CN114679557A

CN114679557A - Recorded data quality inspection method, recorded data quality inspection device, recorded data quality inspection equipment, recording medium and program product

Info

Publication number: CN114679557A
Application number: CN202210271387.9A
Authority: CN
Inventors: 廖鑫
Original assignee: China Construction Bank Corp
Current assignee: China Construction Bank Corp
Priority date: 2022-03-18
Filing date: 2022-03-18
Publication date: 2022-06-28

Abstract

The application discloses a method, a device, equipment, a storage medium and a program product for quality inspection of recorded data. The method comprises the following steps: acquiring recorded data to be subjected to quality inspection; inputting the recorded data to be quality tested into a quality testing model; respectively detecting and processing the recorded data according to the video identification submodel and the audio identification submodel to obtain a first quality inspection result of the recorded data; when at least one of the sound recording quality inspection result and the video recording quality inspection result of the first quality inspection result is an abnormal quality inspection result, sending abnormal recording data to a quality inspection background, wherein the abnormal recording data is recording data corresponding to the abnormal quality inspection result; and when the quality inspection confirmation result generated by the quality inspection background according to the abnormal recorded data is inconsistent with the first quality inspection result, taking the quality inspection confirmation result as a final quality inspection result. According to the embodiment of the application, the recorded data can be subjected to quick quality inspection through the quality inspection model, and abnormal data are subjected to quality inspection confirmation, so that a final quality inspection result is obtained, and the quality inspection efficiency and accuracy of the recorded data are improved.

Description

Recorded data quality inspection method, recorded data quality inspection device, recorded data quality inspection equipment, recording medium and program product

Technical Field

The application belongs to the technical field of artificial intelligence, and particularly relates to a recorded data quality inspection method, device, equipment, storage medium and program product.

Background

At present, along with the economic development of society, the income level of people is gradually improved, the financing consciousness of residents is gradually strengthened, the asset management willingness of people is gradually strengthened, and financial products such as funds, financing, insurance policies and the like become relatively common investment choices.

In order to protect the interests of the customer, the regulatory authorities explicitly stipulate that the sales process needs to be recorded by sound and video recording (double recording) for later review when the sales activity occurs. Therefore, in order to ensure the compliance of recorded data and avoid the related risks of the recorded data, quality inspection of the recorded data is required.

The existing quality inspection method of recorded data mainly comprises the steps that a background quality inspection worker extracts a part of recorded data from a database for storing the recorded data, and quality inspection and verification are carried out on the part of recorded data in a manual identification mode so as to confirm the compliance of the recorded data. However, with the great increase of sales business, the auditing amount of the extraction auditing mode cannot meet the requirement. And quality testing personnel are easy to miss and leak in the process of auditing for a long time.

Disclosure of Invention

The embodiment of the application provides a recorded data quality inspection method, a recorded data quality inspection device, recorded data quality inspection equipment, a storage medium and a program product, and can solve the problems that the existing recorded data quality inspection mode is small in audit quantity and easy to miss.

In a first aspect, an embodiment of the present application provides a method for quality inspection of recorded data, where the recorded data includes recorded data and video data, and the method includes:

acquiring recorded data to be subjected to quality inspection;

inputting the recorded data to be subjected to quality inspection into a quality inspection model, wherein the quality inspection model comprises a video identification submodel and an audio identification submodel;

respectively detecting and processing the recorded data according to the video identification submodel and the audio identification submodel to obtain a first quality inspection result of the recorded data, wherein the first quality inspection result comprises a recording quality inspection result and a video quality inspection result;

when at least one of the sound recording quality inspection result and the video recording quality inspection result of the first quality inspection result is an abnormal quality inspection result, sending abnormal recording data to a quality inspection background, wherein the abnormal recording data is recording data corresponding to the abnormal quality inspection result;

and when a quality inspection confirmation result generated by the quality inspection background according to the abnormal recording data is inconsistent with the first quality inspection result, taking the quality inspection confirmation result as a final quality inspection result.

In some embodiments, the detecting the recorded data according to the video identification submodel and the audio identification submodel to obtain a first quality inspection result of the recorded data includes:

determining video data corresponding to the video identification submodel and sound recording data corresponding to the audio identification submodel from the recording data to be subjected to quality inspection;

processing the video data according to the video identification submodel to obtain a video quality inspection result of the video data;

processing the recording data according to the audio recognition submodel to obtain a recording quality inspection result of the recording data;

and generating the first quality inspection result according to the video quality inspection result and the audio quality inspection result.

In some embodiments, the processing the video data according to the video identification submodel to obtain a video quality inspection result of the video data includes:

determining each service node according to the service type corresponding to the recorded data;

dividing the video data into a plurality of video data according to each service node, wherein the plurality of video data correspond to each service node one to one;

And respectively carrying out video content identification on the plurality of video recording sub-data according to the video identification submodels to obtain video quality inspection results.

In some embodiments, the dividing the video data into a plurality of video data according to each service node includes:

acquiring video characteristics corresponding to each service node;

determining video frames corresponding to the video characteristics from the video data;

and dividing the video data according to each video frame to obtain a plurality of video data.

In some embodiments, the processing the recording data according to the audio identification submodel to obtain a recording quality inspection result of the recording data includes:

determining service keywords according to the service types corresponding to the recorded data, wherein the service keywords comprise necessary keywords and illegal keywords in the service handling process;

carrying out voice recognition processing on the recording data to obtain voice text information;

matching the voice text information with the service keywords to obtain a matching result;

and generating the sound recording quality inspection result according to the matching result.

In some embodiments, after the quality inspection confirmation result is used as the final quality inspection result, the method further comprises:

Acquiring a plurality of recorded data with inconsistent quality inspection confirmation results and the first quality inspection results;

sorting and counting the plurality of recorded data, and determining problem service nodes corresponding to the plurality of recorded data respectively;

and adjusting and optimizing the model parameters of the quality inspection model according to the problem service node so as to update the quality inspection model.

In some embodiments, the adjusting and optimizing the model parameters of the quality inspection model according to the problem service node to update the quality inspection model includes:

adjusting the model parameters according to the problem service node to obtain an adjusted quality inspection model;

inputting a plurality of recorded data to the adjusted quality inspection model to obtain second quality inspection results corresponding to the plurality of recorded data respectively;

calculating a first matching degree between the quality inspection confirmation results of the plurality of recorded data and a plurality of first quality inspection results, and a second matching degree between the quality inspection confirmation results of the plurality of recorded data and a plurality of second quality inspection results;

determining an optimized quality inspection result from the first quality inspection result and the second quality inspection result according to the first matching degree and the second matching degree;

And taking the quality inspection model corresponding to the optimized quality inspection result as an updated quality inspection model.

In a second aspect, an embodiment of the present application provides a recorded data quality inspection apparatus, including:

the acquisition module acquires recorded data to be subjected to quality inspection;

the input module is used for inputting the recorded data to be subjected to quality inspection into a quality inspection model, and the quality inspection model comprises a video identification submodel and an audio identification submodel;

the detection module is used for detecting and processing the recorded data according to the video identification submodel and the audio identification submodel to obtain a first quality inspection result of the recorded data;

the abnormality module is used for sending quality inspection abnormality information to a quality inspection background when the first quality inspection result is an abnormal quality inspection result;

and the confirmation module is used for receiving the quality inspection confirmation result sent by the quality inspection background and replacing the abnormal quality inspection result with the quality inspection confirmation result.

In a third aspect, an embodiment of the present application provides a recorded data quality inspection apparatus, including: a processor and a memory storing computer program instructions;

the recorded data quality inspection method is realized when the processor executes the computer program instructions.

In a fourth aspect, an embodiment of the present application provides a computer storage medium, where computer program instructions are stored on the computer storage medium, and when the computer program instructions are executed by a processor, the method for quality inspection of recorded data as above is implemented.

In a fifth aspect, the present application provides a computer program product, where the computer program product includes computer program instructions, and the computer program instructions, when executed by a processor, implement the recorded data quality inspection method as above.

In all embodiments of the present application, the acquisition, storage, use, processing, etc. of the data complies with relevant regulations of national laws and regulations.

According to the quality inspection method, the device, the equipment, the storage medium and the program product for the recorded data, the recorded data to be quality inspected are input into the quality inspection model, and the recording data and the video data in the recorded data can be respectively detected and processed through the sub-model in the quality inspection model, so that a first quality inspection result of the recorded data is obtained. When the first quality inspection result comprises at least one abnormal quality inspection result, the abnormal recorded data can be sent to a quality inspection background so as to carry out quality inspection confirmation on the abnormal recorded data through the quality inspection background. When the quality inspection confirmation result is the same as the abnormal quality inspection result, the first quality inspection result can be used as a final quality inspection result; if the quality inspection result does not match the abnormal quality inspection result, the quality inspection result can be used as the final quality inspection result. The quality inspection model can be used for carrying out batch automatic quality inspection on a large amount of recorded data, and the first quality inspection result can be directly used as a final quality inspection result when abnormal recorded data do not exist in the quality inspection result. And performing secondary quality inspection confirmation on the recorded data with the abnormality through a quality inspection background to obtain a final quality inspection result. Quality inspection can be carried out through the quality inspection model, and quality inspection efficiency and quality inspection accuracy can be greatly improved. And the personnel of the quality inspection background only need to carry out quality inspection confirmation on the recorded data with the identified abnormality, so that the quality inspection workload of the quality inspection background is reduced, and the personnel cost of the quality inspection background is reduced.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required to be used in the embodiments of the present application will be briefly described below, and it is obvious that the drawings described below are only some embodiments of the present application, and it is obvious for those skilled in the art to obtain other drawings based on the drawings without creative efforts.

Fig. 1 is a schematic flowchart of a recorded data quality inspection method according to an embodiment of the present disclosure;

fig. 2 is a schematic flowchart of a recorded data quality inspection method according to another embodiment of the present application;

fig. 3 is a schematic detailed flowchart of step S220 provided in an embodiment of the present application;

fig. 4 is a schematic detailed flowchart of step S220 provided in another embodiment of the present application;

fig. 5 is a schematic detailed flowchart of step S230 provided in an embodiment of the present application;

fig. 6 is a schematic partial flowchart of a recorded data quality inspection method according to an embodiment of the present disclosure;

fig. 7 is a schematic partial flowchart of a recorded data quality inspection method according to another embodiment of the present application;

fig. 8 is a schematic structural diagram of a recorded data quality inspection apparatus according to an embodiment of the present application;

Fig. 9 is a schematic hardware structure diagram of a recorded data quality inspection apparatus according to an embodiment of the present application.

Detailed Description

Features and exemplary embodiments of various aspects of the present application will be described in detail below, and in order to make objects, technical solutions and advantages of the present application more apparent, the present application will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are merely illustrative of, and not restrictive on, the present application. It will be apparent to one skilled in the art that the present application may be practiced without some of these specific details. The following description of the embodiments is merely intended to provide a better understanding of the present application by illustrating examples thereof.

It is noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.

It should be noted that, in the present application, the embodiments and features of the embodiments may be combined with each other without conflict. The embodiments will be described in detail below with reference to the accompanying drawings.

At present, with the economic development of society, the income level of people is gradually improved, the financial consciousness of residents is gradually strengthened, and partial assets are more willing to be used for investment to earn return. The fund has the characteristic that professional practitioners manage assets, is popular with common residents, and becomes a relatively common investment choice for purchasing the fund. In addition, financial products such as financing and insurance are becoming popular.

Because fund products are various in types, common customers are difficult to identify the difference and risk of each product, and banks need to fully prompt necessary information in a clear and sufficient mode for the customers to pay attention to the fund according to the requirements of a regulatory agency when selling the fund to the customers, so that legal risks, policy risks, market risks and the like related to the sold products are fully prompted, and the requirements of recording and supervising in special sale areas are strictly fulfilled.

However, there are inevitable divulgences in the sales process, such as unclear clear reading speech resulting in unclear understanding of the customers, missing necessary links of the sales resulting in non-compliance of the sales process, and the like. Therefore, in order to ensure the compliance of the sales process, the recorded data needs to be quality checked. The existing quality inspection mode mainly comprises that quality inspectors manually check recorded data from a database, and adopt a manual identification mode to perform quality inspection and audit on the data. However, as the sales traffic increases, the quality inspection efficiency and the quality inspection accuracy of the conventional quality inspection method cannot meet the requirement, and additional personnel cost is required to be generated to match the increase of the traffic.

In order to solve the problem of the prior art, embodiments of the present application provide a method, an apparatus, a device, a storage medium, and a program product for quality inspection of recorded data. First, a method for quality inspection of recorded data provided in the embodiment of the present application is described below. In all the following embodiments of the present application, the acquisition, storage, use, processing, etc. of the data complies with the relevant regulations of the national laws and regulations.

Fig. 1 is a flowchart illustrating a method for quality inspection of recorded data according to an embodiment of the present disclosure. The method comprises the following steps:

s110, acquiring recorded data to be subjected to quality inspection;

s120, inputting the recorded data to be subjected to quality inspection into a quality inspection model, wherein the quality inspection model comprises a video identification submodel and an audio identification submodel;

s130, respectively detecting and processing the recorded data according to the video identification submodel and the audio identification submodel to obtain a first quality inspection result of the recorded data, wherein the first quality inspection result comprises a recording quality inspection result and a video quality inspection result;

s140, when at least one of the sound recording quality inspection result and the video recording quality inspection result of the first quality inspection result is an abnormal quality inspection result, sending abnormal recording data to a quality inspection background, wherein the abnormal recording data is recording data corresponding to the abnormal quality inspection result;

And S150, when a quality inspection confirmation result generated by the quality inspection background according to the abnormal recording data is inconsistent with the first quality inspection result, taking the quality inspection confirmation result as a final quality inspection result.

In this embodiment, by inputting the recorded data to be quality-tested into the quality testing model, the audio recording data and the video recording data in the recorded data can be respectively tested by the sub-model in the quality testing model, so as to obtain a first quality testing result of the recorded data. When the first quality inspection result comprises at least one abnormal quality inspection result, the abnormal recorded data can be sent to a quality inspection background so as to carry out quality inspection confirmation on the abnormal recorded data through the quality inspection background. When the quality inspection confirmation result is the same as the abnormal quality inspection result, the first quality inspection result can be used as a final quality inspection result; if the quality check result does not match the abnormal quality check result, the quality check result can be used as the final quality check result. The quality inspection model can be used for carrying out batch automatic quality inspection on a large amount of recorded data, and the first quality inspection result can be directly used as a final quality inspection result when abnormal recorded data do not exist in the quality inspection result. And performing secondary quality inspection confirmation on the recorded data with the abnormality through a quality inspection background to obtain a final quality inspection result. Quality inspection can be carried out through the quality inspection model, and quality inspection efficiency and quality inspection accuracy can be greatly improved. And the personnel of the quality inspection background only need to carry out quality inspection confirmation on the recorded data with the identified abnormality, thereby reducing the quality inspection workload of the quality inspection background and reducing the personnel cost of the quality inspection background.

When transacting related business for client, business transacting personnel can record audio and video for transacting region to generate record data of business transacting process, and the record data can include record data and video data. The recording device can upload the recorded data to the database for storage when the service is finished. When the recorded data is subjected to quality inspection, the recorded data can be obtained from the database, and the recorded data is subjected to quality inspection. It should be noted that, in the following embodiments, the acquisition, storage, usage, processing, etc. of the data all conform to the relevant regulations of the national laws and regulations.

The following describes a specific implementation manner of each step by taking the recorded data quality inspection device as an execution subject.

In S110, when the quality inspection device performs quality inspection on the recorded data, the recorded data to be quality inspected may be obtained from the database. The recorded data in the database may include quality inspection parameters, and the quality inspection parameters of the recorded data before and after quality inspection are different. That is, the quality testing parameters will change after the recorded data is subjected to quality testing. The quality inspection equipment can determine whether the recorded data is the recorded data which is subjected to quality inspection or the recorded data to be subjected to quality inspection according to the quality inspection parameters of certain recorded data in the database.

In S120, a quality inspection model is preset in the quality inspection device, and after the recorded data to be quality inspected is acquired, the recorded data may be input into the quality inspection model, so as to perform quality inspection on the recorded data through the quality inspection model. The quality inspection model can comprise a video recognition submodel and an audio recognition submodel. The video recognition submodel can perform quality inspection on the video data part of the recorded data, and the audio recognition submodel can perform quality inspection on the audio data part of the recorded data.

In S130, in the quality inspection model, the recorded data is detected by the video identifier sub-model and the audio identifier sub-model, respectively, so as to obtain a first quality inspection result of the recorded data. The first quality inspection result comprises a sound recording quality inspection result and a video recording quality inspection result. The recording quality inspection result is a quality inspection result generated after the audio identifier model performs quality inspection on the recording data; and the video quality inspection result is a quality inspection result generated after the video identification submodel performs quality inspection on the video data.

As an alternative embodiment, referring to fig. 2, in order that the step S130 may include:

s210, determining video data corresponding to the video identification submodel and sound recording data corresponding to the audio identification submodel from the recording data to be quality tested;

S220, processing the video data according to the video identification submodel to obtain a video quality inspection result of the video data;

s230, processing the recording data according to the audio recognition submodel to obtain a recording quality inspection result of the recording data;

s240, generating the first quality inspection result according to the video quality inspection result and the audio quality inspection result.

In this embodiment, quality inspection can be performed on the audio recording data and the video recording data in the recorded data respectively through the two sub-models in the quality inspection model, and corresponding video quality inspection results and audio quality inspection results are generated, and first quality inspection results can be generated according to the video quality inspection results and the audio quality inspection results. The recorded data to be subjected to quality inspection are subjected to quality inspection through the quality inspection model, a large amount of recorded data can be subjected to automatic quality inspection, a corresponding first quality inspection result is obtained, and the quality inspection efficiency is improved.

In S210, after the recording device records the service transaction process, the generated audio recording data and video recording data may be separately stored in the database, or may be stored in the database together.

When the recorded data and the video data are stored in the database in a separated storage mode, the quality inspection equipment can determine the recorded data and the video data according to the reading position in the database when acquiring the recorded data to be subjected to quality inspection. When the audio recording data and the video recording data are stored in the database in a collective storage manner, the quality inspection equipment needs to perform data separation on the recorded data after acquiring the recorded data to be inspected so as to obtain the audio recording data and the video recording data. For example, the quality control device may use ffmpeg (fast Forward mpeg) tool to separate the recorded data.

After the quality inspection equipment determines the video data and the audio data from the recorded data to be inspected, the video data and the audio identification submodel correspond to each other, and the audio data and the audio identification submodel correspond to each other.

In S220, the quality inspection device performs data processing on the video data through the video identification submodel, so as to obtain a video quality inspection result of the video data. The video quality inspection result can be that the video quality inspection passes or the video quality inspection fails. When the video identification submodel performs video quality inspection on the video data, the video identification submodel can identify and process the video data from the multiple transaction nodes according to the types of transaction so as to determine whether the multiple transaction nodes in the video data correspond to the quality inspection rules or not. And when the data corresponding to each transaction node in the video data accord with the quality inspection rule, the video quality inspection result is passed, otherwise, the video quality inspection result is not passed.

As an alternative embodiment, referring to fig. 3, in order for the step S220 to include:

s310, determining each service node according to the service type corresponding to the recorded data;

s320, dividing the video data into a plurality of video subdata according to each service node, wherein the plurality of video subdata correspond to each service node one to one;

S330, respectively carrying out video content identification on the plurality of video recording sub-data according to the video identification submodels to obtain video recording quality inspection results.

In this embodiment, each service node corresponding to the service type may be determined by the service type corresponding to the recorded data, and for each service node, corresponding video data may be divided from the video data, and video content identification may be performed on the video data, so as to determine whether the portion of video content meets the specified requirement of the service node. After the video content of each video data is identified, the video quality inspection results of the video data can be integrated.

In S310, after the quality inspection device acquires the recorded data to be quality inspected, it may determine corresponding service nodes according to the service types corresponding to the recorded data. For example, when a sales transaction of a financial product such as fund, financing or insurance is transacted, the service nodes may include a notice exhibition node, a certificate exhibition node, a user authentication node, an information checking and reminding node, a customer opinion solicitation node, a customer signature node, and the like.

In S320, the video data may be divided according to each service node corresponding to the service type, so as to divide the video data into video data corresponding to the service nodes one to one. By detecting the video data, whether the video data includes the video characteristics corresponding to the service node can be determined, and corresponding video data can be divided from the video data according to the corresponding video characteristics.

For example, taking a certificate display node as an example, in the process of detecting video data, if a video frame containing a qualification certificate displayed to a client by a service clerk is detected, it indicates that the video frame contains video features corresponding to the certificate display node. By taking the video frame as a reference, a time period containing the video frame can be used as video sub-data corresponding to the certificate display node.

It can be understood that after the video features are determined from the video frames in the video data, the video data in the previous period and the video data in the subsequent period may be segmented according to the video frames to serve as the video data. The video duration of the video subdata may be a preset fixed duration or a duration corresponding to a corresponding service node. In the service handling process, two or more video features may be included at the same time, or the time interval between two adjacent video features is small, so that the video data divided according to the video data may have a time period overlapping with other video data. For example, one piece of video data may be the fifth to fifteenth minutes of the video data, and the other piece of video data may be the tenth to twentieth minutes of the video data.

As an alternative embodiment, referring to fig. 4, in order that the step S320 may include:

s410, acquiring video characteristics corresponding to each service node;

s420, determining video frames corresponding to the video characteristics from the video data;

s430, the video data are divided according to the video frames to obtain a plurality of video data.

In this embodiment, after determining the corresponding service node according to the service type, the video features corresponding to each service node may be obtained, the video frames corresponding to each video feature may be determined from the video data, and the video data may be divided from the video data according to the video frames corresponding to each video feature. For example, after the first frame and the last frame of the video corresponding to the video features are determined, the video data can be divided from the video data according to the timestamps of the first frame and the last frame.

In S410, different video features are respectively corresponding to different service nodes. After determining a plurality of corresponding service nodes according to the service types corresponding to the recorded data, the corresponding video characteristics can be respectively determined according to each service node. For example, the video feature may be that a specific behavior feature is included in a certain video frame, or the duration of a certain behavior in consecutive video frames reaches a preset duration, and the like.

In S420, by performing image analysis on the video data, video frames corresponding to respective video features may be determined from the video data china. When the video characteristics are specific behavior characteristics, video frames with the behavior characteristics can be screened from the video data. When the duration of a certain behavior is the preset duration, the head and tail video frames in the duration of the behavior can be screened from the image data, and whether the duration of the behavior reaches the preset duration or not is determined according to timestamps corresponding to the head and tail video frames respectively.

In S430, after determining the video frames corresponding to the video features in the video data, the video data may be divided according to the plurality of video frames, so as to divide the video data into a plurality of video data. Each video data includes video characteristics required by the corresponding service node.

In S330, the quality inspection device may respectively perform video content identification on the plurality of video data through the video identifier sub-model to determine whether the video content in each video data reaches the relevant regulations in the service handling process. For example, taking the certificate display node as an example, the video identification sub-model may perform video content identification on the video recording sub-data corresponding to the certificate display node, determine the location of the user, the orientation of the user, and the certificate location shown by the service handling personnel from the video content, and determine whether the user can see the certificate shown by the service handling personnel in the location. When the user is determined to be located at a position where the user can see the certificate shown by the service handling personnel, the certificate display node can be determined to be in accordance with the service handling flow rule, otherwise, the certificate display node can be determined to be not in accordance with the service handling flow rule.

After the video identification submodel identifies the video content of each video subdata, whether the video content recorded by the video subdata meets the flow regulation of the corresponding service node can be determined. When the video content recorded in each video data item satisfies the relevant specification, it is determined that the video quality inspection result has passed.

In S230, the quality testing device performs data processing on the recorded sound data through the audio identifier model, so as to obtain a recorded sound quality testing result of the recorded sound data. The recording quality inspection result can be that the recording quality inspection passes or the recording quality inspection does not pass.

When the audio recognition submodel performs recording quality inspection on the recording data, keywords in the service handling process can be determined according to the type of the handled service, and the recording data is converted into text data by performing voice recognition on the recording data. And matching the text data with the keywords to obtain a corresponding matching result, and determining a recording quality inspection result according to the matching result. For example, in the business process, it is specified that business transactants need to perform information verification, intention inquiry, risk presentation, and other processes to clients, and users need to perform intention confirmation, risk confirmation, and other processes. By matching the text data with the keywords, whether the business handling personnel and the user execute the process in the business handling process can be determined.

And when the matching degree of the text data and the keywords meets the requirement, determining that the recording quality inspection result is passed. If the text data and the keywords are not matched or the matching degree is low, it can be determined that the corresponding process is missing or abnormal, and the recording quality inspection result is failed.

As an alternative embodiment, referring to fig. 5, in order for the above S230 to include:

s510, determining service keywords according to the service types corresponding to the recorded data, wherein the service keywords comprise necessary keywords and illegal keywords in the service handling process;

s520, performing voice recognition processing on the recording data to obtain voice text information;

s530, matching the voice text information with the service key words to obtain a matching result;

and S540, generating the sound recording quality inspection result according to the matching result.

In this embodiment, the corresponding service keyword may be determined by the service type corresponding to the recorded data, after performing voice recognition processing on the recorded data, corresponding voice text information may be obtained, and the voice text information is matched with the service keyword to obtain a matching result. And generating a recording quality inspection result of the recording data according to the matching result. The recording quality inspection is realized through the voice recognition processing of the quality inspection model, the condition that personnel at a quality inspection background adopt a manual listening mode to perform quality inspection on audio contents in the recorded data can be avoided, and the quality inspection efficiency and the quality inspection accuracy are improved.

In S510, the quality inspection device may determine a service keyword according to a service type corresponding to the recorded data, where the service keyword may include a necessary keyword and an illegal keyword in a service handling process. The necessary keywords are partial words contained in the dialects which are required to be read by business handling personnel in the business handling process. The violation keywords refer to partial words that the business handling personnel should not speak to the user during the handling process, for example, the violation keywords may include warranty, no risk, guaranteed income, or high income.

In S520, after determining the service keyword corresponding to the service type, the audio recognition submodel may perform voice recognition processing on the recording data to convert the recording data into voice text information.

In S530, a matching result of the voice text information and the service keyword may be obtained by matching the voice text information and the service keyword. For example, when the service keywords include necessary keywords and illegal keywords, the matching value is increased when the matching between the voice text information and the service keywords is successful; and when the matching of the voice text information and the illegal keyword is successful, the matching value is reduced. And after the matching of the voice text information and the service key words is completed, a corresponding matching result can be obtained.

In S540, after the voice text information obtained by voice conversion from the recording data is matched with the service keyword and a matching result is obtained, a recording quality inspection result is generated according to the matching result. For example, a matching threshold is preset in the audio recognition submodel, and when the value of the matching result is greater than the matching threshold, the result of the quality inspection of the recording at this time can be determined to be passed; and when the numerical value of the matching result is smaller than the matching threshold, determining that the sound recording quality inspection result does not pass.

It can be understood that the illegal keywords may also include forbidden keywords, and when the matching between the voice text information and the forbidden keywords is successful, the recording quality inspection result may be directly determined as not passing.

In S240, after the video data and the audio data are processed by the video recognition submodel and the audio recognition submodel, a video quality inspection result and a audio quality inspection result can be obtained, respectively, and a first quality inspection result is generated according to the video quality inspection result and the audio quality inspection result. For example, the business personnel complete the corresponding specific actions during the business transaction process, but the risk prompt information language is not informed to the user when the risk prompt is performed to the user. The first quality inspection result is the result of passing the video quality inspection and the result of failing the audio quality inspection.

In S140, after the recorded data is detected and the first quality inspection result is obtained, if one or both of the audio recording quality inspection result and the video recording quality inspection result in the first quality inspection result is an abnormal quality inspection result, it may be determined that the service handling process corresponding to the recorded data is an abnormal service process. The abnormal quality inspection result may include that the audio recording quality inspection result fails or the video recording quality inspection result fails.

The quality inspection equipment can send the abnormal recorded data to the quality inspection background so that the quality inspection background can perform quality inspection confirmation on the abnormal recorded data. The abnormal recording data is the recording data corresponding to the abnormal quality inspection result. After the quality inspection background receives the abnormal recorded data, quality inspection personnel can perform manual quality inspection on the abnormal recorded data and generate a corresponding quality inspection confirmation result.

It can be understood that when the quality inspection device performs batch quality inspection on the recorded data to be inspected, when each first quality inspection result corresponding to one recorded data includes an abnormal quality inspection result, the recorded data can be used as abnormal recorded data and sent to the quality inspection background. The quality inspection equipment can also send a plurality of abnormal recorded data containing abnormal quality inspection results in the first quality inspection result in the preset period to the quality inspection background at preset intervals.

When the quality inspection equipment sends the abnormal recorded data to the quality inspection background, the corresponding first quality inspection result can be sent to the quality inspection background together, so that quality inspection personnel at the quality inspection background can determine abnormal partial data in the abnormal recorded data according to the first quality inspection result and carry out quality inspection confirmation on the partial data, and the quality inspection confirmation efficiency is improved. For example, a quality inspector in the background of quality inspection determines that the video quality inspection result is an abnormal quality inspection result according to the first quality inspection result, and the abnormal quality inspection result indicates that the video content recorded by one or more video data in the video data does not meet the flow regulation of business handling. The quality inspector can confirm the one or more video data and generate corresponding quality inspection confirmation results.

In S150, after the quality inspection background receives the abnormal recording data and generates a corresponding quality inspection confirmation result, the quality inspection confirmation result may be sent to the quality inspection device. After receiving the quality inspection confirmation result, the quality inspection device can determine the abnormal recorded data corresponding to the quality inspection confirmation result and acquire a first quality inspection result corresponding to the abnormal recorded data. When the quality inspection confirmation result is inconsistent with the first quality inspection result, the quality inspection equipment can take the quality inspection confirmation result made by the quality inspection background as the final quality inspection result.

The quality inspection equipment can perform automatic quality inspection on massive recorded data after performing quality inspection on the recorded data to be subjected to quality inspection through the quality inspection model to obtain a first abnormal result, and screens the first quality inspection result containing the abnormal quality inspection result from a plurality of first quality inspection results. For the first quality inspection result with the screened abnormal quality inspection result, the quality inspection equipment can send the corresponding abnormal recorded data to the quality inspection background so as to carry out quality inspection confirmation on the abnormal recorded data by quality inspection personnel, and the result of the quality inspection confirmation is used as a final quality inspection result. The quality testing personnel do not need to carry out quality testing on massive recorded data, and only need to carry out secondary quality testing confirmation on abnormal recorded data with abnormality screened out by the quality testing model, so that the quality testing efficiency and the quality testing accuracy of the recorded data are greatly improved, the task load of the quality testing personnel is reduced, and the corresponding personnel cost is reduced.

As an alternative embodiment, referring to fig. 6, in order to, after the step S140, the method may further include:

s610, acquiring a plurality of recorded data with inconsistent quality inspection confirmation results and the first quality inspection results;

s620, the recorded data are sorted and counted, and problem service nodes corresponding to the recorded data are determined;

S630, according to the problem service node, adjusting and optimizing the model parameters of the quality inspection model to update the quality inspection model.

In this embodiment, after quality inspection is performed on a plurality of recorded data to be quality inspected by the quality inspection model, a first quality inspection result and a quality inspection confirmation result of each recorded data may be obtained. And sorting, counting and analyzing the recorded data with inconsistent first quality inspection results and quality inspection confirmation results, determining problem service nodes with inconsistent first quality inspection results and quality inspection confirmation results, and adjusting and optimizing model parameters of the quality inspection model according to the problem service nodes so as to update the quality inspection model and improve the quality inspection accuracy of the quality inspection model.

In S610, after receiving the quality inspection confirmation result returned by the quality inspection background, if the quality inspection confirmation result is inconsistent with the first quality inspection result, the quality inspection device takes the quality inspection confirmation result as a final quality inspection result. After the quality inspection is performed on the plurality of recorded data to be inspected, the quality inspection device may determine, from the inspected recorded data, a plurality of recorded data whose quality inspection confirmation result is inconsistent with the first quality inspection result.

In S620, after determining the recorded data with the quality inspection confirmation results inconsistent with the first quality inspection result, the quality inspection model may perform sorting statistics on the recorded data, and determine problem service nodes respectively corresponding to the recorded data.

It can be understood that, for a certain recorded data, the first quality inspection result may include an abnormality of multiple service nodes, and in the quality inspection confirmation process for the multiple service nodes in the quality inspection confirmation result, the quality inspection result for part of the service nodes in the multiple service nodes is different from the first quality inspection result. That is, when the first quality inspection result is inconsistent with the quality inspection confirmation result, the quality inspection results of some service nodes may be inconsistent, and the quality inspection results of other service nodes are consistent, the service node corresponding to the recorded data in question is a service node whose quality inspection confirmation result is different from the first quality inspection result. The service node whose quality inspection confirmation result is the same as the first quality inspection result may not be the problem service node.

In S630, after the problem service node corresponding to each piece of recorded data is determined, the reason why the quality inspection confirmation result is inconsistent with the first quality inspection result may be determined according to the problem service node, and the model parameters of the quality inspection model may be adjusted and optimized according to the determined problem reason, so as to update the quality inspection model, thereby improving the quality inspection efficiency and the quality inspection accuracy of the quality inspection model.

Taking the problem service node as the opinion solicitation client node as an example, when the video content of the video data is identified by the quality inspection model and the action characteristic that the user agrees with the point head or agrees with the gesture is not identified, the video quality inspection result in the first quality inspection result obtained by the quality inspection model is not passed. And when the quality inspection background performs quality inspection confirmation on the recorded data, if the user indicates agreement through voice, the recorded video quality inspection result in the quality inspection confirmation result is passed. According to the problem, the quality inspection model can be adjusted, when the corresponding behavior characteristics cannot be identified through video content identification, the recording subdata in the time period corresponding to the recording subdata is determined from the recording data, whether the voice characteristics corresponding to the behavior characteristics exist or not is determined through the voice text information corresponding to the recording subdata, and if the corresponding voice characteristics exist, the service node is determined to accord with the provision of service handling.

As an alternative embodiment, referring to fig. 7, in order for the above S630 to include:

s710, adjusting the model parameters according to the problem service node to obtain an adjusted quality inspection model;

s720, inputting a plurality of recorded data to the adjusted quality inspection model to obtain second quality inspection results corresponding to the recorded data respectively;

s730, calculating a first matching degree between the quality inspection confirmation results of the plurality of recorded data and the plurality of first quality inspection results, and a second matching degree between the quality inspection confirmation results of the plurality of recorded data and the plurality of second quality inspection results;

s740, determining an optimized quality inspection result from the first quality inspection result and the second quality inspection result according to the first matching degree and the second matching degree;

and S750, taking the quality inspection model corresponding to the optimized quality inspection result as an updated quality inspection model.

In this embodiment, after determining the problem service node whose first quality inspection result is inconsistent with the quality inspection confirmation result, the model parameter may be adjusted according to the problem service node to obtain an adjusted quality inspection model. And performing quality inspection on the same batch of recorded data through the adjusted quality inspection model to obtain a plurality of second quality inspection results. According to the first matching degree of the first quality inspection result and the quality inspection confirmation result, the difference between the quality inspection model before adjustment and the quality inspection background for quality inspection can be determined; and according to the second matching degree of the second quality inspection result and the quality inspection confirmation result, the difference between the adjusted quality inspection model and the quality inspection background for quality inspection can be determined. When the second matching degree corresponding to the adjusted quality inspection model is higher, the quality inspection accuracy of the quality inspection model can be improved by adjusting the model parameters, and then the adjusted quality inspection model can be used as an updated model. Otherwise, the quality inspection model before adjustment is still adopted.

In S710, the quality inspection device may adjust the model parameters of the quality inspection model according to the determined problem service node to obtain an adjusted quality inspection model.

In S720, the quality inspection apparatus may input the plurality of recorded data to the adjusted quality inspection model, so as to generate corresponding second quality inspection results according to the plurality of recorded data through the adjusted quality inspection model.

In S730, after determining the second quality inspection results corresponding to the plurality of recorded data, the quality inspection device may determine the first quality inspection results and the quality inspection confirmation results corresponding to the plurality of recorded data.

The quality inspection apparatus may calculate a first matching degree between the quality inspection confirmation results of the plurality of recorded data and the first quality inspection confirmation result, and calculate a second matching degree between the quality inspection confirmation results of the plurality of recorded data and the second quality inspection confirmation result.

In S740, the quality inspection apparatus may determine one of the first matching degree and the second matching degree, which has a higher matching degree, and determine a quality inspection result corresponding to the higher matching degree as an optimized quality inspection result. For example, when the first matching degree is higher than the second matching degree, the first quality inspection result may be determined as an optimized quality inspection result; otherwise, determining the second quality inspection result as the optimized quality inspection result.

It can be understood that, for a plurality of recorded data, there is a case where a first quality inspection result of a part of the recorded data is inconsistent with a quality inspection result, and there is a case where a second quality inspection result of another part of the recorded data is inconsistent with a quality inspection result. When the two recorded data are completely consistent, the quality inspection result generated by the quality inspection model is not changed before and after the adjustment of the model parameters. If the two parts of recorded data only have partial data superposition or do not have superposition, the quality inspection result generated by the quality inspection model changes through adjusting the model parameters. In this case, the quality inspection model with higher quality inspection accuracy can be determined according to the difference between the first quality inspection result and the quality inspection confirmation result and the difference between the second quality inspection result and the quality inspection confirmation result.

In S750, after determining the optimized quality inspection result from the first quality inspection result and the second quality inspection result, the quality inspection model corresponding to the optimized quality inspection result may be used as the updated quality inspection model.

Based on the recorded data quality inspection method provided by the embodiment, correspondingly, the application also provides a specific implementation mode of the recorded data quality inspection device. Please see the examples below.

Referring to fig. 8, a recorded data quality inspection apparatus 800 according to an embodiment of the present application includes the following modules:

An obtaining module 801, obtaining recording data to be quality-tested;

the input module 802 inputs the recorded data to be subjected to quality inspection into a quality inspection model, wherein the quality inspection model comprises a video identification submodel and an audio identification submodel;

the detection module 803 detects the recorded data according to the video identification submodel and the audio identification submodel to obtain a first quality inspection result of the recorded data;

the abnormality module 804 is used for sending abnormal quality inspection information to a quality inspection background when the first quality inspection result is an abnormal quality inspection result;

and the confirmation module 805 is configured to receive the quality inspection confirmation result sent by the quality inspection background, and replace the abnormal quality inspection result with the quality inspection confirmation result.

By inputting the recorded data to be quality-tested into the quality testing model, the recording data and the video data in the recorded data can be respectively detected and processed through the sub-model in the quality testing model, so that a first quality testing result of the recorded data is obtained. When the first quality inspection result comprises at least one abnormal quality inspection result, the abnormal recorded data can be sent to a quality inspection background so as to carry out quality inspection confirmation on the abnormal recorded data through the quality inspection background. When the quality inspection confirmation result is the same as the abnormal quality inspection result, the first quality inspection result can be used as a final quality inspection result; if the quality inspection result does not match the abnormal quality inspection result, the quality inspection result can be used as the final quality inspection result. The quality inspection model can be used for carrying out batch automatic quality inspection on a large amount of recorded data, and the first quality inspection result can be directly used as a final quality inspection result when abnormal recorded data do not exist in the quality inspection result. And performing secondary quality inspection confirmation on the recorded data with the abnormality through a quality inspection background to obtain a final quality inspection result. Quality inspection can be carried out through the quality inspection model, and quality inspection efficiency and quality inspection accuracy can be greatly improved. And the personnel of the quality inspection background only need to carry out quality inspection confirmation on the recorded data with the identified abnormality, thereby reducing the quality inspection workload of the quality inspection background and reducing the personnel cost of the quality inspection background.

The recorded data quality inspection device provided by the embodiment of the invention can realize each process in the method embodiments of fig. 1 to fig. 7, and is not described again to avoid repetition.

Fig. 9 is a schematic diagram illustrating a hardware structure of a recorded data quality inspection apparatus according to an embodiment of the present disclosure.

The recorded data quality testing device may include a processor 901 and memory 902 having stored thereon computer program instructions.

Specifically, the processor 901 may include a Central Processing Unit (CPU), or an Application Specific Integrated Circuit (ASIC), or may be configured to implement one or more Integrated circuits of the embodiments of the present Application.

Memory 902 may include mass storage for data or instructions. By way of example, and not limitation, memory 902 may include a Hard Disk Drive (HDD), floppy Disk Drive, flash memory, optical Disk, magneto-optical Disk, tape, or Universal Serial Bus (USB) Drive or a combination of two or more of these. Memory 902 may include removable or non-removable (or fixed) media, where appropriate. The memory 902 may be internal or external to the integrated gateway disaster recovery device, where appropriate. In a particular embodiment, the memory 902 is a non-volatile solid-state memory.

The memory may include Read Only Memory (ROM), Random Access Memory (RAM), magnetic disk storage media devices, optical storage media devices, flash memory devices, electrical, optical, or other physical/tangible memory storage devices. Thus, in general, the memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., memory devices) encoded with software comprising computer-executable instructions and when the software is executed (e.g., by one or more processors), it is operable to perform operations described with reference to the methods according to an aspect of the present disclosure.

The processor 901 reads and executes the computer program instructions stored in the memory 902 to implement any one of the quality inspection methods of the recorded data in the above embodiments.

In one example, the recorded data quality inspection device may also include a communication interface 903 and a bus 910. As shown in fig. 9, the processor 901, the memory 902, and the communication interface 903 are connected via a bus 910 to complete communication with each other.

The communication interface 903 is mainly used for implementing communication between modules, apparatuses, units and/or devices in this embodiment of the application.

Bus 910 includes hardware, software, or both to couple the components of the recorded data quality inspection device to each other. By way of example, and not limitation, a bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Enhanced Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), a Hypertransport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an infiniband interconnect, a Low Pin Count (LPC) bus, a memory bus, a Micro Channel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a video electronics standards association local (VLB) bus, or other suitable bus or a combination of two or more of these. Bus 910 can include one or more buses, where appropriate. Although specific buses are described and shown in the embodiments of the application, any suitable buses or interconnects are contemplated by the application.

The recorded data quality inspection device can be based on, and thus can realize the recorded data quality inspection method and device described in conjunction with fig. 1 to 8.

In addition, in combination with the recorded data quality inspection method in the foregoing embodiment, the embodiment of the present application may provide a computer storage medium to implement the method. The computer storage medium having computer program instructions stored thereon; the computer program instructions, when executed by a processor, implement any one of the methods for quality inspection of recorded data in the above embodiments.

In addition, a computer program product is provided in the embodiments of the present application, which includes computer program instructions, and when the computer program instructions are executed by a processor, the steps and corresponding contents of the foregoing method embodiments may be implemented.

It is to be understood that the present application is not limited to the particular arrangements and instrumentalities described above and shown in the attached drawings. A detailed description of known methods is omitted herein for the sake of brevity. In the above embodiments, several specific steps are described and shown as examples. However, the method processes of the present application are not limited to the specific steps described and illustrated, and those skilled in the art can make various changes, modifications and additions, or change the order between the steps, after comprehending the spirit of the present application.

The functional blocks shown in the above structural block diagrams may be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, it may be, for example, an electronic circuit, an Application Specific Integrated Circuit (ASIC), suitable firmware, plug-in, function card, or the like. When implemented in software, the elements of the present application are the programs or code segments used to perform the required tasks. The program or code segments can be stored in a machine-readable medium or transmitted by a data signal carried in a carrier wave over a transmission medium or a communication link. A "machine-readable medium" may include any medium that can store or transfer information. Examples of a machine-readable medium include electronic circuits, semiconductor memory devices, ROM, flash memory, Erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, fiber optic media, Radio Frequency (RF) links, and so forth. The code segments may be downloaded via computer networks such as the internet, intranet, etc.

It should also be noted that the exemplary embodiments mentioned in this application describe some methods or systems based on a series of steps or devices. However, the present application is not limited to the order of the above-described steps, that is, the steps may be performed in the order mentioned in the embodiments, may be performed in an order different from the order in the embodiments, or may be performed simultaneously.

Aspects of the present disclosure are described above with reference to flowchart illustrations and/or block diagrams of methods, apparatus and computer program products according to embodiments of the disclosure. It will be understood that each block of the flowchart illustrations and/or block diagrams, and combinations of blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, enable the implementation of the functions/acts specified in the flowchart and/or block diagram block or blocks. Such a processor may be, but is not limited to, a general purpose processor, a special purpose processor, an application specific processor, or a field programmable logic circuit. It will also be understood that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware for performing the specified functions or acts, or combinations of special purpose hardware and computer instructions.

As will be apparent to those skilled in the art, for convenience and brevity of description, the specific working processes of the systems, modules and units described above may refer to corresponding processes in the foregoing method embodiments, and are not described herein again. It should be understood that the scope of the present application is not limited thereto, and any person skilled in the art can easily conceive various equivalent modifications or substitutions within the technical scope of the present application, and these modifications or substitutions should be covered within the scope of the present application.

Claims

1. A method for quality inspection of recorded data, wherein the recorded data comprises audio recording data and video recording data, the method comprising:

acquiring recorded data to be subjected to quality inspection;

2. The method for inspecting the quality of the recorded data according to claim 1, wherein the step of detecting the recorded data according to the video identification submodel and the audio identification submodel to obtain a first quality inspection result of the recorded data comprises the steps of:

Determining video data corresponding to the video identification submodel and recording data corresponding to the audio identification submodel from the recording data to be quality tested;

3. The method for quality inspection of recorded data according to claim 2, wherein the processing the video data according to the video identification submodel to obtain the video quality inspection result of the video data comprises:

dividing the video data into a plurality of video subdata according to each service node, wherein the plurality of video subdata correspond to each service node one to one;

and respectively carrying out video content identification on the plurality of video recording subdata according to the video identification submodels to obtain video quality inspection results.

4. The method for quality inspection of recorded data according to claim 3, wherein the dividing the video data into a plurality of video data according to each service node comprises:

acquiring video characteristics corresponding to each service node;

and dividing the video data according to each video frame to obtain a plurality of video subdata.

5. The method for quality inspection of recorded data according to claim 2, wherein the processing the recorded data according to the audio recognition submodel to obtain the recorded quality inspection result of the recorded data comprises:

performing voice recognition processing on the recording data to obtain voice text information;

6. The method for quality inspection of recorded data according to claim 1, wherein after the step of using the quality inspection confirmation result as a final quality inspection result, the method further comprises:

7. The method of claim 6, wherein the adjusting and optimizing model parameters of the quality control model according to the problem service node to update the quality control model comprises:

8. A recorded data quality inspection apparatus, comprising:

9. A recorded data quality inspection apparatus, characterized by comprising: a processor and a memory storing computer program instructions;

the processor, when executing the computer program instructions, implements the method of quality inspection of recorded data according to any one of claims 1-7.

10. A computer storage medium having computer program instructions stored thereon which, when executed by a processor, implement a method for quality inspection of recorded data according to any one of claims 1 to 7.

11. A computer program product comprising computer program instructions which, when executed by a processor, implement the method of quality testing recorded data according to any one of claims 1 to 7.