CN111696528A - Voice quality inspection method and device, quality inspection equipment and readable storage medium - Google Patents
Voice quality inspection method and device, quality inspection equipment and readable storage medium Download PDFInfo
- Publication number
- CN111696528A CN111696528A CN202010569722.4A CN202010569722A CN111696528A CN 111696528 A CN111696528 A CN 111696528A CN 202010569722 A CN202010569722 A CN 202010569722A CN 111696528 A CN111696528 A CN 111696528A
- Authority
- CN
- China
- Prior art keywords
- quality inspection
- quality
- voice
- violation
- preset
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000007689 inspection Methods 0.000 title claims abstract description 226
- 238000000034 method Methods 0.000 title claims abstract description 57
- 238000012372 quality testing Methods 0.000 claims abstract description 49
- 230000014509 gene expression Effects 0.000 claims description 39
- 238000004590 computer program Methods 0.000 claims description 13
- 238000012545 processing Methods 0.000 abstract description 5
- 230000000694 effects Effects 0.000 abstract description 4
- 238000003908 quality control method Methods 0.000 description 17
- 230000008569 process Effects 0.000 description 10
- 230000006870 function Effects 0.000 description 7
- 238000010276 construction Methods 0.000 description 5
- 238000006243 chemical reaction Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 230000004044 response Effects 0.000 description 4
- 239000012634 fragment Substances 0.000 description 3
- 238000012549 training Methods 0.000 description 3
- 239000000203 mixture Substances 0.000 description 2
- 230000011218 segmentation Effects 0.000 description 2
- 238000012790 confirmation Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000009432 framing Methods 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000036651 mood Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/04—Segmentation; Word boundary detection
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/26—Speech to text systems
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/03—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 characterised by the type of extracted parameters
- G10L25/24—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 characterised by the type of extracted parameters the extracted parameters being the cepstrum
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/48—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use
- G10L25/51—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use for comparison or discrimination
- G10L25/60—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use for comparison or discrimination for measuring the quality of voice signals
Landscapes
- Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Quality & Reliability (AREA)
- Telephonic Communication Services (AREA)
Abstract
The application is applicable to the technical field of voice processing, and provides a voice quality inspection method, a voice quality inspection device, quality inspection equipment and a readable storage medium, wherein the method comprises the following steps: and acquiring a voice text to be quality-tested, inputting the voice text to be quality-tested into a preset quality testing model, determining a target quality testing voice text matched with a quality testing item in the quality testing model, and performing quality testing on the target quality testing voice text according to the quality testing item to obtain a quality testing result. Therefore, the quality testing method and the quality testing device can directly perform automatic quality testing on the voice text to be tested obtained by converting the voice audio to be tested through the preset quality testing model, do not need manual quality testing on the voice audio to be tested, and achieve the effects of improving the accuracy of quality testing results and the quality testing efficiency.
Description
Technical Field
The present application belongs to the field of speech processing technologies, and in particular, to a speech quality inspection method, apparatus, quality inspection device, and readable storage medium.
Background
At present, voice quality inspection is to perform quality inspection on the recording of a call center so as to achieve the purposes of improving customer satisfaction, perfecting customer service, evaluating the work of customer service personnel and the like. Generally, voice quality inspection adopts a manual quality inspection mode, and because the human ear is needed to identify whether the audio frequency recorded by the quality inspection has problems, the problems of inaccurate quality inspection result, low quality inspection efficiency and the like exist.
Disclosure of Invention
The embodiment of the application provides a voice quality inspection method and device, quality inspection equipment and a readable storage medium, and can solve the problems that in the prior art, a quality inspection result is not accurate enough and the quality inspection efficiency is low.
In a first aspect, an embodiment of the present application provides a voice quality inspection method, including:
acquiring a voice text to be quality tested;
inputting the voice text to be quality-tested into a preset quality testing model, and determining a target quality testing voice text matched with a quality testing item in the quality testing model, wherein the quality testing model comprises at least one quality testing item;
and performing quality inspection on the target quality inspection voice text according to the quality inspection item to obtain a quality inspection result.
In a possible implementation manner of the first aspect, before the obtaining the speech text to be quality-checked, the method further includes:
acquiring voice audio to be quality tested;
and converting the voice audio to be tested into a voice text to be tested.
In a possible implementation manner of the first aspect, before the obtaining the speech text to be quality-checked, the method further includes:
obtaining quality inspection items, wherein the quality inspection items comprise violation links, violation contents and violation types;
and obtaining the preset quality inspection model according to the quality inspection item.
In a possible implementation manner of the first aspect, obtaining the preset quality inspection model according to the quality inspection item includes:
searching violation links and quality inspection items with the same violation types;
carrying out clustering analysis on violation contents of quality inspection items with the same violation links and violation types to obtain a violation content set;
and forming the preset quality inspection model according to the violation content set.
In a possible implementation manner of the first aspect, forming a preset quality inspection model corresponding to the quality inspection item according to the illegal content set includes:
and converting the violation content into a regular expression corresponding to the quality inspection item, and forming a preset quality inspection model according to the regular expression.
In a possible implementation manner of the first aspect, after searching for a regular expression matched with the illegal content set according to a preset matching rule and using the regular expression as a preset quality inspection model of the quality inspection item, the method further includes:
identifying generalized characters of the regular expression in the preset quality inspection model;
and replacing generalized characters of the regular expression with fuzzy characters.
In a possible implementation manner of the first aspect, performing quality inspection on the target quality inspection speech text according to the quality inspection item to obtain a quality inspection result, includes:
identifying the violation type of the target quality inspection voice text according to the quality inspection item;
and generating violation scores of the target quality inspection voice text according to the preset weight corresponding to the violation types, and obtaining a quality inspection result according to a comparison result of the violation scores and a preset score threshold.
In a second aspect, an embodiment of the present application provides a voice quality inspection apparatus, including:
the acquisition module is used for acquiring a voice text to be subjected to quality inspection;
the determining module is used for inputting the voice text to be quality tested into a preset quality testing model, determining a target quality testing voice text matched with quality testing items in the quality testing model, and enabling each quality testing item to correspond to one quality testing type;
and the quality inspection module is used for performing quality inspection on the target quality inspection voice text according to the quality inspection item to obtain a quality inspection result.
In a possible implementation manner of the second aspect, the voice quality inspection module further includes:
the audio acquisition module is used for acquiring the voice audio to be subjected to quality inspection;
and the conversion module is used for converting the voice audio to be subjected to quality inspection into a voice text to be subjected to quality inspection.
In a possible implementation manner of the second aspect, the voice quality inspection module further includes:
the quality inspection item acquisition module is used for acquiring quality inspection items, wherein the quality inspection items comprise violation links, violation contents and violation types;
and the quality inspection model construction module is used for obtaining the preset quality inspection model according to the quality inspection item.
In a possible implementation manner of the second aspect, the quality inspection model building module includes:
the searching unit is used for searching illegal links and quality inspection items with the same illegal types;
the clustering unit is used for clustering and analyzing the violation contents of the quality inspection items with the same violation links and violation types to obtain a violation content set;
and the construction unit is used for forming the preset quality inspection model according to the violation content set.
In a possible implementation manner of the second aspect, the construction unit includes:
a conversion subunit, configured to convert the illegal content into a regular expression corresponding to the quality inspection item, and form a preset quality inspection model according to the regular expression
In a possible implementation manner of the second aspect, the building unit further includes:
the identification subunit is used for identifying generalized characters of the regular expression in the preset quality inspection model;
a replacement subunit for replacing the generalized characters of the regular expression with the fuzzy characters
In one possible implementation manner of the second aspect, the quality inspection module includes:
the type identification unit is used for identifying the violation type of the target quality inspection voice text according to the quality inspection item;
and the generating unit is used for generating violation scores of the target quality inspection voice text according to the preset weight corresponding to the violation types and obtaining a quality inspection result according to a comparison result of the violation scores and a preset score threshold.
In a third aspect, an embodiment of the present application provides a quality inspection apparatus, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and when the processor executes the computer program, the method of the first aspect is implemented
In a fourth aspect, the present application provides a readable storage medium, which stores a computer program, and when the computer program is executed by a processor, the method according to the first aspect is implemented
Compared with the prior art, the embodiment of the application has the advantages that:
in the embodiment of the application, the quality control voice text to be detected obtained by converting the voice audio to be detected can be directly subjected to automatic quality control through the preset quality control model, the quality control of the voice audio to be detected does not need to be manually performed, and the effects of improving the accuracy rate of quality control results and the quality control efficiency are achieved.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the embodiments or the prior art descriptions will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and it is obvious for those skilled in the art to obtain other drawings without creative efforts.
Fig. 1 is a schematic flowchart of a voice quality inspection method according to an embodiment of the present application;
fig. 2 is a schematic flowchart of the voice quality inspection method according to the embodiment of the present application before step S101 in fig. 1;
fig. 3 is another schematic flow chart of the voice quality inspection method according to the embodiment of the present application before step S101 in fig. 1;
fig. 4 is a schematic flowchart illustrating a specific process of step S302 in fig. 3 of the voice quality inspection method according to an embodiment of the present application;
fig. 5 is a schematic flowchart illustrating a specific process of step S103 in fig. 1 of a voice quality inspection method according to an embodiment of the present application;
fig. 6 is a schematic structural diagram of a voice quality inspection apparatus according to an embodiment of the present application;
fig. 7 is a schematic structural diagram of a quality inspection apparatus according to an embodiment of the present application.
Detailed Description
In the following description, for purposes of explanation and not limitation, specific details are set forth, such as particular system structures, techniques, etc. in order to provide a thorough understanding of the embodiments of the present application. It will be apparent, however, to one skilled in the art that the present application may be practiced in other embodiments that depart from these specific details. In other instances, detailed descriptions of well-known systems, devices, circuits, and methods are omitted so as not to obscure the description of the present application with unnecessary detail.
It will be understood that the terms "comprises" and/or "comprising," when used in this specification and the appended claims, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.
It should also be understood that the term "and/or" as used in this specification and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.
As used in this specification and the appended claims, the term "if" may be interpreted contextually as "when", "upon" or "in response to" determining "or" in response to detecting ". Similarly, the phrase "if it is determined" or "if a [ described condition or event ] is detected" may be interpreted contextually to mean "upon determining" or "in response to determining" or "upon detecting [ described condition or event ]" or "in response to detecting [ described condition or event ]".
Furthermore, in the description of the present application and the appended claims, the terms "first," "second," "third," and the like are used for distinguishing between descriptions and not necessarily for describing or implying relative importance.
Reference throughout this specification to "one embodiment" or "some embodiments," or the like, means that a particular feature, structure, or characteristic described in connection with the embodiment is included in one or more embodiments of the present application. Thus, appearances of the phrases "in one embodiment," "in some embodiments," "in other embodiments," or the like, in various places throughout this specification are not necessarily all referring to the same embodiment, but rather "one or more but not all embodiments" unless specifically stated otherwise. The terms "comprising," "including," "having," and variations thereof mean "including, but not limited to," unless expressly specified otherwise.
The technical solution provided by the present application is described below by specific examples.
Referring to fig. 1, a schematic flow chart of a voice quality inspection method provided in an embodiment of the present application is shown, by way of example and not limitation, the method may be applied to a quality inspection device, where the quality inspection device includes a terminal device or a server, and the method may include the following steps:
and S101, acquiring a voice text to be subjected to quality inspection.
In a specific application, the voice text to be quality-checked can be a voice text obtained by converting a voice call between a customer service person and a client in the insurance industry.
It should be noted that the voice text to be quality-tested acquired in the embodiment of the present application may be a directly acquired voice text to be quality-tested, and the voice text to be quality-tested may be a voice text to be quality-tested obtained after processing the voice audio to be quality-tested by an external terminal device or a server.
In addition, the voice text to be quality-checked obtained in the embodiment of the present application may be the voice text to be quality-checked obtained after the voice text to be quality-checked is directly obtained and processed.
The following describes how the voice text to be quality-checked obtained in the embodiment of the present application may be the voice text to be quality-checked obtained after the voice text to be quality-checked is directly obtained and processed.
By way of example and not limitation, referring to fig. 2, a flowchart before step S101 in fig. 1 for providing a voice quality inspection method according to an embodiment of the present application is shown, where before obtaining a voice text to be quality inspected, the method further includes:
step S201, voice audio to be tested is obtained.
In specific application, the embodiment of the application can directly acquire the voice audio to be quality tested from the call center, and can also indirectly acquire the voice audio to be quality tested from the relay server, namely, the acquisition source of the voice audio to be quality tested is not limited by the embodiment of the application. In addition, the number of the voice audio to be quality-checked is not limited in the embodiment of the present application, for example, the number of the voice audio to be quality-checked may be 500 voice calls between the customer service staff and the customer.
Step S202, converting the voice audio to be tested into a voice text to be tested.
The voice audio to be quality tested comprises at least one voice audio segment, the voice text to be quality tested comprises at least one voice text segment, and the voice audio segments correspond to the voice text segments one to one.
Illustratively, the specific process of converting the voice audio to be quality-checked into the voice text to be quality-checked may be:
firstly, separating target voice audio in the voice audio to be detected.
It can be understood that, in the embodiments of the present application, the target speech audio is separated by using the difference between the energy of the background noise and the energy of the target speech audio.
Specifically, the target voice audio in the voice audio to be quality tested is separated according to the following 4 steps:
1. and framing the voice audio to be quality tested to obtain an audio frame.
2. The energy of the audio frame is calculated according to the following formula:
wherein E isnIs the energy of an audio frame, N is the time instant, x is a frame sample value, m is the average sound amplitude, and N is the window length.
3. And screening out audio frames with energy larger than an energy threshold value.
4. And forming the target voice audio according to the audio frames with the energy larger than the energy threshold value.
And secondly, performing role object segmentation on the target voice audio to obtain voice audio segments.
Specifically, the character object segmentation is performed on the target voice audio according to the following 4 steps to obtain a voice audio segment:
1. and determining all role objects corresponding to the target voice audio.
For example, the role object of the embodiment of the present application may be a customer service person and a client in the insurance industry.
2. And searching a preset voice characteristic model corresponding to each role object.
The preset voice characteristic model is set in advance according to the voice characteristics of the role object.
For example, the preset speech feature model of the customer service personnel can be obtained by extracting speech feature values of the customer service personnel according to a Mel frequency cepstrum system (MFCC) and inputting the speech feature values into a speech feature model, such as a Gaussian mixture model for training. Correspondingly, the preset speech feature model of the customer can be obtained by extracting speech feature values of customer personnel according to Mel Frequency Cepstrum Coefficient (MFCC) and inputting the speech feature values into a speech feature model, such as a Gaussian mixture model for training
3. And substituting the preset voice characteristic model corresponding to each role object into a preset function to calculate the jump prediction value.
For example, the predetermined function may be a likelihood function plus a penalty term.
4. And taking the moment of the jump prediction value larger than the jump prediction threshold value as a jump point, and segmenting the target voice audio according to the jump point to obtain a voice audio segment.
It can be understood that, in the embodiment of the present application, by predicting the transition point of the speech audio segment, the speech audio segments corresponding to different character objects in the speech audio are segmented.
And thirdly, converting the voice audio fragment into a voice text fragment, and forming a voice text to be quality-checked according to the voice text fragment.
Specifically, the following 4 steps are performed to convert the voice audio segment into a voice text segment, and form a voice text to be quality-checked according to the voice text segment:
1. and extracting the characteristic value of the voice audio segment.
2. And inputting the characteristic value into a preset acoustic model to obtain a voice characteristic vector sequence.
The preset acoustic model is obtained by training according to acoustic data and a voice feature vector sequence in advance.
3. And inputting the voice feature vector sequence into a preset voice model to obtain a character sequence.
4. And forming a voice text to be processed according to the character sequence.
And S102, inputting the voice text to be quality-tested into a preset quality testing model, and determining a target quality testing voice text matched with a quality testing item in the quality testing model.
Wherein the quality inspection model comprises at least one quality inspection item.
It should be noted that the preset quality inspection model in the embodiment of the present application refers to a set of multiple quality inspection items, and each quality inspection item can perform quality inspection on a target quality inspection speech text.
How to obtain the predetermined quality inspection model is described below.
By way of example and not limitation, referring to fig. 3, another schematic flow chart before step S101 in fig. 1 of the voice quality inspection method provided in the embodiment of the present application is shown, before acquiring the voice text to be quality inspected, the method further includes:
and S301, acquiring a quality inspection item.
The quality inspection item comprises a violation link, violation content and a violation type, the violation link comprises but is not limited to a product introduction link, an information checking link, a health notification link, a duty-free statement link, an opening white link, a warranty confirmation link or a hesitation link, and the violation type comprises but is not limited to existence of forbidden words, incomplete expression and/or expression of wrong words and operation loss. For example, the violation link is a product introduction link-60 thousands of common accidents, the violation content is 'accidents cause accidents or disabilities, the highest compensation is fifty thousands of', and the violation type is misstatement. Generally, the words of moods common in point violation content, such as "o", "kay", etc., are removed.
It should be noted that the quality control item in the embodiment of the present application is obtained by manually and directly labeling according to the text of the voice sample, and in a specific application, the acquisition source of the quality control item may be locally stored or may be acquired from a server.
And S302, obtaining a preset quality inspection model according to the quality inspection item.
The preset quality inspection model in the embodiment of the present application refers to a quality inspection set formed by a plurality of quality inspection items.
Specifically, referring to fig. 4, a specific flowchart of the voice quality inspection method provided in the embodiment of the present application in step S302 in fig. 3 is shown, where obtaining the preset quality inspection model according to the quality inspection item includes:
and S401, searching illegal links and quality inspection items with the same illegal types.
And S402, carrying out clustering analysis on violation links and violation contents of quality inspection items with the same violation types to obtain a violation content set.
Specifically, the association degree between the illegal contents is searched, and the illegal content set is formed according to the illegal contents with the association degree larger than the association degree threshold.
For example, the following steps are carried out: the content with violation a is "accident-caused cause of death or disability, the highest indemnity is fifty thousand", the content with violation B is "accident-caused cause of death or disability, the lowest indemnity is forty thousand", the content with violation C is "accident-caused cause of death or disability, the highest indemnity is eighty thousand", the content with violation D is "accident-caused death, the highest indemnity is sixty thousand", wherein each character is used as a basic element of association, and the threshold of association degree is 12. In this way, the association degrees between the illegal content a and the illegal content B and the illegal content C are both 15, and the association degree between the illegal content a and the illegal content D is 10, and it can be seen that the association degrees between the illegal content a and the illegal content B and the illegal content C are greater than the association degree threshold, and the association degree between the illegal content a and the illegal content D is less than the association degree threshold, then the illegal content a, the illegal content B, and the illegal content C are taken as the illegal content set 1, and the illegal content D is taken as the illegal content set 2.
And S403, forming a preset quality inspection model according to the violation content set.
Exemplarily, the illegal content set is converted into a regular expression corresponding to the quality inspection item, and a preset quality inspection model is formed according to the regular expression.
For example, the process of converting the violation content set 1 into a regular expression is: "Accident causes a physical or disability, highest indemnity for five hundred thousand", "Accident causes a physical or disability, lowest indemnity for four hundred thousand", "Accident causes death, highest indemnity for six hundred thousand" is converted into "accident causes a physical or disability, lowest [ high and low ] indemnity for [ forty-five-six ] hundred thousand".
Preferably, because the generalization capability of the regular expression corresponding to the quality inspection item is weak, only the contents appearing in the illegal denier set can be covered, and the generalization capability of the regular expression corresponding to the quality inspection item needs to be improved, that is, the generalization capability of the preset quality inspection model is improved.
For example, the regular expression "accident caused accident or disability, [ high-low ] compensation [ forty-five-six ] hundred thousand" can only be matched against the quality control speech text content "accident caused accident or disability, [ high-low ] compensation [ forty-five-six ] hundred thousand", "accident caused accident or disability, [ low-low compensation forty-hundred thousand" or "accident caused death, high-high compensation sixty-ten thousand" but cannot be matched against "accident caused accident or disability, [ high-low compensation eighty-ten thousand".
In a possible mode, searching a regular expression matched with the illegal content set according to a preset matching rule, and after the regular expression is used as a preset quality inspection model of a quality inspection item, the method further includes:
firstly, identifying generalized characters of a regular expression in a preset quality inspection model.
Wherein the generalized characters are [ ] "
And secondly, replacing generalized characters of the regular expression with fuzzy characters.
Wherein the ambiguous characters may be ". about..
For example, the regular expression "accident caused a cause of death or disability, [ four five six ] one hundred thousand" is replaced with "accident caused a cause of death or disability, [ one hundred thousand ].
Therefore, the generalization capability of the preset quality inspection model can be expanded by replacing the generalization characters of the regular expression with the fuzzy characters.
And S103, performing quality inspection on the target quality inspection voice text according to the quality inspection item to obtain a quality inspection result.
Wherein the quality inspection result comprises pass or fail.
Specifically, referring to fig. 5, which is a detailed flowchart of step S103 in fig. 1 of the voice quality inspection method provided in the embodiment of the present application, performing quality inspection on the target quality inspection voice text according to the quality inspection item to obtain a quality inspection result, including:
and step S501, identifying the violation type of the target quality inspection voice text according to the quality inspection item.
The violation types include, but are not limited to, the presence of forbidden words, incomplete expressions, misexpressions, and lack of speech. The existence of forbidden words means that unfortunate words are obvious and disconcerting words of customer service personnel, incomplete expression means that words expressed by the customer service personnel are absent, language logic is not met, wrong expression means that the words expressed by the customer service personnel conflict with basic speech technology, and speech technology deficiency means that the words expressed by the customer service personnel are obviously absent compared with the basic speech technology.
And step S502, generating violation scores of the target quality inspection voice text according to preset weights corresponding to violation types, and obtaining a quality inspection result according to a comparison result of the violation scores and a preset score threshold.
In this embodiment of the application, a corresponding preset weight is preset for each violation type, for example, a preset weight corresponding to the existence of a forbidden word is 5, a preset weight corresponding to the expression of incomplete is 2, a preset weight corresponding to the expression of error is 2, and a preset weight corresponding to the absence of dialect is 2. Generally, the weight value set for the existence of the forbidden word in the embodiment of the application is greater than the weight value of other violation types.
It is understood that there may be multiple violation types per target quality control phonetic text, with matching relationships to multiple quality control items.
Specifically, in the first step, the violation score of the target quality inspection voice is calculated according to the following formula:
wherein W is the violation score of the target quality inspection voice text, n is the number of all violation types corresponding to the target quality inspection voice text, and A1A violation type corresponding to the target quality inspection voice text, B1A preset weight for the violation type, b1And the coefficient is corresponding to the historical occurrence frequency of the violation type.
And secondly, comparing the obtained violation score with a preset score threshold, wherein if the violation score is larger than the preset score threshold, the quality inspection result is failed, and if the violation score is larger than the preset score threshold, the quality inspection result is passed.
It can be understood that in the process of calculating the violation score of the target quality inspection voice text, the historical occurrence frequency of the violation type corresponding to the target quality inspection voice text is also considered, and the accuracy of calculating the violation score of the target quality inspection voice text is improved.
In the embodiment of the application, the quality control voice text to be detected obtained by converting the voice audio to be detected can be directly subjected to automatic quality control through the preset quality control model, the quality control of the voice audio to be detected does not need to be manually performed, and the effects of improving the accuracy rate of quality control results and the quality control efficiency are achieved.
It should be understood that, the sequence numbers of the steps in the foregoing embodiments do not imply an execution sequence, and the execution sequence of each process should be determined by its function and inherent logic, and should not constitute any limitation to the implementation process of the embodiments of the present application.
Fig. 6 is a block diagram of a voice quality inspection apparatus according to an embodiment of the present application, which corresponds to the voice quality inspection method according to the foregoing embodiment, and only the relevant portions of the voice quality inspection apparatus according to the embodiment of the present application are shown for convenience of description.
Referring to fig. 6, the apparatus includes:
the acquisition module 61 is used for acquiring a voice text to be subjected to quality inspection;
the determining module 62 is configured to input the voice text to be quality tested into a preset quality testing model, and determine a target quality testing voice text matched with quality testing items in the quality testing model, where each quality testing item corresponds to a quality testing type;
and the quality inspection module 63 is used for performing quality inspection on the target quality inspection voice text according to the quality inspection item to obtain a quality inspection result.
In a possible implementation manner, the voice quality inspection module further includes:
the audio acquisition module is used for acquiring the voice audio to be subjected to quality inspection;
and the conversion module is used for converting the voice audio to be subjected to quality inspection into a voice text to be subjected to quality inspection.
In a possible implementation manner, the voice quality inspection module further includes:
the quality inspection item acquisition module is used for acquiring quality inspection items, wherein the quality inspection items comprise violation links, violation contents and violation types;
and the quality inspection model construction module is used for obtaining the preset quality inspection model according to the quality inspection item.
In one possible implementation manner, the quality inspection model building module includes:
the searching unit is used for searching illegal links and quality inspection items with the same illegal types;
the clustering unit is used for clustering and analyzing the violation contents of the quality inspection items with the same violation links and violation types to obtain a violation content set;
and the construction unit is used for forming the preset quality inspection model according to the violation content set.
In one possible implementation, the building unit includes:
a conversion subunit, configured to convert the illegal content into a regular expression corresponding to the quality inspection item, and form a preset quality inspection model according to the regular expression
In a possible implementation manner, the building unit further includes:
the identification subunit is used for identifying generalized characters of the regular expression in the preset quality inspection model;
a replacement subunit for replacing the generalized characters of the regular expression with the fuzzy characters
In one possible implementation, the quality inspection module includes:
the type identification unit is used for identifying the violation type of the target quality inspection voice text according to the quality inspection item;
and the generating unit is used for generating violation scores of the target quality inspection voice text according to the preset weight corresponding to the violation types and obtaining a quality inspection result according to a comparison result of the violation scores and a preset score threshold.
It should be noted that, for the information interaction, execution process, and other contents between the above-mentioned devices/units, the specific functions and technical effects thereof are based on the same concept as those of the embodiment of the method of the present application, and specific reference may be made to the part of the embodiment of the method, which is not described herein again.
Fig. 7 is a schematic structural diagram of a quality inspection apparatus according to an embodiment of the present application. As shown in fig. 7, the quality inspection apparatus 7 of this embodiment includes: at least one processor 70, a memory 71, and a computer program 72 stored in the memory 71 and executable on the at least one processor 70, the processor 70 implementing the steps of the above-described embodiments when executing the computer program 72.
The quality inspection device 7 includes a terminal device or a server.
The Processor 70 may be a Central Processing Unit (CPU), and the Processor 70 may be other general purpose Processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), an off-the-shelf Programmable Gate Array (FPGA) or other Programmable logic device, discrete Gate or transistor logic, discrete hardware components, etc. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
The storage 71 may in some embodiments be an internal storage unit of the quality inspection device 7, such as a hard disk or a memory of the quality inspection device 7.
It will be apparent to those skilled in the art that, for convenience and brevity of description, only the above-mentioned division of the functional units and modules is illustrated, and in practical applications, the above-mentioned function distribution may be performed by different functional units and modules according to needs, that is, the internal structure of the apparatus is divided into different functional units or modules to perform all or part of the above-mentioned functions. Each functional unit and module in the embodiments may be integrated in one processing unit, or each unit may exist alone physically, or two or more units are integrated in one unit, and the integrated unit may be implemented in a form of hardware, or in a form of software functional unit. In addition, specific names of the functional units and modules are only for convenience of distinguishing from each other, and are not used for limiting the protection scope of the present application. The specific working processes of the units and modules in the system may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.
The embodiment of the present application further provides a readable storage medium, where a computer program is stored, and when the computer program is executed by a processor, the computer program can implement the steps in the above method embodiments.
In the above embodiments, the descriptions of the respective embodiments have respective emphasis, and reference may be made to the related descriptions of other embodiments for parts that are not described or illustrated in a certain embodiment.
Those of ordinary skill in the art will appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.
The above-mentioned embodiments are only used for illustrating the technical solutions of the present application, and not for limiting the same; although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; such modifications and substitutions do not substantially depart from the spirit and scope of the embodiments of the present application and are intended to be included within the scope of the present application.
Claims (10)
1. A voice quality inspection method is characterized by comprising the following steps:
acquiring a voice text to be quality tested;
inputting the voice text to be quality-tested into a preset quality testing model, and determining a target quality testing voice text matched with a quality testing item in the quality testing model, wherein the quality testing model comprises at least one quality testing item;
and performing quality inspection on the target quality inspection voice text according to the quality inspection item to obtain a quality inspection result.
2. The voice quality inspection method according to claim 1, wherein before the obtaining of the voice text to be quality inspected, the method further comprises:
acquiring voice audio to be quality tested;
and converting the voice audio to be tested into a voice text to be tested.
3. The voice quality inspection method according to claim 1, wherein before the obtaining of the voice text to be inspected, the method further comprises:
obtaining quality inspection items, wherein the quality inspection items comprise violation links, violation contents and violation types;
and obtaining the preset quality inspection model according to the quality inspection item.
4. The voice quality inspection method according to claim 3, wherein obtaining the predetermined quality inspection model according to the quality inspection item comprises:
searching violation links and quality inspection items with the same violation types;
carrying out clustering analysis on violation contents of quality inspection items with the same violation links and violation types to obtain a violation content set;
and forming the preset quality inspection model according to the violation content set.
5. The voice quality inspection method according to claim 4, wherein forming a preset quality inspection model corresponding to the quality inspection item according to the illegal content set includes:
and converting the violation content into a regular expression corresponding to the quality inspection item, and forming a preset quality inspection model according to the regular expression.
6. The voice quality inspection method according to claim 5, wherein after searching for a regular expression matching the illegal content set according to a preset matching rule and using the regular expression as a preset quality inspection model of the quality inspection item, the method further comprises:
identifying generalized characters of the regular expression in the preset quality inspection model;
and replacing generalized characters of the regular expression with fuzzy characters.
7. The voice quality inspection method according to any one of claims 1 to 6, wherein performing quality inspection on the target quality inspection voice text according to the quality inspection item to obtain a quality inspection result comprises:
identifying the violation type of the target quality inspection voice text according to the quality inspection item;
and generating violation scores of the target quality inspection voice text according to the preset weight corresponding to the violation types, and obtaining a quality inspection result according to a comparison result of the violation scores and a preset score threshold.
8. A voice quality inspection apparatus, comprising:
the acquisition module is used for acquiring a voice text to be subjected to quality inspection;
the determining module is used for inputting the voice text to be quality tested into a preset quality testing model, determining a target quality testing voice text matched with quality testing items in the quality testing model, and enabling each quality testing item to correspond to one quality testing type;
and the quality inspection module is used for performing quality inspection on the target quality inspection voice text according to the quality inspection item to obtain a quality inspection result.
9. A quality inspection apparatus comprising a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor when executing the computer program implements the method of any one of claims 1 to 7.
10. A readable storage medium, storing a computer program, characterized in that the computer program, when executed by a processor, implements the method according to any of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010569722.4A CN111696528B (en) | 2020-06-20 | 2020-06-20 | Voice quality inspection method and device, quality inspection equipment and readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010569722.4A CN111696528B (en) | 2020-06-20 | 2020-06-20 | Voice quality inspection method and device, quality inspection equipment and readable storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111696528A true CN111696528A (en) | 2020-09-22 |
| CN111696528B CN111696528B (en) | 2021-04-23 |
Family
ID=72482571
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010569722.4A Active CN111696528B (en) | 2020-06-20 | 2020-06-20 | Voice quality inspection method and device, quality inspection equipment and readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111696528B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112541774A (en) * | 2020-12-08 | 2021-03-23 | 四川众信佳科技发展有限公司 | AI quality inspection method, device, system, electronic device and storage medium |
| CN112804400A (en) * | 2020-12-31 | 2021-05-14 | 中国工商银行股份有限公司 | Customer service call voice quality inspection method and device, electronic equipment and storage medium |
| CN113065879A (en) * | 2021-04-30 | 2021-07-02 | 中国工商银行股份有限公司 | Data stream quality inspection method and system |
| WO2022126969A1 (en) * | 2020-12-15 | 2022-06-23 | 平安科技（深圳）有限公司 | Service voice quality inspection method, apparatus and device, and storage medium |
Citations (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7395205B2 (en) * | 2001-02-13 | 2008-07-01 | International Business Machines Corporation | Dynamic language model mixtures with history-based buckets |
| CN101946501A (en) * | 2007-12-14 | 2011-01-12 | 高通股份有限公司 | Efficient diffusion dithering using dyadic rationals |
| CN105225290A (en) * | 2014-06-27 | 2016-01-06 | 福特全球技术公司 | Vehicle operating is monitored |
| CN106776806A (en) * | 2016-11-22 | 2017-05-31 | 广东电网有限责任公司佛山供电局 | The methods of marking and system of call center's quality inspection voice |
| US20180350389A1 (en) * | 2017-06-01 | 2018-12-06 | Microsoft Technology Licensing, Llc | Computer-Generated Feedback Of User Speech Meeting Subjective Criteria |
| CN108962282A (en) * | 2018-06-19 | 2018-12-07 | 京北方信息技术股份有限公司 | Speech detection analysis method, apparatus, computer equipment and storage medium |
| CN109446524A (en) * | 2018-10-25 | 2019-03-08 | 第四范式（北京）技术有限公司 | A kind of voice quality detecting method and device |
| CN109740155A (en) * | 2018-12-27 | 2019-05-10 | 广州云趣信息科技有限公司 | A kind of customer service system artificial intelligence quality inspection rule self concludes the method and system of model |
| CN109785606A (en) * | 2019-02-02 | 2019-05-21 | 中能瑞通（北京）科技有限公司 | A kind of power information acquisition system acquisition quality analysis method and device |
| CN109819120A (en) * | 2017-11-20 | 2019-05-28 | 北京智合大方科技有限公司 | One kind carrying out call center's intelligence quality inspection system based on natural language processing technique |
| CN110036384A (en) * | 2017-09-29 | 2019-07-19 | 索尼公司 | Information processing equipment and information processing method |
| CN110197672A (en) * | 2018-02-27 | 2019-09-03 | 招商信诺人寿保险有限公司 | A kind of voice call quality detection method, server, storage medium |
| CN110334241A (en) * | 2019-07-10 | 2019-10-15 | 深圳前海微众银行股份有限公司 | Quality detecting method, device, equipment and the computer readable storage medium of customer service recording |
| CN110600056A (en) * | 2019-08-09 | 2019-12-20 | 深圳市云之音科技有限公司 | Voice quality inspection method and device |
| CN110688401A (en) * | 2019-09-10 | 2020-01-14 | 中国平安财产保险股份有限公司 | Dynamic cache processing method and device, storage medium and electronic equipment |
| CN111049998A (en) * | 2018-10-11 | 2020-04-21 | 上海智臻智能网络科技股份有限公司 | Voice customer service quality inspection method, customer service quality inspection equipment and storage medium |
| CN111294468A (en) * | 2020-02-07 | 2020-06-16 | 普强时代(珠海横琴)信息技术有限公司 | Tone quality detection and analysis system for customer service center calling |
-
2020
- 2020-06-20 CN CN202010569722.4A patent/CN111696528B/en active Active
Patent Citations (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7395205B2 (en) * | 2001-02-13 | 2008-07-01 | International Business Machines Corporation | Dynamic language model mixtures with history-based buckets |
| CN101946501A (en) * | 2007-12-14 | 2011-01-12 | 高通股份有限公司 | Efficient diffusion dithering using dyadic rationals |
| CN105225290A (en) * | 2014-06-27 | 2016-01-06 | 福特全球技术公司 | Vehicle operating is monitored |
| CN106776806A (en) * | 2016-11-22 | 2017-05-31 | 广东电网有限责任公司佛山供电局 | The methods of marking and system of call center's quality inspection voice |
| US20180350389A1 (en) * | 2017-06-01 | 2018-12-06 | Microsoft Technology Licensing, Llc | Computer-Generated Feedback Of User Speech Meeting Subjective Criteria |
| CN110036384A (en) * | 2017-09-29 | 2019-07-19 | 索尼公司 | Information processing equipment and information processing method |
| CN109819120A (en) * | 2017-11-20 | 2019-05-28 | 北京智合大方科技有限公司 | One kind carrying out call center's intelligence quality inspection system based on natural language processing technique |
| CN110197672A (en) * | 2018-02-27 | 2019-09-03 | 招商信诺人寿保险有限公司 | A kind of voice call quality detection method, server, storage medium |
| CN108962282A (en) * | 2018-06-19 | 2018-12-07 | 京北方信息技术股份有限公司 | Speech detection analysis method, apparatus, computer equipment and storage medium |
| CN111049998A (en) * | 2018-10-11 | 2020-04-21 | 上海智臻智能网络科技股份有限公司 | Voice customer service quality inspection method, customer service quality inspection equipment and storage medium |
| CN109446524A (en) * | 2018-10-25 | 2019-03-08 | 第四范式（北京）技术有限公司 | A kind of voice quality detecting method and device |
| CN109740155A (en) * | 2018-12-27 | 2019-05-10 | 广州云趣信息科技有限公司 | A kind of customer service system artificial intelligence quality inspection rule self concludes the method and system of model |
| CN109785606A (en) * | 2019-02-02 | 2019-05-21 | 中能瑞通（北京）科技有限公司 | A kind of power information acquisition system acquisition quality analysis method and device |
| CN110334241A (en) * | 2019-07-10 | 2019-10-15 | 深圳前海微众银行股份有限公司 | Quality detecting method, device, equipment and the computer readable storage medium of customer service recording |
| CN110600056A (en) * | 2019-08-09 | 2019-12-20 | 深圳市云之音科技有限公司 | Voice quality inspection method and device |
| CN110688401A (en) * | 2019-09-10 | 2020-01-14 | 中国平安财产保险股份有限公司 | Dynamic cache processing method and device, storage medium and electronic equipment |
| CN111294468A (en) * | 2020-02-07 | 2020-06-16 | 普强时代(珠海横琴)信息技术有限公司 | Tone quality detection and analysis system for customer service center calling |
Non-Patent Citations (4)
| Title |
|---|
| P. PAGLIERANI 等: ""Uncertainty Evaluation of Objective Speech Quality Measurement in VoIP Systems"", 《IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT》 * |
| 徐东群: "《空气污染对人群健康影响数据清洗及评价方法》", 31 December 2016 * |
| 贵州省贵阳市烤烟种植区划项目组: "《贵州省贵阳市烤烟种植区划》", 31 October 2015 * |
| 贾智慧: ""客服中心语音质检***设计与实现"", 《中国优秀硕士学位论文全文数据库(信息科技辑)》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112541774A (en) * | 2020-12-08 | 2021-03-23 | 四川众信佳科技发展有限公司 | AI quality inspection method, device, system, electronic device and storage medium |
| WO2022126969A1 (en) * | 2020-12-15 | 2022-06-23 | 平安科技（深圳）有限公司 | Service voice quality inspection method, apparatus and device, and storage medium |
| CN112804400A (en) * | 2020-12-31 | 2021-05-14 | 中国工商银行股份有限公司 | Customer service call voice quality inspection method and device, electronic equipment and storage medium |
| CN113065879A (en) * | 2021-04-30 | 2021-07-02 | 中国工商银行股份有限公司 | Data stream quality inspection method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111696528B (en) | 2021-04-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111696528B (en) | Voice quality inspection method and device, quality inspection equipment and readable storage medium | |
| CN112804400B (en) | Customer service call voice quality inspection method and device, electronic equipment and storage medium | |
| CN107093431B (en) | Method and device for quality inspection of service quality | |
| WO2020228173A1 (en) | Illegal speech detection method, apparatus and device and computer-readable storage medium | |
| CN110720124B (en) | Monitoring the use of patient language to identify potential speech and related neurological disorders | |
| US7805300B2 (en) | Apparatus and method for analysis of language model changes | |
| CN113094578B (en) | Deep learning-based content recommendation method, device, equipment and storage medium | |
| CN112468659B (en) | Quality evaluation method, device, equipment and storage medium applied to telephone customer service | |
| KR102019207B1 (en) | Apparatus and method for assessing data quality for text analysis | |
| US6763331B2 (en) | Sentence recognition apparatus, sentence recognition method, program, and medium | |
| CN111611218A (en) | Distributed abnormal log automatic identification method based on deep learning | |
| CN112966082A (en) | Audio quality inspection method, device, equipment and storage medium | |
| CN112671985A (en) | Agent quality inspection method, device, equipment and storage medium based on deep learning | |
| CN114816962B (en) | ATTENTION-LSTM-based network fault prediction method | |
| CN114997169A (en) | Entity word recognition method and device, electronic equipment and readable storage medium | |
| CN114049895B (en) | ASR-based voice quality inspection analysis method and system | |
| CN112966509B (en) | Text quality evaluation method and device, storage medium and computer equipment | |
| CN111723204B (en) | Method and device for correcting voice quality inspection area, correction equipment and storage medium | |
| US11817086B2 (en) | Machine learning used to detect alignment and misalignment in conversation | |
| CN111696527B (en) | Method and device for positioning voice quality inspection area, positioning equipment and storage medium | |
| JP2017211586A (en) | Psychological analyzer, psychological analysis method and program | |
| US11810042B1 (en) | Disclosure quality assurance | |
| Grothendieck | Tracking changes in language | |
| Samaranayaka et al. | Enhancing Call Center Operations Through Semantic Voice Analysis | |
| CN114822517A (en) | Voice annotation quality evaluation method and device, electronic equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CP02 | Change in the address of a patent holder | ||
| CP02 | Change in the address of a patent holder |
Address after: 519031 office 1316, No. 1, lianao Road, Hengqin new area, Zhuhai, Guangdong Patentee after: LONGMA ZHIXIN (ZHUHAI HENGQIN) TECHNOLOGY Co.,Ltd. Address before: Room 417.418.419, building 20, creative Valley, 1889 Huandao East Road, Hengqin New District, Zhuhai City, Guangdong Province Patentee before: LONGMA ZHIXIN (ZHUHAI HENGQIN) TECHNOLOGY Co.,Ltd. |