A method, system, device and storage medium for implementing a multi-language conference

Abstract

The present invention discloses a method, system, device and storage medium for realizing a multi-language conference, and relates to the technical field of machine translation, including obtaining a user's system language and input voice, recognizing and processing the input voice, and outputting a recognized text; translating and processing the recognized text, and outputting subtitle information and record information; analyzing the record information, and outputting an expected correlation value; calculating the correlation value between conflicting information, and deleting the conflicting information; generating a meeting record, and outputting the meeting record; the present invention is used to solve the problem in the prior art that, due to the lack of analysis of multiple people speaking simultaneously in a meeting, when participants mistakenly touch and enter information irrelevant to the meeting, it is impossible to judge and delete the irrelevant information.

Description

Technical Field

[0001] The present invention relates to the technical field of machine translation, and in particular to a method, system, device and storage medium for implementing a multilingual conference. Background Art

[0002] Machine translation, also known as automatic translation, is the process of converting one natural language (source language) into another natural language (target language) using computers; since it was first proposed in the mid-20th century, machine translation has evolved from simple literal translation to today's highly complex and accurate semantic translation. The development of this technology has not only completely changed the way global information is exchanged, but also had a profound impact on economic, political and cultural exchanges.

[0003] The improvement of the existing implementation method for multi-language conferences usually improves the accuracy of language translation to make the translation results more accurate. For example, in the Chinese patent with publication number CN114913857A, a real-time transcription method, system, device and medium based on a multi-language conference system are disclosed. The scheme is to collect voice information, recognize the voice information, translate the recognized text, and then manually modify the translated text to obtain the final translated text. However, the existing improved method lacks analysis of multiple people speaking at the same time in the meeting. When participants accidentally touch and enter information irrelevant to the meeting, it is impossible to judge and delete the irrelevant information. In view of this, it is necessary to improve the existing implementation method of multi-language conferences. Summary of the invention

[0004] The present invention aims to solve at least one of the technical problems in the prior art to a certain extent. By improving the implementation method of multilingual conferences, it is used to solve the problem in the prior art that due to the lack of analysis of multiple people speaking at the same time in the meeting, when participants mistakenly enter information irrelevant to the meeting, it makes it impossible to judge and delete the irrelevant information.

[0005] To achieve the above objectives, in a first aspect, the present invention provides a method for implementing a multi-language conference, comprising:

[0006] Acquire the user's system language and input voice, wherein the input voice includes input time; perform recognition processing on the input voice based on the system language and output recognition text;

[0007] Translate the recognized text, output subtitle information and record information;

[0008] Analyze the record information based on the input time, and output some conflict information based on the analysis result; calculate the correlation between the record information, output multiple record correlation values, calculate and analyze the record correlation values, and output expected correlation values ​​based on the calculation results;

[0009] Calculate the correlation value between the conflicting information and output the conflicting correlation value; analyze the conflicting correlation value based on the expected correlation value, and delete the conflicting information based on the analysis result;

[0010] Generate meeting minutes based on the record information and output the meeting minutes.

[0011] Furthermore, the system language and input voice of the user are obtained, wherein the input voice includes the input time; the input voice is recognized and processed based on the system language, and the output recognition text includes:

[0012] Obtaining the default language of the mobile device used by the user, marked as the system language; the mobile device also includes a voice input device and a display, the voice input device can receive the user's input language and the input time of the input language, the input time includes a start time and an end time;

[0013] The input voice is recognized and processed based on the system language, the input voice is converted into text information, marked as recognized text and output.

[0014] Furthermore, the recognized text is translated, and the output subtitle information and recorded information include:

[0015] Taking the recognition text of any user as an example, the system language corresponding to the recognition text is marked as the original language; the system languages ​​of other users are obtained and marked as other languages;

[0016] Determine whether the other language is the same as the original language. If they are the same, no processing is performed; if they are not the same, the original language is translated into text information in other languages ​​in real time and marked as subtitle information;

[0017] Displaying subtitle information in real time on a display of a mobile device;

[0018] A deep learning model is used to perform contextual semantic understanding of the subtitle information, and the grammar of the subtitle information and some words in the subtitle information are adjusted, and the adjusted subtitle information is marked as recorded information.

[0019] Furthermore, the recorded information is analyzed based on the input time, and some conflict information is output based on the analysis result, including:

[0020] Arrange the input time corresponding to all the record information in ascending order of the start time to obtain a time series;

[0021] Set the sliding window size to 2 and set the sliding window to the leftmost side of the time series;

[0022] Execute a conflict determination loop, the conflict determination loop comprising:

[0023] The start time of the input time on the right side of the sliding window is marked as the first judgment time; the end time of the input time on the left side is marked as the second judgment time;

[0024] Compare the first judgment time with the second judgment time, and when the first judgment time is less than or equal to the second judgment time, move the sliding window to the right by the length of the input time, and repeat the conflict judgment cycle;

[0025] When the first judgment time is greater than the second judgment time, the record information corresponding to the input time in the sliding window is marked as conflict information, and the two conflict information are assigned the same number that is different from other conflict information; the sliding window is moved to the right by the length of the input time, and the conflict judgment loop is repeated; when the sliding window moves to the rightmost side of the time series, the conflict judgment loop is ended.

[0026] Further, calculating the correlation between the record information, outputting a plurality of record correlation values, performing calculation and analysis on the record correlation values, and outputting expected correlation values ​​based on the calculation results include:

[0027] Take a random record information as an example, mark the record information as analysis information; mark all the record information on the left side of the record information in the time series as reference information;

[0028] Calculate the correlation value between the analysis information and each reference information respectively by using the text matching algorithm, and mark the largest calculated correlation value as the record correlation value;

[0029] All record correlation values ​​are sorted in ascending order to obtain a correlation value sequence; the first proportion of record correlation values ​​located on the far left in the correlation value sequence are deleted, the smallest of the remaining record correlation values ​​is marked as the expected correlation value, and the expected correlation value is output.

[0030] Further, calculating the correlation value between the conflicting information and outputting the conflicting correlation value; analyzing the conflicting correlation value based on the expected correlation value, and deleting or not processing the conflicting information based on the analysis result includes:

[0031] Calculate the correlation value of two conflicting information with the same number using a text matching algorithm and mark them as conflict correlation values;

[0032] Determine the relationship between the conflict correlation value and the expected correlation value. When the conflict correlation value is less than the expected correlation value, delete the conflict information with the smallest start time in the conflict information and mark the remaining conflict information back to the record information. When the conflict correlation value is greater than or equal to the expected correlation value, mark all conflict information back to the record information.

[0033] Further, generating a meeting record based on the record information, outputting the meeting record includes:

[0034] Set the default language of the conference system and translate all recorded information into the recorded text in the default language;

[0035] Sort all recorded texts in order of start time, obtain meeting records, and output the meeting records.

[0036] In a second aspect, the present invention also provides a system for implementing a multi-language conference, comprising a voice input module, a translation analysis module and a record generation module, wherein the voice input module is used to obtain a user's system language and input voice, wherein the input voice includes an input time; the input voice is recognized and processed based on the system language, and a recognition text is output;

[0037] The translation analysis module is used to translate the recognized text, output subtitle information and record information;

[0038] It is also used to analyze the record information based on the input time, and output a number of conflict information based on the analysis result; calculate the correlation between the record information, output a plurality of record correlation values, calculate and analyze the record correlation values, and output the expected correlation value based on the calculation result;

[0039] The record generation module is used to calculate the correlation value between conflicting information and output the conflicting correlation value; analyze the conflicting correlation value based on the expected correlation value and delete the conflicting information based on the analysis result; generate a meeting record based on the record information and output the meeting record.

[0040] In a third aspect, the present application provides an electronic device, including a processor and a memory, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the steps in the method described above are executed.

[0041] In a fourth aspect, the present application provides a storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the steps in the method described above are performed.

[0042] The beneficial effects of the present invention are as follows: the present invention obtains input speech and system language, performs speech recognition on the input language based on the system language to obtain recognized text, then performs real-time translation on the recognized text to obtain subtitle information for display, and then adjusts the subtitle information to obtain record information for generating conference records; the advantage of this is that real-time translation of the recognized text can help participants to promptly understand the meaning expressed by the speaker, thereby improving the timeliness of conference translation; and then adjusting the subtitle information to obtain grammatically correct record information can better help participants to understand, thereby improving the intelligence of conference translation.

[0043] The present invention also obtains conflict information by analyzing the input time and calculates the conflict correlation value between the conflict information; calculates the expected correlation value between the recorded information, compares the expected correlation value with the conflict correlation value, performs subsequent processing on the conflict information, and finally generates a meeting record; the advantage of this is that by calculating the expected correlation value and then analyzing the conflict correlation value based on the expected correlation value, it is possible to determine whether the conflict information is information related to the meeting that needs to be recorded, thereby increasing the analysis of multiple people speaking at the same time and improving the practicality of the meeting.

[0044] Other features and advantages of the present application will be described in the following description, and partly become apparent from the description, or be understood by practicing the present application. The purpose and other advantages of the present application can be realized and obtained by the structures specifically pointed out in the written description, claims, and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 is a functional block diagram of the system of the present invention;

[0046] Figure 2 is a flow chart of the steps of the method of the present invention;

[0047] Figure 3 This is a model principle diagram of the text matching algorithm of the present invention. DETAILED DESCRIPTION

[0048] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions 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 are within the scope of protection of the present invention.

[0049] Example 1, please refer to Figure 1 As shown, the present invention provides a system for implementing a multi-language conference, including: a speech input module, a translation analysis module and a record generation module;

[0050] The voice input module is used to obtain the user's system language and input voice, including the input time; recognize and process the input voice based on the system language, and output the recognized text;

[0051] The voice input module is configured with a voice recognition strategy, which includes:

[0052] Obtain the default language of the mobile device used by the user and mark it as the system language; the mobile device also includes a voice input device and a display, and the voice input device can receive the user's input language and the input time of the input language, and the input time includes a start time and an end time;

[0053] Recognize and process the input voice based on the system language, convert the input voice into text information, mark it as recognized text and output it;

[0054] The translation analysis module is used to translate the recognized text and output subtitle information and record information; it is also used to analyze the record information based on the input time and output a number of conflict information based on the analysis result; calculate the correlation between the record information and output a plurality of record correlation values, calculate and analyze the record correlation values ​​and output the expected correlation value based on the calculation result;

[0055] The translation analysis module is configured with translation optimization strategies, which include:

[0056] Taking the recognition text of any user as an example, the system language corresponding to the recognition text is marked as the original language; the system languages ​​of other users are obtained and marked as other languages;

[0057] Determine whether the other language is the same as the original language. If they are the same, no processing is performed; if they are not the same, the original language is translated into text information in other languages ​​in real time and marked as subtitle information;

[0058] Displaying subtitle information in real time on a display of a mobile device;

[0059] Using a deep learning model to understand the context of the subtitle information, adjusting the grammar of the subtitle information and some words in the subtitle information, and marking the adjusted subtitle information as recorded information;

[0060] It should be noted that during the meeting, the speaker's speech needs to be translated and displayed to other meeting participants in a timely manner. Therefore, a real-time translation tool is used to translate the recognized text into other languages. Taking the recognized text as Chinese and the other language as English as an example, if the recognized text is "The weather is nice today", the AI ​​subtitle technology is used for translation, and the subtitle information obtained may be "today weather not bad". The subtitle information can initially help the meeting participants understand the meaning that the speaker wants to express. At this time, the deep learning language model is used to adjust the subtitle information, and the adjusted recorded information may be "The weather is nice today".

[0061] The translation analysis module is also configured with conflict analysis strategies, which include:

[0062] Arrange the input time corresponding to all the record information in ascending order of the start time to obtain a time series;

[0063] Set the sliding window size to 2 and set the sliding window to the leftmost side of the time series;

[0064] Execute a conflict determination loop, the conflict determination loop comprising:

[0065] The start time of the input time on the right side of the sliding window is marked as the first judgment time; the end time of the input time on the left side is marked as the second judgment time;

[0066] Compare the first judgment time with the second judgment time, and when the first judgment time is less than or equal to the second judgment time, move the sliding window to the right by the length of the input time, and repeat the conflict judgment cycle;

[0067] When the first judgment time is greater than the second judgment time, the record information corresponding to the input time in the sliding window is marked as conflict information, and the two conflict information are assigned the same number that is different from other conflict information; the sliding window is moved rightward by the length of the input time, and the conflict judgment cycle is repeated; when the sliding window moves to the rightmost side of the time sequence, the conflict judgment cycle is terminated;

[0068] It should be noted that, if the first judgment time in the sliding window is greater than the second judgment time for the first time, the two conflicting information in the sliding window are assigned a number of 1, and the two conflicting information numbers are the same at this time; when it occurs for the second time, the two conflicting information are assigned a number of 2, and the two conflicting information numbers are the same at this time, but different from the numbers of the two conflicting information for the first time;

[0069] The translation analysis module is also configured with relevant calculation strategies, including:

[0070] Take a random record information as an example, mark the record information as analysis information; mark all the record information on the left side of the record information in the time series as reference information;

[0071] Calculate the correlation value between the analysis information and each reference information respectively by using the text matching algorithm, and mark the largest calculated correlation value as the record correlation value;

[0072] In specific implementation, the text matching algorithm can be configured as the SimNet algorithm. For the principle of the SimNet algorithm, please refer to Figure 3 As shown;

[0073] All record correlation values ​​are sorted in ascending order to obtain a correlation value sequence; the first proportion of record correlation values ​​located on the far left in the correlation value sequence are deleted, the smallest of the remaining record correlation values ​​is marked as the expected correlation value, and the expected correlation value is output.

[0074] In the specific implementation, the first ratio is set to 20%, that is, there are 100 groups of record-related values, and the 20 groups of record-related values ​​on the left are deleted. The purpose of deletion is to avoid the opening remarks or self-introduction part at the beginning of the meeting. This part is irrelevant to the actual meeting content. It is necessary to delete the part with the lowest correlation value to improve the accuracy of subsequent analysis.

[0075] The record generation module is used to calculate the correlation value between the conflicting information and output the conflicting correlation value; analyze the conflicting correlation value based on the expected correlation value and delete the conflicting information based on the analysis result; generate the meeting record based on the record information and output the meeting record;

[0076] The record generation module is configured with conflict calculation strategies, which include:

[0077] Calculate the correlation value of two conflicting information with the same number using a text matching algorithm and mark them as conflict correlation values;

[0078] Determine the relationship between the conflict correlation value and the expected correlation value. When the conflict correlation value is less than the expected correlation value, delete the conflict information with the smallest start time in the conflict information and mark the remaining conflict information back to the record information. When the conflict correlation value is greater than or equal to the expected correlation value, mark all conflict information back to the record information.

[0079] The record generation module is also configured with a record generation strategy, which includes:

[0080] Set the default language of the conference system and translate all recorded information into the recorded text in the default language;

[0081] Sort all recorded texts in order of start time, obtain meeting records, and output the meeting records.

[0082] Example 2, please refer to Figure 2 As shown, the present invention also provides a method for implementing a multi-language conference, comprising:

[0083] Step S1, obtaining the user's system language and input voice, the input voice including the input time; performing recognition processing on the input voice based on the system language, and outputting the recognition text; Step S1 also includes the following sub-steps:

[0084] Step S101, obtaining the default language of the mobile device used by the user, marked as the system language; the mobile device also includes a voice input device and a display, the voice input device can receive the user's input language and the input time of the input language, the input time includes a start time and an end time;

[0085] Step S102: perform recognition processing on the input voice based on the system language, convert the input voice into text information, mark it as recognized text and output it.

[0086] Step S2, translating the recognized text, outputting subtitle information and recording information; Step S2 also includes the following sub-steps:

[0087] Step S201, taking the recognition text of any user as an example, marking the system language corresponding to the recognition text as the original language; obtaining the system languages ​​of other users and marking them as other languages;

[0088] Step S202, determining whether the other language is the same as the original language, if the other language is the same, no processing is performed; if the other language is not the same, translating the original language into text information in the other language in real time and marking it as subtitle information;

[0089] Step S203, displaying subtitle information in real time on the display of the mobile device;

[0090] Step S204: Use a deep learning model to perform contextual semantic understanding on the subtitle information, adjust the grammar of the subtitle information and some words in the subtitle information, and mark the adjusted subtitle information as record information.

[0091] Step S3, analyzing the recorded information based on the input time, and outputting some conflict information based on the analysis result; calculating the correlation between the recorded information, outputting a plurality of recorded correlation values, calculating and analyzing the recorded correlation values, and outputting the expected correlation values ​​based on the calculation result; Step S3 also includes the following sub-steps:

[0092] Step S3011, arranging the input time corresponding to all the record information in ascending order of the start time to obtain a time series;

[0093] Step S3012, setting a sliding window with a window size of 2, and setting the sliding window to the leftmost side of the time series;

[0094] Step S3013: executing a conflict judgment loop, which includes:

[0095] The start time of the input time on the right side of the sliding window is marked as the first judgment time; the end time of the input time on the left side is marked as the second judgment time;

[0096] Compare the first judgment time with the second judgment time, and when the first judgment time is less than or equal to the second judgment time, move the sliding window to the right by the length of the input time, and repeat the conflict judgment cycle;

[0097] When the first judgment time is greater than the second judgment time, the record information corresponding to the input time in the sliding window is marked as conflict information, and the two conflict information are assigned the same number that is different from other conflict information; the sliding window is moved to the right by the length of the input time, and the conflict judgment loop is repeated; when the sliding window moves to the rightmost side of the time series, the conflict judgment loop is ended.

[0098] Step S3021, taking a random piece of record information as an example, marking the record information as analysis information; marking all the record information on the left side of the record information in the time series as reference information;

[0099] Step S3022, using a text matching algorithm to calculate the correlation values ​​between the analysis information and each reference information, and marking the largest calculated correlation value as a record correlation value;

[0100] Step S3023, sort all the record correlation values ​​in ascending order to obtain a correlation value sequence; delete the first proportion of record correlation values ​​located on the far left in the correlation value sequence, mark the smallest of the remaining record correlation values ​​as the expected correlation value, and output the expected correlation value.

[0101] Step S4, calculating the correlation value between the conflicting information and outputting the conflicting correlation value; analyzing the conflicting correlation value based on the expected correlation value, and deleting the conflicting information based on the analysis result; Step S4 also includes the following sub-steps:

[0102] Step S401, using a text matching algorithm to calculate the correlation value of two conflicting information with the same number, and marking them as conflict correlation values;

[0103] Step S402, determine the relationship between the conflict correlation value and the expected correlation value. When the conflict correlation value is less than the expected correlation value, delete the conflict information with the smallest start time in the conflict information, and mark the remaining conflict information back to the record information; when the conflict correlation value is greater than or equal to the expected correlation value, mark all conflict information back to the record information.

[0104] Step S5, generating a meeting record based on the record information, and outputting the meeting record; Step S5 also includes the following sub-steps:

[0105] Step S501, setting the default language of the conference system, and translating all recorded information into recorded text in the default language;

[0106] Step S502: sort all the record texts in the order of the start time to obtain the meeting record, and output the meeting record.

[0107] Embodiment 3, the present application provides an electronic device, including a processor and a memory, the memory storing computer-readable instructions, when the computer-readable instructions are executed by the processor, the steps in the above method are executed. Through the above technical solution, the processor and the memory are interconnected and communicate with each other through a communication bus and / or other forms of connection mechanisms, the memory stores a computer program executable by the processor, when the electronic device is running, the processor executes the computer program, and when executing, the method in any optional implementation of the above embodiment is executed to achieve the following functions: obtain the user's system language and input voice, recognize the input voice, and output the recognized text; translate the recognized text, output subtitle information and record information; analyze the record information, and output the expected correlation value; calculate the correlation value between conflicting information, and delete the conflicting information; generate meeting records, and output meeting records.

[0108] Embodiment 4, the present application provides a storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps in the above method are executed. Through the above technical solution, when the computer program is executed by the processor, the method in any optional implementation of the above embodiment is executed to achieve the following functions: obtaining the user's system language and input voice, recognizing the input voice, and outputting the recognized text; translating the recognized text, outputting subtitle information and record information; analyzing the record information, and outputting the expected correlation value; calculating the correlation value between conflicting information, and deleting the conflicting information; generating meeting records, and outputting the meeting records.

[0109] In the above embodiments of the present application, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.

[0110] It should be understood by those skilled in the art that the embodiments of the present invention can be provided as methods, systems or computer program products. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program codes. Among them, the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, referred to as SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, referred to as EEPROM), erasable programmable read-only memory (Erasable Programmable Read Only Memory, referred to as EPROM), programmable read-only memory (Programmable Read-Only Memory, referred to as PROM), read-only memory (Read-Only Memory, referred to as ROM), magnetic memory, flash memory, disk or optical disk. These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0111] In the embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interfaces, and the indirect coupling or communication connection of devices or units can be electrical, mechanical or other forms.

Claims

1. A method for implementing a multi-language conference, characterized in that: include: Acquire the user's system language and input voice, where the input voice includes the input time; Recognize and process the input speech based on the system language and output the recognized text; Translate the recognized text, output subtitle information and record information; Analyze the recorded information based on the input time, and output some conflict information based on the analysis result; Calculate the correlation between the record information, output multiple record correlation values, calculate and analyze the record correlation values, and output expected correlation values ​​based on the calculation results; Calculate the correlation value between the conflicting information and output the conflicting correlation value; analyze the conflicting correlation value based on the expected correlation value, and delete the conflicting information based on the analysis result; Generate meeting minutes based on the recorded information and output the meeting minutes; Acquire the user's system language and input voice, where the input voice includes the input time; The input speech is recognized and processed based on the system language, and the output recognition text includes: Obtaining the default language of the mobile device used by the user, marked as the system language; the mobile device also includes a voice input device and a display, the voice input device can receive the user's input language and the input time of the input language, the input time includes a start time and an end time; Recognize and process the input voice based on the system language, convert the input voice into text information, mark it as recognized text and output it; Translate the recognized text, output subtitle information and record information including: Taking the recognition text of any user as an example, the system language corresponding to the recognition text is marked as the original language; the system languages ​​of other users are obtained and marked as other languages; Determine whether the other language is the same as the original language. If they are the same, no processing is performed; if they are not the same, the original language is translated into text information in other languages ​​in real time and marked as subtitle information; Displaying subtitle information in real time on a display of a mobile device; Using a deep learning model to understand the context of the subtitle information, adjusting the grammar of the subtitle information and some words in the subtitle information, and marking the adjusted subtitle information as recorded information; The recorded information is analyzed based on the input time, and some conflict information is output based on the analysis results, including: Arrange the input time corresponding to all the record information in ascending order of the start time to obtain a time series; Set the sliding window size to 2 and set the sliding window to the leftmost side of the time series; Execute a conflict determination loop, the conflict determination loop comprising: The start time of the input time on the right side of the sliding window is marked as the first judgment time; the end time of the input time on the left side is marked as the second judgment time; Compare the first judgment time with the second judgment time, and when the first judgment time is less than or equal to the second judgment time, move the sliding window to the right by the length of the input time, and repeat the conflict judgment cycle; When the first judgment time is greater than the second judgment time, the record information corresponding to the input time in the sliding window is marked as conflict information, and the two conflict information are assigned the same number that is different from other conflict information; the sliding window is moved rightward by the length of the input time, and the conflict judgment cycle is repeated; when the sliding window moves to the rightmost side of the time sequence, the conflict judgment cycle is terminated; Calculate the correlation between record information, output multiple record correlation values, calculate and analyze the record correlation values, and output expected correlation values ​​based on the calculation results, including: Take a random record information as an example, mark the record information as analysis information; mark all the record information on the left side of the record information in the time series as reference information; Calculate the correlation value between the analysis information and each reference information respectively by using the text matching algorithm, and mark the largest calculated correlation value as the record correlation value; All the record correlation values ​​are sorted in ascending order to obtain a correlation value sequence; the record correlation values ​​of the first proportion located at the far left in the correlation value sequence are deleted, the smallest of the remaining record correlation values ​​is marked as the expected correlation value, and the expected correlation value is output; Calculate the correlation value between the conflicting information and output the conflicting correlation value; analyze the conflicting correlation value based on the expected correlation value, and delete or do not process the conflicting information based on the analysis result, including: Calculate the correlation value of two conflicting information with the same number using a text matching algorithm and mark them as conflict correlation values; Determine the relationship between the conflict correlation value and the expected correlation value. When the conflict correlation value is less than the expected correlation value, delete the conflict information with the smallest start time in the conflict information and mark the remaining conflict information back to the record information. When the conflict correlation value is greater than or equal to the expected correlation value, mark all conflict information back to the record information.

2. A method for implementing a multi-language conference according to claim 1, characterized in that: Generate meeting records based on the recorded information. The output meeting records include: Set the default language of the conference system and translate all recorded information into the recorded text in the default language; Sort all recorded texts in order of start time, obtain meeting records, and output the meeting records.

3. A system for implementing a multi-language conference, applicable to the method for implementing a multi-language conference as claimed in claim 1 or 2, characterized in that: It includes a voice input module, a translation analysis module and a record generation module. The voice input module is used to obtain the user's system language and input voice, and the input voice includes the input time; the input voice is recognized and processed based on the system language, and the recognition text is output; The translation analysis module is used to translate the recognized text, output subtitle information and record information; It is also used to analyze the record information based on the input time, and output a number of conflict information based on the analysis result; calculate the correlation between the record information, output a plurality of record correlation values, calculate and analyze the record correlation values, and output the expected correlation value based on the calculation result; The record generation module is used to calculate the correlation value between the conflicting information and output the conflict correlation value; Analyze the conflicting correlation values ​​based on the expected correlation values, and delete the conflicting information based on the analysis results; Generate meeting minutes based on the record information and output the meeting minutes.

4. An electronic device, characterized in that: The method comprises a processor and a memory, wherein the memory stores computer-readable instructions. When the computer-readable instructions are executed by the processor, the steps in the method according to claim 1 or 2 are executed.

5. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps in the method according to claim 1 or 2 are executed.

Images