CN105022929A - Cognition accuracy analysis method for personality trait value test - Google Patents
Cognition accuracy analysis method for personality trait value test Download PDFInfo
- Publication number
- CN105022929A CN105022929A CN201510482599.1A CN201510482599A CN105022929A CN 105022929 A CN105022929 A CN 105022929A CN 201510482599 A CN201510482599 A CN 201510482599A CN 105022929 A CN105022929 A CN 105022929A
- Authority
- CN
- China
- Prior art keywords
- cognitive
- parameter
- tester
- information
- test
- 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
Landscapes
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
The invention relates to a cognition accuracy analysis method for a personality trait value test. The method is characterized by comprising the steps of building at least one cognitive dimension of sample model on the basis of at least one statistic model; stimulating a to-be-tested person in a manner of presenting a scene comprising at least one piece of stimulation information; monitoring or collecting physiological information and/or feedback information of the to-be-tested person to carry out statistics and form at least one cognitive dimension of cognition parameter; and comparing the cognition parameter with cognition parameter distribution of the sample model so as to obtain the accuracy of the cognitive dimension. The analysis method overcomes the defects of single test material and environment influence errors, improves the quality of the personality trait value test, and can obtain an accurate personality trait value.
Description
Technical field
The present invention relates to technical field of measurement and test, particularly relate to the cognitive accuracy analysis method of a kind of personal traits value test.
Background technology
Personal traits refers in the factor of composition personality, can cause the behavior of people's behavior and positive guide people, and makes different types of stimulation faced by individual can make the mental structure of identical reflection.On current internet, personality Forecasting Methodology generally adopts the form based on word test paper.Although the personality Forecasting Methodology of word test paper has had abundant achievement in research, as five-factor model personality test (Big Five), the personality test of Cartel 16 factor (Sixteen Personality Factor Questionnaire, 16PF) etc.But user needs the cost plenty of time to carry out answer in this manner, prediction required time depends on the answer speed of exercise question quantity and tester, and prediction steps is various tediously long, tester easily produces psychology of being sick of and conflicting, the accuracy of test result depends on the subjective cooperate degree of tester, and therefore this method testing error is relatively large.
Patent document (CN 103440864A) discloses a kind of voice-based personality characteristics Forecasting Methodology, by setting up voice personality prediction machine learning model in advance, any sound bite utilizing user to provide uses voice personality prediction machine learning model to realize personality characteristics prediction, set up the mapping relations between phonetic feature and personality characteristics factor by statistical learning method, dope every people's grid index.Although length consuming time, effect that this patent overcomes word test paper affect by subjective factor, measures material be not easy to the shortcomings such as acquisition, only use sound materials predicts that the technological means of personality is still dull.Easily by the physiological effect of user and the impact of living environment, there is deviation when reflecting personal traits, differing and reflecting real personal traits surely in phonetic feature.When the living environment of user is noisy, the daily voice of user are always in high-decibel, and when brief term is accustomed to, its real personality is easily ignored by personality prediction machine learning model.Therefore, the test material of this patent is single, the test error that the impact can not getting rid of living environment causes.
Summary of the invention
For the deficiency of prior art, the invention provides the cognitive accuracy analysis method of a kind of personal traits value test, it is characterized in that, described method comprises:
The sample pattern of at least one cognitive dimension is set up based at least one statistical model;
Described tester is stimulated to present the sight mode comprising at least one stimulus information;
Monitoring or gather the physiologic information of described tester and/or feedback information thus statistics forms the cognitive parameter of at least one cognitive dimension;
The cognitive parameter distribution of more described cognitive parameter and described sample pattern thus obtain the accuracy of described cognitive dimension.
According to a preferred implementation, the step of the test model of at least one cognitive dimension of described formation comprises:
The physiologic information parameter of tester described in typing and/or feedback information parameter;
Extract the eigenwert in described physiologic information parameter and/or feedback information parameter;
Sample pattern based on correspondence is added up with each cognitive dimension characteristic of correspondence value thus is formed cognitive parameter.
According to a preferred implementation, the cognitive parameter distribution Region dividing in described sample pattern is at least two the cognitive hierarchical regions can distinguishing cognitive accuracy.
According to a preferred implementation, described method also comprises: based on the presentation speed of the reaction time adjustment stimulus information of described tester, or,
Reaction time based on described tester assesses the conscientiously degree of described tester.
According to a preferred implementation, described method also comprises: the order that the psychological condition adjustment test sight obtained based on the physiologic information by described tester presents, or the order selecting test sight to present according to personal interest by described tester.
According to a preferred implementation, described test sight comprises virtual reality sight and holographic sight, and wherein holographic sight comprises dynamic holographic sight and static holographic sight.
According to a preferred implementation, described stimulus information comprises the stimulus information of detecting a lie that the reaction time based on described tester presents at random.
According to a preferred implementation, described stimulus information comprise the visual stimulus information presented with hologram, the haptic stimulus information presented with braille tactile data and acoustic stimuli information.
According to a preferred implementation, described physiologic information parameter at least comprises eye movement parameter, acoustic wave parameter, facial muscle movements parameter, limb action parameter.
According to a preferred implementation, described sample pattern is one or more in normal distribution model, logistic regression method model, decision-tree model, LEAST SQUARES MODELS FITTING, perceived control model, Boost method model, Hidden Markov Model (HMM), gauss hybrid models, neural network model, degree of deep learning model.
Advantageous Effects of the present invention:
The present invention by measure simultaneously tester multiple physiologic information and stress feedback information, the eye movement of analytical test personnel, brain information, information of acoustic wave and body kinematics informix obtain the cognitive dimension values being combined as personal traits of tester, and obtain the accuracy of the cognitive dimension of tester in conjunction with sample pattern.Analytical approach of the present invention overcomes the shortcoming of single, the environmental impact error of test material, improves the test mass of personal traits value test, thus obtains the value of personal traits accurately of tester.
Accompanying drawing explanation
Fig. 1 is the logic module figure of cognitive accuracy analysis method;
Fig. 2 is the schematic diagram of normal distribution model;
Fig. 3 is cognitive parameter distribution schematic diagram; With
Fig. 4 is the logical schematic of micro-expression recognition method.
Embodiment
Be described in detail below in conjunction with accompanying drawing.
The invention provides a kind of cognitive accuracy analysis method of Personality test value, set up sample pattern by the characteristic parameter measuring the eye movement of sample personnel, acoustic characteristic, brain information and/or body action feature.Present the emulation sight mode comprising at least one stimulus information to tester, simultaneously the physiologic information data such as the eye movement of collecting test personnel under stress situation, information of acoustic wave, brain information and/or body action, and receive the feedback information of tester.Cognitive parameter based on the physiologic information gathered and feedback information statistical test personnel forms test model.The cognitive parameter distribution of the sample pattern of compare test model and sample personnel thus obtain the accuracy of the cognitive dimension of tester.The personal traits value of tester comprehensively can be drawn according to the accuracy of the cognitive dimension of tester.
Stimulus information in the present invention refers to make the external information by stimulating object to produce reaction or feedback information, comprises the information such as dynamic context, image, photo, sound, numeral, word, form.
First the sample personnel of some are selected.Then the eye movement eigenwert of sample personnel is measured respectively, acoustic characteristic value, facial muscle movements eigenwert, brain information feature value, reaction time and other physiologic information eigenwert.Other physiologic information at least also comprises pulse, blood pressure, heart bat, E.E.G frequency band, skin potential.Physiologic information can increase the kind of measurement as required.Physiologic information is used for the psychological condition of discrimination test personnel.If find, the psychological condition of tester is not suitable for test, then adjust the psychological condition of tester.
Eye movement feature comprise measure visual angle, centre visual angle, periphery visual angle, watch that coordinate range, fixation time, fixation time are discrete attentively, direction of gaze, the eyeball residence time, eyeball stop in scope, number of winks, blink speed, closed-eye time, pan distance, pan frequency, pupil footpath, gaze pattern, gaze pattern number of times one or more.
Acoustic characteristic comprises one or more in the size of sound, the frequency of sound, the speed of speaking.
Brain information feature comprise cerebration time delay, xy-Hb signal, Deoxy-Hb signal, Total-Hb signal, interval appointments, channel appointment, maximal value, latent time, one or more in half breadth, mean value, discrete value, median, addition number of times, phase differential, heart rate, FFT composition.
Facial muscle movements feature comprise raise one's eyebrows, frown, skim under the corners of the mouth, the corners of the mouth promotes, be in a pout, sting in lip, of flaccid muscles, muscular tone one or more.
By the cognitive parameter of each cognitive dimension that the eigenwert Using statistics modeling statistics of the eye movement of the sample personnel of measurement, sound wave, brain information, reaction time, facial muscle movements, body kinematics is correlated with.Statistical model comprises one or more of patrolling in normal distribution model, logistic regression method model, decision-tree model, LEAST SQUARES MODELS FITTING, perceived control model, Boost method model, Hidden Markov Model (HMM), gauss hybrid models, neural network model, degree of deep learning model.
In the present invention, the cognitive dimension forming personal traits has 27:
Responsibility-carelessness, loyal-variable, diligent-lazy, optimistic-worried, domination-obey, gregarious-to stay alone, frankly-round and smooth, self-discipline-arbitrarily, dare for-to shrink, careful-crudity, order-disorder, independence-comply with, enthusiasm-cold and detached, trust-suspicious, flexible-ossify, curious-indifferent, adhere to-relax, Li Ta-egoistic, amiable-hostility, patience-impatient, just-eccentric, aesthetic-popular, activity-leisurely and carefree, illusion-reality, sympathy-reason, calm-anxiety, sensitivity-blunt.
There is the relation mapped in the various eigenwert of sample personnel and the cognitive parameter of cognitive dimension.Statistical model counts corresponding cognitive parameter distribution model according to the eigenwert of sample personnel, i.e. sample pattern.
Stimulus information is presented to tester, thus obtains the various eigenwerts of tester under stress situation.Stimulus information comprises acoustic stimuli information, visual stimulus information and haptic stimulus information.Acoustic stimuli information comprises rhythmical music, instruction, the sight sound corresponding with scene of life.Visual stimulus information comprises word, dynamic or static picture/photo.Preferably, visual stimulus information can be dynamic or static hologram, or the sight of virtual reality.Haptic stimulus information comprises braille stimulus information.
Preferably, acoustic stimuli information and visual stimulus information can present simultaneously thus form emulation sight.Emulation sight comprises holographic emulation sight and virtual reality sight.Emulation scenarios is conducive to tester and presents the most real psychological condition, thus feeds back information accurately.
Such as, test macro is the blue and white porcelain image that tester presents the holography of a width 3 D stereo, and the blue and white porcelain artwork of aestheticism represents with the angle of 360 degree.Now monitoring or collecting test personnel physiologic information eigenwert and analyze, just discrimination test personnel whether can feel excited and exciting, and not need tester answer and think deeply.This avoid tester and provide false feedback information based on possible hint after subjective judgement.
Or, build virtual reality scenario, make tester be presented in dinner party scene in the scene of quiet leisure and lively respectively, and play the sound mated with virtual reality scenario.Then request for test personnel feed back the scene liked.Now tester is more easily according to oneself the real idea of sight feedback presented.Instead of the hobby of oneself is judged according to imagination.
According to a preferred implementation, for blind person tester, use tangibly braille and sound to build emulation sight as stimulus information, make blind person tester also can receive stimulus information and feed back real information.
In the process that stimulus information presents, test macro monitoring or the physiologic information of collecting test personnel and/or feedback information thus statistics form the cognitive parameter of at least one dimension.
Particularly, first the physiologic information parameter of test macro typing tester and feedback information.Physiologic information parameter at least comprises eye movement parameter, acoustic wave parameter, facial muscle movements parameter, limb action parameter, reaction time, brain information parameter.Feedback information comprises selection information, action message, the sound sent and the facial expression information that tester makes according to the requirement of stimulus information.Then the eigenwert in physiologic information parameter and/or feedback information parameter is extracted.Eigenwert comprises one or more in eye movement eigenwert, acoustic characteristic value, facial muscle movements eigenwert, limb action eigenwert, reaction time, brain information feature value, feedback characteristic value.A corresponding various features value of cognitive dimension.Eigenwert is formed cognitive parameter according to the cognitive dimension of correspondence and corresponding sample pattern statistics.This cognitive parameter can reflect the degree of awareness of tester to cognitive dimension.Comparative cognition parameter, in the region of the cognitive parameter distribution of sample pattern, obtains corresponding cognitive accuracy.
Embodiment one
The sample pattern that cognitive dimension is selects normal distribution model.As shown in Figure 2, sample pattern is from left to right divided into very poor (5%), poor (20%), general (50%), better (20%), very well (5%) five grade in proportion successively in normal distribution model.Determine that the distributing position of cognitive parameter in normal distribution just can determine the cognitive accuracy of cognitive dimension.
Such as stimulus information is: " I is usually a very active people.”
The feedback information of tester can be divided into " do not meet very much, compare do not meet, uncertain, compare meet and meet very much " five degree, wherein do not meet very much note 1 point, compare and do not meet note 2 points, uncertain note 3 points, relatively meet note 4 points, meet very much note 5 points, wherein (R) oppositely inscribes, and scores as not meeting very much note 5 points, relatively do not meet note 4 points, uncertain note 3 points, compares and meets note 2 points, meets very much note 1 point.
Or the feedback information of tester also can adopt vernier mode to score, and represents different score values with different colors.Be specially: be divided into unit gap by 0 ~ 10 point with 0.1, represent the matching degree with this kind of situation by the depth of color.Such as, represent by blueness and do not meet very much.If tester does not meet being fed back to of a certain stimulus information very much, vernier according to this stimulus information and the very incongruent degree of own situation, can be dragged to the position of blue region respective degrees by tester with mouse.By the score value that position and the tester of computer system record scale should obtain the feedback of this stimulus information.For all the other feedback informations " compare do not meet, uncertain, compare meet and meet very much " four kinds of situations adopt and score in the same way.
If the dimension of test enthusiasm has 6 stimulus informations, so according to the feedback information of tester, the computing method of cognitive parameter are:
The cognitive parameter of enthusiasm dimension is A=(A1+A2+A3+A4+A5+A6)/6 (A4 is for oppositely inscribing).
By adding up the cognitive parameter of all examination questions in each dimension, must assign in itself and sample pattern being compared, determining the distribution situation of cognitive parameter in sample pattern of tested a certain dimension.Thus obtain the cognitive grade of this dimension, i.e. cognitive accuracy.
In cognitive parameter distribution schematic diagram as shown in Figure 3, the accuracy situation of each cognitive dimension of tester.Grey parts is the distribution situation of sample pattern, and the short-term of black represents the distribution situation of the cognitive parameter of the cognitive dimension of tester.For the cognitive dimension of enthusiasm-cold and detached, the cognitive parameter of tester is in the zone of reasonableness of sample pattern, lower to the cognitive accuracy of enthusiasm, is partial to cold and detached.
According to a preferred implementation, in test process, based on the presentation speed of the reaction time adjustment stimulus information of tester, or, based on the conscientious degree that reaction time assessment tester tests platform.In test process, if the interval presentative time of stimulus information is shorter than the reaction time of tester to stimulus information, then easily omit feedback information, and easily cause the nervous flurried psychological condition of tester.Therefore, according to the presentation speed of the reaction time adjustment subsequent stimuli information of tester in normal reaction time range, cognitive parameter more accurately can be obtained.But, if detection system analysis finds that the reaction time of tester is obviously short than the normal reaction time, then need assessment tester to be tested whether half-hearted.In test process, repeatedly present similar information, until conscientious, if the reaction time of tester is all identical, then feedback information is true.If the subsequent reactions time of tester is not in normal reaction time range, then assesses tester half-hearted, get rid of the feedback information of abnormal response time.Which enhance the analysis result of cognitive accuracy.
According to a preferred implementation, the order that the psychological condition adjustment test sight obtained based on the physiologic information by tester presents, or the order selecting test sight to present according to personal interest by tester.Tester likely changes in the Process-centric reason state of test, occurs nervous or uninteresting boring psychological condition.Nervous psychological condition can affect tester equally and feed back false information, thus obtains inaccurate test result.When nervous psychological condition appears in tester, heartbeat can be accelerated, the action occurring frowning, skim under the corners of the mouth.When the physiologic information by tester picks out nervous psychological condition, test macro can adjust presenting sequentially of test sight.The test sight loosened easily making people presents in advance, thus the psychological condition of tester is loosened gradually, avoids because bad psychological condition affects test result.
Or the order allowing tester to select test sight to present according to the hobby of individual in suitable scope, the psychological condition maintaining tester a preferably state, thus makes the analysis result of the cognitive accuracy of tester more accurate.
According to a preferred implementation, stimulus information of the present invention comprises the stimulus information of detecting a lie that the reaction time based on tester presents immediately.Stimulus information of detecting a lie comprises negative testing stimulus information.
Such as: when tester runs into adverse circumstance in the scene of first virtual reality, tester selects precipitant feedback information.But in the scene of second virtual reality, tester, when not having pure assurance, have selected from unadventurous feedback information.The scene of second virtual reality is negative testing scene, contains negative testing stimulus information.If tester does not tell a lie, so feedback information should be select risk.If tester tells a lie, the feedback information so selected is from playing safe.Therefore known, the feedback information of tester in second scenario and the feedback information in first scene inconsistent, tester feedback false information.
Stimulus information can also be word stimulus information.Such as: when word occurs that " I thinks shameful thing sometimes." content, whether request for test personnel feedback meets oneself situation.If tester's feedack for meet very much, illustrate that the attitude of tester is honest.If tester selects not meet very much, so tester is dishonest.
If find, tester is dishonest, and test macro then repeats to detect a lie the stimulus information near stimulus information at random.Test macro selects the high feedback information of repetition rate as real feedback information.
Embodiment two
The present embodiment take Hidden Markov Model (HMM) as the sample pattern that statistical model sets up at least one cognitive dimension.Hidden Markov model (Hidden Markov Model, HMM) is statistical model, and it is used for description one containing the Markovian process implying unknown parameter.It is the implicit parameter determining this process from observable parameter.Then these parameters are utilized for further analysis.
Hidden Markov model (HMM) usually can describe with five units, comprises 2 state sets and 3 probability matrixs:
1. implicit state S
Meeting Markov property between these states, is actual the state implied in Markov model.These states cannot obtain by directly observing usually.(such as S1, S2, S3 etc.)
2. Observable state O
Being associated with implicit state in a model, obtaining by directly observing.(such as O1, O2, O3 etc., the number of Observable state not necessarily wants consistent with the number of implicit state.)
3. initial state probabilities matrix π
Represent the probability matrix of implicit state at initial time t=1, (such as during t=1, P (S1)=p1, P (S2)=P2, P (S3)=p3, then initial state probabilities matrix π=[p1 p2 p3].
4. implicit state transition probability matrix A.
Describe the transition probability between each state in HMM model.
Wherein Aij=P (Sj|Si), 1≤i, j≤N. represent under t, state are the condition of Si, are the probability of Sj in t+1 moment state.
5. observer state transition probability matrix B (English Confusion Matrix by name, literal translates as confusion matrix is not too easy to from literal understanding).
Make N represent implicit state number, M represents Observable state number, then:
Bij＝P(Oi|Sj)，1≤i≤M，1≤j≤N.
Represent under t, implicit state are Sj conditions, the probability of observation state Oi.
Generally, succinct expression hidden Markov model is carried out by λ=(A, B, π) tlv triple.
As mentioned above, the sample personnel of a large amount of different background, Different Culture, different sexes, all ages and classes are chosen.Sample personnel are tested and sets up the sample pattern of Hidden Markov in conjunction with test result.
Tester is presented to the sight comprising at least one stimulus information, stimulate tester to produce physiological reaction and/or psychoreaction.The physiologic information parameter of typing tester and/or feedback information parameter.Extract the eigenwert in physiologic information parameter and/or feedback information parameter.
Wherein, physiological reaction comprises micro-expression shape change of tester.
Micro-expression is attempting to constrain, hide a kind of facial expression fast, not easily discovered showed in real feelings process people, its duration is generally only 1/25 to 1/5 second, also have researcher to be defined as lower than 1/2s by micro-expression duration, often ignore by people.But confirm the face mapping behavioral study of people, the hidden emotion of people more fully comes out by micro-expression, and it more can express the true emotional of people.
Identify the method for micro-expression of tester comprise contingency model method, optical flow method, based on the recognition methods of difference slicing capacity figure and Gabor transformation and machine learning method.
As shown in Figure 4, micro-expression identifies and is divided into 2 stages, training stage and test phases by optical flow method automatically.In the training stage, first training video is read in, then to video decode, then to frame flag neutral expression, micro-expression and red expression, and the calculating of these frames is used to determine that the threshold value of the micro-expression of light stream and light stream strain threshold (set the scope of test phase whereby.) at test phase, read in test video, first mark neutral expression, in test video, then calculate light stream and the light stream strain value of each frame and neutral expression's frame, if meet setting range, then judge that this expression frame is micro-expression frame.
Adopt the region segmentation of facial image to determine micro-expression in conjunction with optical flow method, step comprises:
1, adopt Adaboost-Haar sorter location human eye, calculate the position of human eye barycenter, and detect the barycenter link position of two; Then by the barycenter line of alignment two by all frame recordings to 70 frames, in order to increase stability, coupling upper left corner skin pixels.
2, facial image compartition is become 8 regions, i.e. the mouth of forehead, right and left eyes, left and right cheek, left and right and chin.
3, calculate light stream strain and threshold value, in order to reduce the impact of macro sheet feelings, grand expressive parts being separated and rejects.
The micro-expression identified follows following 2 standards, 1) strain size must be obviously large than peripheral region.2) the strain duration must be greater than 0.2S.Wherein, the mean strain in each region
wherein, N is totalframes,
it is the strain value of the region R of f frame.Every n frame detects local peaking, the peak value of a micro-expression in every n frame
the condition be detected is
meet at all values of peak value ± 4 frame
wherein, α ∈ (0,1).
Machine learning method
Face is divided the three-dimensional gradient direction histogram of extraction 12 interested regions and regional thereof as descriptor; Then, the three-dimensional gradient orientation histogram descriptor that the action in each region is used to based on the quantity reciprocal of local describes.
If a sets of video data is defined as v (x, y, t), be respectively δ along x-axis, y-axis, the axial spatio-temporal gradient note of t
vx(x, y, t), δ
vy(x, y, t) and δ
vt(x, y, t), each group spatio-temporal gradient is to (δ
vx, δ
vy), (δ
vy, δ
vt) and (δ
vx, δ
vt) amplitude be respectively
Location θ
xy(x, y, t), θ
yt(x, y, t), θ
xt(x, y, t) is designated as respectively
The histogram of these 6 signals of each frame of micro-expression is normalized, is detected by k mean cluster at each human face region that the foundation of micro-expression muscle, muscle amplitude of variation are maximum, the precision of the release of muscle.Adopt machine learning method can identify the onset of micro-expression in 13 accurately, apex, and 3 stages of offset.
The present invention according to micro-expression of above-mentioned micro-expression recognition method effective typing sample personnel, and sets up the hidden Markov sample pattern of micro-expression.
Utilize the physiologic information eigenwert of Hidden Markov Model (HMM) statistical test personnel and/or feedback information eigenwert thus obtain the cognitive parameter of each cognitive dimension of tester, this partial content is identical with the content of embodiment one, no longer repeat specification.The cognitive parameter of sample pattern is divided at least two cognitive grades, such as, better with poor.Also four cognitive grades " good, better, poor, poor " can be divided into.The cognitive parameter distribution region of cognitive parameter same cognitive dimension in sample pattern of each cognitive dimension of contrast test personnel, thus obtain the cognitive parameter level of the cognitive dimension of tester, i.e. cognitive accuracy.
Tester is being carried out in the process of personal traits value test, carrying out conscientious attitude, the emotional change of analytical test personnel by micro-expression of analytical test personnel and whether lie.Thus mood, the attitude of the adjustment tester that takes appropriate measures and make the honest feedback information of tester.
If find that tester's is unhappy to micro-Expression analysis of tester, there is low, gloomy mood, then adjust the order that stimulus information presents, the mood of adjustment tester.Such as, can transfer tester's active mood, the stimulus information containing music or game presents in advance, thus make the mood that tester recovers positive.
If find that the attitude of tester is half-hearted to micro-Expression analysis of tester, test macro sends information to tester.If micro-expression of tester changes into conscientiously, then continue to present stimulus information.Otherwise, in test process, repeatedly present similar information, until tester is conscientious.Then, be again presented on the stimulus information that tester presents half-hearted period, obtain the real feedback information of tester, thus ensure that the test result of tester is more accurate.
If find that lie phenomenon appears in tester to micro-Expression analysis of tester, tester is given a warning or points out.Require that tester's honesty treats the test of personal traits value.Warning can be that speech form or written form present.Or the voice containing warning message and word present to tester simultaneously.When presenting warning message, the flicker of light can be coordinated.Meanwhile, after warning sends, test macro inserts at random and presents stimulus information of detecting a lie in orderly stimulus information, monitors and micro-expression information of analytical test personnel simultaneously.When all to show tester be honest for the feedback result of stimulus information of detecting a lie and the analysis result of micro-expression, test macro preserves feedback information eigenwert and/or the physiologic information eigenwert of tester, statistical study obtains the cognitive parameter of each cognitive dimension of tester, obtains the cognitive accuracy of cognitive dimension.
Embodiment three
Respectively sample pattern is set up to the eye movement characteristic sum acoustic characteristic of sample personnel.Separately the eye movement characteristic sum acoustic characteristic of tester is tested, the cognitive parameter obtaining each the cognitive dimension obtained by eye movement feature respectively and the cognitive parameter of each cognitive dimension obtained by acoustic characteristic.Analyze identical cognitive dimension, the cognitive parameter obtained by eye movement feature and the weighted mean value of cognitive parameter to be obtained by acoustic characteristic, the total cognitive parameter namely obtained.Is compared in the cognitive parameter distribution region of the total cognitive parameter of a certain cognitive dimension and sample pattern and obtain the cognitive accuracy of tester.
The method being recorded cognitive parameter by eye movement feature is described.
The eye movement image of collecting sample personnel, sets up the eye movement sample pattern be made up of eye movement.Extract the eigenwert of each sample in eye movement sample pattern.Set up eye movement feature database, this eye movement feature database comprises the eigenwert of each eye movement sample of extraction and the region-of-interest information of each eye movement sample and/or cognitive parameter.
Extract the eye movement eigenwert of tester, comprising region-of-interest information and/or the cognitive parameter of the eye movement of this tester of extraction.By the eye movement eigenwert of tester and each eigenwert be stored in eye movement feature database are compared, the eye movement eigenwert of tester is mated with most similar features value.And the region-of-interest information of the eye movement eigenwert of coupling and/or cognitive Parameter analysis are obtained cognitive parameter.Transmit and store cognitive parameter.
Eye movement sample can be the image that the eyes by directly taking people obtain.Or the artificial eye movement by producing eyes and periphery modeling thereof.When producing specific artificial eye movement, to form the size of the pupil of eyes, iris, the white of the eye, eyelid and eyelashes, shape, color, texture and correlativity and such as illumination, glasses etc. external condition carry out modeling.To produce multiple eye movement images of the image comprising eyes and periphery thereof at random.
The method being recorded cognitive parameter by acoustic characteristic is described.
1) the cognitive parameter machine learning model of sound wave is set up:
The multinomial perceptional factors score value that cognitive parametric measurement obtains the true benchmark scoring of each the cognitive dimension as sample personnel is carried out for the multiple sample personnel selected.Gather the sound wave fragment of multiple sample personnel normal articulation.Pre-service is carried out to sound wave fragment and extracts multinomial sound wave prosodic features.Extract the multinomial statistical characteristics of acoustic characteristic.Set up and comprise the sound wave cognitive parameter machine learning model of sound wave prosodic features to the mapping relations of the cognitive parameter of cognitive dimension.The multinomial perceptional factors score value of each sample personnel and multinomial statistical characteristics corresponding to sound wave fragment every sound wave prosodic features are inputted respectively the cognitive parameter machine learning model of sound wave to train.
2) the cognitive parameter prediction of cognitive dimension:
The normal articulation sound wave of collecting test personnel obtains sound wave fragment to be predicted.Pre-service is carried out to sound wave fragment and extracts multinomial sound wave prosodic features and corresponding multinomial statistical nature.The cognitive parameter machine learning model of multinomial statistical nature input sound wave of multinomial sound wave prosodic features and correspondence is carried out the regretional analysis of perceptional factors score value and obtain every sound wave prosodic features and multinomial perceptional factors score value corresponding to statistical nature.Respectively by all perceptional factors score values weighted sum corresponding for each sound wave prosodic features, finally obtain the cognitive parameter of tester and store output.
If the weight being recorded the cognitive parameter A of a certain cognitive dimension by eye movement feature is W
1, the weight being recorded the cognitive parameter B of this cognitive dimension by acoustic characteristic is W
2.The total cognitive parameter W of certain then this cognitive dimension
12
By the total cognitive parameter of a certain cognitive dimension and this cognitive dimension, compare in the cognitive parameter distribution region in sample pattern, obtains the cognitive grade corresponding with total cognitive parameter, namely obtain the cognitive accuracy of tester in this cognitive dimension.
Cognitive accuracy analysis method of the present invention is not only applicable to test person class, is also applicable to carry out the test of personal traits value to the intelligent machine of personification or intelligent software.
Artificially routine with intelligent machine, present acoustic stimuli information to intelligent robot, and require robot feedback information.According to the cognitive parameter of 27 cognitive dimensions of the feedback information statistical machine people of robot, and the cognitive parameter of the sample pattern of cognitive parameter and the mankind is compared, determine the distributing position of cognitive parameter in sample pattern of robot, thus determine the cognitive accuracy of each cognitive dimension of robot.Can the personal traits of analysis robot according to the test result of the cognitive accuracy of robot, thus the setting in adjustment interior data storehouse, mould the multiple robot with different personality.Such as be partial to enthusiasm, loyalty, responsibility, optimism, gregarious, amiable, responsive personal traits for accompanying the robot of solitary elder to be applicable to having.And be applicable to having for the treatment of the robot of domestic hygiene and be partial to responsibility, loyalty, self-discipline, careful, orderly, calm, independently personal traits.
It should be noted that; above-mentioned specific embodiment is exemplary; those skilled in the art can find out various solution under the inspiration of the disclosure of invention, and these solutions also all belong to open scope of the present invention and fall within protection scope of the present invention.It will be understood by those skilled in the art that instructions of the present invention and accompanying drawing thereof are illustrative and not form limitations on claims.Protection scope of the present invention is by claim and equivalents thereof.
Claims (10)
1. a cognitive accuracy analysis method for personal traits value test, it is characterized in that, described method comprises:
The sample pattern of at least one cognitive dimension is set up based at least one statistical model;
Described tester is stimulated to present the sight mode comprising at least one stimulus information;
Monitoring or gather the physiologic information of described tester and/or feedback information thus statistics forms the cognitive parameter of at least one cognitive dimension;
The cognitive parameter distribution of more described cognitive parameter and described sample pattern thus obtain the accuracy of described cognitive dimension.
2. cognitive accuracy analysis method as claimed in claim 1, is characterized in that, the step of the test model of at least one cognitive dimension of described formation comprises:
The physiologic information parameter of tester described in typing and/or feedback information parameter;
Extract the eigenwert in described physiologic information parameter and/or feedback information parameter;
Sample pattern based on correspondence is added up with each cognitive dimension characteristic of correspondence value thus is formed cognitive parameter.
3. cognitive accuracy analysis method as claimed in claim 2, is characterized in that, the cognitive parameter distribution Region dividing in described sample pattern is at least two the cognitive hierarchical regions can distinguishing cognitive accuracy.
4. cognitive accuracy analysis method as claimed in claim 3, it is characterized in that, described method also comprises: based on the presentation speed of the reaction time adjustment stimulus information of described tester, or,
Reaction time based on described tester assesses the conscientiously degree of described tester.
5. cognitive accuracy analysis method as claimed in claim 3, it is characterized in that, described method also comprises: the order that the psychological condition adjustment test sight obtained based on the physiologic information by described tester presents, or the order selecting test sight to present according to personal interest by described tester.
6. cognitive accuracy analysis method as claimed in claim 5, is characterized in that, described test sight comprises virtual reality sight and holographic sight, and wherein holographic sight comprises dynamic holographic sight and static holographic sight.
7. cognitive accuracy analysis method as claimed in claim 4, is characterized in that, described stimulus information comprises the stimulus information of detecting a lie that the reaction time based on described tester presents at random.
8. the cognitive accuracy analysis method as described in one of aforementioned claim, is characterized in that, described stimulus information comprises the visual stimulus information presented with hologram, the haptic stimulus information presented with braille tactile data and acoustic stimuli information.
9. cognitive accuracy analysis method as claimed in claim 8, is characterized in that, described physiologic information parameter at least comprises eye movement parameter, acoustic wave parameter, facial muscle movements parameter, limb action parameter.
10. cognitive accuracy analysis method as claimed in claim 8, it is characterized in that, described sample pattern comprises one or more in normal distribution model, logistic regression method model, decision-tree model, LEAST SQUARES MODELS FITTING, perceived control model, Boost method model, Hidden Markov Model (HMM), gauss hybrid models, neural network model, degree of deep learning model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510482599.1A CN105022929B (en) | 2015-08-07 | 2015-08-07 | A kind of cognition accuracy analysis method of personal traits value test |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510482599.1A CN105022929B (en) | 2015-08-07 | 2015-08-07 | A kind of cognition accuracy analysis method of personal traits value test |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105022929A true CN105022929A (en) | 2015-11-04 |
| CN105022929B CN105022929B (en) | 2018-05-08 |
Family
ID=54412895
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510482599.1A Active CN105022929B (en) | 2015-08-07 | 2015-08-07 | A kind of cognition accuracy analysis method of personal traits value test |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105022929B (en) |
Cited By (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106599558A (en) * | 2016-12-05 | 2017-04-26 | 武汉智普天创科技有限公司 | Cognitive assessment method and system based on virtual reality |
| CN106667506A (en) * | 2016-12-21 | 2017-05-17 | 上海与德信息技术有限公司 | Method and device for detecting lies on basis of electrodermal response and pupil change |
| CN106691476A (en) * | 2017-01-16 | 2017-05-24 | 清华大学 | Image recognition Mentality Analysis System based on features of eye movements |
| CN106843500A (en) * | 2017-02-27 | 2017-06-13 | 南通大学 | Human-subject test rehabilitation training system based on the dynamic tracer technique of eye |
| CN107348962A (en) * | 2017-06-01 | 2017-11-17 | 清华大学 | A kind of personal traits measuring method and equipment based on brain-computer interface technology |
| CN108364131A (en) * | 2018-02-09 | 2018-08-03 | 合不合（厦门）网络科技有限公司 | The automatic identification of personality type is carried out using neural network and divides the method for group |
| CN108523906A (en) * | 2018-04-27 | 2018-09-14 | 合肥工业大学 | Personality analysis method and system, storage medium based on pulse characteristics |
| CN109192277A (en) * | 2018-08-29 | 2019-01-11 | 沈阳康泰电子科技股份有限公司 | A kind of psychological characteristics measure based on general effective question and answer scale |
| CN109345104A (en) * | 2018-09-25 | 2019-02-15 | 博跃恒民康(深圳)智能化科技有限公司 | A kind of user capability method for improving, system and storage medium |
| CN109445578A (en) * | 2018-10-12 | 2019-03-08 | 上海陈天桥脑疾病研究所 | A kind of cognitive ability assessment system and method |
| CN109498039A (en) * | 2018-12-25 | 2019-03-22 | 北京心法科技有限公司 | Personality assessment's method and device |
| CN109672930A (en) * | 2018-12-25 | 2019-04-23 | 北京心法科技有限公司 | Personality association type emotional arousal method and apparatus |
| CN109683840A (en) * | 2018-12-25 | 2019-04-26 | 北京心法科技有限公司 | The exciting method and device of plyability cognitive information processing mechanism |
| WO2019134609A1 (en) * | 2018-01-05 | 2019-07-11 | 陈锦焜 | Scene generation system, and scene generation method |
| CN111460245A (en) * | 2019-01-22 | 2020-07-28 | 刘宏军 | Multi-dimensional crowd characteristic measuring method |
| CN111461153A (en) * | 2019-01-22 | 2020-07-28 | 刘宏军 | Crowd characteristic deep learning method |
| CN112137627A (en) * | 2020-09-10 | 2020-12-29 | 北京津发科技股份有限公司 | Self-adaptive method and system for human factor evaluation and training |
| CN113507892A (en) * | 2019-03-18 | 2021-10-15 | 欧姆龙健康医疗事业株式会社 | Biological information acquisition device and biological information acquisition method |
| TWI758036B (en) * | 2020-12-25 | 2022-03-11 | 中原大學 | Building method, assessment form and evaluation method of child-based personality trait |
| CN114870191A (en) * | 2022-07-08 | 2022-08-09 | 北京智精灵科技有限公司 | Cognitive assessment improving method and system based on personality difference |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US3905132A (en) * | 1974-11-07 | 1975-09-16 | Us Navy | Hidden knowledge detector |
| US5762611A (en) * | 1996-11-12 | 1998-06-09 | The United States Of America As Represented By The Secretary Of The Navy | Evaluation of a subject's interest in education, training and other materials using brain activity patterns |
| CN1739451A (en) * | 2005-07-21 | 2006-03-01 | 高春平 | Method and device for monitoring psycological and professional test truth |
| US20130209977A1 (en) * | 2012-02-09 | 2013-08-15 | Anthrotronix, Inc. | Performance Assessment Tool |
| CN103440864A (en) * | 2013-07-31 | 2013-12-11 | 湖南大学 | Personality characteristic forecasting method based on voices |
| CN104127194A (en) * | 2014-07-14 | 2014-11-05 | 华南理工大学 | Depression evaluating system and method based on heart rate variability analytical method |
-
2015
- 2015-08-07 CN CN201510482599.1A patent/CN105022929B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US3905132A (en) * | 1974-11-07 | 1975-09-16 | Us Navy | Hidden knowledge detector |
| US5762611A (en) * | 1996-11-12 | 1998-06-09 | The United States Of America As Represented By The Secretary Of The Navy | Evaluation of a subject's interest in education, training and other materials using brain activity patterns |
| CN1739451A (en) * | 2005-07-21 | 2006-03-01 | 高春平 | Method and device for monitoring psycological and professional test truth |
| US20130209977A1 (en) * | 2012-02-09 | 2013-08-15 | Anthrotronix, Inc. | Performance Assessment Tool |
| CN103440864A (en) * | 2013-07-31 | 2013-12-11 | 湖南大学 | Personality characteristic forecasting method based on voices |
| CN104127194A (en) * | 2014-07-14 | 2014-11-05 | 华南理工大学 | Depression evaluating system and method based on heart rate variability analytical method |
Non-Patent Citations (1)
| Title |
|---|
| 任立生: "《设计心理学》", 31 August 2005 * |
Cited By (26)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106599558A (en) * | 2016-12-05 | 2017-04-26 | 武汉智普天创科技有限公司 | Cognitive assessment method and system based on virtual reality |
| CN106667506A (en) * | 2016-12-21 | 2017-05-17 | 上海与德信息技术有限公司 | Method and device for detecting lies on basis of electrodermal response and pupil change |
| CN106691476A (en) * | 2017-01-16 | 2017-05-24 | 清华大学 | Image recognition Mentality Analysis System based on features of eye movements |
| CN106843500A (en) * | 2017-02-27 | 2017-06-13 | 南通大学 | Human-subject test rehabilitation training system based on the dynamic tracer technique of eye |
| CN107348962B (en) * | 2017-06-01 | 2019-10-18 | 清华大学 | A kind of personal traits measurement method and equipment based on brain-computer interface technology |
| CN107348962A (en) * | 2017-06-01 | 2017-11-17 | 清华大学 | A kind of personal traits measuring method and equipment based on brain-computer interface technology |
| CN109999311A (en) * | 2018-01-05 | 2019-07-12 | 陈锦焜 | The method of scenarios generation system and scenarios generation |
| WO2019134609A1 (en) * | 2018-01-05 | 2019-07-11 | 陈锦焜 | Scene generation system, and scene generation method |
| CN108364131A (en) * | 2018-02-09 | 2018-08-03 | 合不合（厦门）网络科技有限公司 | The automatic identification of personality type is carried out using neural network and divides the method for group |
| CN108523906A (en) * | 2018-04-27 | 2018-09-14 | 合肥工业大学 | Personality analysis method and system, storage medium based on pulse characteristics |
| CN109192277B (en) * | 2018-08-29 | 2021-11-02 | 沈阳康泰电子科技股份有限公司 | Psychological characteristic measuring method based on universal effective question-answering ruler |
| CN109192277A (en) * | 2018-08-29 | 2019-01-11 | 沈阳康泰电子科技股份有限公司 | A kind of psychological characteristics measure based on general effective question and answer scale |
| CN109345104A (en) * | 2018-09-25 | 2019-02-15 | 博跃恒民康(深圳)智能化科技有限公司 | A kind of user capability method for improving, system and storage medium |
| CN109445578A (en) * | 2018-10-12 | 2019-03-08 | 上海陈天桥脑疾病研究所 | A kind of cognitive ability assessment system and method |
| CN109683840A (en) * | 2018-12-25 | 2019-04-26 | 北京心法科技有限公司 | The exciting method and device of plyability cognitive information processing mechanism |
| CN109672930A (en) * | 2018-12-25 | 2019-04-23 | 北京心法科技有限公司 | Personality association type emotional arousal method and apparatus |
| CN109498039A (en) * | 2018-12-25 | 2019-03-22 | 北京心法科技有限公司 | Personality assessment's method and device |
| CN111460245A (en) * | 2019-01-22 | 2020-07-28 | 刘宏军 | Multi-dimensional crowd characteristic measuring method |
| CN111461153A (en) * | 2019-01-22 | 2020-07-28 | 刘宏军 | Crowd characteristic deep learning method |
| CN111461153B (en) * | 2019-01-22 | 2023-08-04 | 刘宏军 | Crowd feature deep learning method |
| CN113507892A (en) * | 2019-03-18 | 2021-10-15 | 欧姆龙健康医疗事业株式会社 | Biological information acquisition device and biological information acquisition method |
| CN112137627A (en) * | 2020-09-10 | 2020-12-29 | 北京津发科技股份有限公司 | Self-adaptive method and system for human factor evaluation and training |
| CN112137627B (en) * | 2020-09-10 | 2021-08-03 | 北京津发科技股份有限公司 | Intelligent human factor evaluation and training method and system |
| TWI758036B (en) * | 2020-12-25 | 2022-03-11 | 中原大學 | Building method, assessment form and evaluation method of child-based personality trait |
| CN114870191A (en) * | 2022-07-08 | 2022-08-09 | 北京智精灵科技有限公司 | Cognitive assessment improving method and system based on personality difference |
| CN114870191B (en) * | 2022-07-08 | 2022-10-28 | 北京智精灵科技有限公司 | Cognitive assessment improving method and system based on personality difference |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105022929B (en) | 2018-05-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105022929A (en) | Cognition accuracy analysis method for personality trait value test | |
| Lisetti et al. | Automatic facial expression interpretation: Where human-computer interaction, artificial intelligence and cognitive science intersect | |
| Berry et al. | Some components and consequences of a babyface. | |
| CN110507335A (en) | Inmate's psychological health states appraisal procedure and system based on multi-modal information | |
| CN106205633B (en) | It is a kind of to imitate, perform practice scoring system | |
| Winkielman et al. | Affect and processing dynamics | |
| WO2019024247A1 (en) | Data exchange network-based online teaching evaluation system and method | |
| CN108175425B (en) | Analysis processing device and cognitive index analysis method for potential value test | |
| CN108175424B (en) | Test system for cognitive ability value test | |
| Krumhuber et al. | The role of facial movements in emotion recognition | |
| CN108065942B (en) | Method for compiling stimulation information aiming at oriental personality characteristics | |
| CN109199410B (en) | Speech cognition assessment method based on eye movement | |
| CN106485085A (en) | A kind of intellect service robot health identification system and method | |
| CN109528217A (en) | A kind of mood detection and method for early warning based on physiological vibrations analysis | |
| CN115312195A (en) | Health assessment method for calculating individual psychological abnormality based on emotion data | |
| Guo | Size-invariant facial expression categorization and associated gaze allocation within social interaction space | |
| CN112464774A (en) | Emotion identification method in video learning based on eye movement measurement | |
| Grebot et al. | Uncanny valley hypothesis and hierarchy of facial features in the human likeness continua: An eye-tracking approach. | |
| Buck et al. | Measuring the dynamic stream of display: Spontaneous and intentional facial expression and communication. | |
| CN111028853A (en) | Spoken language expressive force evaluation method and system | |
| Massoz | Non-invasive, automatic, and real-time characterization of drowsiness based on eye closure dynamics | |
| CN112450932B (en) | Psychological disorder detection system and method | |
| CN117547271B (en) | Psychological diathesis intelligent evaluation analyzer | |
| WO2024062935A1 (en) | Information processing device, system, information processing method, and program | |
| Xu et al. | Improvement of Air Traffic Controller Scheduling Based on Multi-mode Fatigue Information |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |