CN106548788A - A kind of intelligent emotion determines method and system - Google Patents
A kind of intelligent emotion determines method and system Download PDFInfo
- Publication number
- CN106548788A CN106548788A CN201510613689.XA CN201510613689A CN106548788A CN 106548788 A CN106548788 A CN 106548788A CN 201510613689 A CN201510613689 A CN 201510613689A CN 106548788 A CN106548788 A CN 106548788A
- Authority
- CN
- China
- Prior art keywords
- audio
- frequency information
- information section
- frequency
- emotion
- 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
- 230000008451 emotion Effects 0.000 title claims abstract description 301
- 238000000034 method Methods 0.000 title claims abstract description 42
- 230000002159 abnormal effect Effects 0.000 claims abstract description 207
- 238000004458 analytical method Methods 0.000 claims abstract description 31
- 230000001143 conditioned effect Effects 0.000 claims abstract description 23
- 230000006854 communication Effects 0.000 claims abstract description 15
- 238000004891 communication Methods 0.000 claims abstract description 14
- 230000004044 response Effects 0.000 claims description 21
- 238000012360 testing method Methods 0.000 claims description 14
- 230000001186 cumulative effect Effects 0.000 claims description 13
- 238000001514 detection method Methods 0.000 claims description 11
- 230000004913 activation Effects 0.000 claims description 8
- 230000002085 persistent effect Effects 0.000 claims description 6
- 239000000203 mixture Substances 0.000 claims description 3
- 238000005516 engineering process Methods 0.000 abstract description 6
- 238000002347 injection Methods 0.000 description 11
- 239000007924 injection Substances 0.000 description 11
- 230000008569 process Effects 0.000 description 8
- 230000008859 change Effects 0.000 description 6
- 230000002452 interceptive effect Effects 0.000 description 4
- 230000004048 modification Effects 0.000 description 4
- 238000012986 modification Methods 0.000 description 4
- 241000894007 species Species 0.000 description 4
- 238000013480 data collection Methods 0.000 description 3
- 230000002996 emotional effect Effects 0.000 description 2
- 230000003340 mental effect Effects 0.000 description 2
- 230000036651 mood Effects 0.000 description 2
- 238000013518 transcription Methods 0.000 description 2
- 230000035897 transcription Effects 0.000 description 2
- 241001269238 Data Species 0.000 description 1
- 241001413866 Diaphone Species 0.000 description 1
- 208000019022 Mood disease Diseases 0.000 description 1
- 235000021167 banquet Nutrition 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000036760 body temperature Effects 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 239000003795 chemical substances by application Substances 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 210000004247 hand Anatomy 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 239000007943 implant Substances 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000002045 lasting effect Effects 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 239000004575 stone Substances 0.000 description 1
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
- 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
- G10L25/63—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 estimating an emotional state
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Hospice & Palliative Care (AREA)
- Psychiatry (AREA)
- General Health & Medical Sciences (AREA)
- Child & Adolescent Psychology (AREA)
- Signal Processing (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Telephonic Communication Services (AREA)
Abstract
Embodiments provide a kind of intelligent emotion and determine method and system, obtain the audio-frequency information of personnel to be detected and user's communication;From each audio-frequency information section for constituting audio-frequency information, abnormal emotion audio-frequency information section is determined, wherein, abnormal emotion audio-frequency information section meets the pre-conditioned audio-frequency information section of correspondence by the preset audio information for characterizing personnel's abnormal emotion to be detected for including;When abnormal emotion audio-frequency information section is capable of determining that, determine that the corresponding personnel to be detected of audio-frequency information section have abnormal emotion.The personnel to be detected that more objective, analysis is obtained are more accurate with the presence or absence of abnormal emotion.The present invention relates to field of computer technology.
Description
Technical field
The present invention relates to field of computer technology, more particularly to a kind of intelligent emotion determines method and system.
Background technology
With society development science and technology progress, bring abundant material life, then social competition also with
Become fierce.People produce negative emotion in keen competition unavoidably, not only affect work also to affect
Personal is healthy, below by taking the contact staff for receiving calls as an example.
Contact staff is by receiving calls as user's solve problem, user's shape shape and color that longevity of service runs into
Color, psychology and body bear immense pressure and are difficult to mediate, and easily produce abnormal emotion, may so lead
Work efficiency drop is caused, with generation situations such as user's confusing communication.In prior art, for contact staff
The method and channel that pressure is discongested all waits to expand, and mainly passes through traditional to the motion management of contact staff
The mental status assessment of the communication way of person to person's face-to-face interview and questionnaire form made of paper, often cannot be true
Reflect the real mental verbs of employee.
In order to solve this problem, in prior art propose emotion record analyses bootstrap technique, mainly include as
Lower step:The mood data for contact staff's record is read by motion management platform, issuing;By intelligent handss
Table gathers the physiological datas such as the heart rate of user, body temperature;By the analysis to these data, contact staff is obtained
Currently whether there is abnormal emotion.
Emotion record analyses bootstrap technique of the prior art realizes the analysis to abnormal emotion and guides, but its
The acquisition of abnormal emotion data is depended on contact staff oneself typing and professional equipment to physiological data
Collection information, the mood data of contact staff oneself typing are unavoidable excessively subjective, and physiological data is by customer service people
The health of member it is different and different, therefore, so analyze the abnormal emotion that obtains not accurate enough.
The content of the invention
Embodiments provide a kind of intelligent emotion and determine method and system, to solve in prior art
Inaccurate problem is determined to abnormal emotion.
Based on the problems referred to above, a kind of intelligent emotion provided in an embodiment of the present invention determines method, including:
Obtain the audio-frequency information of personnel to be detected and user's communication;
From each audio-frequency information section for constituting the audio-frequency information, abnormal emotion audio-frequency information section is determined,
Wherein, the abnormal emotion audio-frequency information section is by including for characterizing personnel's exception to be detected
The preset audio information of emotion meets the pre-conditioned audio-frequency information section of correspondence;
When the abnormal emotion audio-frequency information section is capable of determining that, determines that the audio-frequency information section is corresponding and treat
There is abnormal emotion in testing staff.
A kind of intelligent emotion provided in an embodiment of the present invention determines system, including:
Audio collection module, for obtaining the audio-frequency information of personnel to be detected and user's communication;
Speech waveform analysis module, for, from each audio-frequency information section for constituting the audio-frequency information, determining different
Reason thread audio-frequency information section, wherein, the abnormal emotion audio-frequency information section is described for characterizing by including
The preset audio information of personnel's abnormal emotion to be detected meets the pre-conditioned audio-frequency information section of correspondence;And work as
When being capable of determining that the abnormal emotion audio-frequency information section, the corresponding people to be detected of the audio-frequency information section is determined
There is abnormal emotion in member.
The beneficial effect of the embodiment of the present invention includes:
A kind of intelligent emotion provided in an embodiment of the present invention determines method and system, obtains personnel to be detected and uses
The audio-frequency information of family call;From each audio-frequency information section for constituting audio-frequency information, determine that abnormal emotion audio frequency is believed
Breath section, wherein, abnormal emotion audio-frequency information section is by including for characterizing personnel's abnormal emotion to be detected
Preset audio information meets the pre-conditioned audio-frequency information section of correspondence;When be capable of determining that abnormal emotion audio frequency believe
During breath section, determine that the corresponding personnel to be detected of audio-frequency information section have abnormal emotion.The embodiment of the present invention is provided
Intelligent emotion determine method, by obtaining the audio-frequency information of personnel, talking to be detected and audio-frequency information being carried out
Analyze to determine people to be detected with the presence or absence of abnormal emotion, as the language of people tends to embody current feelings
Thread, therefore audio-frequency information can objectively embody the current emotional state of personnel to be detected, in prior art
The acquisition of abnormal emotion data is depended on contact staff oneself typing and professional equipment to physiological data
Collection information is compared, more objective, and the personnel to be detected that so analysis is obtained whether there is abnormal emotion more
Accurately.And the acquisition of abnormal emotion data needs to complete by professional external equipment auxiliary in prior art,
Increased cost.
Description of the drawings
Fig. 1 is the flow chart that a kind of intelligent emotion provided in an embodiment of the present invention determines method;
Fig. 2 is the flow chart that a kind of intelligent emotion that the embodiment of the present invention 1 is provided determines method;
Fig. 3 is the flow chart of a kind of intelligent emotion bootstrap technique that the embodiment of the present invention 2 is provided;
Fig. 4 is intelligent AC module provided in an embodiment of the present invention, exchange injection module and sensitive language
The flow chart of collection module cooperating;
Fig. 5 is the structural representation that a kind of intelligent emotion provided in an embodiment of the present invention determines one of system;
Fig. 6 be a kind of intelligent emotion provided in an embodiment of the present invention determine system two structural representation.
Specific embodiment
Embodiments provide a kind of intelligent emotion and determine method and system, below in conjunction with Figure of description
The preferred embodiments of the present invention are illustrated, it will be appreciated that preferred embodiment described herein is only used for
The description and interpretation present invention, is not intended to limit the present invention.And in the case where not conflicting, in the application
Embodiment and embodiment in feature can be mutually combined.
The embodiment of the present invention provides a kind of intelligent emotion and determines method, as shown in figure 1, including:
S101, the audio-frequency information for obtaining personnel to be detected and user's communication.
In S102, each audio-frequency information section of the audio-frequency information obtained from composition S101, abnormal emotion sound is determined
Frequency message segment,
Wherein, abnormal emotion audio-frequency information section is by including for characterizing the pre- of personnel's abnormal emotion to be detected
If audio-frequency information meets the pre-conditioned audio-frequency information section of correspondence.
S103, when abnormal emotion audio-frequency information section is capable of determining that, determine that audio-frequency information section is corresponding to be checked
There is abnormal emotion in survey personnel.
Below in conjunction with the accompanying drawings, the method and relevant device for being provided to the present invention with specific embodiment is retouched in detail
State.
Embodiment 1:
In the embodiment of the present invention 1, there is provided a kind of intelligent emotion determines method, as shown in Fig. 2 specifically including
Following steps:
S201, the audio-frequency information for obtaining personnel to be detected and user's communication.
By taking the contact staff for receiving calls as an example, using contact staff as personnel to be detected, can obtain in real time
Contact staff's terminal attends a banquet microphone audio input as the audio frequency of the personnel to be detected and user's communication for obtaining
Information, and obtain the voice document comprising audio-frequency information.
Further, the embodiment of the present invention realizes a kind of intelligent emotion and determines system, and the system can have many
Individual functional module composition, the executive agent of this step can be the audio collection module in system.
S202, the audio-frequency information obtained in S201 is converted to into structurized audio frequency text.
In this step, text index can be set up by existing speech transcription mode, by non-structured language
Sound file is converted to structurized text message i.e. audio frequency text, is that subsequently audio information analysis are processed
Lay the first stone.
Further, for determining system for intelligent emotion provided in an embodiment of the present invention, this step is held
Row main body can be audio collection module.It is follow-up to can be by the voice in system the step of audio information analysis
Waveform analyses module is performed, therefore, the audio frequency text that this step can be generated by audio collection module is sent out
Give speech waveform analysis module, and also can by employee's job number of personnel correspondence to be detected, login time,
The information such as work accumulative duration, history degrees of emotion are sent to speech waveform analysis module, so as to speech waveform
Analysis module is used when being analyzed to audio-frequency information.
Further, due to analysis be personnel to be detected audio-frequency information, and different places have different sides
Speech, the embodiment of the present invention is additionally contemplates that the problem of accent, actually used provided in an embodiment of the present invention
When intelligent emotion determines method, in order that audio-frequency information to be converted to structurized audio frequency text more accurate
Really, also need to be adapted to reference to each region accent, optimize acoustic model so as to extensively can cover
Accent, also needs with reference to professional knowledge and hotline service scope, carries out speech model and optimizes to lift voice
Transcription accuracy rate.
S203, the audio frequency text to obtaining in S202 are parsed, and determine each audio-frequency information Duan Yin
The preset audio information that corresponding each audio frequency text chunk is included respectively in frequency text,
Wherein, preset audio packet is containing at least one following information:Keyword message, emotion detection information
With quiet duration information.
Further, for determining system for intelligent emotion provided in an embodiment of the present invention, this step is held
Row main body can be speech waveform analysis module.Speech waveform analysis module can detect base in certain audio output
Frequently, the amplitude of variation such as pitch, there is provided be likely to occur the prediction of anxious state of mind, it is possible to navigate to anxious state of mind
Positional information of the audio frequency in whole piece voice;Detect and analyze the change of word speed, mute time etc..
These information, when being embodied as, can be generated the index file of XML format, the index file can
With including following one or more information:
The personnel to be detected of the both call sides that audio frequency text includes and the voice of user;
Exceed the word speed of average speed in the audio frequency of word speed information, that is, both call sides in short-term text chunk;
In the sound end information of call, average word speed information, that is, arbitrary audio frequency text chunk, call is double
(unit can be for beginning and ending time of each say per of side, word speed information:Word/second), and arbitrary sound
In frequency text chunk, (unit can be the respective average word speed information of both call sides:Word/second);
The frequency that crest amplitude information and/or crest amplitude occur in audio frequency text chunk;
Audio frequency text chunk medium pitch information, can be highest frequency and max volume information;
Fundamental frequency information in audio frequency text chunk.
Further, the division for constituting each audio-frequency information section i.e. audio frequency text chunk of audio-frequency information can be with root
Carry out according to being actually needed, can be using the audio section of setting duration as an audio-frequency information section, it is also possible to one
Secondary call is an audio-frequency information section.
After index file is generated, the default sound that each audio frequency text chunk is included can be determined from index file
Frequency information.That is, speech waveform analysis module can be directed to the preset audio letter for needing analysis and retrieving
Breath (keyword message, emotion detection information and it is long when silence information in one or more), from index text
Enter line retrieval in part, and return the audio-frequency information that there may be abnormal emotion of concern.
It is embodied as performing following steps:
Step one, when preset audio packet when containing keyword message, for each in each audio-frequency information section
Audio-frequency information section, the audio-frequency information section corresponding audio frequency text chunk is compared with preset keyword;It is determined that
The preset keyword included in the audio frequency text chunk, and preset keyword occur initial time and terminate when
Between;
In step one, preset keyword can be word, word, phrase etc., can arrange according to actual needs
The word for characterizing abnormal emotion is preset keyword.When multiple keywords are provided with, keyword can be generated
List, each audio frequency text chunk is compared with the keyword in list respectively, obtains comprising any in list
The audio frequency text chunk list of one or more keywords, and the appearance in audio frequency text chunk respectively of each keyword
Initial time and terminate the time.
Step 2, when preset audio packet detection information containing emotion, for each audio-frequency information section in it is every
Individual audio-frequency information section, determine characterize in the corresponding audio frequency text chunk of the audio-frequency information section emotion Testing index as
The desired value of next or multiple indexs:The word speed information of both call sides, amplitude information, frequency information, sound
Amount information, fundamental frequency information;In determining the audio frequency text chunk, desired value reaches the These parameters for corresponding to metrics-thresholds,
And the desired value of These parameters reach correspondence metrics-thresholds when occur in the audio frequency text chunk it is initial when
Between and terminate the time;
In step 2, emotion Testing index can be characterized by following one or more indexs:Both call sides
Word speed information, amplitude information, frequency information, information volume, fundamental frequency information etc., it is right that each index has
The metrics-thresholds answered, each metrics-thresholds can be set according to the history voice data of the personnel to be detected,
Namely by history voice data empirically condition, the setting of alarm threshold is carried out.So, by index
Threshold value can characterize the personnel to be detected, and in the emotion safety range of non-abnormal emotion subaudio frequency, (fundamental frequency is relative to be become
Change degree, word speed thresholding, change persistent period etc., wherein, fundamental frequency can refer to the basic tone of sound, base
Frequency contrast can refer to the change contrast of the basic basis tone of two sections of voices, and the change persistent period can refer to that fundamental frequency is sent out
Have lasting duration during changing altogether, for example:Certain continue for 10 minutes, occur one after 2 hours again
Secondary pitch variation, continue for 15 minutes again), then by each index of each audio frequency text chunk in index file
Desired value compare with corresponding metrics-thresholds, determine each audio frequency text chunk whether in the safe model of the emotion
In enclosing, if exceeding emotion safety range, then it is assumed that mood disorderss occur.
Step 3, when preset audio packet contain quiet duration information when, for each audio-frequency information section in it is every
Individual audio-frequency information section, according to per words diaphone of both call sides in the corresponding audio frequency text chunk of the audio-frequency information section
The initial time of frequency, end time and persistent period, determine personnel to be detected described in the audio frequency text chunk
The quiet duration that correspondence audio frequency is included;Include in determining the audio frequency text chunk meet preset duration it is quiet when
It is long, and initial time and the termination time of the quiet duration generation for including.
Further, in this step, preset audio packet is containing keyword message, emotion detection information and quiet
One or more information in sound duration information, the index for characterizing emotion Testing index can be both call sides
One or more index in word speed information, amplitude information, frequency information, information volume, fundamental frequency information,
It can be seen that, can be according to actual needs using above- mentioned information as not when the Audio Information Retrieval of abnormal emotion is carried out
Same search condition carries out logical combination, for example:Comprising keyword A and the exception not comprising keyword B
Audio-frequency information of emotion etc..
S204, from each audio-frequency information section for constituting the audio-frequency information, determine abnormal emotion audio-frequency information section,
Wherein, the abnormal emotion audio-frequency information section is by including for characterizing personnel's exception to be detected
The preset audio information of emotion meets the pre-conditioned audio-frequency information section of correspondence.
Further, according to the result in S203, in this step, can be true by speech waveform analysis module
Make abnormal emotion audio-frequency information section.For characterizing the preset audio information of personnel's abnormal emotion to be detected
Can be above-mentioned preset audio information.
This step can be specifically embodied as:
The audio-frequency information section comprising preset keyword is determined from each audio-frequency information section;And/or
Determine that the desired value of the pre-set level for characterizing emotion Testing index reaches correspondence from each audio-frequency information section
The audio-frequency information section of metrics-thresholds;And/or
Determine that quiet duration meets the audio-frequency information section of preset duration from each audio-frequency information section.
S205, when the abnormal emotion audio-frequency information section is capable of determining that, determine that audio-frequency information section is corresponding
There is abnormal emotion in personnel to be detected.
Embodiment 2:
With embodiment 1 correspondingly, the embodiment of the present invention 2 provides a kind of intelligent emotion bootstrap technique, is implementing
In example 1, when determining that personnel to be detected have abnormal emotion, by personnel to be detected in prior art
Transmission makes laughs picture or joke guiding the abnormal emotion for analyze personnel to be detected, so cannot from it is basic must
To the source of personnel's abnormal emotion to be detected, cannot also treat testing staff carries out profound interactive guiding.Cause
This, the embodiment of the present invention 2 provides a kind of intelligent emotion bootstrap technique, is determining that it is different that personnel to be detected are present
During reason thread, testing staff is treated by way of intelligent interaction carries out profound interactive guiding, can be timely
It was found that the source of personnel's abnormal emotion to be detected, and alleviate the abnormal emotion of personnel to be detected well.
A kind of intelligent emotion bootstrap technique that the embodiment of the present invention 2 is provided, as shown in figure 3, including following step
Suddenly:
S301, the audio-frequency information for obtaining personnel to be detected and user's communication.
S302, the audio-frequency information obtained in S301 is converted to into structurized audio frequency text.
S303, the audio frequency text to obtaining in S302 are parsed, and determine each audio-frequency information Duan Yin
The preset audio information that corresponding each audio frequency text chunk is included respectively in frequency text,
Wherein, preset audio packet is containing at least one following information:Keyword message, emotion detection information
With quiet duration information;
S304, from constitute audio-frequency information each audio-frequency information section in, determine abnormal emotion audio-frequency information section,
Wherein, abnormal emotion audio-frequency information section is by including for characterizing the pre- of personnel's abnormal emotion to be detected
If audio-frequency information meets the pre-conditioned audio-frequency information section of correspondence.
Further, the specific embodiment of step S301~step S304 is walked in may refer to embodiment 1
Rapid S201~step S204.
S305, for each abnormal emotion audio-frequency information section, determine that the abnormal emotion audio-frequency information section is corresponding
Credibility,
Wherein, when in abnormal emotion audio-frequency information section, it is more that correspondence desired value meets pre-conditioned index,
Corresponding credibility is higher, pre-conditioned including following at least two condition:Comprising preset keyword, quiet
It is right that duration meets preset duration, the word speed information of both call sides reaches correspondence metrics-thresholds, amplitude information reaches
Answer metrics-thresholds, frequency information to reach correspondence metrics-thresholds, information volume and reach correspondence metrics-thresholds, fundamental frequency
Information reaches correspondence metrics-thresholds.
In this step, the corresponding credibility of abnormal emotion audio-frequency information section can be the abnormal emotion audio-frequency information
Segment table levies the credibility of abnormal emotion, then in each condition arranged for judgement abnormal emotion audio-frequency information section
In, the condition for meeting is more, and the audio-frequency information section is higher for the credibility of abnormal emotion audio-frequency information section.
Further, for determining system for intelligent emotion provided in an embodiment of the present invention, this step is held
Row main body can be early warning pushing module.Speech waveform analysis module determine abnormal emotion audio-frequency information section it
Afterwards, the corresponding audio frequency text chunk of abnormal emotion audio-frequency information section determined can be sent to early warning and pushes mould
Block, determines the credibility of each audio frequency text chunk for receiving by early warning pushing module.
S306, corresponding credibility is distinguished based on each abnormal emotion audio-frequency information section, from each abnormal emotion
In audio-frequency information section, it is determined that optimum abnormal emotion audio-frequency information section.
Further, for determining system for intelligent emotion provided in an embodiment of the present invention, this step is held
Row main body can be early warning pushing module.Early warning pushing module can be according to the abnormal emotion audio-frequency information for receiving
The corresponding audio frequency text chunk of section, determines optimum abnormal emotion audio-frequency information section, so as to subsequently according to the optimum
The determination of abnormal emotion audio-frequency information section carries out the content of intelligent AC with personnel to be detected.
This step can be specifically embodied as following steps:
In judging each abnormal emotion audio-frequency information section, if there is an abnormal emotion audio-frequency information section and cause this
The credibility highest of one abnormal emotion audio-frequency information section;
If existing, the abnormal emotion audio-frequency information section is defined as into optimum abnormal emotion audio-frequency information section;
Otherwise, if there are multiple abnormal emotion audio-frequency information sections causes the plurality of abnormal emotion audio-frequency information section
Credibility is equal and highest, then in judging the plurality of abnormal emotion audio-frequency information section, if there is an exception
Emotion audio-frequency information section causes the corresponding cumulative activation of time of origin of an abnormal emotion audio-frequency information section
Duration is most long;
If existing, the abnormal emotion audio-frequency information section is defined as into optimum abnormal emotion audio-frequency information section;
Otherwise, cause the plurality of abnormal emotion audio-frequency information section right if there are multiple abnormal emotion audio-frequency information sections
Answer cumulative activation duration equal and most long, then judge the time of origin of the plurality of abnormal emotion audio-frequency information section
In, the worst abnormal emotion audio-frequency information of the emotion that characterized of history degrees of emotion in the range of the correspondence time period
Section;
The abnormal emotion audio-frequency information section is defined as into optimum abnormal emotion audio-frequency information section.
Further, it is determined that during optimum abnormal emotion audio-frequency information section, first can according to credibility, if
Exist multiple audio-frequency information sections credibility is equal and highest, then can consider the tired of the plurality of audio-frequency information section
Meter operating time, the cumulative activation duration can start to correspondence audio frequency for the last working of personnel to be detected
The stream time of message segment time of origin, it is assumed that there are the first audio-frequency information section and the second audio-frequency information
Section, the two credibility is equal and highest, and the first audio-frequency information section time of origin is morning 9:00 to 9:30 (on
Work hours at noon are 9:00, then correspondence cumulative operation time is 0 to half an hour), the second audio-frequency information section
Time of origin is afternoon 4:00 to 4:20 (work hours in the afternoon be 1:00, then correspondence cumulative operation time
For 3 hours to 20 minutes 3 hours), then judge according to cumulative operation time, the second audio-frequency information section
For optimum audio-frequency information section, if the corresponding cumulative operation time of multiple audio-frequency information sections is also equal, then,
During the time of origin of multiple abnormal emotion audio-frequency information sections can be considered, history emotion in the range of the correspondence time period
The situation of the emotion characterized by rank, it is assumed that the first audio-frequency information section and the second audio-frequency information section, the two is credible
Spend equal and highest, and cumulative operation time is identical, and the time of origin of the first audio-frequency information section it is corresponding when
Between scope (9:00~10:00) time of origin correspondence of the history degrees of emotion less than the second audio-frequency information section in
Time range (4:00~6:00) history degrees of emotion in, then characterize the second worst sound of history emotion
Frequency message segment is optimum abnormal emotion audio-frequency information section.
S307, according in the optimum abnormal emotion audio-frequency information section determined in S306, correspondence desired value meets
The species of the abnormal emotion characterized by pre-conditioned index, determines the content of intelligent AC.
Further, the species of the abnormal emotion of different index characterizations is different, for example:Keyword index may
Personnel's abnormal emotion to be detected is characterized from user, long pointer may characterize personnel's exception to be detected when quiet
Emotion therefore, it can, according in optimum abnormal emotion audio-frequency information section, correspond to desired value from fatigue etc.
Meet the species of the abnormal emotion characterized by pre-conditioned index, determine the content of intelligent AC.
Further, for determining system for intelligent emotion provided in an embodiment of the present invention, this step is held
Row main body can be early warning pushing module.
S308, the platform being located to the personnel to be detected that there is abnormal emotion are sent with the intelligent AC for determining
Content intelligent AC dialog box, and receive the response message of personnel to be detected.
Further, for determining system for intelligent emotion provided in an embodiment of the present invention, this step is held
Row main body can be intelligent AC module, and early warning pushing module determines the content of intelligent AC, by intelligently handing over
Flow module is exchanged with personnel to be detected.
Intelligent AC dialogue can be realized by the language database enriched, by key search technology reality
The now input to employee to be detected carries out intelligent response.In language library, same keyword adopts a plurality of reply data
Correspondence, when intelligent response is carried out, can be from random selection output in a plurality of reply data.
The form of communication of intelligent AC module can include it is various, for example:Written communication pattern and speech exchange
Pattern, when being embodied as, according to the selection of contact staff, confirms AC mode, and perform this step it
Before, determine that employee to be detected is currently at the idle condition of inoperative, when being embodied as, can provide and exempt to beat
Button is disturbed, it is currently working condition or idle shape that employee to be detected can show itself by interruption-free button
State, in order to avoid during voice output, affect normal customer service subsequent duty.
Further, the corresponding intelligent AC dialogue open mode of method provided in an embodiment of the present invention is passive
Pattern, by emotion warning module automatic push, can also open aggressive mode, as needed by people to be detected
Member is autonomous to be selected to enter, to realize the self-oriented management of emotion.
Whether S309, each response message for receiving, include specifying in judging the response message
Keyword.
Further, the nominal key in this step can be with identical with the preset keyword in above-mentioned steps
Can be with difference.For each response message for receiving, can compare with nominal key.
Further, for determining system for intelligent emotion provided in an embodiment of the present invention, this step is held
Row main body can be sensitive language collection module.
S310, after intelligent AC terminates, the nominal key included according to the response message for receiving
The order of severity and number of times are that personnel to be detected determine this degrees of emotion.
During intelligent AC, system is likely to be received the response message that a plurality of personnel to be detected send, can
With the order of severity of nominal key that included according to a plurality of response message and number of times as personnel to be detected
Determine this degrees of emotion, using in next abnormal emotion determination process as history degrees of emotion.
Further, for determining system for intelligent emotion provided in an embodiment of the present invention, this step is held
Row main body can be sensitive language collection module.
Further, intelligent emotion provided in an embodiment of the present invention determines that system can also include exchange injection mould
Block, exchange injection module provide a kind of exchange injection way, that is to say, that, it is allowed to backstage authority user can
To choose whether to be injected during carrying out Intelligent dialogue with system in personnel to be detected, can after implant operation
Exchanged with carrying out recessiveness in real time with personnel to be detected by artificial replacement intelligent AC system, and by selecting to move back
Go out injection, automatically switch back into intelligent AC system.Personnel's unaware to be detected in whole communication process.
Further, during intelligent AC is carried out with personnel to be detected, when the response of personnel to be detected is
When needing the problem of routine processes (for example:The problems such as consulting date, time), correspondence program can be called
Problem is processed, and routine processes result is returned to into personnel to be detected.
Further, intelligent AC module, exchange injection module and sensitive language collection module coordinate work
The flow process of work can be shown in Fig. 4, as shown in figure 4, Fig. 4 is to carry out the process of intelligent AC, step with user
S401 can equivalent to the response message of personnel to be detected is received in step S308 the step of, step S402
Can be implemented by sensitive language collection module equivalent to step S309, step S403 and step S404,
Step S405 and step S408 can be implemented by sensitive language collection module, step S406 and step S407
Can be implemented by exchange injection module, step S409~step S414 can be implemented by intelligent AC module,
Wherein, the problem of routine processes is needed in step S409 for example:The problems such as consulting date, time.
Based on same inventive concept, the embodiment of the present invention additionally provides a kind of intelligent emotion and determines system, due to
To a kind of aforementioned intelligent emotion, the principle of these devices and system institute solve problem determines that method is similar, therefore should
The enforcement of device and system may refer to the enforcement of preceding method, repeats part and repeats no more.
A kind of intelligent emotion provided in an embodiment of the present invention determines one of system, as shown in figure 5, including:
Audio collection module 501, for obtaining the audio-frequency information of personnel to be detected and user's communication;
Speech waveform analysis module 502, for from each audio-frequency information section for constituting the audio-frequency information, really
Determine abnormal emotion audio-frequency information section, wherein, the abnormal emotion audio-frequency information section is by including for characterizing
The preset audio information of personnel's abnormal emotion to be detected meets the pre-conditioned audio-frequency information section of correspondence;With
And when the abnormal emotion audio-frequency information section is capable of determining that, determine that the audio-frequency information section is corresponding to be checked
There is abnormal emotion in survey personnel.
Further, the audio collection module 501, is additionally operable in the speech waveform analysis module 502
From each audio-frequency information section for constituting the audio-frequency information, before determining abnormal emotion audio-frequency information section, by institute
State audio-frequency information and be converted to structurized audio frequency text;And the audio frequency text is solved
Analysis, determines that each audio-frequency information section distinguishes corresponding each audio frequency text chunk bag in the audio frequency text
The preset audio information for containing, wherein, the preset audio packet is containing at least one following information:Keyword
Information, emotion detection information and quiet duration information;
The speech waveform analysis module 502, specifically for determining comprising pre- from each audio-frequency information section
If the audio-frequency information section of keyword;And/or sign emotion Testing index is determined from each audio-frequency information section
The desired value of pre-set level reaches the audio-frequency information section of correspondence metrics-thresholds;And/or from each audio-frequency information section
The middle audio-frequency information section for determining that quiet duration meets preset duration.
Further, the audio collection module 501, specifically for when the preset audio packet is containing pass
During key word information, for each the audio-frequency information section in each audio-frequency information section, will be the audio-frequency information section corresponding
Audio frequency text chunk is compared with preset keyword;The preset keyword included in determining the audio frequency text chunk,
And initial time and the termination time of preset keyword appearance;
When the preset audio packet detection information containing emotion, for each sound in each audio-frequency information section
Frequency message segment, determine characterize in the corresponding audio frequency text chunk of the audio-frequency information section emotion Testing index as next
The desired value of individual or multiple indexs:The word speed information of both call sides, amplitude information, frequency information, volume letter
Breath, fundamental frequency information;In determining the audio frequency text chunk, desired value reaches the These parameters for corresponding to metrics-thresholds, with
And the desired value of These parameters reaches the initial time occurred in the audio frequency text chunk during correspondence metrics-thresholds
With the time of termination;
When the preset audio packet contains quiet duration information, for each sound in each audio-frequency information section
Frequency message segment, according to per words correspondence audio frequency of both call sides in the corresponding audio frequency text chunk of the audio-frequency information section
Initial time, end time and persistent period, determine personnel's correspondence to be detected described in the audio frequency text chunk
The quiet duration that audio pack contains;What is included in determining the audio frequency text chunk meets the quiet duration of preset duration,
And initial time and the termination time of the quiet duration generation for including.
Further, the system also includes:Early warning pushing module 503;
The early warning pushing module 503, in the speech waveform analysis module 502, described in constituting
In each audio-frequency information section of audio-frequency information, after determining abnormal emotion audio-frequency information section, for each abnormal feelings
Thread audio-frequency information section, determines the corresponding credibility of abnormal emotion audio-frequency information section, wherein, when the exception
In emotion audio-frequency information section, correspondence desired value meets that pre-conditioned index is more, and corresponding credibility is higher,
It is described pre-conditioned including following at least two condition:When meeting default comprising preset keyword, quiet duration
Long, the word speed information of both call sides reaches correspondence metrics-thresholds, amplitude information and reaches correspondence metrics-thresholds, frequency
Rate information reaches correspondence metrics-thresholds, information volume and reaches correspondence metrics-thresholds, fundamental frequency information and reaches and correspondingly refer to
Mark threshold value;And corresponding credibility is distinguished based on each abnormal emotion audio-frequency information section, from the different reason
In thread audio-frequency information section, it is determined that optimum abnormal emotion audio-frequency information section.
Further, the early warning pushing module 503, specifically for judging each abnormal emotion audio frequency letter
In breath section, if there is an abnormal emotion audio-frequency information section and cause an abnormal emotion audio-frequency information section
Credibility highest;If existing, the abnormal emotion audio-frequency information section is defined as into optimum abnormal emotion audio frequency letter
Breath section;Otherwise, if there are multiple abnormal emotion audio-frequency information sections causes the plurality of abnormal emotion audio-frequency information section
Credibility is equal and highest, then in judging the plurality of abnormal emotion audio-frequency information section, if exist one it is different
Reason thread audio-frequency information section causes the corresponding accumulative work of the time of origin of an abnormal emotion audio-frequency information section
Make duration most long;If existing, the abnormal emotion audio-frequency information section is defined as into optimum abnormal emotion audio frequency letter
Breath section;Otherwise, if there are multiple abnormal emotion audio-frequency information sections causes the plurality of abnormal emotion audio-frequency information section
Correspondence cumulative activation duration is equal and most long, then judge the time of origin of the plurality of abnormal emotion audio-frequency information section
In, the worst abnormal emotion audio-frequency information of the emotion that characterized of history degrees of emotion in the range of the correspondence time period
Section;The abnormal emotion audio-frequency information section is defined as into optimum abnormal emotion audio-frequency information section.
Further, the system also includes:Intelligent AC module 504 and sensitive language collection module 505;
The early warning pushing module 503, is additionally operable to it is determined that after optimum abnormal emotion audio-frequency information section, root
According in the optimum abnormal emotion audio-frequency information section, correspondence desired value meets the pre-conditioned index institute table
The species of the abnormal emotion levied, determines the content of intelligent AC;
The intelligent AC module 504, it is flat for what is be located to the personnel to be detected that there is abnormal emotion
Platform sends the intelligent AC dialog box of the content with the intelligent AC for determining, and receives the personnel to be detected
Response message;
The sensitive language collection module 505, for each response message for receiving, judging should
Whether nominal key is included in bar response message;After intelligent AC terminates, disappeared according to the response for receiving
The order of severity and number of times of the nominal key that breath includes is that the personnel to be detected determine this emotion
Rank.
A kind of intelligent emotion provided in an embodiment of the present invention determines the two of system, as shown in fig. 6, carrying with Fig. 5
For a kind of intelligent emotion determine that one of system is compared, also include:Exchange injection module 601, the exchange
Injection module 601, for during intelligent AC, when selecting to start injection, receives backstage and has
The voice of authority user carries out real-time, interactive with personnel to be detected, when not selecting or exiting injection, by intelligence
Energy AC system carries out real-time, interactive with personnel to be detected.
The function of above-mentioned each unit may correspond to the respective handling step in flow process shown in Fig. 1 to Fig. 4, here
Repeat no more.
A kind of intelligent emotion provided in an embodiment of the present invention determines method and system, obtains personnel to be detected and uses
The audio-frequency information of family call;From each audio-frequency information section for constituting audio-frequency information, determine that abnormal emotion audio frequency is believed
Breath section, wherein, abnormal emotion audio-frequency information section is by including for characterizing personnel's abnormal emotion to be detected
Preset audio information meets the pre-conditioned audio-frequency information section of correspondence;When be capable of determining that abnormal emotion audio frequency believe
During breath section, determine that the corresponding personnel to be detected of audio-frequency information section have abnormal emotion.The embodiment of the present invention is provided
Intelligent emotion determine method, by obtaining the audio-frequency information of personnel, talking to be detected and audio-frequency information being carried out
Analyze to determine people to be detected with the presence or absence of abnormal emotion, as the language of people tends to embody current feelings
Thread, therefore audio-frequency information can objectively embody the current emotional state of personnel to be detected, in prior art
The acquisition of abnormal emotion data is depended on contact staff oneself typing and professional equipment to physiological data
Collection information is compared, more objective, and the personnel to be detected that so analysis is obtained whether there is abnormal emotion more
Accurately.And the acquisition of abnormal emotion data needs to complete by professional external equipment auxiliary in prior art,
Increased cost.
Through the above description of the embodiments, those skilled in the art can be understood that the present invention
Embodiment can be realized by hardware, it is also possible to come real by the mode of software plus necessary general hardware platform
It is existing.Based on such understanding, the technical scheme of the embodiment of the present invention can be embodied in the form of software product
Come, the software product can be stored in a non-volatile memory medium (can be CD-ROM, USB flash disk,
Portable hard drive etc.) in, use including some instructions so that computer equipment (can be personal computer,
Server, or the network equipment etc.) perform method described in each embodiment of the invention.
It will be appreciated by those skilled in the art that accompanying drawing is the schematic diagram of a preferred embodiment, the mould in accompanying drawing
Block or flow process are not necessarily implemented necessary to the present invention.
It will be appreciated by those skilled in the art that the module in device in embodiment can be described according to embodiment
Carry out being distributed in the device of embodiment, it is also possible to carry out that respective change is disposed other than the present embodiment one
Or in multiple devices.The module of above-described embodiment can merge into a module, it is also possible to be further split into
Multiple submodule.
The embodiments of the present invention are for illustration only, do not represent the quality of embodiment.
Obviously, those skilled in the art can carry out various changes and modification without deviating from the present invention's to the present invention
Spirit and scope.So, if these modifications of the present invention and modification belong to the claims in the present invention and its wait
Within the scope of technology, then the present invention is also intended to comprising these changes and modification.
Claims (12)
1. a kind of intelligent emotion determines method, it is characterised in that include:
Obtain the audio-frequency information of personnel to be detected and user's communication;
From each audio-frequency information section for constituting the audio-frequency information, abnormal emotion audio-frequency information section is determined,
Wherein, the abnormal emotion audio-frequency information section is by including for characterizing personnel's exception to be detected
The preset audio information of emotion meets the pre-conditioned audio-frequency information section of correspondence;
When the abnormal emotion audio-frequency information section is capable of determining that, determines that the audio-frequency information section is corresponding and treat
There is abnormal emotion in testing staff.
2. the method for claim 1, it is characterised in that from constituting each of the audio-frequency information
In audio-frequency information section, before determining abnormal emotion audio-frequency information section, also include:
The audio-frequency information is converted to into structurized audio frequency text;
The audio frequency text is parsed, determines each audio-frequency information section in audio frequency text text
The preset audio information that corresponding each audio frequency text chunk is included respectively in part, wherein, the preset audio information
Comprising at least one following information:Keyword message, emotion detection information and quiet duration information;
From each audio-frequency information section for constituting the audio-frequency information, abnormal emotion audio-frequency information section is determined, specifically
Including:
The audio-frequency information section comprising preset keyword is determined from each audio-frequency information section;And/or
Determine that the desired value of the pre-set level for characterizing emotion Testing index reaches from each audio-frequency information section
The audio-frequency information section of correspondence metrics-thresholds;And/or
Determine that quiet duration meets the audio-frequency information section of preset duration from each audio-frequency information section.
3. method as claimed in claim 2, it is characterised in that the audio frequency text is solved
Analysis, determines that each audio-frequency information section distinguishes corresponding each audio frequency text chunk bag in the audio frequency text
The preset audio information for containing, specifically includes:
When preset audio packet when containing keyword message, for each audio frequency in each audio-frequency information section
Message segment, the audio-frequency information section corresponding audio frequency text chunk is compared with preset keyword;Determine the sound
The preset keyword included in frequency text chunk, and initial time and the termination time of preset keyword appearance;
When the preset audio packet detection information containing emotion, for each sound in each audio-frequency information section
Frequency message segment, determine characterize in the corresponding audio frequency text chunk of the audio-frequency information section emotion Testing index as next
The desired value of individual or multiple indexs:The word speed information of both call sides, amplitude information, frequency information, volume letter
Breath, fundamental frequency information;In determining the audio frequency text chunk, desired value reaches the These parameters for corresponding to metrics-thresholds, with
And the desired value of These parameters reaches the initial time occurred in the audio frequency text chunk during correspondence metrics-thresholds
With the time of termination;
When the preset audio packet contains quiet duration information, for each sound in each audio-frequency information section
Frequency message segment, according to per words correspondence audio frequency of both call sides in the corresponding audio frequency text chunk of the audio-frequency information section
Initial time, end time and persistent period, determine personnel's correspondence to be detected described in the audio frequency text chunk
The quiet duration that audio pack contains;What is included in determining the audio frequency text chunk meets the quiet duration of preset duration,
And initial time and the termination time of the quiet duration generation for including.
4. the method as described in any one of claim 1-3, it is characterised in that from constituting the audio frequency
In each audio-frequency information section of information, after determining abnormal emotion audio-frequency information section, also include:
For each abnormal emotion audio-frequency information section, determine that the abnormal emotion audio-frequency information section is corresponding credible
Degree,
Wherein, when, in the abnormal emotion audio-frequency information section, correspondence desired value meets pre-conditioned index and gets over
Many, corresponding credibility is higher, described pre-conditioned including following at least two condition:Comprising preset critical
Word, quiet duration meet preset duration, the word speed information of both call sides reaches correspondence metrics-thresholds, amplitude letter
Breath reaches correspondence metrics-thresholds, frequency information and reaches correspondence metrics-thresholds, information volume and reaches correspondence index threshold
Value, fundamental frequency information reach correspondence metrics-thresholds;
Corresponding credibility is distinguished based on each abnormal emotion audio-frequency information section, is believed from each abnormal emotion audio frequency
In breath section, it is determined that optimum abnormal emotion audio-frequency information section.
5. method as claimed in claim 4, it is characterised in that based on each abnormal emotion audio-frequency information section
The corresponding credibility of difference, from each abnormal emotion audio-frequency information section, it is determined that optimum abnormal emotion audio frequency
Message segment, specifically includes:
In judging each abnormal emotion audio-frequency information section, if there is an abnormal emotion audio-frequency information section makes
Obtain the credibility highest of an abnormal emotion audio-frequency information section;
If existing, the abnormal emotion audio-frequency information section is defined as into optimum abnormal emotion audio-frequency information section;
Otherwise, if there are multiple abnormal emotion audio-frequency information sections causes the plurality of abnormal emotion audio-frequency information section
Credibility is equal and highest, then in judging the plurality of abnormal emotion audio-frequency information section, if there is an exception
Emotion audio-frequency information section causes the corresponding cumulative activation of time of origin of an abnormal emotion audio-frequency information section
Duration is most long;
If existing, the abnormal emotion audio-frequency information section is defined as into optimum abnormal emotion audio-frequency information section;
Otherwise, cause the plurality of abnormal emotion audio-frequency information section right if there are multiple abnormal emotion audio-frequency information sections
Answer cumulative activation duration equal and most long, then judge the time of origin of the plurality of abnormal emotion audio-frequency information section
In, the worst abnormal emotion audio-frequency information of the emotion that characterized of history degrees of emotion in the range of the correspondence time period
Section;
The abnormal emotion audio-frequency information section is defined as into optimum abnormal emotion audio-frequency information section.
6. method as claimed in claim 4, it is characterised in that it is determined that optimum abnormal emotion audio frequency letter
After breath section, also include:
According in the optimum abnormal emotion audio-frequency information section, correspondence desired value meets the pre-conditioned finger
The species of the characterized abnormal emotion of mark, determines the content of intelligent AC;
The platform being located to the personnel to be detected that there is abnormal emotion is sent with the intelligent AC for determining
Content intelligent AC dialog box, and receive the response message of the personnel to be detected;
For each response message for receiving, in judging the response message, whether include designated key
Word;
After intelligent AC terminates, the serious journey of the nominal key included according to the response message for receiving
Degree and number of times are that the personnel to be detected determine this degrees of emotion.
7. a kind of intelligent emotion determines system, it is characterised in that include:
Audio collection module, for obtaining the audio-frequency information of personnel to be detected and user's communication;
Speech waveform analysis module, for, from each audio-frequency information section for constituting the audio-frequency information, determining different
Reason thread audio-frequency information section, wherein, the abnormal emotion audio-frequency information section is described for characterizing by including
The preset audio information of personnel's abnormal emotion to be detected meets the pre-conditioned audio-frequency information section of correspondence;And work as
When being capable of determining that the abnormal emotion audio-frequency information section, the corresponding people to be detected of the audio-frequency information section is determined
There is abnormal emotion in member.
8. system as claimed in claim 7, it is characterised in that the audio collection module, is additionally operable to
In the speech waveform analysis module from each audio-frequency information section for constituting the audio-frequency information, it is determined that abnormal feelings
Before thread audio-frequency information section, the audio-frequency information is converted to into structurized audio frequency text;And to institute
State audio frequency text to be parsed, determine each audio-frequency information section in the audio frequency text respectively
The preset audio information that corresponding each audio frequency text chunk is included, wherein, the preset audio packet is containing at least
A kind of following information:Keyword message, emotion detection information and quiet duration information;
The speech waveform analysis module, specifically for determining from each audio-frequency information section comprising default pass
The audio-frequency information section of key word;And/or the default of sign emotion Testing index is determined from each audio-frequency information section
The desired value of index reaches the audio-frequency information section of correspondence metrics-thresholds;And/or from each audio-frequency information section really
Fixed quiet duration meets the audio-frequency information section of preset duration.
9. system as claimed in claim 8, it is characterised in that the audio collection module is concrete to use
In when preset audio packet when containing keyword message, believe for each audio frequency in each audio-frequency information section
Breath section, the audio-frequency information section corresponding audio frequency text chunk is compared with preset keyword;Determine the audio frequency
The preset keyword included in text chunk, and initial time and the termination time of preset keyword appearance;
When the preset audio packet detection information containing emotion, for each sound in each audio-frequency information section
Frequency message segment, determine characterize in the corresponding audio frequency text chunk of the audio-frequency information section emotion Testing index as next
The desired value of individual or multiple indexs:The word speed information of both call sides, amplitude information, frequency information, volume letter
Breath, fundamental frequency information;In determining the audio frequency text chunk, desired value reaches the These parameters for corresponding to metrics-thresholds, with
And the desired value of These parameters reaches the initial time occurred in the audio frequency text chunk during correspondence metrics-thresholds
With the time of termination;
When the preset audio packet contains quiet duration information, for each sound in each audio-frequency information section
Frequency message segment, according to per words correspondence audio frequency of both call sides in the corresponding audio frequency text chunk of the audio-frequency information section
Initial time, end time and persistent period, determine personnel's correspondence to be detected described in the audio frequency text chunk
The quiet duration that audio pack contains;What is included in determining the audio frequency text chunk meets the quiet duration of preset duration,
And initial time and the termination time of the quiet duration generation for including.
10. the system as described in any one of claim 7-9, it is characterised in that also include:Early warning is pushed
Module;
The early warning pushing module, in the speech waveform analysis module, from the composition audio-frequency information
Each audio-frequency information section in, after determining abnormal emotion audio-frequency information section, believe for each abnormal emotion audio frequency
Breath section, determines the corresponding credibility of abnormal emotion audio-frequency information section, wherein, when the abnormal emotion audio frequency
In message segment, correspondence desired value meets that pre-conditioned index is more, and corresponding credibility is higher, described pre-
If condition includes following at least two condition:Meet preset duration, lead to comprising preset keyword, quiet duration
The word speed information of words both sides reaches correspondence metrics-thresholds, amplitude information and reaches correspondence metrics-thresholds, frequency information
Reach correspondence metrics-thresholds, information volume and reach correspondence metrics-thresholds, fundamental frequency information and reach correspondence metrics-thresholds;
And corresponding credibility is distinguished based on each abnormal emotion audio-frequency information section, believe from each abnormal emotion audio frequency
In breath section, it is determined that optimum abnormal emotion audio-frequency information section.
11. systems as claimed in claim 10, it is characterised in that the early warning pushing module, specifically
For judging in each abnormal emotion audio-frequency information section, if there is an abnormal emotion audio-frequency information section makes
Obtain the credibility highest of an abnormal emotion audio-frequency information section;If existing, the abnormal emotion audio frequency is believed
Breath section is defined as optimum abnormal emotion audio-frequency information section;Otherwise, if there are multiple abnormal emotion audio-frequency information sections
So that the credibility of the plurality of abnormal emotion audio-frequency information section is equal and highest, then the plurality of abnormal emotion is judged
In audio-frequency information section, if there is an abnormal emotion audio-frequency information section and cause an abnormal emotion audio frequency letter
The corresponding cumulative activation duration of time of origin of breath section is most long;If existing, by the abnormal emotion audio-frequency information
Section is defined as optimum abnormal emotion audio-frequency information section;Otherwise, make if there are multiple abnormal emotion audio-frequency information sections
Obtain the plurality of abnormal emotion audio-frequency information section correspondence cumulative activation duration equal and most long, then judge the plurality of different
In the time of origin of reason thread audio-frequency information section, in the range of the correspondence time period feelings characterized by history degrees of emotion
The worst abnormal emotion audio-frequency information section of thread;The abnormal emotion audio-frequency information section is defined as into optimum abnormal emotion
Audio-frequency information section.
12. systems as claimed in claim 10, it is characterised in that also include:Intelligent AC module and
Sensitive language collection module;
The early warning pushing module, is additionally operable to it is determined that after optimum abnormal emotion audio-frequency information section, according to institute
State in optimum abnormal emotion audio-frequency information section, correspondence desired value meets what the pre-conditioned index was characterized
The species of abnormal emotion, determines the content of intelligent AC;
The intelligent AC module, the platform for being located to the personnel to be detected that there is abnormal emotion are sent out
The intelligent AC dialog box of the content with the intelligent AC for determining is sent, and receives answering for the personnel to be detected
Answer message;
The sensitive language collection module, for each response message for receiving, judges that this should
Whether nominal key is included in answering message;After intelligent AC terminates, according in the response message for receiving
Including nominal key the order of severity and number of times be that the personnel to be detected determine this degrees of emotion.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510613689.XA CN106548788B (en) | 2015-09-23 | 2015-09-23 | Intelligent emotion determining method and system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510613689.XA CN106548788B (en) | 2015-09-23 | 2015-09-23 | Intelligent emotion determining method and system |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106548788A true CN106548788A (en) | 2017-03-29 |
CN106548788B CN106548788B (en) | 2020-01-07 |
Family
ID=58365640
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510613689.XA Active CN106548788B (en) | 2015-09-23 | 2015-09-23 | Intelligent emotion determining method and system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106548788B (en) |
Cited By (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107293309A (en) * | 2017-05-19 | 2017-10-24 | 四川新网银行股份有限公司 | A kind of method that lifting public sentiment monitoring efficiency is analyzed based on customer anger |
CN107609736A (en) * | 2017-08-09 | 2018-01-19 | 广州思涵信息科技有限公司 | A kind of teaching diagnostic analysis system and method for integrated application artificial intelligence technology |
CN108648768A (en) * | 2018-04-16 | 2018-10-12 | 广州市菲玛尔咨询服务有限公司 | A kind of consulting recommendation method and its management system |
CN108735233A (en) * | 2017-04-24 | 2018-11-02 | 北京理工大学 | A kind of personality recognition methods and device |
CN109087670A (en) * | 2018-08-30 | 2018-12-25 | 西安闻泰电子科技有限公司 | Mood analysis method, system, server and storage medium |
CN109145101A (en) * | 2018-09-06 | 2019-01-04 | 北京京东尚科信息技术有限公司 | Interactive method, device and computer readable storage medium |
CN109587360A (en) * | 2018-11-12 | 2019-04-05 | 平安科技(深圳)有限公司 | Electronic device should talk with art recommended method and computer readable storage medium |
CN109785123A (en) * | 2019-01-21 | 2019-05-21 | 中国平安财产保险股份有限公司 | A kind of business handling assisted method, device and terminal device |
CN110393539A (en) * | 2019-06-21 | 2019-11-01 | 合肥工业大学 | Psychological abnormality detection method, device, storage medium and electronic equipment |
CN111599379A (en) * | 2020-05-09 | 2020-08-28 | 北京南师信息技术有限公司 | Conflict early warning method, device, equipment, readable storage medium and triage system |
CN112235468A (en) * | 2020-10-16 | 2021-01-15 | 绍兴市寅川软件开发有限公司 | Audio processing method and system for voice customer service evaluation |
CN113053385A (en) * | 2021-03-30 | 2021-06-29 | 中国工商银行股份有限公司 | Abnormal emotion detection method and device |
CN113222458A (en) * | 2021-05-31 | 2021-08-06 | 上海工程技术大学 | Urban rail transit driver safety risk assessment model and system |
CN113515636A (en) * | 2021-09-13 | 2021-10-19 | 阿里健康科技(中国)有限公司 | Text data processing method and electronic equipment |
CN114693319A (en) * | 2022-04-21 | 2022-07-01 | 广州美保科技有限公司 | Customer service quality management improving system and method based on artificial intelligence |
CN115547501A (en) * | 2022-11-24 | 2022-12-30 | 国能大渡河大数据服务有限公司 | Employee emotion perception method and system combining working characteristics |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102623009A (en) * | 2012-03-02 | 2012-08-01 | 安徽科大讯飞信息技术股份有限公司 | Abnormal emotion automatic detection and extraction method and system on basis of short-time analysis |
CN102625005A (en) * | 2012-03-05 | 2012-08-01 | 广东天波信息技术股份有限公司 | Call center system with function of real-timely monitoring service quality and implement method of call center system |
CN102831891A (en) * | 2011-06-13 | 2012-12-19 | 富士通株式会社 | Processing method and system for voice data |
CN103491251A (en) * | 2013-09-24 | 2014-01-01 | 深圳市金立通信设备有限公司 | Method and terminal for monitoring user calls |
CN103634472A (en) * | 2013-12-06 | 2014-03-12 | 惠州Tcl移动通信有限公司 | Method, system and mobile phone for judging mood and character of user according to call voice |
CN104036776A (en) * | 2014-05-22 | 2014-09-10 | 毛峡 | Speech emotion identification method applied to mobile terminal |
-
2015
- 2015-09-23 CN CN201510613689.XA patent/CN106548788B/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102831891A (en) * | 2011-06-13 | 2012-12-19 | 富士通株式会社 | Processing method and system for voice data |
CN102623009A (en) * | 2012-03-02 | 2012-08-01 | 安徽科大讯飞信息技术股份有限公司 | Abnormal emotion automatic detection and extraction method and system on basis of short-time analysis |
CN102625005A (en) * | 2012-03-05 | 2012-08-01 | 广东天波信息技术股份有限公司 | Call center system with function of real-timely monitoring service quality and implement method of call center system |
CN103491251A (en) * | 2013-09-24 | 2014-01-01 | 深圳市金立通信设备有限公司 | Method and terminal for monitoring user calls |
CN103634472A (en) * | 2013-12-06 | 2014-03-12 | 惠州Tcl移动通信有限公司 | Method, system and mobile phone for judging mood and character of user according to call voice |
CN104036776A (en) * | 2014-05-22 | 2014-09-10 | 毛峡 | Speech emotion identification method applied to mobile terminal |
Cited By (19)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108735233A (en) * | 2017-04-24 | 2018-11-02 | 北京理工大学 | A kind of personality recognition methods and device |
CN107293309A (en) * | 2017-05-19 | 2017-10-24 | 四川新网银行股份有限公司 | A kind of method that lifting public sentiment monitoring efficiency is analyzed based on customer anger |
CN107609736A (en) * | 2017-08-09 | 2018-01-19 | 广州思涵信息科技有限公司 | A kind of teaching diagnostic analysis system and method for integrated application artificial intelligence technology |
CN108648768A (en) * | 2018-04-16 | 2018-10-12 | 广州市菲玛尔咨询服务有限公司 | A kind of consulting recommendation method and its management system |
CN109087670A (en) * | 2018-08-30 | 2018-12-25 | 西安闻泰电子科技有限公司 | Mood analysis method, system, server and storage medium |
CN109145101A (en) * | 2018-09-06 | 2019-01-04 | 北京京东尚科信息技术有限公司 | Interactive method, device and computer readable storage medium |
CN109587360A (en) * | 2018-11-12 | 2019-04-05 | 平安科技(深圳)有限公司 | Electronic device should talk with art recommended method and computer readable storage medium |
CN109785123A (en) * | 2019-01-21 | 2019-05-21 | 中国平安财产保险股份有限公司 | A kind of business handling assisted method, device and terminal device |
CN110393539A (en) * | 2019-06-21 | 2019-11-01 | 合肥工业大学 | Psychological abnormality detection method, device, storage medium and electronic equipment |
CN110393539B (en) * | 2019-06-21 | 2021-11-23 | 合肥工业大学 | Psychological anomaly detection method and device, storage medium and electronic equipment |
CN111599379A (en) * | 2020-05-09 | 2020-08-28 | 北京南师信息技术有限公司 | Conflict early warning method, device, equipment, readable storage medium and triage system |
CN111599379B (en) * | 2020-05-09 | 2023-09-29 | 北京南师信息技术有限公司 | Conflict early warning method, device, equipment, readable storage medium and triage system |
CN112235468A (en) * | 2020-10-16 | 2021-01-15 | 绍兴市寅川软件开发有限公司 | Audio processing method and system for voice customer service evaluation |
CN113053385A (en) * | 2021-03-30 | 2021-06-29 | 中国工商银行股份有限公司 | Abnormal emotion detection method and device |
CN113222458A (en) * | 2021-05-31 | 2021-08-06 | 上海工程技术大学 | Urban rail transit driver safety risk assessment model and system |
CN113515636A (en) * | 2021-09-13 | 2021-10-19 | 阿里健康科技(中国)有限公司 | Text data processing method and electronic equipment |
CN114693319A (en) * | 2022-04-21 | 2022-07-01 | 广州美保科技有限公司 | Customer service quality management improving system and method based on artificial intelligence |
CN115547501A (en) * | 2022-11-24 | 2022-12-30 | 国能大渡河大数据服务有限公司 | Employee emotion perception method and system combining working characteristics |
CN115547501B (en) * | 2022-11-24 | 2023-04-07 | 国能大渡河大数据服务有限公司 | Employee emotion perception method and system combining working characteristics |
Also Published As
Publication number | Publication date |
---|---|
CN106548788B (en) | 2020-01-07 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106548788A (en) | A kind of intelligent emotion determines method and system | |
Low et al. | Automated assessment of psychiatric disorders using speech: A systematic review | |
US8145474B1 (en) | Computer mediated natural language based communication augmented by arbitrary and flexibly assigned personality classification systems | |
Ivanov et al. | Recognition of personality traits from human spoken conversations | |
Klaylat et al. | Emotion recognition in Arabic speech | |
US20190253558A1 (en) | System and method to automatically monitor service level agreement compliance in call centers | |
CN114503115A (en) | Generating rich action items | |
US8694307B2 (en) | Method and apparatus for temporal speech scoring | |
US8687792B2 (en) | System and method for dialog management within a call handling system | |
US11461863B2 (en) | Idea assessment and landscape mapping | |
Burgoon et al. | Choosing between micro and macro nonverbal measurement: Application to selected vocalic and kinesic indices | |
JP2021036292A (en) | Information processing program, information processing method, and information processing device | |
Valenti et al. | Using topic modeling to infer the emotional state of people living with Parkinson’s disease | |
Huber et al. | Automatically analyzing brainstorming language behavior with Meeter | |
US10602974B1 (en) | Detection and management of memory impairment | |
Irvine et al. | Rewarding chatbots for real-world engagement with millions of users | |
CN114242109A (en) | Intelligent outbound method and device based on emotion recognition, electronic equipment and medium | |
Jordanous et al. | What makes a musical improvisation creative? | |
Khawaja et al. | Think before you talk: An empirical study of relationship between speech pauses and cognitive load | |
Zhou et al. | The role of different types of conversations for meeting success | |
Guha et al. | A sentiment analysis of the PhD experience evidenced on Twitter | |
JP2012242528A (en) | Talk evaluation device, method and program | |
KR20230047104A (en) | Digital Devices and Applications for the Treatment of Social Communication Disorders | |
Züger et al. | Sensing and supporting software developers' focus | |
Tran et al. | VPASS: Voice Privacy Assistant System for Monitoring In-home Voice Commands |
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 |