Invention content
The purpose of the present invention is to provide a kind of method, apparatus, equipment and the computer-readable storages of evaluation customer serviceMedium improves the authenticity and coverage area of evaluation data to realize job evaluation of the reflection to contact staff of objective reality.
To achieve the above object, an embodiment of the present invention provides following technical solutions：
A method of evaluation customer service, including：
Obtain the voice data of customer service and client's call；
Voice segmentation is carried out to the voice data, obtains the corresponding customer service voices segment of the customer service；
The customer service voices characteristic in the customer service voices segment is extracted, by the customer service voices characteristic and is presetObject voice feature data in database is matched, and determination is maximum right with the customer service voices characteristic matching degreeAs voice feature data；
Will be with the maximum object voice feature data of the customer service voices characteristic matching degree, corresponding default scoring is reallyIt is set to the evaluation of the customer service.
Wherein, described to obtain customer service with before the voice data of client's call, further include：
Real-time recording is carried out to the communication process of the customer service and the client, obtains corresponding audio file；
The audio file is detected into line activating, distinguishes the quiet data and voice data in the audio file.
Wherein, it is described will with the maximum object voice feature data of the customer service voices characteristic matching degree, it is correspondingDefault scoring is determined as after the evaluation of the customer service, further includes：
It stores the scoring is corresponding with the customer service to preset evaluation table, and the evaluation table is subjected to visualization exhibitionShow.
Wherein, further include：
Voice segmentation is carried out to the voice data, obtains the corresponding customer voice segment of the client；
Extract the customer voice characteristic in the customer voice segment, by the customer voice characteristic with it is describedObject voice feature data in database is matched, and determination is maximum right with the customer voice characteristic matching degreeAs voice feature data；
It will be with the maximum object voice feature data of the customer voice characteristic matching degree, corresponding default satisfactionIt is determined as the satisfaction of the client.
Wherein, further include：
According to the second value shared by customer service evaluation shared the first weights and customer satisfaction, described in comprehensive assessmentThe service quality of customer service.
A kind of device of evaluation customer service, including：
Acquisition module, the voice data for obtaining customer service and client's call；
Voice divides module, for carrying out voice segmentation to the voice data, obtains the corresponding customer service language of the customer serviceTablet section；
Matching module, for extracting the customer service voices characteristic in the customer service voices segment, by the customer service voicesCharacteristic is matched with the object voice feature data in presetting database, and is determined and the customer service voices characteristicThe maximum object voice feature data of matching degree；
Evaluation module, for will with the maximum object voice feature data of the customer service voices characteristic matching degree, it is rightThe default scoring answered is determined as the evaluation of the customer service.
Wherein, further include：
Recording module carries out real-time recording for the communication process to the customer service and the client, obtains corresponding soundFrequency file；
Discriminating module distinguishes the quiet data in the audio file for being detected into line activating to the audio fileAnd voice data.
Wherein, further include：
Display module, for storing the scoring is corresponding with the customer service to preset evaluation table, and by the evaluationTable is visualized.
A kind of equipment of evaluation customer service, including：
Memory, for storing computer program；
Processor realizes the side of the evaluation customer service described in above-mentioned any one when for executing the computer programThe step of method.
A kind of computer readable storage medium is stored with computer program on the computer readable storage medium, describedThe step of method of the evaluation customer service described in above-mentioned any one is realized when computer program is executed by processor.
By above scheme it is found that it is provided in an embodiment of the present invention it is a kind of evaluation customer service method, including：Obtain visitorThe voice data of clothes and client's call；Voice segmentation is carried out to the voice data, obtains the corresponding customer service voices of the customer serviceSegment；The customer service voices characteristic in the customer service voices segment is extracted, by the customer service voices characteristic and present countIt is matched, and is determined and the maximum object of customer service voices characteristic matching degree according to the object voice feature data in libraryVoice feature data；Will with the maximum object voice feature data of the customer service voices characteristic matching degree, it is corresponding defaultScoring is determined as the evaluation of the customer service.
As it can be seen that the method is analyzed by the voice data generated with client's call to customer service, customer service people is extractedSound bite of the member in service process by carrying out characteristic matching to the sound bite, and then determines the clothes of contact staffBusiness scoring, to realize automatic Evaluation customer service, both facilitates client, also simplifies the operation stream of evaluation customer serviceJourney；Also, it is evaluated according to the voice data that customer service is generated with client's call, is capable of the reflection of objective reality to contact staffJob evaluation, will not have leakage the case where commenting, to expand the coverage area of evaluation data, work people for assessment customer serviceThe service quality of member provides authentic and valid foundation, is also beneficial to enterprise and carries out corresponding behave improvement according to evaluation data.
Correspondingly, device, equipment and the computer-readable storage of a kind of evaluation customer service provided in an embodiment of the present inventionMedium similarly has above-mentioned technique effect.
Specific implementation mode
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, completeSite preparation describes, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based onEmbodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every otherEmbodiment shall fall within the protection scope of the present invention.
The embodiment of the invention discloses a kind of method, apparatus, equipment and the computer-readable storage mediums of evaluation customer serviceMatter improves the authenticity and coverage area of evaluation data to realize job evaluation of the reflection to contact staff of objective reality.
Referring to Fig. 1, a kind of method of evaluation customer service provided in an embodiment of the present invention, including：
S101, the voice data for obtaining customer service and client's call；
Specifically, during customer service and client carry out voice communication, customer service and client's call can be obtained in real timeVoice data can also carry out real-time recording to the voice communication of customer service and client, after customer service and client's end of conversation, obtainVoice data in recording file.
S102, voice segmentation is carried out to voice data, obtains the corresponding customer service voices segment of customer service；
It should be noted that the voice segmentation is under conditions of no priori, i.e., in speaker's quantity and identityIn the case of all unknown, the speaker for being automatically found one section of words human speech sound of speaking more changes point, and classifies to these voice segmentsMark, makes identical personal data of speaking be noted as same class.Can so tell in one section of recording has several personages to existDialogue, which voice is which personage says.So can be corresponding by the customer service in voice data according to voice cutting techniquesCustomer service voices segment is discernable.
Customer service voices characteristic in S103, extraction customer service voices segment, by customer service voices characteristic and present countIt is matched, and is determined and the maximum object voice of customer service voices characteristic matching degree according to the object voice feature data in libraryCharacteristic；
Specifically, the voice feature data generally comprises：Sound spectrum data, prosodic features data and sound quality featureData.Object voice feature data in the database is the corresponding voice feature data of different moods pre-saved, exampleSuch as the corresponding sound spectrum data of angry mood, prosodic features data and sound quality characteristic, the corresponding sound spectrum of tranquil moodCharacteristic, prosodic features data and sound quality characteristic, the corresponding sound spectrum data of happy mood, prosodic features numberAccording to sound quality characteristic etc..Therefore it may act as with the maximum object voice feature data of customer service voices characteristic matching degreeEmotional reactions of customer service during communicating with client.
S104, will be with the maximum object voice feature data of customer service voices characteristic matching degree, corresponding default scoringIt is determined as the evaluation of customer service.
It should be noted that the corresponding object voice feature data of each mood correspondence in the database is preset withCorresponding scoring, thus will be determined as with the maximum object voice feature data of customer service voices characteristic matching degree customer service with visitorAfter emotional reactions in the communication process of family, corresponding default score can be determined as the evaluation of customer service.
Such as：Irascible mood, angry mood, exciting mood, tranquil mood, happy feelings are previously stored in the databaseThe corresponding irascible voice feature data of thread, excited emoticon, angry voice feature data, excited speech characteristic, calmnessVoice feature data, happy voice feature data, excited voice feature data, wherein irascible voice feature data, angry voiceCharacteristic, excited speech characteristic, tranquil voice feature data, happy voice feature data, excited voice feature dataIt is referred to as object voice feature data.
Wherein, each object voice feature data is corresponding with different default scorings respectively.Such as：Irascible phonetic feature numberIt it is 0 point according to corresponding default scoring, the corresponding default scoring of angry voice feature data is 2 points, excited speech characteristic pairThe default scoring answered is 4 points, the corresponding default scoring of tranquil voice feature data is 6 points, happy voice feature data is correspondingDefault scoring is 8 points, the corresponding default scoring of excited voice feature data is 10 points.Therefore when the emotional reactions for determining customer serviceIt afterwards, can be by the corresponding default service evaluation to score as the customer service of its emotional reactions.It is of course also possible to be each objectVoice feature data is corresponding respectively to preset different code of points, with the service quality of more objective appraisal customer service staff.
It can be seen that the method for evaluation customer service provided in this embodiment, passes through the voice number for generation of conversing customer service and clientAccording to being analyzed, sound bite of the contact staff in service process is extracted, by carrying out characteristic matching to the sound bite,And then determine the service scoring of contact staff, to realize automatic Evaluation customer service, client is both facilitated, is also simplifiedEvaluate the operating process of customer service；Also, it is evaluated according to the voice data that customer service is generated with client's call, it can be objectiveJob evaluation of the true reflection to contact staff, will not have the case where leakage is commented, to expand the covering of evaluation dataRange, the service quality to assess customer service staff provide authentic and valid foundation, are also beneficial to enterprise according to review numberAccording to the corresponding behave improvement of progress.
The embodiment of the invention discloses the methods of another evaluation customer service, relative to a upper embodiment, the present embodimentFurther instruction and optimization have been made to technical solution.
Referring to Fig. 2, the method for another evaluation customer service provided in an embodiment of the present invention, including：
S201, to the communication process real-time recording of customer service and client, obtain corresponding audio file；
S202, audio file is activated and is detected, distinguish the quiet data and voice data in audio file；
In the present embodiment, real-time recording can be carried out to the voice communication of customer service and client, and obtains corresponding audioFile since there are mute parts and phonological component in audio file, therefore detects audio file into line activating, can distinguishQuiet data in audio file and voice data, and then quiet data is deleted from the audio file, so as to more accurateGround obtains the voice data of customer service and client's call.
S203, the voice data for obtaining customer service and client's call；
S204, voice segmentation is carried out to voice data, obtains the corresponding customer service voices segment of customer service；
Customer service voices characteristic in S205, extraction customer service voices segment, by customer service voices characteristic and present countIt is matched, and is determined and the maximum object voice of customer service voices characteristic matching degree according to the object voice feature data in libraryCharacteristic；
S206, will be with the maximum object voice feature data of customer service voices characteristic matching degree, corresponding default scoringIt is determined as the evaluation of customer service.
It can be seen that the method for evaluation customer service provided in this embodiment, is carried out real by the communication process to customer service and clientShi Luyin obtains corresponding audio file；Audio file is detected into line activating, distinguish quiet data in audio file andVoice data, and then sound bite of the contact staff in service process is extracted, by carrying out feature to the sound biteMatch, and then determine the service scoring of contact staff, to realize automatic Evaluation customer service, both facilitates client, it is also simpleThe operating process of evaluation customer service is changed；Also, it is evaluated according to the voice data that customer service is generated with client's call, it canJob evaluation of the reflection of objective reality to contact staff will not have the case where leakage is commented, to expand evaluation dataCoverage area, the service quality to assess customer service staff provide authentic and valid foundation, are also beneficial to enterprise according to commentingValence mumber is according to the corresponding behave improvement of progress.
Based on above-mentioned any embodiment, it should be noted that it is described will be with the customer service voices characteristic matching degree mostBig object voice feature data, corresponding default scoring are determined as after the evaluation of the customer service, further include：
It stores the scoring is corresponding with the customer service to preset evaluation table, and the evaluation table is subjected to visualization exhibitionShow.
Such as：It in the table by the storage corresponding with the scoring that it is obtained of the name of customer service staff, and can by presettingIt is shown depending on changing window, the work at present quality of oneself is clearly recognized convenient for customer service staff, also allows for administratorThe work of member summarised assessment difference customer service staff.
Based on above-mentioned any embodiment, it should be noted that can also be recorded to the satisfaction for the client that comes to visit, specificallyFor：
Voice segmentation is carried out to the voice data, obtains the corresponding customer voice segment of the client；
Extract the customer voice characteristic in the customer voice segment, by the customer voice characteristic with it is describedObject voice feature data in database is matched, and determination is maximum right with the customer voice characteristic matching degreeAs voice feature data；
It will be with the maximum object voice feature data of the customer voice characteristic matching degree, corresponding default satisfactionIt is determined as the satisfaction of the client.
Specifically, similar with the evaluation detailed process of service of customer service staff.
It should be noted that the corresponding object voice feature data of each mood correspondence in the database is preset withCorresponding satisfaction, thus will be determined as with the maximum object voice feature data of customer voice characteristic matching degree client withAfter emotional reactions in customer service work people's communication process, corresponding default satisfaction can be determined as to the client to this serviceSatisfaction.
Such as：Irascible mood, angry mood, exciting mood, tranquil mood, happy feelings are previously stored in the dataThe corresponding irascible voice feature data of thread, excited emoticon, angry voice feature data, excited speech characteristic, calmnessVoice feature data, happy voice feature data, excited voice feature data, wherein irascible voice feature data, angry voiceCharacteristic, excited speech characteristic, tranquil voice feature data, happy voice feature data, excited voice feature dataIt is referred to as object voice feature data.
Wherein, each object voice feature data is corresponding with different default satisfactions respectively.Such as：Irascible phonetic featureThe corresponding default satisfaction of data is 0, the corresponding default satisfaction of angry voice feature data is 20%, excited speech characteristicAccording to corresponding default satisfaction be 40%, the corresponding default satisfaction of tranquil voice feature data is 60%, happy phonetic featureThe corresponding default satisfaction of data is 80%, the corresponding default satisfaction of excited voice feature data is 100%.Therefore works as and determineAfter the emotional reactions of client, the satisfaction that the corresponding default satisfaction of its emotional reactions can service this as the clientDegree.It is of course also possible to for each object voice feature data, corresponding default different satisfaction is regular respectively, with more objective noteRecord satisfaction of the client to service.
Wherein, the second value shared by the first shared weights and customer satisfaction can also be evaluated according to customer service,The service quality of customer service described in comprehensive assessment.
Such as：Customer service can be evaluated and customer satisfaction is divided according to 5 to 5 weight, then primaryDuring session services, what customer service obtained is evaluated as 10 points, and corresponding customer satisfaction is 80%, and corresponding comprehensive assessment is divided into5.4 point.It is possible thereby to the service quality of customer service staff is carried out longitudinal comparison, for choose outstanding contact staff provide according toAccording to.
Based on above-mentioned any embodiment, it should be noted that it is described that voice segmentation is carried out to the voice data, including：
The voice data is split according to preset time span, obtains multiple sound bites；
Cluster operation is carried out to the multiple sound bite using probability distribution clustering method, it is corresponding to obtain the customer serviceCluster labels；
The cluster labels are arranged and merged sequentially in time, the corresponding customer service voices piece of the customer service is obtainedSection.
Wherein, described that cluster operation is carried out to the multiple sound bite using probability distribution clustering method, described in acquisitionThe corresponding cluster labels of customer service, including：
Any two sound bite is determined as sound bite group, the corresponding information of every group of sound bite group is calculated and loses letterNumerical value, and determine that the corresponding information of all sound bite groups loses the minimum value in functional value；
Judge whether the minimum value is more than preset threshold value；
If so, stopping clustering and being arranged and merged according to cluster result, the corresponding cluster mark of the customer service is obtainedLabel；
If it is not, then merging two sound bites in the corresponding sound bite group of the minimum value, and execute instituteIt states and any two sound bite is determined as sound bite group, calculate the corresponding information of every group of sound bite group and lose functional value,And determine the step of corresponding information of all sound bite groups loses the minimum value in functional value.
Based on above-mentioned any embodiment, it should be noted that the customer service voices characteristic in extraction customer service voices segment,Including：
The PCM data in the customer service voices segment is extracted, the customer service voices characteristic is extracted from the PCM dataAccording to.
It, can be described in each dimension extraction of the PCM data specifically, in order to improve the accuracy of subsequent match stepCustomer service voices characteristic, and the vector set of various dimensions is formed, in order to carry out the matching of characteristic.
It is wherein, described to match customer service voices characteristic with the object voice feature data in presetting database,And the determining and maximum object voice feature data of customer service voices characteristic matching degree, including：
Sound spectrum data, prosodic features data and the sound quality characteristic for including by the customer service audio characteristic dataSound spectrum data, prosodic features data and the sound quality characteristic for including with the multi-object audio characteristic respectively are one by oneIt is matched；
Sound spectrum data, prosodic features data and the sound quality characteristic for including according to the customer service audio characteristic dataAccording to the sound spectrum data, prosodic features data and sound quality characteristic for including with the multi-object audio characteristic respectivelySound spectrum data, prosodic features data and the sound quality characteristic that matching degree and the multi-object audio characteristic includeAccording to corresponding preset weights, determine and the maximum multi-object audio characteristic of the customer service audio characteristic data matching degree.
It should be noted that the calculating of matching degree can pass through weighted average, neural network model or clustering algorithm etc.Mode is calculated, and only one such embodiment is calculated by weighted average, therefore this specification is herein no longerIt repeats.
Specifically, the sound spectrum data include：MFCC features and GFCC features, the prosodic features data packetIt includes：Pitch features, short-time energy feature (Short Term Energy) feature, ZCR features and word speed (Speed) feature；InstituteStating sound quality characteristic includes：Formant (Formants) feature.
Wherein, MFCC is the abbreviation of Mel frequency cepstral coefficients, and Mel frequencies are extracted based on human hearing characteristic,, with Hz frequencies at nonlinear correspondence relation, Mel frequency cepstral coefficients (MFCC) are then to utilize this relationship between them, meter for itObtained Hz spectrum signatures；GFCC is characterized as the aural signature based on Gammatone filters.
Wherein, the fundamental frequency (fundamental frequency) of Pitch features and sound is related, reflects pitch information；ZCR (zero-crossing rate, zero-crossing rate) feature refers to the ratio of the sign change of a signal, such as signal is from positive numberBecome negative or reversed, is the main feature classify to tapping sound；
Wherein, formant (Formants) feature refers to some regions of the energy Relatively centralized in the frequency spectrum of sound, altogetherIt shakes the peak not still determinant of sound quality, and reflects the physical features of sound channel (resonant cavity).
Below to it is provided in an embodiment of the present invention it is a kind of evaluation customer service device be introduced, one kind described belowEvaluate customer service device with it is above-described it is a kind of evaluate customer service method can be cross-referenced.
Referring to Fig. 3, a kind of device of evaluation customer service provided in an embodiment of the present invention, including：
Acquisition module 301, the voice data for obtaining customer service and client's call；
Voice divides module 302, for carrying out voice segmentation to the voice data, obtains the corresponding customer service of the customer serviceSound bite；
Matching module 303, for extracting the customer service voices characteristic in the customer service voices segment, by the customer service languageSound characteristic is matched with the object voice feature data in presetting database, and is determined and the customer service voices characteristicAccording to the maximum object voice feature data of matching degree；
Evaluation module 304, for will with the maximum object voice feature data of the customer service voices characteristic matching degree,Corresponding default score is determined as the evaluation of the customer service.
Wherein, further include：
Recording module carries out real-time recording for the communication process to the customer service and the client, obtains corresponding soundFrequency file；
Discriminating module distinguishes the quiet data in the audio file for being detected into line activating to the audio fileAnd voice data.
Wherein, further include：
Display module, for storing the scoring is corresponding with the customer service to preset evaluation table, and by the evaluationTable is visualized.
Wherein, further include：
Customer voice segment acquisition module obtains the client and corresponds to for carrying out voice segmentation to the voice dataCustomer voice segment；
Customer voice characteristic matching module, for extracting the customer voice characteristic in the customer voice segmentAccording to, the customer voice characteristic is matched with the object voice feature data in the database, and determine and instituteState the maximum object voice feature data of customer voice characteristic matching degree；
Customer satisfaction determining module, being used for will be special with the maximum object voice of the customer voice characteristic matching degreeData are levied, corresponding default satisfaction is determined as the satisfaction of the client.
Wherein, further include：
Evaluation module takes for second shared by customer service evaluation shared the first weights and customer satisfactionIt is worth, the service quality of customer service described in comprehensive assessment.
Below to it is provided in an embodiment of the present invention it is a kind of evaluation customer service equipment be introduced, one kind described belowEvaluate customer service equipment with it is above-described it is a kind of evaluate customer service method and device can be cross-referenced.
Referring to Fig. 4, a kind of equipment of evaluation customer service provided in an embodiment of the present invention, including：
Memory 401, for storing computer program；
Processor 402 realizes the evaluation customer service clothes described in above-mentioned any embodiment when for executing the computer programThe step of method of business.
A kind of computer readable storage medium provided in an embodiment of the present invention is introduced below, one kind described belowComputer readable storage medium with it is above-described it is a kind of evaluation customer service method, apparatus and equipment can be cross-referenced.
A kind of computer readable storage medium is stored with computer program on the computer readable storage medium, describedThe step of method of the evaluation customer service as described in above-mentioned any embodiment is realized when computer program is executed by processor.
Each embodiment is described by the way of progressive in this specification, the highlights of each of the examples are with otherThe difference of embodiment, just to refer each other for identical similar portion between each embodiment.
The foregoing description of the disclosed embodiments enables those skilled in the art to implement or use the present invention.Various modifications to these embodiments will be apparent to those skilled in the art, as defined hereinGeneral Principle can be realized in other embodiments without departing from the spirit or scope of the present invention.Therefore, of the inventionIt is not intended to be limited to the embodiments shown herein, and is to fit to and the principles and novel features disclosed herein phase oneThe widest range caused.