CN117973910A - Performance evaluation method, device and storage medium based on voiceprint and matching keywords - Google Patents

Abstract

The invention relates to a performance assessment method, a device and a storage medium based on voiceprints and matching keywords, which are applied to the technical field of call recording and comprise the following steps: the method comprises the steps of formulating a standard assessment index, obtaining a conversation record of an employee, carrying out voiceprint recognition and matching according to voiceprint characteristics of the employee, obtaining an audio paragraph of the employee, converting the audio paragraph of the employee into a text paragraph, matching the text paragraph with keywords in the formulated index, and carrying out performance assessment through rule calculation automation; compared with the existing manual checking mode, the method does not need a large amount of manual matching and statistics, reduces the manual and time cost, can dynamically change checking rules, obtains checking results by intelligent calculation of the system, and can accurately match electric pin staff according to voiceprints at the same time, thereby improving the accuracy.

Description

Performance evaluation method, device and storage medium based on voiceprint and matching keywords

Technical Field

The invention relates to the technical field of call recording, in particular to a performance assessment method, a performance assessment device and a storage medium based on voiceprints and matching keywords.

Background

In some of the existing companies, staff talk with clients through telephone, and in order to evaluate the performance of staff, the talk content is usually recorded;

in the existing scheme, special assessment personnel are arranged, and the assessment personnel manually assess the performance of electric sales personnel through the number of key words hit on the recorded content by listening to the call record.

The performance assessment is carried out by personnel listening to the record, so that on one hand, the labor cost is high, the subjectivity is too strong, the performance assessment cannot be completed efficiently, and on the other hand, a great amount of time is required.

Disclosure of Invention

Therefore, the invention aims to provide a performance evaluation method, a device and a storage medium based on voiceprints and matching keywords, so as to solve the problems that in the prior art, performance scores of staff are evaluated by manually listening to call records to identify keywords, on one hand, labor cost is high, subjectivity is too high, performance evaluation cannot be completed efficiently, and on the other hand, a large amount of time is required.

According to a first aspect of an embodiment of the present invention, there is provided a performance assessment method based on voiceprints and matching keywords, the method comprising:

Creating performance classification business types, building corresponding keywords according to different business types, and building performance evaluation rules of the corresponding business types according to the keywords;

acquiring voiceprint characteristics of each employee;

acquiring call records of staff and clients and corresponding service types, uploading the call records to a server, and matching the call records with voiceprint features of all staff by the server according to the service types to acquire audio paragraphs of the corresponding staff;

Converting the audio paragraphs of the corresponding staff into text paragraphs;

And extracting keywords of the corresponding business types in the text paragraphs, and obtaining the performance score of the employee according to the extracted keywords and the performance evaluation rules of the corresponding business types.

Preferably, the method comprises the steps of,

The different service types include: and receiving the sound recording and recording the conversation.

Preferably, the first and second channels are arranged in a row,

The step of obtaining the voiceprint characteristics of each employee comprises the following steps:

the staff records the speaking according to a preset section of characters, performs voice conversion according to voice print audio recorded by the staff, matches the voice print audio with the preset section of characters according to the result of the conversion, extracts voice print characteristics of the staff if the matching is successful, and re-records if the matching is failed.

Preferably, the method comprises the steps of,

The server matches the call record with the voiceprint features of each employee according to the service type, and the obtaining of the audio paragraph of the corresponding employee comprises:

If the service type is call recording, respectively extracting audio paragraphs corresponding to two audio tracks by identifying audio tracks of staff and clients in the call recording, selecting any audio paragraph corresponding to one audio track to match with voiceprint features of the staff, if matching, indicating that the audio paragraph corresponding to the audio track is the audio paragraph of the staff, and if not, the audio paragraph corresponding to the other audio track is the audio paragraph of the staff.

Preferably, the method comprises the steps of,

If the service type is the reception record, preprocessing the call record to remove interference factors;

and carrying out voice conversion on the call record after the interference factors are removed to distinguish each paragraph, carrying out audio cutting on the call record through the starting time and the ending time of each paragraph to obtain a plurality of audio clips, respectively matching each audio paragraph with the voiceprint characteristics of the staff, and extracting the audio paragraphs of the staff.

Preferably, the method comprises the steps of,

After the audio paragraphs of the corresponding staff are acquired, the audio paragraphs of the corresponding staff are stored after text transfer recognition, and meanwhile, the audio paragraphs of the staff are uploaded to cloud storage asynchronously based on HLS.

Preferably, the method comprises the steps of,

The mobile application of the staff opens the access authority of the microphone and the call recording authority, the mobile application slices the call recording during the call between the staff and the client, and the sliced aac format file is uploaded to the server, wherein the uploading interval is that the file is uploaded every X seconds.

Preferably, the method comprises the steps of,

After the call recording is finished, the server receives the end identification and the service type field transmitted by the mobile application, and asynchronously sets a timing detection task to process the waiting for the completion of the processing of the fragmented audio.

According to a second aspect of an embodiment of the present invention, there is provided a performance assessment apparatus based on voiceprints and matching keywords, the apparatus comprising:

the rule setting module: the system is used for creating performance classification business types, corresponding keywords are established according to different business types, and performance evaluation rules of the corresponding business types are established according to the keywords;

voiceprint feature acquisition module: the voice print feature acquiring module is used for acquiring voice print features of each employee;

An audio paragraph acquisition module: the method comprises the steps that call records of staff and clients and corresponding service types are obtained, the call records are uploaded to a server, and the server matches the call records with voiceprint features of all staff according to the service types to obtain audio paragraphs of corresponding staff;

and a text conversion module: the audio paragraph conversion module is used for converting the audio paragraph of the corresponding employee into a text paragraph;

And a performance scoring module: and the system is used for extracting keywords of the corresponding business types in the text paragraphs, and obtaining the performance score of the employee according to the extracted keywords and the performance evaluation rules of the corresponding business types.

According to a third aspect of embodiments of the present invention, there is provided a storage medium storing a computer program which, when executed by a master, implements the steps of the above-described method.

The technical scheme provided by the embodiment of the invention can comprise the following beneficial effects:

According to the application, through formulating standard assessment indexes, through acquiring call records of staff, and carrying out voiceprint recognition and matching according to voiceprint characteristics of the staff, audio paragraphs of the staff are acquired, the audio paragraphs of the staff are converted into text paragraphs, the text paragraphs are matched with keywords in the formulated indexes, and performance assessment is carried out automatically through rule calculation; compared with the existing manual checking mode, the method does not need a large amount of manual matching and statistics, reduces the manual and time cost, can dynamically change checking rules, obtains checking results by intelligent calculation of the system, and can accurately match electric pin staff according to voiceprints at the same time, thereby improving the accuracy.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention as claimed.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and together with the description, serve to explain the principles of the invention.

Fig. 1 is a flow diagram illustrating a voiceprint and matching keyword based performance assessment method according to an illustrative embodiment;

Fig. 2 is a system diagram of a voiceprint and matching keyword based performance assessment apparatus shown in accordance with another illustrative embodiment;

in the accompanying drawings: the system comprises a 1-rule setting module, a 2-voiceprint feature acquisition module, a 3-audio paragraph acquisition module, a 4-text conversion module and a 5-performance scoring module.

Detailed Description

Reference will now be made in detail to exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, the same numbers in different drawings refer to the same or similar elements, unless otherwise indicated. The implementations described in the following exemplary examples do not represent all implementations consistent with the invention. Rather, they are merely examples of apparatus and methods consistent with aspects of the invention as detailed in the accompanying claims.

Example 1

Fig. 1 is a flow chart illustrating a voiceprint and matching keyword based performance assessment method according to an exemplary embodiment, the method comprising:

S1, creating performance classification business types, building corresponding keywords according to different business types, and building performance evaluation rules of the corresponding business types according to the keywords;

s2, obtaining voiceprint characteristics of each employee;

S3, acquiring call records of staff and clients and corresponding service types, uploading the call records to a server, and matching the call records with voiceprint features of all staff by the server according to the service types to acquire audio paragraphs of corresponding staff;

s4, converting the audio paragraphs of the corresponding staff into text paragraphs;

S5, extracting keywords of the corresponding business types in the text paragraphs, and obtaining performance scores of the staff according to the extracted keywords and performance evaluation rules of the corresponding business types;

It can be appreciated that the background is managed by the login server first, and the performance classification associated service types are created: such as reception recording, call recording, etc.; for different service types, corresponding keywords are created, corresponding service types are selected, and an assessment rule is created, wherein the rule can be configured to have an assessment score or hit rate corresponding to the hit number of the keywords and a corresponding assessment score, for example, the number of the hits of the keywords is 5, the corresponding assessment score is 10, the number of the hits of the keywords is 10, and the corresponding assessment score is 20; or the hit rate of the key words is 20 percent, and the corresponding assessment score is 10; logging in a mobile application, clicking and logging in voiceprint, displaying a section of text by a system, recording the speech of an employee according to the text content, performing voice-to-text conversion by the mobile application according to voiceprint audio recorded by the employee, matching the voice-to-text with the text according to the result of the text conversion, extracting voiceprint characteristics of the employee if the matching is successful, and re-recording if the matching is failed until the logging is successful, wherein the step is to detect whether the voice-to-text of the employee is clear and effective; the mobile equipment needs to open the access right of the microphone and the call recording right, the mobile application slices the call recording, and uploads the sliced aac-format files to the information system once every ten seconds, and the uploading interval time is dynamically configurable; the information system receives the uploaded fragmented files, and asynchronously matches the voiceprint mp3 files which are input by staff in advance with the uploaded fragmented files, so that the unvented sounds of the staff and the clients are separated by matching, and the specific process is as follows for two different service types (reception recording and call recording):

Call recording:

The method comprises the steps of carrying out recognition processing on call records through an FFmpeg open source program, obtaining different sound tracks in the call records, distinguishing sound tracks of electric sales staff from sound tracks of clients by the sound tracks (sound channels), wherein the sound tracks are divided into two types (namely clients and electric sales staff) at most because the sound tracks have a mode of 1 to 1, and then carrying out MFCC feature extraction on the sound tracks in the sound tracks through the sound tracks after the sound tracks are cut and then matching, so that the attribution of the two sound tracks can be obtained, namely, the sound tracks corresponding to the sound tracks can be known to be the electric sales staff. The MFCC features preserve some content that is semantically related, filtering out irrelevant information such as background noise. The MFCC is characterized by the use of a set of key coefficients used to create mel-cepstral, so that its cepstral and human non-linear auditory system are more closely related, and the MFCC features are extracted as follows: framing and windowing an audio signal, carrying out Fourier transform on the signal, mapping a frequency spectrum to a Mel scale, taking logarithm, and carrying out DCT (discrete cosine transform) to obtain MFCC (frequency division multiplexing) characteristics;

Reception record:

the method comprises the steps that a recording file needs to be preprocessed in advance in the complex scene of the to-be-recorded, so that other influencing factors such as background noise are removed, the to-be-recorded is mono, paragraphs are needed to be distinguished from voice-to-text in advance in the to-be-recorded, audio cutting is carried out according to the starting time and the ending time of the paragraphs, feature extraction and comparison are carried out on audio fragments one by one after cutting, and finally a matching result is obtained, namely, the audio paragraphs of electric marketing personnel are obtained;

After matching, performing text transfer recognition on voice paragraphs judged as staff, and storing, and asynchronously uploading segmented audio to cloud storage based on HLS;

Triggering the matching keyword event to perform asynchronous processing after the call or the reception recording is finished, and taking interface efficiency into consideration because the real-time requirement of performance assessment problems is not high, using a message queue to perform asynchronous processing to a certain extent;

After the recording is finished, the information system receives an ending identification and a service classification field transmitted by the mobile application, and asynchronously sets a timing detection task to process whether the fragmented audio is processed or not;

if the processing is finished, collecting the text conversion result of the conversation identified as the employee into text paragraphs;

matching the service classification field uploaded through the mobile application interaction to a corresponding service type, and performing matching hit with keywords associated with the performance classification;

Triggering according to rules corresponding to the keyword matching result and the service distribution, calculating a score and storing the score;

It is worth to say that, based on the slice file uploaded before, the m3u8 file of the audio can be subjected to voice live broadcast, and the live broadcast delay can be controlled within 20 seconds; if staff disagrees with the examination result, the on-demand manual rechecking can be performed based on the voice slice file;

the performance assessment and the calculation of average value according to the sum of the keyword matching scores of each call in a certain time dimension or the calculation of corresponding result value according to the additional public representation;

The calculation can be triggered at fixed time according to the background configuration of the server, and the manual trigger calculation is supported;

If the calculation result is in doubt, the manual modification of the performance score can be supported after the manual review;

According to the application, through formulating standard assessment indexes, through acquiring call records of staff, and carrying out voiceprint recognition and matching according to voiceprint characteristics of the staff, audio paragraphs of the staff are acquired, the audio paragraphs of the staff are converted into text paragraphs, the text paragraphs are matched with keywords in the formulated indexes, and performance assessment is carried out automatically through rule calculation; compared with the existing manual checking mode, the method does not need a large amount of manual matching and statistics, reduces the manual and time cost, can dynamically change checking rules, obtains checking results by intelligent calculation of the system, and can accurately match electric pin staff according to voiceprints at the same time, thereby improving the accuracy.

Embodiment two:

Fig. 2 is a system diagram of a voiceprint and matching keyword based performance assessment apparatus shown in accordance with another exemplary embodiment, the apparatus comprising:

rule setting module 1: the system is used for creating performance classification business types, corresponding keywords are established according to different business types, and performance evaluation rules of the corresponding business types are established according to the keywords;

voiceprint feature acquisition module 2: the voice print feature acquiring module is used for acquiring voice print features of each employee;

Audio paragraph acquisition module 3: the method comprises the steps that call records of staff and clients and corresponding service types are obtained, the call records are uploaded to a server, and the server matches the call records with voiceprint features of all staff according to the service types to obtain audio paragraphs of corresponding staff;

The text conversion module 4: the audio paragraph conversion module is used for converting the audio paragraph of the corresponding employee into a text paragraph;

Performance scoring module 5: the system comprises a text paragraph, a performance evaluation rule and a keyword extraction module, wherein the text paragraph is used for extracting keywords of corresponding service types in the text paragraph, and obtaining performance scores of the staff according to the extracted keywords and the performance evaluation rule of the corresponding service types;

It can be understood that the rule setting module 1 is used for creating performance classification service types, corresponding keywords are established according to different service types, and performance evaluation rules of the corresponding service types are established according to the keywords; the voiceprint feature acquisition module 2 is used for acquiring voiceprint features of each employee; the audio paragraph obtaining module 3 is used for obtaining call records of staff and clients and corresponding service types, uploading the call records to a server, and matching the call records with voiceprint features of all staff according to the service types by the server to obtain audio paragraphs of corresponding staff; the text conversion module 4 is used for converting the audio paragraph of the corresponding employee into a text paragraph; the performance scoring module 5 is used for extracting keywords of the corresponding business types in the text paragraphs, and obtaining the performance score of the employee according to the extracted keywords and the performance evaluation rules of the corresponding business types; according to the embodiment, through formulating standard assessment indexes, through acquiring call records of staff, carrying out voiceprint recognition and matching according to voiceprint characteristics of the staff, acquiring audio paragraphs of the staff, converting the audio paragraphs of the staff into text paragraphs, matching the text paragraphs with keywords in the formulated indexes, and carrying out performance assessment through rule calculation automation; compared with the existing manual checking mode, the method does not need a large amount of manual matching and statistics, reduces the manual and time cost, can dynamically change checking rules, obtains checking results by intelligent calculation of the system, and can accurately match electric pin staff according to voiceprints at the same time, thereby improving the accuracy.

Embodiment III:

The present embodiment provides a storage medium storing a computer program which, when executed by a master controller, implements each step in the above method;

It is to be understood that the above-mentioned storage medium may be a read-only memory, a magnetic disk or an optical disk, etc.

It is to be understood that the same or similar parts in the above embodiments may be referred to each other, and that in some embodiments, the same or similar parts in other embodiments may be referred to.

It should be noted that in the description of the present invention, the terms "first," "second," and the like are used for descriptive purposes only and are not to be construed as indicating or implying relative importance. Furthermore, in the description of the present invention, unless otherwise indicated, the meaning of "plurality" means at least two.

Any process or method descriptions in flow charts or otherwise described herein may be understood as representing modules, segments, or portions of code which include one or more executable instructions for implementing specific logical functions or steps of the process, and further implementations are included within the scope of the preferred embodiment of the present invention in which functions may be executed out of order from that shown or discussed, including substantially concurrently or in reverse order, depending on the functionality involved, as would be understood by those reasonably skilled in the art of the present invention.

It is to be understood that portions of the present invention may be implemented in hardware, software, firmware, or a combination thereof. In the above-described embodiments, the various steps or methods may be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, may be implemented using any one or combination of the following techniques, as is well known in the art: discrete logic circuits having logic gates for implementing logic functions on data signals, application specific integrated circuits having suitable combinational logic gates, programmable Gate Arrays (PGAs), field Programmable Gate Arrays (FPGAs), and the like.

Those of ordinary skill in the art will appreciate that all or a portion of the steps carried out in the method of the above-described embodiments may be implemented by a program to instruct related hardware, where the program may be stored in a computer readable storage medium, and where the program, when executed, includes one or a combination of the steps of the method embodiments.

In addition, each functional unit in the embodiments of the present invention may be integrated in one processing module, or each unit may exist alone physically, or two or more units may be integrated in one module. The integrated modules may be implemented in hardware or in software functional modules. The integrated modules may also be stored in a computer readable storage medium if implemented in the form of software functional modules and sold or used as a stand-alone product.

The above-mentioned storage medium may be a read-only memory, a magnetic disk or an optical disk, or the like.

In the description of the present specification, a description referring to terms "one embodiment," "some embodiments," "examples," "specific examples," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present invention. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiments or examples. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

While embodiments of the present invention have been shown and described above, it will be understood that the above embodiments are illustrative and not to be construed as limiting the invention, and that variations, modifications, alternatives and variations may be made to the above embodiments by one of ordinary skill in the art within the scope of the invention.

Claims (10)

1. The performance evaluation method based on voiceprint and matching keywords is characterized by comprising the following steps:

Creating performance classification business types, building corresponding keywords according to different business types, and building performance evaluation rules of the corresponding business types according to the keywords;

acquiring voiceprint characteristics of each employee;

acquiring call records of staff and clients and corresponding service types, uploading the call records to a server, and matching the call records with voiceprint features of all staff by the server according to the service types to acquire audio paragraphs of the corresponding staff;

Converting the audio paragraphs of the corresponding staff into text paragraphs;

And extracting keywords of the corresponding business types in the text paragraphs, and obtaining the performance score of the employee according to the extracted keywords and the performance evaluation rules of the corresponding business types.

2. The method of claim 1, wherein the step of determining the position of the substrate comprises,

The different service types include: and receiving the sound recording and recording the conversation.

3. The method of claim 2, wherein the step of determining the position of the substrate comprises,

The step of obtaining the voiceprint characteristics of each employee comprises the following steps:

the staff records the speaking according to a preset section of characters, performs voice conversion according to voice print audio recorded by the staff, matches the voice print audio with the preset section of characters according to the result of the conversion, extracts voice print characteristics of the staff if the matching is successful, and re-records if the matching is failed.

4. The method of claim 2, wherein the step of determining the position of the substrate comprises,

The server matches the call record with the voiceprint features of each employee according to the service type, and the obtaining of the audio paragraph of the corresponding employee comprises:

If the service type is call recording, respectively extracting audio paragraphs corresponding to two audio tracks by identifying audio tracks of staff and clients in the call recording, selecting any audio paragraph corresponding to one audio track to match with voiceprint features of the staff, if matching, indicating that the audio paragraph corresponding to the audio track is the audio paragraph of the staff, and if not, the audio paragraph corresponding to the other audio track is the audio paragraph of the staff.

5. The method of claim 4, wherein the step of determining the position of the first electrode is performed,

If the service type is the reception record, preprocessing the call record to remove interference factors;

and carrying out voice conversion on the call record after the interference factors are removed to distinguish each paragraph, carrying out audio cutting on the call record through the starting time and the ending time of each paragraph to obtain a plurality of audio clips, respectively matching each audio paragraph with the voiceprint characteristics of the staff, and extracting the audio paragraphs of the staff.

6. The method according to claim 4 or 5, wherein,

After the audio paragraphs of the corresponding staff are acquired, the audio paragraphs of the corresponding staff are stored after text transfer recognition, and meanwhile, the audio paragraphs of the staff are uploaded to cloud storage asynchronously based on HLS.

7. The method of claim 1, wherein the step of determining the position of the substrate comprises,

The mobile application of the staff opens the access authority of the microphone and the call recording authority, the mobile application slices the call recording during the call between the staff and the client, and the sliced aac format file is uploaded to the server, wherein the uploading interval is that the file is uploaded every X seconds.

8. The method of claim 7, wherein the step of determining the position of the probe is performed,

After the call recording is finished, the server receives the end identification and the service type field transmitted by the mobile application, and asynchronously sets a timing detection task to process the waiting for the completion of the processing of the fragmented audio.

9. Performance assessment device based on voiceprint and matching keywords, characterized in that the device comprises:

the rule setting module: the system is used for creating performance classification business types, corresponding keywords are established according to different business types, and performance evaluation rules of the corresponding business types are established according to the keywords;

voiceprint feature acquisition module: the voice print feature acquiring module is used for acquiring voice print features of each employee;

An audio paragraph acquisition module: the method comprises the steps that call records of staff and clients and corresponding service types are obtained, the call records are uploaded to a server, and the server matches the call records with voiceprint features of all staff according to the service types to obtain audio paragraphs of corresponding staff;

and a text conversion module: the audio paragraph conversion module is used for converting the audio paragraph of the corresponding employee into a text paragraph;

And a performance scoring module: and the system is used for extracting keywords of the corresponding business types in the text paragraphs, and obtaining the performance score of the employee according to the extracted keywords and the performance evaluation rules of the corresponding business types.

10. A storage medium storing a computer program which, when executed by a master, implements the steps of the voiceprint and matching keyword based performance assessment method of any one of claims 1 to 8.