Artificial intelligence-based dual review assistance method, apparatus, device, and medium

Abstract

The application relates to the technical field of artificial intelligence, and discloses a double-recording examination auxiliary method and device based on artificial intelligence, equipment and a medium, which comprises the following steps: receiving a double-recording key and a double-recording abstract sent by a double-recording device, and acquiring double-recording information from a cloud storage system according to the double-recording key; cutting the double-recording information according to the double-recording abstract to obtain at least one double-recording segment, and extracting audio information in the double-recording segment; converting the audio information to obtain text information, and associating the double-recording abstract with the text information; receiving a quality inspection request sent by a control end, and sending the text information to the control end according to the quality inspection request. The application guarantees the simplification and accuracy of the text information, improves the efficiency of quality inspection personnel in checking the double-recording information and monitoring whether the on-site personnel business operation is standardized, and reduces the fatigue strength of the quality inspection personnel.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of artificial intelligence, and in particular to a double-recording examination assisting method and device based on artificial intelligence, equipment and medium. BACKGROUND

[0002] The double-recording information is data information generated by recording the business handling of the on-site personnel, which is used to shoot the double-recording information by a double-recording device, and save the double-recording information in a cloud storage system, so as to monitor the operation specification degree and business level of the on-site personnel, and save the evidence of the business-related operation and behavior. Therefore, the current quality inspection personnel needs to watch the double-recording information throughout the process, so as to identify the non-compliance of the business handled by the on-site personnel.

[0003] However, the inventor realizes that, since the duration of the double-recording information is usually long, the quality inspection personnel needs to watch the double-recording video throughout the process, which not only has a high fatigue intensity, but also has a low quality inspection efficiency. SUMMARY

[0004] The purpose of the present application is to provide a double-recording examination assisting method and device based on artificial intelligence, equipment and medium, which is used to solve the problem of long duration of the double-recording information, which needs the quality inspection personnel to watch the double-recording video throughout the process, resulting in high fatigue intensity and low quality inspection efficiency.

[0005] To achieve the above-mentioned purpose, the present application provides a double-recording examination assisting method based on artificial intelligence, which runs in a double-recording background server connected with a double-recording device and a cloud storage system respectively, comprising:

[0006] Receiving the double-recording secret key and the double-recording abstract sent by the double-recording device, and obtaining the double-recording information from the cloud storage system according to the double-recording secret key; wherein the double-recording information is video data and audio information shot for the whole business, the double-recording secret key is the secret key information generated by the double-recording device when uploading the shot double-recording information to the cloud storage system, and the double-recording abstract describes the time period of each business link under the whole business;

[0007] According to the double-recording abstract, the double-recording information is cut to obtain at least one double-recording segment, and the audio information in the double-recording segment is extracted;

[0008] The audio information is converted to obtain text information, and the double-recording abstract is associated with the text information; wherein the double-recording segment is video data and audio information shot for a specific business link under the whole business;

[0009] Receiving the quality inspection request sent by the control end, and sending the text information to the control end according to the quality inspection request.

[0010] According to the double-recording key, the method further comprises:

[0011] According to the double-recording summary, a text block is constructed in a preset database.

[0012] After the double-recording summary is associated with the text information, the method further comprises:

[0013] The text block corresponding to the double-recording summary associated with the text information in the database is identified, the text information with the preset format is displayed, and / or the text information associated with the timestamp is saved to the text module.

[0014] According to the double-recording summary, a text block is constructed in a preset database.

[0015] A cache block for saving data is constructed in the database.

[0016] The double-recording time period in the double-recording summary is extracted, the double-recording time period is entered into the cache header of the cache block, and the cache block is converted into the text block; wherein the double-recording time period represents the start time and the end time of the double-recording segment, and the cache header is description information for describing the cached data in the text block.

[0017] The start time in the double-recording time period is extracted, and the text blocks in the database are arranged in ascending order according to the start time.

[0018] After the double-recording summary is constructed in the preset database, the method further comprises:

[0019] The attribute information in the double-recording summary is extracted, the attribute information is entered into the configuration file of the database, and the double-recording information is associated with the database.

[0020] According to the double-recording summary, the double-recording information is cut to obtain at least one double-recording segment, and the audio information in the double-recording segment is extracted.

[0021] The double-recording time period in the double-recording summary is extracted; wherein the double-recording time period represents the start time and the end time of the business link.

[0022] According to the double-recording time period, the double-recording information is cut to form at least one double-recording segment, and the audio information in the double-recording segment is extracted.

[0023] According to the double-recording summary, a text block is constructed in a preset database.

[0015] A cache block for saving data is constructed in the database.

[0016] The double-recording time period in the double-recording summary is extracted, the double-recording time period is entered into the cache header of the cache block, and the cache block is converted into the text block; wherein the double-recording time period represents the start time and the end time of the double-recording segment, and the cache header is description information for describing the cached data in the text block.

[0017] The start time in the double-recording time period is extracted, and the text blocks in the database are arranged in ascending order according to the start time.

[0018] After the double-recording summary is constructed in the preset database, the method further comprises:

[0019] The attribute information in the double-recording summary is extracted, the attribute information is entered into the configuration file of the database, and the double-recording information is associated with the database.

[0020] According to the double-recording summary, the double-recording information is cut to obtain at least one double-recording segment, and the audio information in the double-recording segment is extracted.

[0021] The double-recording time period in the double-recording summary is extracted; wherein the double-recording time period represents the start time and the end time of the business link.

[0022] According to the double-recording time period, the double-recording information is cut to form at least one double-recording segment, and the audio information in the double-recording segment is extracted.

[0023] According to the double-recording summary, a text block is constructed in a preset database.

[0015] A cache block for saving data is constructed in the database.

[0016] The double-recording time period in the double-recording summary is extracted, the double-recording time period is entered into the cache header of the cache block, and the cache block is converted into the text block; wherein the double-recording time period represents the start time and the end time of the double-recording segment, and the cache header is description information for describing the cached data in the text block.

[0017] The start time in the double-recording time period is extracted, and the text blocks in the database are arranged in ascending order according to the start time.

[0018] After the double-recording summary is constructed in the preset database, the method further comprises:

[0019] The attribute information in the double-recording summary is extracted, the attribute information is entered into the configuration file of the database, and the double-recording information is associated with the database.

[0020] According to the double-recording summary, the double-recording information is cut to obtain at least one double-recording segment, and the audio information in the double-recording segment is extracted.

[0021] The double-recording time period in the double-recording summary is extracted; wherein the double-recording time period represents the start time and the end time of the business link.

[0022] According to the double-recording time period, the double-recording information is cut to form at least one double-recording segment, and the audio information in the double-recording segment is extracted.

[0023]

[0024] uploading the audio information to the cloud storage system, receiving an audio key generated by the cloud storage system according to the audio information;

[0025] judging whether the number of obtained audio keys is consistent with the number of the double-recording time periods;

[0026] if yes, recording the audio key in a text block recording the double-recording time period in the database;

[0027] if no, deleting the audio key and ending;

[0028] receiving a conversion request sent by a control end, extracting the audio key from the database according to the conversion request, and downloading the audio information from the cloud storage system according to the audio key;

[0029] the conversion of the audio information to obtain text information, and the association of the double-recording summary with the text information, comprising:

[0030] calling a preset conversion engine to convert the audio information to obtain text information;

[0031] obtaining an audio key corresponding to the audio information and deleting the audio information, identifying a text block recording the audio key in the database, recording the text information in the text block, and associating the text information with the double-recording summary in the text block.

[0032] In the above scheme, after the association of the double-recording summary with the text information, the method further comprises:

[0033] annotating a keyword in the text information, so that the keyword is displayed in a preset format in the text information; wherein the keyword comprises a keyword word, a sensitive word and a violation word; wherein the keyword word represents a key node in the business link; and the format comprises a key format for annotating the keyword word, a sensitive format for annotating the sensitive word, and a violation format for annotating the violation word.

[0034] In the above scheme, after the association of the double-recording summary with the text information, the method further comprises:

[0035] obtaining audio information corresponding to the text information in which the keyword is located, obtaining a playing time of the keyword in the audio information, and making a timestamp of the playing time, and associating the timestamp with the keyword in the text information;

[0036] after the association of the timestamp with the keyword in the text information, the method further comprises:

[0037] uploading the timestamp and the keyword into a blockchain.

[0038] To achieve the above object, the application further provides a double-recording review auxiliary device based on artificial intelligence, which is installed in a double-recording background server connected with a double-recording device and a cloud storage system respectively, and comprises:

[0039] a double-recording input module, configured to receive a double-recording key and a double-recording abstract sent by the double-recording device and acquire double-recording information from the cloud storage system according to the double-recording key, wherein the double-recording information is video data and audio information shot for an overall service, the double-recording key is key information generated when the double-recording device uploads the shot double-recording information to the cloud storage system, and the double-recording abstract describes time periods of each service link under the overall service;

[0040] an audio extraction module, configured to cut the double-recording information to obtain at least one double-recording clip according to the double-recording abstract and extract audio information in the double-recording clip;

[0041] a text conversion module, configured to convert the audio information to obtain text information and associate the double-recording abstract with the text information, wherein the double-recording clip is video data and audio information shot for a specific service link under the overall service;

[0042] a text output module, configured to receive a quality inspection request sent by a control end, and send the text information to the control end according to the quality inspection request.

[0043] To achieve the above object, the application further provides a computer device, which comprises a memory, a processor and a computer program stored in the memory and executable on the processor, and the processor of the computer device implements the steps of the double-recording review auxiliary method based on artificial intelligence when executing the computer program.

[0044] To achieve the above object, the application further provides a computer readable storage medium, which stores a computer program, and the computer program stored in the readable storage medium implements the steps of the double-recording review auxiliary method based on artificial intelligence when executed by a processor.

[0045] The application provides a double-recording review auxiliary method, device, equipment and medium based on artificial intelligence, at least one double-recording segment is obtained by cutting the double-recording information according to the double-recording abstract, and the audio information in the double-recording segment is extracted, so that the subsequent obtained text information only reflects the content of each business link in the double-recording information, and does not involve the content of other irrelevant businesses, ensuring the simplification and accuracy of the text information, and eliminating the content of irrelevant business links in the obtained audio information, ensuring the simplification and accuracy of the obtained audio information.

[0046] The text information is obtained by converting the audio information, and the double-recording abstract is associated with the text information, so that the quality inspection personnel only need to quickly browse the content in the text information to know whether the double-recording information has a preset keyword, and specifically find the content in the double-recording information according to the keyword, thereby improving the efficiency of the quality inspection personnel in checking the double-recording information and monitoring whether the on-site personnel business operation is standardized, and reducing the fatigue strength of the quality inspection personnel. BRIEF DESCRIPTION OF DRAWINGS

[0047] Figure 1 The flowchart of the double-recording review auxiliary method of the application based on artificial intelligence is shown in the first embodiment.

[0048] Figure 2 The environmental application schematic diagram of the double-recording review auxiliary method in the second embodiment of the double-recording review auxiliary method based on artificial intelligence is shown.

[0049] Figure 3 The specific method flowchart of the double-recording review auxiliary method in the second embodiment of the double-recording review auxiliary method based on artificial intelligence is shown.

[0050] Figure 4 The program module schematic diagram of the double-recording review auxiliary device in the third embodiment of the double-recording review auxiliary device based on artificial intelligence is shown.

[0051] Figure 5 The hardware structure schematic diagram of the computer equipment in the fourth embodiment of the computer equipment is shown. DETAILED DESCRIPTION

[0052] In order to make the purpose, technical scheme and advantages of the application clearer, the application is further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the application, and are not used to limit the application. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of the application.

[0053] The application provides a double-recording review auxiliary method, device and equipment based on artificial intelligence and a medium.

[0054] Embodiment one:

[0055] Please refer to Figure 1 The double-recording review auxiliary method based on artificial intelligence of the embodiment is run in a double-recording background server connected with a double-recording device and a cloud storage system respectively, and includes the following steps.

[0056] S101: receiving a double-recording secret key and a double-recording abstract sent by a double-recording device, and obtaining double-recording information from a cloud storage system according to the double-recording secret key; wherein the double-recording information is video data and audio information of an overall business, the double-recording secret key is secret key information generated when the double-recording device uploads the double-recording information to the cloud storage system, and the double-recording abstract describes time periods of each business link under the overall business.

[0057] S103: cutting the double-recording information to obtain at least one double-recording segment according to the double-recording abstract, and extracting audio information in the double-recording segment.

[0058] S104: converting the audio information to obtain text information, and associating the double-recording abstract with the text information; wherein the double-recording segment is video data and audio information of a specific business link under the overall business.

[0059] S108: receiving a quality inspection request sent by a control end, and sending the text information to the control end according to the quality inspection request.

[0060] In the exemplary embodiment, the dual-recording information is uploaded to the cloud storage system for storage by receiving the dual-recording key sent by the dual-recording device and obtaining the dual-recording information from the cloud storage system according to the dual-recording key, and the dual-recording device does not store the dual-recording information, and the dual-recording information can be obtained from the cloud storage system only by the server / terminal that obtains the dual-recording key, so that the dual-recording device existing in a complex place as a separate device will not cause the dual-recording information to be leaked due to being accessed or cracked by an unauthorized client, ensuring the safety of customer information and avoiding the risk of customer information leakage caused by the unauthorized client invading the dual-recording device to cause the dual-recording information to be leaked. The dual-recording key and the dual-recording digest are data information generated based on the http protocol.

[0061] By cutting the dual-recording information according to the dual-recording digest to obtain at least one dual-recording segment, and extracting the audio information in the dual-recording segment, the subsequent obtained text information only reflects the content of each business link in the dual-recording information, and does not involve the content of other irrelevant businesses, ensuring the simplification and accuracy of the text information, and further eliminating the irrelevant business link content in the obtained audio information, ensuring the simplification and accuracy of the obtained audio information.

[0062] By converting the audio information to obtain text information, and associating the dual-recording digest with the text information, the quality inspection personnel only need to quickly browse the content in the text information to know whether the dual-recording information has a preset keyword, and specifically find the content in the dual-recording information according to the keyword, improving the efficiency of the quality inspection personnel in checking the dual-recording information and monitoring whether the on-site personnel's business operation is standardized, and reducing the fatigue intensity of the quality inspection personnel.

[0063] By sending the text information to the control end according to the quality inspection request sent by the control end, the text information corresponding to the target digest in the quality inspection request is obtained, and the technical effect of configurable display of the text information of the dual-recording information is realized.

[0064] Embodiment two:

[0065] This embodiment is a specific application scenario of the above-mentioned embodiment one, and through this embodiment, the method provided by the application can be more clearly and specifically described.

[0066] In the following, the method provided by the embodiment is specifically described by taking, in a dual-recording background server running a dual-recording review auxiliary method, cutting dual-recording information according to a dual-recording digest to obtain at least one dual-recording segment, extracting audio information in the dual-recording segment, and converting the audio information to obtain text information. It should be noted that this embodiment is only exemplary and does not limit the scope of the embodiments of the application.

[0067] Figure 2 An environment application schematic diagram of the dual-recording review assistance method according to Embodiment Two of the present application is schematically shown.

[0068] In an exemplary embodiment, a dual-recording background server 2 of the dual-recording review assistance method is connected to a dual-recording device 3, a cloud storage system 4 and a control terminal 5 respectively through a network; the dual-recording background server 2 can provide services through one or more networks, and the network can include various network devices such as routers, switches, multiplexers, hubs, modems, bridges, repeaters, firewalls, proxy devices and / or the like. The network 3 can include physical links such as coaxial cable links, twisted pair cable links, fiber optic links, combinations thereof and / or the like. The network can include wireless links such as cellular links, satellite links, Wi-Fi links and / or the like; the control terminal 5 can be a computer device such as a smartphone, a tablet computer, a notebook computer, a desktop computer and / or the like.

[0069] Figure 3 is a specific method flowchart of a dual-recording review assistance method provided by an embodiment of the present application, which specifically includes steps S201 to S208.

[0070] S201: receiving a dual-recording secret key and a dual-recording summary sent by a dual-recording device, and obtaining dual-recording information from a cloud storage system according to the dual-recording secret key; wherein the dual-recording information is video data and audio information shot for an overall service, the dual-recording secret key is secret key information generated by the dual-recording device when uploading the shot dual-recording information to the cloud storage system, and the dual-recording summary describes time periods of each service link under the overall service.

[0071] To ensure the security of the dual-recording information and avoid the risk of customer information leakage caused by unauthorized clients invading the dual-recording device and causing the dual-recording information to leak, in this step, the dual-recording information is uploaded to the cloud storage system for storage by receiving the dual-recording secret key sent by the dual-recording device and obtaining the dual-recording information from the cloud storage system according to the dual-recording secret key, and the dual-recording information is not saved in the dual-recording device, and the dual-recording information can only be obtained from the cloud storage system by a server / terminal that has obtained the dual-recording secret key, so that the dual-recording device existing in a complex place as an independent device will not cause the dual-recording information to leak due to being accessed or cracked by unauthorized clients, and the security of customer information is ensured. The dual-recording secret key and the dual-recording summary are data information generated based on an http protocol.

[0072] S202: constructing a text block in a preset database according to the dual-recording summary.

[0073] To ensure that the corresponding double recording information fragments of each business link can correspond to the double recording summary and be sequentially saved in the database, the step is implemented by constructing a text block in the database according to the double recording summary, realizing management and sorting of the text information saved in the database through the text block, so that the quality inspection personnel can obtain the text information of the double recording fragment required according to the double recording summary, and the text information generated according to the double recording fragment can be sequentially displayed.

[0074] In a preferred embodiment, the text block is constructed in the preset database according to the double recording summary, comprising:

[0075] S21: A cache block for saving data is constructed in the database.

[0076] The step is implemented by constructing a cache block for saving subsequent double recording fragments in an independent database, so as to independently save each double recording fragment, so as to sort and select each double recording fragment.

[0077] S22: The double recording time period in the double recording summary is extracted, the double recording time period is entered into the cache header of the cache block, and the cache block is converted into the text block; wherein the double recording time period represents the start time and end time of the double recording fragment, and the cache header is description information for describing the cache data in the text block.

[0078] In this step, the buffer head contains all management data related to the buffer state, including fast number, block length, access counter, etc. These data are not directly stored after the buffer head, but in an independent area of the physical memory, represented by a corresponding pointer in the buffer head structure.

[0079] S23: The start time in the double recording time period is extracted, and the text blocks in the database are arranged in ascending order according to the start time.

[0080] In this step, the text blocks in the database are arranged in ascending order according to the start time, so that the quality inspection personnel can obtain the text information required from the text blocks according to the time logic.

[0081] Further, the double recording summary further comprises attribute information, the attribute information comprising a double recording number and a video batch number of the double recording information, the double recording number being a unique identifier of the double recording device, and the video batch number being a unique number of the double recording information.

[0082] After the text block is constructed in the preset database according to the double recording summary, the method further comprises:

[0083] S24: extracting attribute information in the double recording summary, inputting the attribute information into a configuration file of the database, and associating the double recording information with the database.

[0084] This step enables the quality inspection personnel to quickly obtain the required double recording equipment and the double recording information recorded by the double recording equipment by associating the database with the double recording information by using the attribute information.

[0085] S203: cutting the double recording information according to the double recording summary to obtain at least one double recording clip, and extracting audio information in the double recording clip.

[0086] To eliminate the irrelevant business link content in the obtained audio information and ensure the accuracy of the obtained audio information, this step enables the subsequent obtained text information to only reflect the content about each business link in the double recording information and not involve other irrelevant business content by cutting the double recording information according to the double recording summary to obtain at least one double recording clip and extracting audio information in the double recording clip, thereby ensuring the accuracy and precision of the text information.

[0087] In a preferred embodiment, the cutting the double recording information according to the double recording summary to obtain at least one double recording clip and extracting audio information in the double recording clip comprises:

[0088] S31: extracting a double recording time period in the double recording summary; wherein the double recording time period represents the start time and the end time of the business link.

[0089] S32: cutting the double recording information according to the double recording time period to form at least one double recording clip, and extracting audio information in the double recording clip.

[0090] In this embodiment, the double recording information is cut according to the double recording time period, so that the subsequent obtained text information only reflects the content about each business link in the double recording information and does not involve other irrelevant business content, thereby ensuring the accuracy and precision of the text information. Further, the double recording information can be cut by using a video lossless segmentation software Bandicut to obtain the double recording clip.

[0091] S204: converting the audio information to obtain text information, and associating the double recording summary with the text information; wherein the double recording clip is video data and audio information shot for a specific business link under the overall business.

[0092] In order to improve the efficiency of the quality inspection personnel in checking the double recording information, the text information is obtained by converting the audio information, and the double recording summary is associated with the text information, so that the quality inspection personnel only needs to quickly browse the content in the text information to know whether there is a preset keyword in the double recording information, and according to the keyword, the specific content in the double recording information is found, the efficiency of the quality inspection personnel in checking the double recording information and monitoring whether the on-site personnel's business operation is standardized is improved, and the fatigue strength of the quality inspection personnel is reduced.

[0093] In a preferred embodiment, before the audio information is converted to obtain text information, and the double recording summary is associated with the text information, the method further comprises:

[0094] S41: uploading the audio information to the cloud storage system and receiving the audio key generated by the cloud storage system according to the audio information.

[0095] In order to ensure the security of the obtained audio information, after obtaining the audio information, it is uploaded to the cloud storage system to obtain the corresponding audio key, so as to reduce the time of saving the audio information in the server, and avoid the situation that the unauthorized client or malicious client accesses or cracks the server, resulting in client information leakage.

[0096] S42: determining whether the number of obtained audio keys is consistent with the number of double recording time periods.

[0097] In order to avoid errors in the step of extracting double recording segments according to double recording time periods, so that the obtained audio information cannot correspond to the double recording time periods, and the final text information cannot match the corresponding double recording segments, in this step, the number of audio keys and double recording time periods are compared by judging whether the number of obtained audio keys is consistent with the number of double recording time periods, so as to ensure that the obtained audio information can correspond to the double recording time periods one by one.

[0098] S43: if yes, the audio key is recorded in the text block in the database recording the double recording time period.

[0099] S44: if no, delete the audio key and end.

[0100] S45: receiving the conversion request sent by the control end, extracting the audio key from the database according to the conversion request, and downloading the audio information from the cloud storage system according to the audio key.

[0101] In this step, the conversion request is a request information generated based on the http protocol.

[0102] In a preferred embodiment, the converting the audio information into text information and associating the dual recording summary with the text information comprises:

[0103] S46: calling a preset conversion engine to convert the audio information into text information.

[0104] In this step, the ASR engine is used as the conversion engine, and the ASR engine is a technology for converting human speech into text.

[0105] S47: obtaining an audio key corresponding to the audio information and deleting the audio information, identifying a text block in the database recording the audio key, entering the text information into the text block, and associating the text information with the dual recording summary in the text block.

[0106] In this step, by using the audio key as an intermediate link, the text information associated with the audio key is entered into the text block recording the audio key, so that only the text information and the audio key are saved in the server, and the audio information and the dual recording information are not saved, thereby ensuring the security of the business dual recording information and the audio.

[0107] S205: labeling the keywords in the text information, and displaying the keywords in the text information in a preset format.

[0108] In order to facilitate the quality inspection personnel to quickly find the part that may appear in the text information The risk of non-compliance, this step labels the keywords in the text information, and displays the keywords in the text information in a preset format, so that the keywords are labeled in a special format, so that the quality inspection personnel can quickly find and review the relevant dual recording fragments.

[0109] In a preferred embodiment, the keywords include keyword words, sensitive words, and violation words; wherein the keyword words represent the key nodes in the business process, such as transaction amount and transaction object; the sensitive words refer to words with insulting and discriminatory nature; and the violation words refer to words that do not comply with current laws and regulations.

[0110] The format includes a key format for labeling keyword words, such as labeling the keyword words with a specified color;

[0111] a sensitive format for labeling the sensitive words, such as bolding the sensitive words;

[0112] a violation format for labeling the violation words, such as tilting the violation words.

[0113] The key word in the text information is marked, and the key word is displayed in a preset format in the text information, including:

[0114] S51: The key word in the text information is identified by a preset key regular expression, and the format of the key word is set as a key format;

[0115] S52: The sensitive word in the text information is identified by a preset sensitive regular expression, and the format of the sensitive word is set as a sensitive format;

[0116] S53: The rule violation word in the text information is identified by a preset rule violation regular expression, and the format of the rule violation word is set as a rule violation format.

[0117] In this embodiment, the regular expression describes a pattern of string matching, which can be used to check whether a string contains a certain substring, replace the matching substring, or extract a substring that meets a certain condition from a certain string; therefore, by using the regular expression, the key regular expression with the key word, the sensitive regular expression with the sensitive word, and the rule violation regular expression with the rule violation word are constructed, the entire text information is checked, and the formats of the key word, the sensitive word, and the rule violation word are set as the key format, the sensitive format, and the rule violation format respectively, so that the quality inspection personnel can quickly capture the key word, the sensitive word, and the rule violation word, and the quality inspection efficiency of the quality inspection personnel on the text information is improved.

[0118] S206: Obtain the audio information corresponding to the text information where the key word is located, obtain the playing time of the key word in the audio information, and make a timestamp of the playing time, and associate the timestamp with the key word in the text information.

[0119] To facilitate the quality inspection personnel to quickly obtain the part of the double recording segment playing the key word, and further improve the double recording checking efficiency; in this step, the audio information corresponding to the text information where the key word is located is obtained, the playing time of the key word in the audio information is obtained, and the timestamp of the playing time is made, and the timestamp is associated with the key word in the text information, so that the quality inspection personnel can obtain the position of playing the key word in the double recording segment through the timestamp associated with the key word, and the efficiency of double recording information checking is improved.

[0120] In a preferred embodiment, the audio information corresponding to the text information where the key word is located is obtained, and the playing time of the key word in the audio information is obtained, including:

[0121] S61: dividing the text information into at least one text segment according to the punctuation marks as the divider,

[0122] In this step, the punctuation marks can be set as "period", and / or "semicolon", and / or "comma", and / or "question mark", and / or "exclamation mark".

[0123] S62: obtaining the text segment with the keyword and identifying the playing time of each text segment in the audio information,

[0124] In this step, the playing time of the audio data corresponding to the text segment in the audio information when the text segment is obtained by the conversion engine (ASR engine) is set as the playing time of the keyword in the audio information; wherein the conversion engine will make the audio (each audio corresponds to a word / phrase) in the audio information and the word / phrase in the text information one-to-one corresponding when converting the audio information into text information; therefore, the audio data corresponding to the text segment with the keyword can be obtained through the conversion engine, wherein the audio data records the audio of all words in the text segment, and the position of the audio data in the audio information, i.e. the playing time, is obtained, so as to obtain the playing time of the keyword in the audio information.

[0125] Further, the playing time is converted into the timestamp by a timestamp conversion tool (for example: UNIX Time Stamp), and the timestamp is associated with the text information by taking the timestamp as the hyperlink of the keyword.

[0126] Preferably, after the timestamp is associated with the keyword in the text information, the method further comprises:

[0127] uploading the timestamp and the keyword to the blockchain.

[0128] It should be noted that the corresponding summary information is obtained based on the timestamp and the keyword, specifically, the summary information is obtained by hashing the timestamp and the keyword, such as using sha256s algorithm. Uploading the summary information to the blockchain can ensure its security and fairness and transparency to the user. The user equipment can download the summary information from the blockchain to verify whether the timestamp and the keyword are tampered. The blockchain referred to in this example is a new application mode of distributed data storage, peer-to-peer transmission, consensus mechanism, encryption algorithm and other computer technologies. Blockchain, in essence, is a decentralized database, a string of data blocks associated using cryptography, each containing a batch of network transaction information for verifying its validity (anti-fake) and generating the next block. The blockchain can include a blockchain underlying platform, a platform product service layer, and an application service layer.

[0129] S207: Identify the text block corresponding to the double-entry summary associated with the text information in the database, save the text information with the preset format displaying the keyword and / or the text information associated with the timestamp to the text module.

[0130] To ensure that the text information can be saved in the database in order, so that the quality inspection personnel can check according to the order of each business link in the overall business, this step arranges the text information according to the order of the double-entry summary in the database by identifying the text block corresponding to the double-entry summary associated with the text information in the database, and saving the text information in the buffer area of the text block, improving the order of the text information saving, so that the arrangement of the text information sent to the quality inspection personnel is consistent with the logical order of each specific business link in the overall business, improving the user experience of the quality inspection personnel.

[0131] S208: Receive the quality inspection request sent by the control end, and send the text information to the control end according to the quality inspection request.

[0132] To enable the quality inspection personnel to obtain the corresponding text information according to their needs and check, this step sends the text information to the control end according to the quality inspection request sent by the control end, obtains the text information corresponding to the target summary in the quality inspection request, and realizes the technical effect of configurable display of double-entry information. The quality inspection request is a request information generated based on the http protocol.

[0133] In a preferred embodiment, the receiving of the quality inspection request sent by the control end and the sending of the text information to the control end according to the quality inspection request comprise:

[0134] S81: extract the target information recorded in the quality inspection request, obtain the database corresponding to the target information and set it as a target library; wherein the target information records the double transcript summary of the text information required by the control terminal;

[0135] S82: according to the order of the text blocks in the target library, extract the text information in each text block in turn, and return the text information to the control terminal.

[0136] In this embodiment, the target information is used to achieve the technical purpose of the control terminal to obtain text information according to its needs. By sending the text information saved in the text block to the control terminal in turn according to the order of the text block, the quality inspection personnel can obtain the corresponding text in turn according to the playing time sequence of the double transcript information, which provides convenience for the quality inspection personnel to check the double transcript information.

[0137] Embodiment three:

[0138] Please refer to Figure 4 The double transcript review auxiliary device 1 based on artificial intelligence in this embodiment is installed in a double transcript background server 2 connected with a double transcript device 3 and a cloud storage system 4 respectively, and comprises:

[0139] A double transcript input module 11 is used to receive the double transcript key and double transcript summary sent by the double transcript device, and obtain the double transcript information from the cloud storage system according to the double transcript key; wherein the double transcript information is video data and audio information shot for the whole business, the double transcript key is key information generated when the double transcript device uploads the shot double transcript information to the cloud storage system, and the double transcript summary describes the time period of each business link under the whole business;

[0140] An audio extraction module 13 is used to cut the double transcript information to obtain at least one double transcript segment according to the double transcript summary, and extract the audio information in the double transcript segment;

[0141] A text conversion module 14 is used to convert the audio information to obtain text information, and associate the double transcript summary with the text information; wherein the double transcript segment is video data and audio information shot for a specific business link under the whole business;

[0142] A text output module 18 is used to receive the quality inspection request sent by the control terminal, and send the text information to the control terminal according to the quality inspection request.

[0143] Optionally, the double transcript review auxiliary device 1 further comprises:

[0144] A block construction module 12 is used to construct a text block in a preset database according to the double transcript summary.

[0145] Optionally, the block construction module 12 comprises:

[0146] A cache construction unit 121, configured to construct a cache block for saving data in the database.

[0147] A block conversion unit 122, configured to extract a double recording time period in the double recording summary, and record the double recording time period in a cache header of the cache block, so as to convert the cache block into the text block; wherein the double recording time period represents a start time and an end time of the double recording clip, and the cache header is description information for describing cache data in the text block.

[0148] A block arrangement unit 123, configured to extract the start time in the double recording time period, and arrange the text blocks in the database in ascending order according to the start time.

[0149] A library association unit 124, configured to extract attribute information in the double recording summary, and record the attribute information in a configuration file of the database, so as to associate the double recording information with the database.

[0150] Optionally, the audio extraction module 13 comprises:

[0151] A double recording extraction unit 131, configured to extract a double recording time period in the double recording summary; wherein the double recording time period represents a start time and an end time of the business link.

[0152] An audio extraction unit 132, configured to cut the double recording information according to the double recording time period to form at least one double recording clip, and extract audio information in the double recording clip.

[0153] Optionally, the text conversion module 14 comprises:

[0154] An audio uploading unit 141, configured to upload the audio information to the cloud storage system, and receive an audio key generated by the cloud storage system according to the audio information.

[0155] A quantity checking unit 142, configured to determine whether the number of obtained audio keys is consistent with the number of double recording time periods.

[0156] A key writing unit 143, configured to record the audio key in a text block in which the double recording time period is recorded in the database.

[0157] A key deleting unit 144, configured to delete the audio key and end.

[0158] The audio downloading unit 145 is configured to receive a conversion request sent by the control terminal, extract the audio key from the database according to the conversion request, and download the audio information from the cloud storage system according to the audio key.

[0159] The audio converting unit 146 is configured to convert the audio information into text information by invoking a preset conversion engine.

[0160] The text writing unit 147 is configured to obtain an audio key corresponding to the audio information, delete the audio information, identify a text block in the database recording the audio key, record the text information in the text block, and associate the text information with the double-recording abstract in the text block.

[0161] Optionally, the double-recording review auxiliary device 1 further comprises:

[0162] The format modifying module 15 is configured to mark a keyword in the text information, and display the keyword in the text information in a preset format.

[0163] Optionally, the format modifying module 15 further comprises:

[0164] The keyword modifying unit 151 is configured to identify a keyword in the text information by a preset keyword regular expression, and set a format of the keyword as a keyword format.

[0165] The sensitive modifying unit 152 is configured to identify a sensitive word in the text information by a preset sensitive regular expression, and set a format of the sensitive word as a sensitive format.

[0166] The violation modifying unit 153 is configured to identify a violation word in the text information by a preset violation regular expression, and set a format of the violation word as a violation format.

[0167] Optionally, the double-recording review auxiliary device 1 further comprises:

[0168] The timestamp associating module 16 is configured to obtain audio information corresponding to the text information in which the keyword is located, obtain a playing time of the keyword in the audio information, make a timestamp of the playing time, and associate the timestamp with the keyword in the text information.

[0169] Optionally, the timestamp associating module 16 further comprises:

[0170] The text segmenting unit 161 is configured to segment the text information into at least one text segment according to a punctuation mark in the text information as a segment symbol.

[0171] The time acquisition unit 162 is configured to acquire the text segments with the keyword and identify the playing time of each text segment in the audio information.

[0172] Optionally, the dual-recording review auxiliary device 1 further comprises:

[0173] The key text writing module 17 is configured to identify the text block corresponding to the dual-recording summary associated with the text information in the database, display the text information with the keyword in the preset format, and save the text information associated with the timestamp to the text module.

[0174] Optionally, the dual-recording review auxiliary device 1 further comprises:

[0175] The text output module 18 is configured to receive the quality inspection request sent by the control terminal, and send the text information to the control terminal according to the quality inspection request.

[0176] Optionally, the text output module 18 comprises:

[0177] The target identification unit 181 is configured to extract the target information recorded in the quality inspection request, acquire the database corresponding to the target information and set it as a target database; wherein the target information records the dual-recording summary of the text information required by the control terminal.

[0178] The sorting feedback unit 182 is configured to extract the text information in each text block in sequence according to the sorting of the text blocks in the target database, and return the text information to the control terminal.

[0179] The embodiments of the present application can acquire and process related data based on artificial intelligence technology. Artificial intelligence (AI) is the use of digital computers or computer-controlled machines to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0180] The basic technology of artificial intelligence generally includes technologies such as sensors, special artificial intelligence chips, cloud computing, distributed storage, big data processing, operation / interaction systems, mechatronics, etc. The software technology of artificial intelligence mainly includes computer vision technology, robot technology, biometric technology, speech processing technology, natural language processing technology, and machine learning / deep learning, etc.

[0181] In the present embodiment, at least one dual-recording segment is obtained by cutting dual-recording information according to a dual-recording summary based on a natural language model of artificial intelligence technology, and audio information in the dual-recording segment is extracted and converted to obtain text information.

[0182] Embodiment Four

[0183] To achieve the above object, the present application further provides a computer device 6, components of the dual-recording review assisting device of Embodiment Three can be distributed in different computer devices, the computer device 6 can be a smart phone, a tablet computer, a notebook computer, a desktop computer, a rack server, a blade server, a tower server or a cabinet server (including a stand-alone server, or a server cluster composed of multiple application servers) and the like executing programs. The computer device of the present embodiment at least includes but is not limited to: a memory 61, a processor 62, etc. which can be connected to each other in communication through a system bus, as shown in the figure. It should be pointed out that Figure 5 Figure 5 Only the computer device with components is shown, but it should be understood that it is not required to implement all the shown components, and more or fewer components can be alternatively implemented.

[0184] In the present embodiment, the memory 61 (i.e. the readable storage medium) includes a flash memory, a hard disk, a multimedia card, a card-type memory (e.g. an SD or DX memory, etc.), a random access memory (RAM), a static random access memory (SRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read-only memory (PROM), a magnetic memory, a magnetic disk, an optical disk, etc. In some embodiments, the memory 61 can be an internal storage unit of the computer device, such as a hard disk or a memory of the computer device. In other embodiments, the memory 61 can also be an external storage device of the computer device, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the computer device. Of course, the memory 61 can also include both the internal storage unit and the external storage device of the computer device. In the present embodiment, the memory 61 is generally used to store the operating system and various application software installed on the computer device, such as the program code of the dual-recording review assisting device of Embodiment Three, etc. In addition, the memory 61 can also be used to temporarily store various data that have been output or will be output.

[0185] The processor 62 in some embodiments can be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chips. The processor 62 is generally used to control the overall operation of the computer device. In the present embodiment, the processor 62 is used to run the program code or process data stored in the memory 61, such as running the dual-recording review assisting device, to implement the dual-recording review assisting method of Embodiments One and Two.​

[0186] Embodiment five:

[0187] To achieve the above object, the application further provides a computer readable storage medium, such as a flash memory, a hard disk, a multimedia card, a card memory (for example, an SD or DX memory, etc.), a random access memory (RAM), a static random access memory (SRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read-only memory (PROM), a magnetic memory, a magnetic disk, an optical disk, a server, an App application store, etc., which stores a computer program, and the program is executed by the processor 62 to realize the corresponding function. The computer readable storage medium of the embodiment is used to store the computer program for realizing the double recording review auxiliary method, and the program is executed by the processor 62 to realize the double recording review auxiliary method of the embodiment one and the embodiment two.

[0188] The above embodiment numbers of the application are only for description, and do not represent the advantages and disadvantages of the embodiments.

[0189] Through the description of the above embodiments, those skilled in the art can clearly understand that the above embodiment methods can be realized by means of software and necessary general hardware platforms, of course, they can also be realized by hardware, but in many cases, the former is a better embodiment.

[0190] The above is only the preferred embodiment of the application, and does not limit the patent scope of the application, and any equivalent structure or equivalent process transformation, or direct or indirect application in other related technical fields, which is made by using the content of the application specification and drawings, is also included in the patent protection scope of the application.

Claims

1. An artificial intelligence-based dual-recording review assistance method, running in a dual-recording backend server connected to both the dual-recording device and a cloud storage system, characterized in that, include: The system receives a dual-recording key and a dual-recording digest sent by a dual-recording device, and retrieves dual-recording information from the cloud storage system based on the dual-recording key. The dual-recording information consists of video and audio data captured for the overall business, the dual-recording key is the key information generated by the dual-recording device when it uploads the captured dual-recording information to the cloud storage system, and the dual-recording digest describes the time period of each business segment under the overall business. A text block is constructed in a preset database based on the dual-recording summary; Based on the dual recording summary, the dual recording information is segmented to obtain at least one dual recording segment, and the audio information in the dual recording segment is extracted; The audio information is converted into text information, and the dual-recording summary is associated with the text information; wherein, the dual-recording segment refers to video data and audio information captured for a specific business segment under the overall business; the conversion of the audio information into text information and the association of the dual-recording summary with the text information includes: The preset conversion engine is invoked to convert the audio information into text information; Obtain the audio key corresponding to the audio information and delete the audio information; identify the text block in the database that records the audio key; enter the text information into the text block; and associate the text information with the dual recording summary in the text block. The keywords in the text information are marked so that the keywords are displayed in the text information in a preset format; the format includes a key format for marking keywords, a sensitive format for marking sensitive words, and a violation format for marking illegal words; Obtain the audio information corresponding to the text information containing the keyword, obtain the playback time of the keyword in the audio information, create a timestamp of the playback time, and associate the timestamp with the keyword in the text information; Identify the text block corresponding to the dual-recording summary associated with the text information in the database, and save the text information displaying the keywords in a preset format, and / or the text information associated with the timestamp, into the text block; wherein, the keywords include keywords, sensitive words, and illegal words; wherein, the keywords represent key nodes in the business process; Receive a quality inspection request sent by the control terminal, and send the text information to the control terminal according to the quality inspection request; The receiving control terminal sends a quality inspection request, and according to the quality inspection request, the text information is sent to the control terminal, including: Extract the target information recorded in the quality inspection request, obtain the database corresponding to the target information and set it as the target database; wherein, the target information records the dual-recording summary of the text information required by the control terminal; According to the sorting of the text blocks in the target library, extract the text information from each text block in sequence, and return the text information to the control terminal.

2. The artificial intelligence-based dual-recording review assistance method according to claim 1, characterized in that, The step of constructing a text block in a preset database based on the dual-recording summary includes: Build a cache block in the database to store data; Extract the dual-recording time period from the dual-recording summary, and record the dual-recording time period into the cache header of the cache block, so that the cache block is converted into the text block; wherein, the dual-recording time period represents the start time and end time of the dual-recording segment, and the cache header is descriptive information used to describe the cached data in the text block; Extract the start time from the dual-recording time period, and sort the text blocks in the database in ascending order according to the start time; After constructing the text block in a preset database based on the dual-recording summary, the method further includes: Extract the attribute information from the dual-recording summary, and enter the attribute information into the configuration file of the database to associate the dual-recording information with the database.

3. The artificial intelligence-based dual-recording review assistance method according to claim 1, characterized in that, The step of cutting the dual recording information according to the dual recording summary to obtain at least one dual recording segment, and extracting audio information from the dual recording segment, includes: Extract the dual recording time period from the dual recording summary; wherein the dual recording time period represents the start and end time of the business process; The dual recording information is segmented according to the dual recording time period to form at least one dual recording segment, and the audio information in the dual recording segment is extracted.

4. The artificial intelligence-based dual-recording review assistance method according to claim 3, characterized in that, Before converting the audio information into text information and associating the dual-recording summary with the text information, the method further includes: Upload the audio information to the cloud storage system and receive the audio key generated by the cloud storage system based on the audio information; Determine whether the number of obtained audio keys matches the number of dual recording time periods; If so, the audio key is entered into the text block in the database that contains the dual recording time period; If not, delete the audio key and end; The system receives a conversion request from the control terminal, extracts the audio key from the database based on the conversion request, and downloads the audio information from the cloud storage system based on the audio key.

5. The artificial intelligence-based dual-recording review assistance method according to claim 1, characterized in that, After associating the timestamp with the keywords in the text information, the method further includes: The timestamp and the keyword are uploaded to the blockchain.

6. An artificial intelligence-based dual-recording review auxiliary device, installed in a dual-recording backend server connected to both a dual-recording device and a cloud storage system, characterized in that, include: The dual-recording input module is used to receive the dual-recording key and dual-recording summary sent by the dual-recording device, and to retrieve the dual-recording information from the cloud storage system according to the dual-recording key; wherein, the dual-recording information is video data and audio information captured for the overall business, the dual-recording key is the key information generated by the dual-recording device when uploading the captured dual-recording information to the cloud storage system, and the dual-recording summary describes the time period of each business link under the overall business; The block construction module is used to construct text blocks in a preset database based on the dual recording summary; the audio extraction module is used to cut the dual recording information according to the dual recording summary to obtain at least one dual recording segment, and to extract audio information from the dual recording segment. A text conversion module is used to convert the audio information into text information and associate the dual-recording summary with the text information; wherein, the dual-recording segment refers to video data and audio information captured for a specific business process under the overall business; the conversion of the audio information into text information and the association of the dual-recording summary with the text information includes: calling a preset conversion engine to convert the audio information into text information; obtaining the audio key corresponding to the audio information and deleting the audio information; identifying the text block in the database that records the audio key; entering the text information into the text block; and associating the text information with the dual-recording summary in the text block. The format modification module is used to mark keywords in the text information so that the keywords are displayed in the text information in a preset format; the format includes a key format for marking keywords, a sensitive format for marking sensitive words, and a violation format for marking illegal words; The timestamp association module is used to obtain the audio information corresponding to the text information containing the keyword, obtain the playback time of the keyword in the audio information, create a timestamp of the playback time, and associate the timestamp with the keyword in the text information; The key text writing module is used to identify the text block corresponding to the dual-recording summary associated with the text information in the database, and save the text information displaying the keyword in the preset format, and / or the text information associated with the timestamp, into the text block; wherein, the keyword includes keywords, sensitive words, and illegal words; wherein, the keywords represent key nodes in the business process; A text output module is used to receive a quality inspection request sent by a control terminal, and send the text information to the control terminal according to the quality inspection request; the process of receiving the quality inspection request sent by the control terminal and sending the text information to the control terminal according to the quality inspection request includes: extracting target information recorded in the quality inspection request, obtaining a database corresponding to the target information and setting it as a target database; wherein, the target information records a dual-recording summary of the text information required by the control terminal; extracting the text information from each text block in sequence according to the order of the text blocks in the target database, and returning the text information to the control terminal.

7. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor of the computer device executes the computer program, it implements the steps of the artificial intelligence-based dual-recording review assistance method according to any one of claims 1 to 5.

8. A computer-readable storage medium storing a computer program thereon, characterized in that, When the computer program stored in the readable storage medium is executed by a processor, it implements the steps of the artificial intelligence-based dual-recording review assistance method according to any one of claims 1 to 5.

Images