not to be construed as indicating or implying their relative importance or implicitly indicating the number of technical features indicated. . Thus, features defining "first" or "second" may include at least one of the features, either explicitly or implicitly. In addition, the technical solutions between the various embodiments may be combined with each other, but must be based on the realization of those skilled in the art, and when the combination of the technical solutions is contradictory or impossible to implement, it should be considered that the combination of the technical solutions does not exist. Nor is it within the scope of protection required by this application.

Referring to FIG. 1 , it is a schematic diagram of an optional application environment of each embodiment of the present application.

In this embodiment, the present application is applicable to an application environment including, but not limited to, the terminal device 1, the application server 2, and the network 3. The terminal device 1 may be a mobile phone, a smart phone, a notebook computer, a digital broadcast receiver, a PDA (personal digital assistant), a PAD (tablet computer), a PMP (portable multimedia player), a navigation device, an in-vehicle device, etc. Mobile devices, etc., as well as fixed terminals such as digital TVs, desktop computers, notebooks, broadband phones, servers, and the like. The application server 2 may be a computing device such as a rack server, a blade server, a tower server, or a rack server. The application server 2 may be a standalone server or a server cluster composed of multiple servers. The network 3 may be an intranet, an Internet, a Global System of Mobile communication (GSM), a Wideband Code Division Multiple Access (WCDMA), a 4G network, Wireless or wired networks such as 5G networks, Bluetooth, Wi-Fi, and call networks.

The application server 2 can be respectively connected to one or more of the terminal devices 1 through the network 3 for data transmission and interaction.

Referring to FIG. 2, it is a schematic diagram of an optional hardware architecture of the application server 2 of the present application.

In this embodiment, the application server 2 may include, but is not limited to, the memory 11, the processor 12, and the network interface 13 being communicably connected to each other through a system bus. It is to be noted that FIG. 2 only shows the application server 2 with components 11-13, but it should be understood that not all illustrated components may be implemented, and more or fewer components may be implemented instead.

The memory 11 includes at least one type of readable storage medium including a flash memory, a hard disk, a multimedia card, a card type memory (eg, SD or DX memory, etc.), a random access memory (RAM), a static Random access memory (SRAM), read only memory (ROM), electrically erasable programmable read only memory (EEPROM), programmable read only memory (PROM), magnetic memory, magnetic disk, optical disk, and the like. In some embodiments, the memory 11 may be an internal storage unit of the application server 2, such as a hard disk or memory of the application server 2. In other embodiments, the memory 11 may also be an external storage device of the application server 2, such as a plug-in hard disk equipped on the application server 2, a smart memory card (SMC), and a secure digital number. (Secure Digital, SD) card, flash card, etc. Of course, the memory 11 can also include both the internal storage unit of the application server 2 and its external storage device. In this embodiment, the memory 11 is generally used to store an operating system installed in the application server 2 and various types of application software, such as program code of the meeting minutes generation system 100. Further, the memory 11 can also be used to temporarily store various types of data that have been output or are to be output.

The processor 12 may be a Central Processing Unit (CPU), controller, microcontroller, microprocessor, or other data processing chip in some embodiments. The processor 12 is typically used to control the overall operation of the application server 2, such as performing control and processing related to data interaction or communication with the terminal device 1. In this embodiment, the processor 12 is configured to run program code or process data stored in the memory 11, such as running the conference minutes generating system and the like.

The network interface 13 may comprise a wireless network interface or a wired network interface, which is typically used to establish a communication connection between the application server 2 and other electronic devices. In this embodiment, the network interface 13 is mainly used to connect the application server 2 to one or more of the terminal devices 1 through the network 3, and the application server 2 and the one or more terminals. A data transmission channel and a communication connection are established between the devices 1.

So far, the hardware structure and functions of the devices related to this application have been described in detail. Hereinafter, various embodiments of the present application will be made based on the above description.

First, the present application proposes a meeting minutes generation system 100.

Referring to FIG. 3, it is a program module diagram of the first embodiment of the meeting minutes generation system 100 of the present application.

In this embodiment, the meeting minutes generating system 100 includes a series of computer program instructions stored in the memory 11, and when the computer program instructions are executed by the processor 12, the meeting minutes generating operation of the embodiments of the present application can be implemented. . In some embodiments, the meeting minutes generation system 100 can be divided into one or more modules based on the particular operations implemented by the various portions of the computer program instructions. For example, in FIG. 3, the meeting minutes generation system 100 can be divided into a content acquisition module 101, an extraction module 102, and a generation module 103. among them:

The content obtaining module 101 is configured to obtain audio record information of a conference, and extract, from the audio record information, the content of each speaker's speech according to the voice feature of each speaker.

In an embodiment, after the conference call starts, the application server 2 collects the conference voice content through each terminal device 1, receives the voice content sent by each terminal device 1 and saves the voice content, and the voice content can be saved into a specified audio format, such as MP3. , wma, wav, etc.

Specifically, when the participant on the side of the terminal device 1 starts speaking, the terminal device 1 collects the voice content through a sound collecting device (for example, a microphone). The terminal device 1 can send the collected voice content to the application server 2 in real time or periodically, or when the participant on the side of the terminal device 1 ends a speech, the terminal device 1 will continuously collect the voice. The content is sent to the application server 2. After receiving the voice content sent by the terminal device 1, the application server 2 saves the voice content.

The content obtaining module 101 can obtain the audio record information of the conference, because the full voice content of the conference is saved on the application server 2. In the present embodiment, the audio recording information is preferably the voice content of the conference. In other embodiments of the present application, if the conference call is a video conference call, the conference record received and saved by the application server 2 is audio and video (voice and video screen) content, and at this time, the audio record information acquired by the content acquisition module 101 Also preferred is the voice content of the conference.

The voice characteristics of each speaker (participant) can be pre-acquired prior to the meeting. Specifically, each participant is preset with a unique ID number. Before the meeting, the voice characteristics of each participant are pre-admitted, and then an identity index table is established according to the voice characteristics and ID number of each participant. The identity index table stores the correspondence between the voice characteristics of each participant and the ID of each participant, thereby enabling confirmation of the membership of the participant. The participants can be from the local or remote speakers.

In an embodiment, the speaker's voice feature may be generated into a speaker model, and the speaker model and the corresponding speaker ID number are stored in the identity index table.

After completing the identity index table of the participant, when it is necessary to analyze a certain piece of voice content in the audio record information belonging to the speaker's speech content, the speaker sound feature of the segment of the voice content needs to be extracted first, and the sound feature is extracted. Compare with each speaker model in the identity index table and get a matching score. If the matching score reaches a preset score, it indicates that the speaker model corresponding to the sound feature parameter exists in the index table, thereby obtaining the speaker ID number and confirming the speaker identity. Otherwise, it indicates that there is no speaker model corresponding to the sound feature in the index table, and a new speaker model and a new ID number are generated according to the sound feature, and stored in the identity index table, so as to facilitate the search for matching.

When performing matching scoring, a UBM model (general background model) and an i-vector extraction algorithm can be used for matching scoring. For example, the i-vector value is calculated from the two pieces of speech content as the sound characteristics of the speaker of the two pieces of speech content. For the two calculated i-vector values, the input is scored by the dot-product algorithm or the PLDA algorithm. If the score exceeds a certain threshold, it is considered that the two speech contents belong to the same speaker. .

By mapping the voice content of each piece of the audio record information to the ID number of the participant, the content acquisition module 101 may extract each audio from the audio record information according to the voice feature of each speaker. The content of a speech by the speaker.

The extracting module 102 is configured to perform keyword extraction on the content of the speech of each of the speakers.

In an embodiment, the voice content of each speaker may be converted into a corresponding text before keyword extraction. Optionally, when the converted text content has multiple segments, the extraction module 102 may first sort the multiple segments of text content in a certain order. For example, the multi-segment text content can be sorted according to the time axis (eg, according to the order in which the text content is generated, the number of sentences, the serial number, etc.).

The extraction module 102 can employ a TF-IDF algorithm to extract keywords for each of the speakers' speech content. The TF-IDF algorithm can be used to assess how important a word is in a spoken text. The importance of a word increases proportionally with the number of times it appears in the text. When performing TF-IDF calculation, the TF-IDF value of a certain word is obtained by word frequency (TF) and inverse document frequency (IDF), and the TF-IDF value is higher if the word is more important to the spoken text. The bigger. Therefore, the extraction module 102 can rank the TF-IDF value in the first few words as the keyword of the utterance text. For example, a word with the TF-IDF value ranked in the top five is used as a keyword for the spoken text.

The generating module 103 is configured to generate a meeting minutes corresponding to the meeting according to the extracted keywords.

In an embodiment, the generating module 103 may generate a meeting minutes based on the extracted keywords in combination with the speaking content to which each keyword belongs. In other implementation manners of the present application, the generating module 103 may further take the speaker's intonation (generally, the higher the intonation of the voice content, correspondingly, the higher the importance of the voice content) as a consideration parameter to generate The meeting minutes.

In an embodiment, the generating module 103 may further process the generated meeting minutes by using an NLP natural language algorithm to generate a more fluent and standardized meeting minutes. The NLP analysis engine based on the NLP natural language algorithm can pre-collect and store a large amount of real corpus, so that the linguistic behavior of the words in the meeting minutes can be revised.

Referring to FIG. 4, it is a program module diagram of a second embodiment of the meeting minutes generating system 100 of the present application. In this embodiment, the meeting minutes generating system 100 includes a series of computer program instructions stored in the memory 11, and when the computer program instructions are executed by the processor 12, the meeting minutes generating operation of the embodiments of the present application can be implemented. . In some embodiments, the meeting minutes generation system 100 can be divided into one or more modules based on the particular operations implemented by the various portions of the computer program instructions. For example, in FIG. 4, the meeting minutes generation system 100 can be divided into a content acquisition module 101, an extraction module 102, a generation module 103, a feature creation module 104, and a transmission module 105. The program modules 101-103 are the same as the first embodiment of the meeting minutes generation system 100 of the present application, and the feature creation module 104 and the transmission module 105 are added thereto. among them:

The feature establishing module 104 is configured to acquire a voice sample of each of the speakers, and extract a sound feature of each of the speakers from a voice sample of each of the speakers.

Specifically, before the conference is performed, each participant is required to perform a conference check-in by voice to obtain a voice sample, thereby realizing pre-admission of the voice of each participant and performing sound feature extraction.

The sending module 105 is configured to send the meeting minutes generated by the generating module 103 to the preset user by mail or fax, or provide a link to the preset user to obtain the meeting minutes. The preset user may be a participant or other pre-designated person.

In an embodiment, the sending module 105 may also encrypt the meeting minutes to ensure data security before storing or transmitting the meeting minutes. For example, the meeting minutes are compressed and encrypted, and the decompression password is a designated password or a password known or agreed by each participant.

In addition, the present application also proposes a method for generating a meeting minutes.

Referring to Fig. 5, it is a schematic flow chart of the implementation of the first embodiment of the method for generating meeting minutes of the present application. In this embodiment, the order of execution of the steps in the flowchart shown in FIG. 5 may be changed according to different requirements, and some steps may be omitted.

Step S502: Acquire audio record information of a conference, and extract the content of each speaker's speech from the audio record information according to the voice feature of each speaker.

Specifically, when the participant on the side of the terminal device 1 starts speaking, the terminal device 1 collects the voice content through a sound collection device (for example, a microphone). The terminal device 1 can send the collected voice content to the application server 2 in real time or periodically, or when the participant on the side of the terminal device 1 ends a speech, the terminal device 1 will continuously collect the voice. The content is sent to the application server 2. After receiving the voice content sent by the terminal device 1, the application server 2 saves the voice content.

Since the full voice content of the conference is saved on the application server 2, the audio record information of the conference can be obtained from the application server 2. In the present embodiment, the audio recording information is preferably the voice content of the conference. In other embodiments of the present application, if the conference call is a video conference call, the conference record received and saved by the application server 2 is audio and video (voice and video picture) content, and at this time, the acquired audio record information is also preferably the same. The voice content of the meeting.

By mapping the voice content of each piece of the audio record information with the ID number of the participant, the voice of each speaker can be extracted from the audio record information according to the voice feature of each speaker. The content of the speech.

Step S504, performing keyword extraction on the content of the speech of each of the speakers.

In an embodiment, the voice content of each speaker may be converted into a corresponding text before keyword extraction. Optionally, when there are multiple segments of the converted text content, the plurality of pieces of text content may be first sorted in a certain order. For example, the multi-segment text content can be sorted according to the time axis (eg, according to the order in which the text content is generated, the number of sentences, the serial number, etc.).

In an embodiment, a TF-IDF algorithm may be employed to extract keywords for each of the speakers' speech content. The TF-IDF algorithm can be used to assess how important a word is in a spoken text. The importance of a word increases proportionally with the number of times it appears in the text. When performing TF-IDF calculation, the TF-IDF value of a certain word is obtained by word frequency (TF) and inverse document frequency (IDF), and the TF-IDF value is higher if the word is more important to the spoken text. The bigger. Therefore, the first few words of the TF-IDF value can be used as the keywords of the speech text. For example, a word with the TF-IDF value ranked in the top five is used as a keyword for the spoken text.

Step S506, generating a meeting minutes corresponding to the meeting according to the extracted keywords.

In an embodiment, the meeting minutes may be generated based on the extracted keywords in combination with the speaking content to which each keyword belongs. In other implementation manners of the present application, the speaker's intonation (generally, the higher the intonation of the voice content, correspondingly, the higher the importance of the voice content) may be further taken as a consideration parameter to generate the conference. summary.

In an embodiment, the generated meeting minutes may be further processed by an NLP natural language algorithm to generate a more fluent and standardized meeting minutes. The NLP analysis engine based on the NLP natural language algorithm can pre-collect and store a large amount of real corpus, so that the linguistic behavior of the words in the meeting minutes can be revised.

Through the above steps S502-S506, the conference minutes generating method proposed by the present application firstly acquires audio record information of the conference, and extracts each of the speakers from the audio record information according to the voice feature of each speaker. The content of the speech; secondly, performing keyword extraction on the content of the speech of each of the speakers; further, generating a meeting minutes corresponding to the meeting according to the extracted keywords; and finally, generating the meeting minutes by mail Or send it to the preset user in the form of a fax, or provide a link to the preset user to obtain the meeting minutes. In this way, it is possible to automatically summarize and generate meeting minutes according to the meeting content records, so that the participants can review the content of the meeting. The participants in the meeting can focus more on the content and process of the meeting. After the meeting, the meeting summary is streamlined and accurate. It can also be used for reference and reference by other people in need. Compared with traditional manual recording, this solution is more efficient and accurate, and saves human resource costs.

Referring to FIG. 6, it is a schematic diagram of an implementation process of a second embodiment of a method for generating a meeting minutes of the present application. In this embodiment, the order of execution of the steps in the flowchart shown in FIG. 6 may be changed according to different requirements, and some steps may be omitted.

Step S500: Acquire a voice sample of each of the speakers, and extract a sound feature of each of the speakers from a voice sample of each of the speakers.

In one embodiment, the speaker's voice characteristics may be generated into a speaker model, and the speaker model and the corresponding speaker ID number are stored in the identity index table.

Step S508, sending the meeting minutes to the preset user by mail or fax, or providing a link to the preset user to obtain the meeting minutes. The preset user may be a participant or other pre-designated person.

In an embodiment, the meeting minutes may also be encrypted prior to storing or transmitting the meeting minutes to ensure data security. For example, compress and encrypt the meeting minutes, decompress the password as a specified password or a password known or agreed by each participant.

Through the above steps S500-S508, the method for generating meeting minutes proposed by the present application firstly acquires a voice sample of each of the speakers, and extracts each of the speakers from the voice samples of each of the speakers. a sound feature; secondly, acquiring audio record information of the conference, and extracting the content of each speaker from the audio record information according to the voice feature of each speaker; and, for each of the speakers The content of the speech of the person is extracted by the keyword; further, the meeting minutes corresponding to the meeting are generated according to the extracted keywords; finally, the generated meeting minutes are sent to the preset user by mail or fax, or The preset user provides a link to obtain the meeting minutes. In this way, it is possible to automatically summarize and generate meeting minutes according to the meeting content records, so that the participants can review the content of the meeting. The participants in the meeting can focus more on the content and process of the meeting. After the meeting, the meeting summary is streamlined and accurate. It can also be used for reference and reference by other people in need. Compared with traditional manual recording, this solution is more efficient and accurate, and saves human resource costs.

The serial numbers of the embodiments of the present application are merely for the description, and do not represent the advantages and disadvantages of the embodiments.

Through the description of the above embodiments, those skilled in the art can clearly understand that the foregoing embodiment method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be through hardware, but in many cases, the former is better. Implementation. Based on such understanding, the technical solution of the present application, which is essential or contributes to the prior art, may be embodied in the form of a software product stored in a storage medium (such as ROM/RAM, disk, The optical disc includes a number of instructions for causing a terminal device (which may be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to perform the methods described in various embodiments of the present application.

The above is only a preferred embodiment of the present application, and is not intended to limit the scope of the patent application, and the equivalent structure or equivalent process transformations made by the specification and the drawings of the present application, or directly or indirectly applied to other related technical fields. The same is included in the scope of patent protection of this application.

Claims

A method for generating a meeting minutes, which is applied to an application server, wherein the method comprises:

Obtaining audio record information of a conference, and extracting, according to the voice feature of each speaker, the content of each speaker's speech from the audio record information;

Keyword extraction for each of the speakers' speeches; and

A meeting minutes corresponding to the meeting are generated according to the extracted keywords.
The method of generating a meeting minutes according to claim 1, wherein the method further comprises:

Acquiring a voice sample of each of the speakers, and extracting a sound feature of each of the speakers from a voice sample of each of the speakers.
The method of generating a meeting minutes according to claim 1, wherein the step of extracting the content of each of the speakers from the audio recording information according to the sound characteristics of each speaker comprises:

Setting an ID number for each of the speakers, and establishing a speaker model according to the voice characteristics of each of the speakers;

Extracting a voice feature of the speaker from the first piece of voice in the audio record information;

Comparing the extracted sound features with the plurality of speaker models and obtaining a matching score; and

The ID number of the speaker of the first segment of speech is determined according to the level of the matching score.
The method of generating a meeting minutes according to claim 2, wherein the step of extracting the content of each of the speakers from the audio recording information according to the sound characteristics of each speaker comprises:

Setting an ID number for each of the speakers, and establishing a speaker model according to the voice characteristics of each of the speakers;

Extracting a voice feature of the speaker from the first piece of voice in the audio record information;

Comparing the extracted sound features with the plurality of speaker models and obtaining a matching score; and

The ID number of the speaker of the first segment of speech is determined according to the level of the matching score.
The method of generating a meeting minutes according to claim 1, wherein the step of performing keyword extraction on the content of each of the speakers of the speaker comprises:

Converting the content of each speaker's speech into text content;

Calculating a TF-IDF value of each word in the text content by a TF-IDF algorithm; and

The words whose top TF-IDF values are ranked are identified as keywords of the speech content and extracted.
The method of generating a meeting minutes according to claim 2, wherein the step of performing keyword extraction on the content of each of the speakers of the speaker comprises:

Converting the content of each speaker's speech into text content;

Calculating a TF-IDF value of each word in the text content by a TF-IDF algorithm; and

The words whose top TF-IDF values are ranked are identified as keywords of the speech content and extracted.
The method of generating a meeting minutes according to claim 1, wherein the step of generating a meeting minutes corresponding to the meeting according to the extracted keywords comprises:

Generating conference subject matter based on the extracted keywords; and

The conference subject matter is processed by a natural language algorithm to generate a conference minutes corresponding to the conference.
The method of generating a meeting minutes according to claim 1, wherein the method further comprises:

Sending the meeting minutes to the preset user by mail or fax, or providing a link to the preset user to obtain the meeting minutes.
An application server, comprising: a memory, a processor, on the memory, a conference minutes generating system executable on the processor, wherein the meeting minutes generating system is used by the processor The following steps are implemented during execution:

Obtaining audio record information of a conference, and extracting, according to the voice feature of each speaker, the content of each speaker's speech from the audio record information;

Keyword extraction for each of the speakers' speeches; and

A meeting minutes corresponding to the meeting are generated according to the extracted keywords.
The application server according to claim 9, wherein said meeting minutes generating system is further implemented when said processor executes:

Acquiring a voice sample of each of the speakers, and extracting a sound feature of each of the speakers from a voice sample of each of the speakers.
The application server according to claim 9, wherein the step of extracting the content of each speaker's speech from the audio recording information according to the sound feature of each speaker comprises:

Setting an ID number for each of the speakers, and establishing a speaker model according to the voice characteristics of each of the speakers;

Extracting a voice feature of the speaker from the first piece of voice in the audio record information;

Comparing the extracted sound features with the plurality of speaker models and obtaining a matching score; and

The ID number of the speaker of the first segment of speech is determined according to the level of the matching score.
The application server according to claim 9, wherein the step of performing keyword extraction on the content of the speech of each of the speakers includes:

Converting the content of each speaker's speech into text content;

Calculating a TF-IDF value of each word in the text content by a TF-IDF algorithm; and

The words whose top TF-IDF values are ranked are identified as keywords of the speech content and extracted.
The application server according to claim 9, wherein the step of generating a meeting minutes corresponding to the meeting according to the extracted keywords comprises:

Generating conference subject matter based on the extracted keywords; and

The conference subject matter is processed by a natural language algorithm to generate a conference minutes corresponding to the conference.
The application server according to claim 9, wherein said meeting minutes generating system is further implemented when said processor executes:

Sending the meeting minutes to the preset user by mail or fax, or providing a link to the preset user to obtain the meeting minutes.
A computer readable storage medium storing a meeting minutes generation system, the meeting minutes generation system being executable by at least one processor to cause the at least one processor to perform the following steps:

Obtaining audio record information of a conference, and extracting, according to the voice feature of each speaker, the content of each speaker's speech from the audio record information;

Keyword extraction for each of the speakers' speeches; and

A meeting minutes corresponding to the meeting are generated according to the extracted keywords.
The computer readable storage medium of claim 15 wherein said meeting minutes generation system is further implemented when said processor is executed:

Acquiring a voice sample of each of the speakers, and extracting a sound feature of each of the speakers from a voice sample of each of the speakers.
The computer readable storage medium according to claim 15, wherein the step of extracting the content of each of the speakers from the audio recording information according to the sound characteristics of each speaker, specifically comprises :

Setting an ID number for each of the speakers, and establishing a speaker model according to the voice characteristics of each of the speakers;

Extracting a voice feature of the speaker from the first piece of voice in the audio record information;

Comparing the extracted sound features with the plurality of speaker models and obtaining a matching score; and

The ID number of the speaker of the first segment of speech is determined according to the level of the matching score.
The computer readable storage medium according to claim 15, wherein the step of performing keyword extraction on the content of the speech of each of the speakers comprises:

Converting the content of each speaker's speech into text content;

Calculating a TF-IDF value of each word in the text content by a TF-IDF algorithm; and

The words whose top TF-IDF values are ranked are identified as keywords of the speech content and extracted.
The computer readable storage medium of claim 15, wherein the step of generating a meeting minutes corresponding to the meeting based on the extracted keywords comprises:

Generating conference subject matter based on the extracted keywords; and

The conference subject matter is processed by a natural language algorithm to generate a conference minutes corresponding to the conference.
The computer readable storage medium of claim 15 wherein said meeting minutes generation system is further implemented when said processor is executed:

Sending the meeting minutes to the preset user by mail or fax, or providing a link to the preset user to obtain the meeting minutes.