CN101662550B - Method and system for service quality detection for call center - Google Patents
Method and system for service quality detection for call center Download PDFInfo
- Publication number
- CN101662550B CN101662550B CN200910170363A CN200910170363A CN101662550B CN 101662550 B CN101662550 B CN 101662550B CN 200910170363 A CN200910170363 A CN 200910170363A CN 200910170363 A CN200910170363 A CN 200910170363A CN 101662550 B CN101662550 B CN 101662550B
- Authority
- CN
- China
- Prior art keywords
- recording
- service
- user
- quality
- parameter
- 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.)
- Active
Links
Images
Abstract
The invention discloses a method and a system for service quality detection for a call center. The method comprises the following steps of: obtaining service sound recording of the call center; extracting the attribute information in the service sound recording; determining a parameter value used for quality detection according to the attribute information and the corresponding relationship between the attribute information and a parameter for quality detection; and determining the service quality assessment value of the telephone traffic service according to the parameter value. The method can objectively detect the service quality of the telephone traffic service each time and has high realizing efficiency and high universality.
Description
Technical field
The present invention relates to communication technical field, refer to a kind of operator's service quality automatic testing method and system that is used for call center system especially.
Background technology
The call center is by the customer service of higher-end businesses such as large-scale telecom operators and bank, large-scale outsourcing Calling Center Development as promoting level of customer service and the basic means of linking up efficient at first.Along with the continuous transformation of modern commerce operation mode, the call center has begun from original incoming call (Inbound) call center just to make also that to (Outbound) Calling Center Development of breathing out the call center changes to the profit center from original cost centre.But because the particularity of " product " (being generally operator's service voice) that the call center produces also just lets the call center on product quality detection problem, be faced with higher difficulty.
Existing call center quality inspection system generally is to adopt manually-operated mode to accomplish, and the service of extracting the operator by quality inspection personnel is recorded and analyzed, evaluates, for example: for operator's service marking etc.This artificial mode that detects that extracts needs input great amount of manpower and material resources, financial resources to realize the detection to operator's Service Phone.In fact; Quality inspection personnel can not be monitored analysis to each Service Phone; Also just can not there be time enough to go to accomplish understanding to the whole service scenario of operator; The one-sidedness and the imperfection that cause the quality inspection result add factors such as the subjective judge of quality inspection personnel, thereby it are inaccurate, injustice to make that the quality testing result understands.
Summary of the invention
The embodiment of the invention provides a kind of call center services quality determining method and system, realizes the automatic detection to the call center services quality, and testing result is objective.
A kind of call center services quality determining method comprises:
Obtain the service recording of call center;
Extract the attribute information in the said service recording, said attribute information comprises at least two kinds in the following message: the traffic content of the sex of service duration, user recording and emotion and operator's recording;
According to said attribute information, reach the corresponding relation of the parameter of said attribute information and quality testing, confirm to be used for the parameter value of quality testing;
The said parameter value of confirming to be used for quality testing comprises at least two kinds of following deterministic process:
Confirm the service duration of said service recording,, search the threshold interval at the service duration place of the service recording that gets access to, confirm the service duration parameter value of service recording based on the corresponding relation of the service duration threshold interval that is provided with the service duration parameters;
Determine the operator's recording that comprises in the said service recording, according to the traffic content of said operator's recording and preestablish the traffic content and the corresponding relation of content quality parameter, confirm the content quality parameter value of correspondence;
Determine the user recording that comprises in the said service recording; Sex and emotion based on the pairing recording of said user recording user; And the various emotions of the different sexes that is provided with in advance and the corresponding relation of user feeling parameter, confirm corresponding user feeling parameter value;
According to the said parameter value of determining, confirm the service quality test and appraisal value of this traffic service.
A kind of call center services quality detecting system comprises:
The recording acquisition module, the service recording that is used to obtain the call center;
Extraction module is used for extracting the attribute information of said service recording, and said attribute information comprises at least two kinds in the following message: the traffic content of the sex of service duration, user recording and emotion and operator's recording;
The parameter determination module is used for according to said attribute information, and the corresponding relation of the parameter of said attribute information and quality testing, confirms to be used for the parameter value of quality testing; Said parameter determination module specifically comprises: duration is analyzed submodule, is confirmed submodule, quality analysis submodule, and emotion is analyzed submodule; Said duration is analyzed the service duration that submodule is used for confirming said service recording; According to the corresponding relation of the service duration threshold interval that is provided with the service duration parameters; Search the threshold interval at the service duration place of the service recording that gets access to, confirm the service duration parameter value of service recording; Said definite submodule is used for determining respectively operator's recording and user recording that service recording that said recording acquisition module gets access to comprises; Said quality analysis submodule is used for the traffic content of recording according to said operator and preestablishes the traffic content and the corresponding relation of content quality parameter, confirms corresponding content quality parameter value; Said emotion is analyzed submodule, is used for sex and emotion according to the pairing recording of said user recording user, and the various emotions of the different sexes that is provided with in advance and the corresponding relation of user feeling parameter, confirms corresponding user feeling parameter value;
The quality determination module is used for according to said parameter value, confirms the service quality of this traffic service.
Call center services quality determining method and system that the embodiment of the invention provides; Through at least two kinds of attribute information that the service of obtaining the call center is recorded; For example: the traffic content of the sex of service duration, user recording and emotion and operator's recording etc. is carried out Comprehensive Assessment, according to predefined test and appraisal rule, confirms the test and appraisal value of Comprehensive Assessment; Because the service recording to all can be adopted identical standard rating, its evaluating result is objective; And can realize automatically that its implementation efficiency is high; Can set different test and appraisal rules, its flexibility height and highly versatile to different service.
Said method has taken into full account service duration, user's emotion and operator's service content quality; Through from operator and user both sides' dialogue; The voice data of automatic mining unstructured obtains expression content and expresses mode (being user feeling), and operator's content quality is carried out requirement; Simultaneously user's feeling in service analyzed; On the basis of understanding user's impression, deepen understanding, to improve service level in service to the user; That is to say that this evaluating result is the impression of having taken all factors into consideration operator's service and having been served the user, can realize thoroughly evaluating the traffic service.
Unit or system that the realization of said method can be fully has a processing capacity by computer etc. accomplish automatically, do not need artificial the participation, have accelerated the execution speed of call center's quality testing, reduced the cost input, have improved monitoring efficiency.
Description of drawings
Fig. 1 is the flow chart of call center services quality determining method in the embodiment of the invention;
Fig. 2 is the concrete realization flow figure of call center services quality determining method in the embodiment of the invention one;
Fig. 3 is for carrying out the flow chart that content quality is analyzed to operator's recording in the embodiment of the invention one;
Fig. 4 is for carrying out the flow chart that emotion is analyzed to user recording in the embodiment of the invention one;
Fig. 5 is the structural representation of call center services quality detecting system in the embodiment of the invention.
Embodiment
The call center services quality determining method that the embodiment of the invention provides; Extract the recording of service each time of call center; According to service duration parameter, user feeling parameter and the content quality parameter of service recording each time, the service quality of Comprehensive Assessment service recording.Its flow process is as shown in Figure 1, and promptly in the actual detected process, to service recording each time, execution in step is following:
Step S1: the service recording of obtaining the call center.
The call center generally can be provided with the recording module of attending a banquet, and is used for operator's service is recorded.So the service that can from the recording database, get access to traffic service is each time recorded.
Step S2: extract the attribute information in the service recording of obtaining.
Attribute information comprises at least two kinds in the following message: the traffic content of the sex of service duration, user recording and emotion and operator's recording.
Step S3: according to the attribute information that extracts, reach the corresponding relation of the parameter of attribute information and quality testing, confirm to be used for the parameter value of quality testing.
According to the kind of attribute information, corresponding, can determine at least two kinds of parameter values that are used for quality testing.
Then confirm to be used for the process of the parameter value of quality testing, comprise at least two kinds in the following deterministic process accordingly according to the attribute information that extracts:
(i) service duration of definite service recording according to service duration threshold interval that is provided with and the corresponding relation of serving duration parameters, is searched the threshold interval at the service duration place of the service recording that gets access to, the service duration parameter value that the service of confirming is recorded.
(ii) determine the operator's recording that comprises in the service recording, according to the traffic content of operator's recording and preestablish the traffic content and the corresponding relation of content quality parameter, confirm the content quality parameter value of correspondence;
(iii) determine the user recording that comprises in the service recording; Sex and emotion according to the pairing recording of user recording user; And the various emotions of the different sexes that is provided with in advance and the corresponding relation of user feeling parameter, confirm corresponding user feeling parameter value.
Step S4:, confirm the service quality test and appraisal value of this traffic service according to the parameter value of determining that is used for quality testing.Specifically comprise:
Based on the parameter value of the determining shared separately weight proportion in service quality test and appraisal value that is provided with, calculate the service quality test and appraisal value of this traffic service.
Embodiment one
Carrying out the traffic quality testing with three kinds of attribute information of traffic content of the sex that adopts service duration, user recording and emotion and operator's recording below is that example is elaborated.Its flow process is as shown in Figure 2, comprises the following steps:
Step S10: the service recording of obtaining the call center.
, recording comprises that the operator records and user recording from getting access to the service recording of traffic service each time the database.
When adopting single-threaded the recording, operator's recording and user recording are together; When adopting the two-wire journey to record, operator and user are recorded respectively, then operator's recording and user recording are two independent parts.
Preferable, after getting access to the service recording, denoising is carried out in the service recording.For example: can be through (Matrix Laboratory, wavelet transformation noise-eliminating method Matlab) carry out denoising Processing to the service recording that gets access to based on the matrix experiment chamber.
Carry out Daubechies (dbN based on Matlab; N representes the exponent number of small echo) wavelet de-noising, through the threshold value in this pressure noise-eliminating method is set, the high frequency coefficient in the wavelet decomposition structure all is set to 0; Promptly filter all HFSs; Then signal is carried out wavelet reconstruction,, obtain the effect of comparatively level and smooth service recording to reach the high frequency noise of eliminating beyond the voice.
Step S11:, confirm corresponding service duration parameter value according to the service duration of the service recording that gets access to.
According to the corresponding relation of the service duration threshold interval that is provided with the service duration parameters, search the threshold interval at the service duration place of the service recording that gets access to, confirm the service duration parameter value of service recording.
For example, can preestablish the pairing service duration of each service duration threshold interval and get score value,, search the threshold interval at its place, thereby confirm its corresponding score value, i.e. service duration parameter according to the service duration of the service that gets access to recording.As: service duration threshold interval 60～120 seconds, corresponding must be divided into 90 fens; 120～150 seconds corresponding must be divided into 95 fens; Or the like.
Step S12: determine the operator's recording and the user recording that comprise in the service recording of obtaining respectively.
When above-mentioned service recording of obtaining is when adopting single-threaded the recording, from the service recording, extract operator's recording and user recording.Above-mentioned service recording of obtaining is when adopting the two-wire journey to record, and then only need determine two parts recording that wherein comprises, and which part is that the operator records, and that part of is user recording, and need not to extract again.
Step S13: the service quality analysis is carried out in operator's recording, confirmed corresponding content quality parameter value.
Operator's recording is converted into the recording text; The keyword of specified services in advance that comprises in the text according to recording accounts between the proportional band of ratio and setting of the total vocabulary in the recording text corresponding relation with the content quality parameter, confirms corresponding content quality parameter value.As shown in Figure 3, specifically comprise the following steps:
Step S131: operator's recording is converted into recording text (being the text of TXT form) through the speech text switching software.
Step S132: the quantity of the preassigned service keyword that comprises in the recording text that statistics is changed out.
Wherein, preassigned service keyword can comprise: product professional knowledge vocabulary, service expression, communication skill vocabulary or the like.For example: the product professional knowledge vocabulary such as key sentence of name of product and product information; Service expressions such as term of courtesy, the language of thanking you, the language of apologizing, Opening Sentences, conclusion; Language is prohibited in service, communication skill vocabulary such as the guiding term of communication skill aspect, communication term.
Preferable, can specify different service keywords to different service, this service specified keyword can be stored in the knowledge base, uses during in order to statistics.
Step S133: the vocabulary total amount that comprises in the recording text that statistics is changed out.
Step S134: the quantity of the service keyword that counting statistics goes out accounts for the ratio of the vocabulary total amount in the recording text.Promptly calculate the probability that the service keyword occurs, so that confirm content quality based on this probability of occurrence.
Step S135: according between the proportional band that is provided with the corresponding relation of content quality parameter, confirm the content quality parameter value of this service recording.
For example, can be provided with in advance between some proportional bands and corresponding content quality gets score value, account for according to the quantity of the service keyword that calculates between the proportional band at ratio value place of the vocabulary total amount in the recording text, search corresponding content quality and get score value.As: 50-60% between the proportional band, corresponding must be divided into 90 fens; What 60-70% was corresponding between the proportional band must be divided into 95 fens; Or the like.
Step S14: user recording is carried out the emotion analysis, confirm corresponding user feeling parameter value.
According to predetermined sex probability threshold value and the user recording pitch signal cumulative density function value that calculates, confirm the pairing recording of user recording user's sex; Then; Interval according to the affective characteristics parameter of recording user's the user recording that calculates of sex and the pairing affective characteristics parameter distribution of the various emotions of different sexes that counts; Confirm this recording user's emotion, for example: emotion such as angry, frightened, glad, tranquil, sad, surprised; According to the sex and the emotion of the user recording of determining, search the corresponding relation of the various emotions and the user feeling parameter of predefined different sexes again, confirm the user feeling parameter value of this user recording.
Wherein, the user feeling parameter also can get score value through emotion and embody, and for example: preestablishing tranquil corresponding emotion, to get score value be 80, and glad corresponding score value is 90 etc.
Step S15:, confirm the service quality test and appraisal value of this traffic service according to the service duration parameter value of determining, content quality parameter value and user feeling parameter value.Specifically can adopt the test and appraisal rule of setting, confirm the service quality of this traffic service.
According to above-mentioned service duration parameter value, content quality parameter value and the user feeling parameter value of obtaining, and the weight proportion separately set in the rule of test and appraisal, calculate the test and appraisal value of the final service quality of this traffic service.Be the product of calculation services duration parameters value, content quality parameter value and user feeling parameter value and each self-corresponding weight proportion, and each sum of products that calculates, obtain the test and appraisal value of the service quality of this traffic service.
For example: service duration parameter value, content quality parameter value and the shared weight proportion of user feeling parameter value are 1: 2: 2, then can be following according to the test and appraisal value of weight proportion calculation services quality:
Above-mentioned call center services quality determining method of the present invention also comprises:
Step S16: obtain the operation information that the operator carries out the traffic service.
The operator who from the operation system of call center, obtains storage carries out the operation information of traffic service, includes each operator's on attending a banquet operation note in the operation information, can determine the corresponding operator of each service recording according to operation note.
Step S17:, confirm the operator of execution this traffic service corresponding with the service recording of obtaining according to the operation information of the traffic service that gets access to.
Promptly can the service of at every turn obtaining recording and the operator who operates be carried out relatedly, set up incidence relation.
Step S18: after determining the service quality of each traffic service, the service quality evaluating result of determining is stored.So that call during follow-up the needs.
This step can be carried out after step S15, promptly obtains the service quality of each traffic service among the storing step S15, and the service quality of recording can be inquired about, extracts, served at every turn to realization.
Also can after step S17, carry out, promptly obtain the operator that corresponding this traffic of execution of service recording that confirm and that obtain is served among service quality test and appraisal value and the step S17 of each traffic service among the storing step S15.Thereby can realize inquiring about, extract, add up certain operator's of storage the service quality of all service recording, so that each operator's service quality is carried out comprehensively, objectively evaluated.
Above-mentioned steps S14 carries out the emotion analysis to confirm corresponding user feeling parameter to user recording, and its concrete realization flow such as Fig. 4 show, comprise the following steps:
Step S141: according to predetermined sex probability threshold value and the said user recording pitch signal cumulative density function value that calculates, the user recording user's sex of recording is judged, confirmed the pairing recording of user recording user's sex.
Can utilize the difference of the pitch signal frequency statistics model of different sexes that sex is separated.
At first, the cumulative density function value P (f of the Frequency point of the probability difference maximum of pitch signal sex on the cumulative function curve of needs calculating user recording
T).
Then; Judge according to the pitch signal cumulative density function value of the user recording user's sex of recording; Be specially: with the cumulative density function value that calculates, compare, confirm the pairing recording of user recording user's sex with the cumulative distribution threshold value of determining in advance.That is: if the cumulative density function value that obtains, confirms then that the pairing recording of user recording user's sex is the male sex greater than the cumulative distribution threshold value; Otherwise, confirm that the pairing recording of user recording user's sex is the women.
Wherein, cumulative distribution threshold value basis is confirmed the statistics of the cumulative density function value of the maximum Frequency point of the probability difference of pitch signal sex on the cumulative distribution function curve of the user of different sexes.For example: can confirm that the cumulative distribution threshold value is:
Promptly choose the maximum Frequency point f of probability difference of sex on the cumulative distribution function curve
TAs threshold frequency, the average of the gender's that the selected threshold frequency is corresponding cumulative density function value is as the cumulative distribution threshold value.
Wherein, f
TBe the maximum Frequency point of the probability difference of sex on the cumulative function curve;
P
m(f
T) be the male sex's pitch signal cumulative distribution function;
P
f(f
T) be women's pitch signal cumulative distribution function.
For example: above-mentioned cumulative distribution function specifically can calculate through following column count Parzen window method, and its process is following:
To regions
, the pitch signal probability density function can pass through formula (1) and calculate:
Wherein, the probability density of
expression pitch signal;
V
NBe regions
Coverage; Wherein,
Be one and comprise f in an interior regions;
N is all number of samples;
K
NFor falling into regions
In number of samples.
Fall into the zone
Number of samples K
NCan be through a window function of definition
Represent, then according to the Parzen window estimation technique, even f=is (f
1, f
2..., f
n)
l, define a window function
Wherein, i=1,2 ..., n.
In fact
is that to have defined one with functional value 1 be the center with the initial point, and rib length is 1 n dimension hypercube.The translation through coordinate and the convergent-divergent of yardstick can change the position and the size of hypercube.If sample f
iFall into f is that the center is with h
NBe the long zone of rib
Then counting when interior is 1, otherwise counting is 0.Can utilize window function
to realize above-mentioned agreement, that is:
Then fall into the zone
Number of samples K
NComputing formula (2) be:
Wherein, i is a positive integer;
f
iValue for certain sample point in the window function of definition;
h
NThe rib of expression normal state window function forms is long.
Probability of use density function normal state window function
its computing formula (3) is as follows:
With formula (3) substitution formula (2), obtain falling into the zone
Number of samples K
NComputing formula (4) as follows:
With formula (4) substitution formula (1), can obtain pitch signal probability density function concrete computing formula (5) as follows:
Formula (5) is and utilizes the normal state window that the probability density function of fundamental frequency is carried out the computing formula that Parzen window method is estimated, utilizes formula (5) both can estimate the probability density of voice signal fundamental frequency; This probability density function is carried out integral operation, can obtain the fundamental frequency cumulative distribution function.
According to the gene frequency cumulative distribution function that obtains, both can calculated rate point f
TThe corresponding male sex's pitch signal cumulative density function value P
m(f
T) and Frequency point f
TCorresponding women's pitch signal cumulative density function value P
f(f
T); Also can calculate the cumulative density function value P (f of the pitch signal of user recording
T).
Wherein, the male sex's pitch signal cumulative density function value P
m(f
T) and Frequency point f
TCorresponding women's pitch signal cumulative density function value P
f(f
T), can confirm through the cumulative density function value of the known voice sample data of sex being carried out statistical analysis.
Step S142: interval according to the affective characteristics parameter of the user recording that calculates, the recording user's that determines sex and the pairing affective characteristics parameter distribution of the various emotions of different sexes that counts, confirm this recording user's emotion.Specifically comprise:
Wherein, the pairing affective characteristics parameter distribution of the various emotions of different sexes is interval, carries out confirming behind the classification analysis according to the voice sample of the different emotions of each sex of statistic of classification in advance.
Specifically can adopt based on the three dimensions of emotion theoretical; Sex and the known voice sample of emotion are carried out statistic of classification and storage; To the voice sample of the various emotions of different sexes, for example: angry, frightened, glad, tranquil, sad, surprised etc., add up the distributed area at the affective characteristics parameter place of every kind of emotion; Thereby confirm the affective characteristics parameter distributions interval of the various emotions of different sexes, for example: the average of fundamental tone and formant and variance distributed area.
When need confirming the pairing recording of user recording user's emotion, calculate the affective characteristics parameter of user recording, and the distributed area at the affective characteristics parameter place of confirming to calculate, confirm that according to distributed area corresponding emotion is this recording user's emotion.For example: calculate the fundamental tone of user recording and the average and the variance of formant, generally can select to calculate the average and the variance on first three rank of formant; And search average and the distributed area at variance place of fundamental tone and the formant of the user recording that calculates, confirm that according to distributed area corresponding emotion is this recording user's emotion.
Step S143: according to the sex and the emotion of the user recording of determining, search the corresponding relation of the every kind of emotion and the user feeling parameter of predefined different sexes, confirm the user feeling parameter value of this user recording.
For example: the user feeling parameter value specifically can get score value through user feeling and characterize.Promptly can preestablish every kind of pairing user feeling of emotion and get score value,, search corresponding user feeling and get score value according to the user feeling of determining, for example: user feeling difference correspondence such as angry, frightened, glad, tranquil, sad, surprised must score value.
According to above-mentioned call center services quality determining method of the present invention, can make up a kind of call center services quality detecting system, as shown in Figure 5, comprising: recording acquisition module 10, extraction module 20, parameter determination module 30 and quality determination module 40.
Recording acquisition module 10, the service recording that is used to obtain the call center.
Accordingly, parameter determination module 30 specifically comprises: duration is analyzed submodule 301, is confirmed submodule 302, quality analysis submodule 303 and emotion are analyzed submodule 304.
Duration is analyzed submodule 301; Be used for confirming the service duration of service recording; According to the corresponding relation of the service duration threshold interval that is provided with the service duration parameters, search the threshold interval at the service duration place of the service recording that gets access to, confirm the service duration parameter value of service recording.
Preferable, quality analysis submodule 303 further can comprise: conversion unit 3031, computing unit 3032 and quality are confirmed unit 3033.
Quality is confirmed unit 3033, be used for according between predefined proportional band with the corresponding relation of content quality parameter, search between the proportional band at the ratio value place that computing unit 3032 calculates, confirm the corresponding content quality parameter value of this service recording.
Emotion is analyzed submodule 304, is used for sex and emotion according to the pairing recording of user recording user, and the various emotions of the different sexes that is provided with in advance and the corresponding relation of user feeling parameter, confirms corresponding user feeling parameter value.
Preferable, emotion is analyzed submodule 304, further can comprise: sex confirms that unit 3041, emotion confirm that unit 3042 and parameter confirm unit 3043.
Sex is confirmed unit 3041, is used for according to predetermined sex probability threshold value and the user recording pitch signal cumulative density function value that calculates, confirms the pairing recording of user recording user's sex.
Emotion is confirmed unit 3042, is used for according to the affective characteristics parameter of recording user's sex, the user recording that calculates and the pairing affective characteristics parameter distribution of the various emotions of different sexes that counts in advance intervally, confirms this recording user's emotion.
Parameter is confirmed unit 3043, is used for sex and emotion according to the user recording of determining, searches the corresponding relation of the various emotions and the user feeling parameter of predefined different sexes, confirms the user feeling parameter value of this user recording.
Said system also comprises: noise elimination module 50 is used for through the wavelet transformation de-noising denoising being carried out in the service recording that recording acquisition module 10 obtains.
Said system also comprises: message processing module 60 and memory module 70.
Preferable, a voice sample collection module can also be set, for emotion is analyzed the voice sample of various emotions that submodule 304 provides the recording different sexes of collection in the said system.
Above-mentioned call center services quality detecting system of the present invention can be realized through software, also can realize through hardware, and for example, can be a kind of information processing chip.
Be a concrete calculating case explanation below.
Make a telephone call to the call center when inquiring about the service of certain business information when a male user, and the recording module of attending a banquet can be recorded to the service of this time traffic.The recording acquisition module can get access to the service recording of this time from the recording module of attending a banquet; Preferable, can carry out denoising Processing by noise elimination module.
Analyze the service duration parameter value that submodule is confirmed this time service recording by duration then.For example: the service duration of this traffic service is 70 seconds, and in the service duration parameter of setting, 60～120 seconds corresponding that score value be 90 minutes, and then the score of this service duration was 90 fens.
Confirm operator's recording and the user recording that wherein comprises respectively by confirming submodule.Then, the quality analysis submodule carries out the service quality analysis to operator's recording, confirms corresponding content quality parameter value; For example: user's emotion is tranquil, and in the pairing affective characteristics parameter distribution of the various emotions of the different sexes interval that is provided with, sex be the male sex, and corresponding to get score value be 80 minutes to emotion for tranquil, and then the user feeling of this traffic service must be divided into 80 fens.
Emotion is analyzed submodule user recording is carried out the emotion analysis, confirms corresponding user feeling parameter value; For example: the ratio that the service keyword that comprises in the operator recording accounts for its vocabulary total amount is 18%, between the service keyword proportional band of setting in, 15%～20% is 80 minutes, then to get score value be 80 minutes to the content quality of this traffic service.
The quality determination module is used for according to the above-mentioned service duration parameter value that obtains, content quality parameter value and user feeling parameter value, confirms the service quality test and appraisal value of this traffic service.The weight proportion of service duration parameter, content quality parameter and user feeling parameter is 1: 2: 2 in the test and appraisal rule of for example setting, and then the service quality test and appraisal value of this traffic service does
Promptly 82 minutes.
The above; Be merely the preferable embodiment of the present invention; But protection scope of the present invention is not limited thereto; Any technical staff who is familiar with the present technique field variation that can expect easily, replaces or is applied to other similar devices in the technical scope that the present invention discloses, all should be encompassed within protection scope of the present invention.Therefore, protection scope of the present invention should be as the criterion with the protection range of claims.
Claims (12)
1. a call center services quality determining method is characterized in that, comprising:
Obtain the service recording of call center;
Extract the attribute information in the said service recording, said attribute information comprises at least two kinds in the following message: the traffic content of the sex of service duration, user recording and emotion and operator's recording;
According to said attribute information, reach the corresponding relation of the parameter of said attribute information and quality testing, confirm to be used for the parameter value of quality testing; The said parameter value of confirming to be used for quality testing comprises at least two kinds of following deterministic process:
Confirm the service duration of said service recording,, search the threshold interval at the service duration place of the service recording that gets access to, confirm the service duration parameter value of service recording based on the corresponding relation of the service duration threshold interval that is provided with the service duration parameters;
Determine the operator's recording that comprises in the said service recording, according to the traffic content of said operator's recording and preestablish the traffic content and the corresponding relation of content quality parameter, confirm the content quality parameter value of correspondence;
Determine the user recording that comprises in the said service recording; Sex and emotion based on the pairing recording of said user recording user; And the various emotions of the different sexes that is provided with in advance and the corresponding relation of user feeling parameter, confirm corresponding user feeling parameter value;
According to the said parameter value of determining, confirm the service quality test and appraisal value of this traffic service.
2. the method for claim 1 is characterized in that, and is said according to said parameter value, confirms that the service quality test and appraisal value of this traffic service specifically comprises:
Based on the said parameter value shared separately weight proportion in service quality test and appraisal value that is provided with, calculate the service quality test and appraisal value of this traffic service.
3. the method for claim 1 is characterized in that, confirms the record process of pairing content quality parameter value of said operator, specifically comprises:
Said operator recording is converted into the recording text, and the quantity of the preassigned service keyword that comprises in the said recording text that goes out of counting statistics accounts for the ratio value of the vocabulary total amount in the said recording text that counts;
According between predefined proportional band with the corresponding relation of content quality parameter, search between the proportional band at said ratio value place, confirm the corresponding content quality parameter value of this service recording.
4. the method for claim 1 is characterized in that, confirms the process of the pairing user feeling parameter value of said user recording, specifically comprises:
According to predetermined sex probability threshold value and the said user recording pitch signal cumulative density function value that calculates, confirm the pairing recording of said user recording user's sex;
Interval according to the affective characteristics parameter of recording user's sex, the user recording that calculates and the pairing affective characteristics parameter distribution of the various emotions of different sexes that counts in advance, confirm this recording user's emotion;
According to the sex and the emotion of the user recording of determining, search the corresponding relation of the various emotions and the user feeling parameter of predefined different sexes, confirm the user feeling parameter value of this user recording.
5. like the described method of the arbitrary claim of claim 2-4; It is characterized in that; Also comprise: obtain the operation information that carries out the traffic service according to the operator of storage, confirm the operator of execution this traffic service corresponding according to said operation information with said service recording;
After the service quality of then said definite this traffic service, also comprise: the said service quality test and appraisal value that will determine and the operator who carries out this traffic service store.
6. method as claimed in claim 5 is characterized in that, before the attribute information in the said service recording of said extraction, also comprises: through the wavelet transformation de-noising denoising is carried out in the said service recording of obtaining.
7. a call center services quality detecting system is characterized in that, comprising:
The recording acquisition module, the service recording that is used to obtain the call center;
Extraction module is used for extracting the attribute information of said service recording, and said attribute information comprises at least two kinds in the following message: the traffic content of the sex of service duration, user recording and emotion and operator's recording;
The parameter determination module is used for according to said attribute information, and the corresponding relation of the parameter of said attribute information and quality testing, confirms to be used for the parameter value of quality testing; Said parameter determination module specifically comprises: duration is analyzed submodule, is confirmed submodule, quality analysis submodule, and emotion is analyzed submodule; Said duration is analyzed the service duration that submodule is used for confirming said service recording; According to the corresponding relation of the service duration threshold interval that is provided with the service duration parameters; Search the threshold interval at the service duration place of the service recording that gets access to, confirm the service duration parameter value of service recording; Said definite submodule is used for determining respectively operator's recording and user recording that service recording that said recording acquisition module gets access to comprises; Said quality analysis submodule is used for the traffic content of recording according to said operator and preestablishes the traffic content and the corresponding relation of content quality parameter, confirms corresponding content quality parameter value; Said emotion is analyzed submodule, is used for sex and emotion according to the pairing recording of said user recording user, and the various emotions of the different sexes that is provided with in advance and the corresponding relation of user feeling parameter, confirms corresponding user feeling parameter value;
The quality determination module is used for according to said parameter value, confirms the service quality of this traffic service.
8. system as claimed in claim 7 is characterized in that, said quality analysis submodule specifically comprises:
Conversion unit is used for said operator's recording is converted into the recording text;
Computing unit, the quantity that is used for the preassigned service keyword that said recording text that counting statistics goes out comprises accounts for the ratio value of the vocabulary total amount in the said recording text that counts;
Quality is confirmed the unit, be used for according between predefined proportional band with the corresponding relation of content quality parameter, search between the proportional band at said ratio value place, confirm the corresponding content quality parameter value of this service recording.
9. system as claimed in claim 7 is characterized in that, said emotion is analyzed submodule, specifically comprises:
Sex is confirmed the unit, is used for according to predetermined sex probability threshold value and the said user recording pitch signal cumulative density function value that calculates, confirms the pairing recording of said user recording user's sex;
Emotion is confirmed the unit, is used for according to the affective characteristics parameter of recording user's sex, the user recording that calculates and the pairing affective characteristics parameter distribution of the various emotions of different sexes that counts in advance intervally, confirms this recording user's emotion;
Parameter is confirmed the unit, is used for sex and emotion according to the user recording of determining, searches the corresponding relation of the various emotions and the user feeling parameter of predefined different sexes, confirms the user feeling parameter value of this user recording.
10. system as claimed in claim 7 is characterized in that, said quality determination module specifically is used for: at the shared separately weight proportion of service quality test and appraisal value, calculate the service quality test and appraisal value of this traffic service according to the said parameter that is provided with.
11. system as claimed in claim 7 is characterized in that, also comprises:
Noise elimination module is used for through the wavelet transformation de-noising denoising being carried out in the said service recording of obtaining.
12. like the described system of the arbitrary claim of claim 7-11, it is characterized in that, also comprise:
Message processing module is used to obtain the operation information that carries out the traffic service based on the operator of storage, confirms the operator of execution this traffic service corresponding with said service recording based on said operation information;
Memory module is used for said service quality test and appraisal value of determining and the operator that carries out this traffic service are stored.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200910170363A CN101662550B (en) | 2009-09-11 | 2009-09-11 | Method and system for service quality detection for call center |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200910170363A CN101662550B (en) | 2009-09-11 | 2009-09-11 | Method and system for service quality detection for call center |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101662550A CN101662550A (en) | 2010-03-03 |
| CN101662550B true CN101662550B (en) | 2012-10-03 |
Family
ID=41790321
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN200910170363A Active CN101662550B (en) | 2009-09-11 | 2009-09-11 | Method and system for service quality detection for call center |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101662550B (en) |
Families Citing this family (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103067621A (en) * | 2011-10-18 | 2013-04-24 | 殷程 | Home seat management system of call center and method thereof |
| CN103701999B (en) * | 2012-09-27 | 2017-03-01 | 中国电信股份有限公司 | The monitoring method of call center's voice call and system |
| CN103905657B (en) * | 2012-12-28 | 2017-09-29 | 中国移动通信集团江苏有限公司 | A kind of method and system of monitoring calling service quality |
| CN104158987B (en) * | 2014-07-29 | 2017-03-29 | 北京佳讯飞鸿电气股份有限公司 | A kind of method for lifting artificial telephone traffic system service quality |
| CN104680428B (en) * | 2015-03-16 | 2018-03-16 | 朗新科技股份有限公司 | A kind of construction method of power network CSAT model |
| CN106847305B (en) * | 2015-12-04 | 2020-09-18 | 中国移动通信集团公司 | Method and device for processing recording data of customer service telephone |
| CN107093431B (en) * | 2016-02-18 | 2020-07-07 | 中国移动通信集团辽宁有限公司 | Method and device for quality inspection of service quality |
| CN105975514A (en) * | 2016-04-28 | 2016-09-28 | 朱宇光 | Automatic quality testing method and system |
| 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 |
| CN106453971B (en) * | 2016-11-22 | 2019-04-16 | 广东电网有限责任公司佛山供电局 | The acquisition methods and call center's quality inspection system of call center's quality inspection voice |
| CN107154257B (en) * | 2017-04-18 | 2021-04-06 | 苏州工业职业技术学院 | Customer service quality evaluation method and system based on customer voice emotion |
| CN107293309B (en) * | 2017-05-19 | 2021-04-30 | 四川新网银行股份有限公司 | Method for improving public opinion monitoring efficiency based on client emotion analysis |
| CN107341685A (en) * | 2017-05-24 | 2017-11-10 | 百度在线网络技术（北京）有限公司 | Data analysing method and device |
| CN107833059B (en) * | 2017-11-03 | 2021-06-18 | 广州杰赛科技股份有限公司 | Service quality evaluation method and system for customer service |
| CN108874619B (en) * | 2018-05-14 | 2023-09-22 | 广州宏同信息技术有限公司 | Information monitoring method, storage medium and server |
| CN108833722B (en) * | 2018-05-29 | 2021-05-11 | 平安科技（深圳）有限公司 | Speech recognition method, speech recognition device, computer equipment and storage medium |
| CN111311041B (en) * | 2018-12-12 | 2024-02-20 | 马上消费金融股份有限公司 | Data extraction method and device and related device |
| CN109639914A (en) * | 2019-01-08 | 2019-04-16 | 深圳市沃特沃德股份有限公司 | Intelligent examining method, system and computer readable storage medium |
| CN112714217A (en) * | 2019-10-25 | 2021-04-27 | 中兴通讯股份有限公司 | Telephone traffic quality inspection method, device, storage medium and server |
| CN110839112A (en) * | 2019-11-18 | 2020-02-25 | 广东电网有限责任公司佛山供电局 | Problem voice detection method and device |
| CN112256836A (en) * | 2020-11-04 | 2021-01-22 | 中国建设银行股份有限公司 | Recording data processing method and device and server |
| CN113438375B (en) * | 2021-05-24 | 2022-09-27 | 商客通尚景科技（上海）股份有限公司 | Method for maintaining seat state |
| CN113571143A (en) * | 2021-07-26 | 2021-10-29 | 首都医科大学附属北京天坛医院 | Audio information processing method and device |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101005540A (en) * | 2006-01-19 | 2007-07-25 | 北京讯鸟软件有限公司 | Method for realizing sitting service level evaluation of calling center service |
| CN101150435A (en) * | 2007-10-24 | 2008-03-26 | 中兴通讯股份有限公司 | A system and method for customer comment feedback by using multimedia |
-
2009
- 2009-09-11 CN CN200910170363A patent/CN101662550B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101005540A (en) * | 2006-01-19 | 2007-07-25 | 北京讯鸟软件有限公司 | Method for realizing sitting service level evaluation of calling center service |
| CN101150435A (en) * | 2007-10-24 | 2008-03-26 | 中兴通讯股份有限公司 | A system and method for customer comment feedback by using multimedia |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101662550A (en) | 2010-03-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101662550B (en) | Method and system for service quality detection for call center | |
| CN112804400B (en) | Customer service call voice quality inspection method and device, electronic equipment and storage medium | |
| CN110910901B (en) | Emotion recognition method and device, electronic equipment and readable storage medium | |
| CN108108902B (en) | Risk event warning method and device | |
| CN102623011B (en) | Information processing apparatus, information processing method and information processing system | |
| CN102623009B (en) | Abnormal emotion automatic detection and extraction method and system on basis of short-time analysis | |
| US20100332287A1 (en) | System and method for real-time prediction of customer satisfaction | |
| US20090112588A1 (en) | Method for segmenting communication transcripts using unsupervsed and semi-supervised techniques | |
| CN104301554A (en) | Device and method used for detecting service quality of customer service staff | |
| CN110942229A (en) | Service quality evaluation method and device, electronic equipment and storage medium | |
| JP2007041593A (en) | Method and apparatus for extracting voiced/unvoiced classification information using harmonic component of voice signal | |
| CN108022587A (en) | Audio recognition method, device, computer equipment and storage medium | |
| Dubey et al. | Non-intrusive speech quality assessment using several combinations of auditory features | |
| CN111080109B (en) | Customer service quality evaluation method and device and electronic equipment | |
| US20060200346A1 (en) | Speech quality measurement based on classification estimation | |
| US20100088323A1 (en) | Method and apparatus for visualization of interaction categorization | |
| Pappas et al. | Anger detection in call center dialogues | |
| CN110929011A (en) | Conversation analysis method, device and equipment | |
| JP2010266522A (en) | Apparatus, method and program for dividing dialogue state and recording medium | |
| CN114443930A (en) | News public opinion intelligent monitoring and analyzing method, system and computer storage medium | |
| CN112562736A (en) | Voice data set quality evaluation method and device | |
| McDougall et al. | Application of the ‘TOFFA’framework to the analysis of disfluencies in forensic phonetic casework | |
| CN112233693A (en) | Sound quality evaluation method, device and equipment | |
| CN111160778A (en) | Outbound project auditing and evaluating method and system based on big data and computer equipment | |
| Morovati | Increase the accuracy of speech signal categories in high noise environments |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |