CN105893784A - Method for generating character test questionnaire based on image and surveying interactive method - Google Patents
Method for generating character test questionnaire based on image and surveying interactive method Download PDFInfo
- Publication number
- CN105893784A CN105893784A CN201610484985.9A CN201610484985A CN105893784A CN 105893784 A CN105893784 A CN 105893784A CN 201610484985 A CN201610484985 A CN 201610484985A CN 105893784 A CN105893784 A CN 105893784A
- Authority
- CN
- China
- Prior art keywords
- image
- user
- concept
- personality
- questionnaire
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 50
- 238000012360 testing method Methods 0.000 title claims abstract description 33
- 230000002452 interceptive effect Effects 0.000 title claims abstract description 5
- 238000012800 visualization Methods 0.000 claims abstract description 8
- 238000012216 screening Methods 0.000 claims abstract description 5
- 238000011835 investigation Methods 0.000 claims description 21
- 238000012549 training Methods 0.000 claims description 13
- 239000000284 extract Substances 0.000 claims description 10
- 238000003066 decision tree Methods 0.000 claims description 7
- 230000000007 visual effect Effects 0.000 claims description 6
- 238000013528 artificial neural network Methods 0.000 claims description 4
- 230000008569 process Effects 0.000 description 5
- 230000008901 benefit Effects 0.000 description 4
- 238000004590 computer program Methods 0.000 description 4
- 230000008859 change Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 241000282326 Felis catus Species 0.000 description 2
- 230000006399 behavior Effects 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 238000002372 labelling Methods 0.000 description 2
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 description 1
- 241001481166 Nautilus Species 0.000 description 1
- 241000282373 Panthera pardus Species 0.000 description 1
- 241000287531 Psittacidae Species 0.000 description 1
- 206010047571 Visual impairment Diseases 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 239000012141 concentrate Substances 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000010998 test method Methods 0.000 description 1
- 238000013519 translation Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H10/00—ICT specially adapted for the handling or processing of patient-related medical or healthcare data
- G16H10/20—ICT specially adapted for the handling or processing of patient-related medical or healthcare data for electronic clinical trials or questionnaires
Landscapes
- Health & Medical Sciences (AREA)
- Engineering & Computer Science (AREA)
- Epidemiology (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Primary Health Care (AREA)
- Public Health (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a method for generating a character test questionnaire based on an image and a surveying interactive method. The method for generating the character test questionnaire comprises the following steps: acquiring an image set which the user likes and a character nature truth set of the user and establishing a first relationship between the image and the user; extracting a concept and establishing the relationship between the image and the concept; extracting image features of each image in the image set and establishing a second relationship of the image and the user; confirming a concept set shared by a user set under a given user set according to the relationship between the image and the concept and the relationship of the image and the user; screening out the concept set at specific character distinguishing degree; and screening out the image with representative character from the image set under each concept to be taken as an image option of a visualization problem; generating the questionnaire through the screened-out concept and the image option under the concept. According to the embodiment of the invention, the accurate user character can be acquired within a shorter time period and the cross-language property is better.
Description
Technical field
The present embodiments relate to image procossing, data mining and psychological techniques field, specifically
Relate to a kind of personality based on image test questionnaire generate method and investigate mutual method.
Background technology
Personality is a psychological characteristic, is devoted to complicated and diversified for mankind behavior, with a small amount of
, specific characteristic stable, that can measure go explain.With the attribute phase in demography
Ratio, character trait goes to explain and the difference of prediction behavior of men from the inherent Psychological Angle of people.
Currently used widest personality model is: big five models (Big Five, BF), is also five factors
Model (Five-Factor Model, FFM).Big five models (Big Five, BF) by the personality of people,
It is divided into five aspects: open (openness), conscious (Conscientiousness), extroversion
(Extraversion), affine (Agreeableness) and neurotic (Neuroticism).Based on big five
The character study of model (Big Five, BF), has had a lot of application, such as: occupation is auxiliary
Help, targeted ads, personalized recommendation system, disease identification and prevention, even man-machine interaction.
Tradition personality test, such as: big five personality test questionnaires (Big-Five Inventory,
BFI), wait standard lattice questionnaire, be by measuring dividing of individual five dimensions in personality
Number investigates individual personality.This class questionnaire, the series of problems of questionnaire is special by psychology
Family sets；The test mode of questionnaire is: experimenter reads and understands the containing of exercise question of textual form
Justice, then the thinking personality attribute that goes out of itself, and according to self attributes and text exercise question
Similarity degree marking between implication, from very different meaning to agreeing to very much.
Although big five personality test questionnaire (Big-Five Inventory, BFI) are by extensively
Use, but existing text based personality test questionnaire, but have and limit to as follows:
(1) often answering a problem, experimenter will first read and understand the implication of this problem, when
The when of more problems, experimenter can be caused the big five properties of the biggest burden, particularly standard
Lattice questionnaire is usually the NEO-PI-R-240 of 240 problems
(Neuroticism-Extraversion-Openness Personality Inventory Revised).Do
The when of such questionnaire, experimenter needs to think deeply the personality attribute that itself goes out, so
Similarity degree between attribute and the implication of problem of later evaluation oneself, but, experimenter is not
Being scholar or the expert having research to personality, more generally, experimenter does not very understand self
Personality attribute, this is that investigation personality attribute causes the biggest error.(2) text based is adjusted
Interrogating volume, expressed obvious investigation and be intended to, for promoting self-phychology, experimenter inclines
Mainstream values is more conformed to in the answer making them, more likable.Such as at BFI-10
Having a problem in that in questionnaire " I am more lazy ", most subjects can select " disagreeing ".
This error being very easy to cause investigation, the accuracy of impact investigation.(3) text based is adjusted
Interrogating volume is language-sensitive, and psychological professional is carrying out drawing of questionnaire for oneself country
The when of entering, it is impossible to directly translate the meaning of stem, in addition it is also necessary to consider the personality of concrete stem
Implication, such as: in the BFI of english version, it is neurotic (Neuroticism) that " calm " is used to tolerance
This dimension personality, if directly " calm " being translated into German, corresponding word is
" ruhlg ", however " ruhlg " in German, not only this is one-dimensional with neurotic (Neuroticism)
Degree has relation, also relevant with export-oriented (Extraversion) this dimension.
In view of this, the special proposition present invention.
Summary of the invention
In view of the above problems, it is proposed that the present invention is to provide one to solve above-mentioned at least in part
A kind of based on image the personality test questionnaire of problem generates method.Additionally, also provide for one
Plant the method that investigation is mutual.
To achieve these goals, according to an aspect of the invention, it is provided techniques below side
Case:
A kind of personality based on image test questionnaire generates method, and described method at least includes:
Obtain image set and the personality attribute truth set of described user that user likes, and build
Described image and first relation of described user；
Image in the described image set liking described user carries out based on degree of depth study general
Read and extract, build the relation of described image and described concept；
Extract the characteristics of image of each image in described image set, and determine according to described characteristics of image
Image feature vector and user characteristics set, and it is described to utilize described image feature vector to expand
Image and first relation of user, build the second relation of image and user；
According to described image and the relation of concept and described image and second relation of user, determine
Given user gathers lower user and collects total concept set；
Described user characteristics set and described user are integrated total concept set cooperation as training set, and
Using the personality attribute truth set of described user as target, training returns device, and use based on
The gradient at visual angle promotes decision Tree algorithms, filters out the concept set of concrete personality discrimination；
Use clustering algorithm, the concept set of the concrete personality discrimination filtered out described in utilization,
In image collection from each concept, screening has the representational image of personality, asks as visualization
The image option of topic；
By the described image option under the described concept filtered out and described concept, generate described
Personality based on image test questionnaire.
Preferably, the image in the described described image set liking described user is carried out based on deeply
The concept of degree study is extracted, and specifically includes:
Image in the described image set that described user is liked by employing deep neural network is carried out
Concept is extracted.
Preferably, described employing gradient based on visual angle promotes decision Tree algorithms, filters out concrete
The concept set of personality discrimination, specifically includes:
Initialize according to below equation and return by force device:
Wherein, described F0Represent and initially return by force device；Described piRepresent that user is through text based
Personality tests the personality attribute true value obtained；Described N represents training sample number；
Residual error r is calculated according to below equationi:
ri=pi-Fm-1(xi)
Wherein, described m represents iterations；DescribedDescribed K table
Show personality conceptual point number；Described d represents the characteristic dimension of described image；
According to below equation residual error returned and obtains weak recurrence device:
Weak recurrence device is added to strong recurrence in device, obtains the new device that returns by force:
Fm(x)=Fm-1(x)+vT(x；Vm,Rm,Am)
Wherein, described v represents weight；
Determine that final recurrence device is:
Wherein, described
Described T represents that base returns device；Described I () represents indicator function；Described J represents leafy node number
Amount；Described VmRepresent the feature of the conceptual point chosen；Described RmRepresent the classification that study is arrived
Pattern；Described AmRepresent the leafy node value of output；Described M represents that the base of use returns device
Number；Described θ represents training set；
The described final recurrence device being determined by, filters out described concrete personality discrimination
Concept set.
To achieve these goals, according to another aspect of the present invention, a kind of profit is additionally provided
The questionnaire generated by method described in technique scheme carries out investigating mutual method.Described
Investigate mutual method to include:
Set the investigation form of described questionnaire as a given basket, set containing corresponding
The image of concept is as the option of problem；
Set the interactive form of described questionnaire as: described user according to self to described image
Understanding, select meet investigation stem image；
According to user's answer, utilize described recurrence device, calculate big five character traits of described user.
Compared with prior art, technique scheme at least has the advantages that
The embodiment of the present invention, by using any of the above-described technical scheme, uses image to ask as investigation
The option of volume, it is provided that can accurate and effective and user-friendly personality test questionnaire, Ke Yiyou
Effect ground alleviates the burden of user when filling in questionnaires, and promotes user-friendliness, it is possible in the short period
Inside obtain accurate user's personality, have preferably across language performance, reduce personality and survey
The error of examination.Solve the offset issue of text personality method of testing, and language-sensitive problem.
Compare the generation method of existing text based personality test questionnaire: psychological professional sets
Determine exercise question and option that text describes, on the one hand experimenter can be allowed the most free of a burden by image
Understanding, it is not necessary to carry out the translation for language, allow experimenter complete investigation at short notice and ask
Volume；On the other hand, arranged by the questionnaire problem of data-driven, it is possible to make personality investigate
There is interpretability.
Certainly, the arbitrary product implementing the present invention is not necessarily required to realize above-described institute simultaneously
There is advantage.
Other features and advantages of the present invention will illustrate in the following description, and, part
Ground becomes apparent from description, or understands by implementing the present invention.The present invention's
Purpose and further advantage can be special by institute in the description write, claims and accompanying drawing
The method do not pointed out realizes and obtains.
Accompanying drawing explanation
Accompanying drawing, as the part of the present invention, is used for providing further understanding of the invention,
The schematic description and description of the present invention is used for explaining the present invention, but does not constitute the present invention
Improper restriction.Obviously, the accompanying drawing in describing below is only some embodiments, for ability
For the those of ordinary skill of territory, on the premise of not paying creative work, it is also possible to according to this
A little accompanying drawings obtain other accompanying drawings.In the accompanying drawings:
Fig. 1 is to test questionnaire according to the personality based on image shown in an exemplary embodiment
The schematic flow sheet of generation method；
Fig. 2 a is to test investigation according to the personality based on image shown in another exemplary embodiment
In questionnaire generation method, concept extracts schematic diagram；
Fig. 2 b is to test investigation according to the personality based on image shown in another exemplary embodiment
Questionnaire generation method clusters schematic diagram；
Fig. 3 is to test questionnaire according to the personality based on image shown in an exemplary embodiment
Investigation interaction figure；
Fig. 4 is to test questionnaire according to the personality based on image shown in an exemplary embodiment
Investigation exchange method schematic flow sheet.
These accompanying drawings and word describe and are not intended as limiting by any way the design model of the present invention
Enclose, but be that those skilled in the art illustrate idea of the invention by reference specific embodiment.
Detailed description of the invention
Below in conjunction with the accompanying drawings and the embodiment of the present invention is solved the technical problem that by specific embodiment,
The technical scheme used and the technique effect of realization carry out clear, complete description.Obviously,
Described embodiment is only a part of embodiment of the application, is not whole embodiments.
Based on the embodiment in the application, those of ordinary skill in the art are not paying creative work
Under premise, other equivalents all obtained or the embodiment of obvious modification all fall within the guarantor of the present invention
In the range of protecting.The embodiment of the present invention can be according to the multiple difference being defined and covered by claim
Mode embodies.
It should be noted that in the following description, understand for convenience, give many tools
Body details.However, it will be apparent that the realization of the present invention can not have these details.
Also, it should be noted in the case of the most clearly limiting or do not conflict, in the present invention
Each embodiment and technical characteristic therein can be mutually combined and form technical scheme.
The major technique design of the present invention is: setting a series of visualization problem, each can
Depending on change problem all round some specific concepts, generate correspondence according to the concept contained in image
Personality test exercise question and option.Image option under problem is to have personality generation under these concepts
The image of table.Utilize user's answer option to visualization problem, analyze big the five of user
Character trait (Big Five Traits).
The embodiment of the present invention proposes a kind of personality based on image test questionnaire and generates method.
As it is shown in figure 1, the method may include that step S100 is to step S160.
S100: obtain image set and the personality attribute truth set of user that user likes, and structure
Build the first relation of image and user.
Wherein, any one image in the image set that user likes all belongs to the personality of this user
Property true value is labelling.Personality attribute true value is carried out the test of text based personality by user and produces.
Membership credentials index with the relation representing image and user；If using tlv triple representation, can
To be expressed as<image, user, relation classification>.
S110: the image in the image set liking user carries out concept based on degree of depth study and carries
Take, build the relation of image and concept.
In this step, the things comprised in concept representative image.Based on contained by the image extracted
Concept set, be organized into relation index, represent the relation of image and concept；If use tlv triple
Representation, can be expressed as<image, concept set, relation classification>.
Image in image set that user liked is exemplified below carry out learning based on the degree of depth
Concept extract process.
Based on GoogLeNet network structure, figure net (ImageNet) carries out pre-training,
Obtain deep neural network disaggregated model.Use the image set that user is liked by deep neural network
In image carry out concept extraction.
The image that input picture is concentrated, (extensive visualization identifies challenge to be output as ILSVRC
Match) Section 2 challenge: 1000 kinds provided in classifying and positioning go out in present image
Existing credibility.For each image, only choose probability more than before 0.1, and ranking 5
Classification, as shown in Figure 2 a.Wherein, by Egyptian Mau, timber, lamp, flower leopard and Buddha's warrior attendant parrot
The concept that nautilus contains as this image.In order to expand the quantity of concept, by initial 1000 kinds
Figure net (ImageNet) classification is denoted as the 1st level of concept.According to word net (WordNet)
Hierarchical structure, by the concept of the 1st level, be upwards generalized to the 2nd level, the 3rd level,
4th level.Word net is a complete english vocabulary semantic net, and it is by english vocabulary, especially
It is a word concept, according to whether having implication relation, builds hierarchical structure.Finally extract
1789 concepts.
S120: extract the characteristics of image of each image in image set, and determine figure according to characteristics of image
As characteristic vector and user characteristics set, and image feature vector is utilized to expand image and user
The first relation, build second relation of image and user.
Wherein, characteristics of image includes color characteristic, textural characteristics, pattern features and content characteristic.
These mark sheets are shown as image feature vector.User is expressed as its all images liked
The set of image feature vector, as user characteristics set.
After utilizing image feature vector to expand first relation of image and user, image is with user's
Relation is<image, image feature vector, user, relation classification>.
S130: according to the second relation of the relation of image and concept and image with user, determine to
Determine user to gather lower user and collect total concept set.
Such as: according to<image, concept set, image-conceptual relation classification>and<image, figure
As characteristic vector, user, image-customer relationship classification>, derive:<user collects, user
Collection characteristic set, user collects total concept set, and user collects total image set, and image-user is closed
It is classification, image-conceptual relation classification >.
S140: user characteristics set and user are integrated total concept set cooperation as training set, and with
The personality attribute truth set of user returns device as target, training, and uses based on visual angle
Gradient promotes decision Tree algorithms, filters out the concept set of concrete personality discrimination.
As example, describe gradient based on visual angle in detail and promote the process of decision Tree algorithms.
For each user, from the pictures that he likes, extract K concept,
This K concept can regard as K angle (the i.e. K character concept angle of the personality of this user
Degree), namely user x [x(1) ... x(i)... x(k)] T, each of which
x(i)=[xi(d-1)+1..., xid], the spy of the image under i-th concept in expression user's pictures
Levying, d represents the characteristic dimension of image.
The embodiment of the present invention is using gradient based on visual angle lifting decision tree (vGBDT) algorithm
Time, each take turns select base return device when, not only consider adjusting and optimizing parameter and output leaf
Child node, also calculates in epicycle iteration, will use which conceptual point.Wherein, vGBDT
The form that the base used returns device T is as follows:
Wherein, I () represents indicator function；J represents leafy node quantity；M represents iteration time
Number；VmRepresent the feature of the conceptual point chosen；RmRepresent the classification mode that study is arrived；AmTable
Show the leafy node value of output.
Assume piRepresent that user tests the personality attribute true value obtained through text based personality, excellent
Selection of land, pi∈[-4,4]。Wherein N represents training sample number.
Initialize and return by force device, i.e.
Residual error r is calculated according to below equationi:
ri=pi-Fm-1(xi)
Wherein, m represents iterations；xiRepresent user；I=1 ..., N.
According to below equation residual error returned and obtains weak recurrence device:
Weak recurrence device is added to strong recurrence in device, obtains new returning by force device, it may be assumed that
Fm(x)=Fm-1(x)+vT(x；Vm,Rm,Am)
Wherein, v represents weight.
As m=M, F=Fm。
Final recurrence device is:
Wherein, M represents that the base of use returns the number of device, namely the personality concept angle used
The number of degree；θ represents training set.
S150: use clustering algorithm, utilizes the concept set of the concrete personality discrimination filtered out,
In image collection from each concept, screening has the representational image of personality, asks as visualization
The image option of topic.
This step is for the image collection containing the identical concept filtered out, preferably by conviction
Propagation clustering (Affinity Propagation Clustering) algorithm carries out concrete concept hypograph choosing
The selection of item.
S160: by the image option under the concept filtered out and concept, generates based on image
Personality test questionnaire.
As example, after using vGBDT algorithm, M candidate concepts will be selected out,
This also represents the stem securing M exercise question.Such as select " cat " this concept, then originally
The stem of topic just can be expressed as: the following image containing " cat ", and which you like best and open.
For each concept, all belief propagation algorithm is used to screen its corresponding image set,
Select the cluster centre of J AP algorithm, namely J representative image.To be used for
M-th concept that predict user's personality true value p, that select, is designated asFor arbitrarily
ImageVGBDT algorithm can be used to allocate them to M base and to return deviceIn one.If J is 5, then can be byImage in concept set, is polymerized to 5
Class.Result such as Fig. 2 b, image is noted as 1-5.In order to prevent the image as exercise question option
Being falsely dropped, and have influence on the response of user, we use belief propagation (Affinity
Propagation, AP) algorithm, rightCluster, only with in maximum kind, lean on most
The image being noted as 1-5 of nearly cluster centre, such as Fig. 2 b.So far, by vGBDT algorithm
M the problem determined, and J the option determined by AP algorithm under M problem is
Complete, such as Fig. 3, complete product process based on image personality test questionnaire.
The investigation method of traditional text survey document, is to set multiple exercise questions, and experimenter is readding
When reading and understand, can bear bigger burden, experimenter also needs to certain personality science element
Support, the character trait of self can be told, and contrast with stem, answer similar or agree to
Degree；The stem of text questionnaire problem can express and significantly investigate intention, experimenter's meeting
Conscious make the answer favourable for self, conceal truth, cause a deviation；Text
Questionnaire or language-sensitive, in terms of language, relatively difficult.
The embodiment of the present invention is asked by setting the visualization of a series of " select you favorite image "
Topic, and collect user the answer option of problem is done, thus a kind of property based on image is provided
The investigation exchange method of lattice test questionnaire.As shown in Figure 4, the method may include that
S400: set the investigation form of questionnaire as a given basket, set containing phase
Answer the image option as problem of concept.
S410: set the interactive form of questionnaire as: user according to self understanding to image,
Select to meet the image of investigation stem.
S420: according to user's answer, utilizes and returns device, calculates big five character traits of user.
Using the embodiment of the present invention, user is only with answering a problem: " with hypograph, which
Opening is that you are favorite？" as shown in Figure 3.User no longer has obstacle for the understanding of questionnaire,
The investigation of questionnaire is intended to hide after images, and user can not directly do by image is conscious
The answer deviated.
In order to assess the present invention, test of heuristics result and the user of the present invention are used tradition by us
Personality test Questionnaire results contrasts, and use root-mean-square error is as module, all
Square error is between 1.232-1.796.
Although each step is carried out by above-described embodiment according to the mode of above-mentioned precedence
Describe, it will be recognized to those skilled in the art that for the effect realizing the present embodiment, no
With step between perform not necessarily in such order, its can simultaneously (parallel) execution or
Performing with reverse order, these simply change all within protection scope of the present invention.
The technical scheme provided the embodiment of the present invention above is described in detail.Although
Apply concrete individual example herein principle and the embodiment of the present invention are set forth, but,
The explanation of above-described embodiment is only applicable to help to understand the principle of the embodiment of the present invention；Meanwhile, right
For those skilled in the art, according to the embodiment of the present invention, in detailed description of the invention and should
All can make a change within the scope of with.
It should be noted that referred to herein to flow chart or block diagram be not limited solely to herein
Shown form, it can also divide and/or combine.
It can further be stated that: labelling and word in accompanying drawing are intended merely to be illustrated more clearly that this
Invention, is not intended as the improper restriction to scope.
Term " includes " or any other like term is intended to comprising of nonexcludability,
So that include that the process of a series of key element, method, article or equipment/device not only wrap
Include those key elements, but also include other key element being not expressly set out, or also include these
The key element that process, method, article or equipment/device are intrinsic.
Each step of the present invention can realize with general calculating device, and such as, they can
To concentrate on single calculating device, such as: personal computer, server computer, hands
Holding equipment or portable set, laptop device or multi-processor device, it is also possible to be distributed in
On the network that multiple calculating devices are formed, they can perform institute with the order being different from herein
The step illustrated or describe, or they are fabricated to respectively each integrated circuit modules, or
Multiple modules in them or step are fabricated to single integrated circuit module realize.Therefore,
The invention is not restricted to any specific hardware and software or it combines.
The method that the present invention provides can use PLD to realize, it is also possible to implements
For computer software or program module, (it includes performing particular task or realizing specific abstract
The routine of data type, program, object, assembly or data structure etc.), such as according to this
Inventive embodiment can be a kind of computer program, runs this computer program and makes
Computer performs to be used for demonstrated method.Described computer program includes computer-readable
Storage medium, this medium comprises computer program logic or code section, is used for realizing described
Method.Described computer-readable recording medium can be mounted built-in medium in a computer
Or the removable medium that can disassemble from basic computer is (such as: use hot plug
The storage device of technology).Described built-in medium includes but not limited to rewritable non-volatile deposit
Reservoir, such as: RAM, ROM, flash memory and hard disk.Described removable medium include but
Be not limited to: optical storage media (such as: CD-ROM and DVD), magnetic-optical storage medium (such as:
MO), magnetic storage medium (such as: tape or portable hard drive), have built-in rewritable non-easily
Lose property memorizer media (such as: storage card) and have built-in ROM media (such as:
ROM box).
The above, the only detailed description of the invention in the present invention, but protection scope of the present invention
It is not limited thereto, any is familiar with the people of this technology in the technical scope that disclosed herein,
It is appreciated that the conversion or replacement expected, all should contain within the scope of the comprising of the present invention, therefore,
Protection scope of the present invention should be as the criterion with the protection domain of claims.
Claims (4)
1. personality based on an image test questionnaire generates method, it is characterised in that
Described method at least includes:
Obtain image set and the personality attribute truth set of described user that user likes, and build
Described image and first relation of described user；
Image in the described image set liking described user carries out based on degree of depth study general
Read and extract, build the relation of described image and described concept；
Extract the characteristics of image of each image in described image set, and determine according to described characteristics of image
Image feature vector and user characteristics set, and it is described to utilize described image feature vector to expand
Image and first relation of user, build the second relation of image and user；
According to described image and the relation of concept and described image and second relation of user, determine
Given user gathers lower user and collects total concept set；
Described user characteristics set and described user are integrated total concept set cooperation as training set, and
Using the personality attribute truth set of described user as target, training returns device, and use based on
The gradient at visual angle promotes decision Tree algorithms, filters out the concept set of concrete personality discrimination；
Use clustering algorithm, the concept set of the concrete personality discrimination filtered out described in utilization,
In image collection from each concept, screening has the representational image of personality, asks as visualization
The image option of topic；
By the described image option under the described concept filtered out and described concept, generate described
Personality based on image test questionnaire.
Method the most according to claim 1, it is characterised in that described to described user
Image in the described image set liked carries out concept based on degree of depth study and extracts, and specifically includes:
Image in the described image set that described user is liked by employing deep neural network is carried out
Concept is extracted.
Method the most according to claim 1, it is characterised in that described employing is based on regarding
The gradient at angle promotes decision Tree algorithms, filters out the concept set of concrete personality discrimination, specifically
Including:
Initialize according to below equation and return by force device:
Wherein, described F0Represent and initially return by force device；Described piRepresent that user is through text based
Personality tests the personality attribute true value obtained；Described N represents training sample number；
Residual error r is calculated according to below equationi:
ri=pi-Fm-1(xi)
Wherein, described m represents iterations；DescribedDescribed K table
Show personality conceptual point number；Described d represents the characteristic dimension of described image；
According to below equation residual error returned and obtains weak recurrence device:
Weak recurrence device is added to strong recurrence in device, obtains the new device that returns by force:
Fm(x)=Fm-1(x)+vT(x；Vm,Rm,Am)
Wherein, described v represents weight；
Determine that final recurrence device is:
Wherein, described
Described T represents that base returns device；Described I () represents indicator function；Described J represents leafy node number
Amount；Described VmRepresent the feature of the conceptual point chosen；Described RmRepresent the classification that study is arrived
Pattern；Described AmRepresent the leafy node value of output；Described M represents that the base of use returns device
Number；Described θ represents training set；
The described final recurrence device being determined by, filters out described concrete personality discrimination
Concept set.
4. one kind utilizes the questionnaire that in the claims 1-3, arbitrary described method generates
Carry out investigating mutual method, it is characterised in that the mutual method of described investigation includes:
Set the investigation form of described questionnaire as a given basket, set containing corresponding
The image of concept is as the option of problem；
Set the interactive form of described questionnaire as: described user according to self to described image
Understanding, select meet investigation stem image；
According to user's answer, utilize described recurrence device, calculate big five character traits of described user.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610484985.9A CN105893784A (en) | 2016-06-28 | 2016-06-28 | Method for generating character test questionnaire based on image and surveying interactive method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610484985.9A CN105893784A (en) | 2016-06-28 | 2016-06-28 | Method for generating character test questionnaire based on image and surveying interactive method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105893784A true CN105893784A (en) | 2016-08-24 |
Family
ID=56719380
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610484985.9A Pending CN105893784A (en) | 2016-06-28 | 2016-06-28 | Method for generating character test questionnaire based on image and surveying interactive method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105893784A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106796613A (en) * | 2016-09-14 | 2017-05-31 | 深圳市大疆创新科技有限公司 | Control method, control device and electronic installation |
| CN108537624A (en) * | 2018-03-09 | 2018-09-14 | 西北大学 | A kind of tourist service recommendation method based on deep learning |
| CN109271544A (en) * | 2018-07-11 | 2019-01-25 | 中国科学院自动化研究所 | Automatically the method and device of artist's masterpiece is selected |
| CN110084318A (en) * | 2019-05-07 | 2019-08-02 | 哈尔滨理工大学 | A kind of image-recognizing method of combination convolutional neural networks and gradient boosted tree |
| CN113506601A (en) * | 2021-07-13 | 2021-10-15 | 北京美医医学技术研究院有限公司 | Big data-based skin tolerance analysis system |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1969552A (en) * | 2004-06-17 | 2007-05-23 | 皇家飞利浦电子股份有限公司 | Personalized summaries using personality attributes |
| GB2462449A (en) * | 2008-08-07 | 2010-02-10 | Greencannon Ltd | System and method for mood assessment |
| CN103106346A (en) * | 2013-02-25 | 2013-05-15 | 中山大学 | Character prediction system based on off-line writing picture division and identification |
-
2016
- 2016-06-28 CN CN201610484985.9A patent/CN105893784A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1969552A (en) * | 2004-06-17 | 2007-05-23 | 皇家飞利浦电子股份有限公司 | Personalized summaries using personality attributes |
| GB2462449A (en) * | 2008-08-07 | 2010-02-10 | Greencannon Ltd | System and method for mood assessment |
| CN103106346A (en) * | 2013-02-25 | 2013-05-15 | 中山大学 | Character prediction system based on off-line writing picture division and identification |
Non-Patent Citations (1)
| Title |
|---|
| JITAO SANG 等: "Visual BFI: an Exploratory Study for Image-based Personality Test", 《ARXIV》 * |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106796613A (en) * | 2016-09-14 | 2017-05-31 | 深圳市大疆创新科技有限公司 | Control method, control device and electronic installation |
| CN106796613B (en) * | 2016-09-14 | 2021-10-08 | 深圳市大疆创新科技有限公司 | Control method, control device and electronic device |
| CN108537624A (en) * | 2018-03-09 | 2018-09-14 | 西北大学 | A kind of tourist service recommendation method based on deep learning |
| CN108537624B (en) * | 2018-03-09 | 2021-08-03 | 西北大学 | Deep learning-based travel service recommendation method |
| CN109271544A (en) * | 2018-07-11 | 2019-01-25 | 中国科学院自动化研究所 | Automatically the method and device of artist's masterpiece is selected |
| CN109271544B (en) * | 2018-07-11 | 2020-11-13 | 中国科学院自动化研究所 | Method and device for automatically selecting painter representatives |
| CN110084318A (en) * | 2019-05-07 | 2019-08-02 | 哈尔滨理工大学 | A kind of image-recognizing method of combination convolutional neural networks and gradient boosted tree |
| CN113506601A (en) * | 2021-07-13 | 2021-10-15 | 北京美医医学技术研究院有限公司 | Big data-based skin tolerance analysis system |
| CN113506601B (en) * | 2021-07-13 | 2024-01-16 | 北京美医医学技术研究院有限公司 | Skin tolerance analysis system based on big data |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Baggio et al. | Quantitative methods in tourism: A handbook | |
| Recker | Scientific research in information systems: a beginner's guide | |
| Van Buuren | Flexible imputation of missing data | |
| Gelman et al. | Beyond subjective and objective in statistics | |
| Hussain et al. | Classification, clustering and association rule mining in educational datasets using data mining tools: A case study | |
| Robbins | Fixed categories in a portable landscape: the causes and consequences of land-cover categorization | |
| CN108804677B (en) | Deep learning problem classification method and system combining multi-level attention mechanism | |
| O'Donovan et al. | Exploratory font selection using crowdsourced attributes | |
| Wagstaff | Intelligent clustering with instance-level constraints | |
| Van der Lans et al. | Cross-national logo evaluation analysis: An individual-level approach | |
| CN105893784A (en) | Method for generating character test questionnaire based on image and surveying interactive method | |
| Szabo et al. | Review and use of learning theories within computer science education research: Primer for researchers and practitioners | |
| Rogel-Salazar | Data science and analytics with Python | |
| Johnston et al. | Applied Unsupervised Learning with Python: Discover hidden patterns and relationships in unstructured data with Python | |
| Mutar et al. | Smoke detection based on image processing by using grey and transparency features | |
| Herlambang et al. | Intelligent computing system to predict vocational high school student learning achievement using Naï ve Bayes algorithm | |
| Prosise | Applied machine learning and AI for engineers | |
| Buscema et al. | Intelligent Data Mining in Law Enforcement Analytics: New Neural Networks Applied to Real Problems | |
| Han | Research on personalized recommendation method of educational resources based on learner behavior analysis | |
| Whitlock et al. | Data science for ai leaders | |
| Fischetti et al. | R: Predictive Analysis | |
| Pavlekovic et al. | Comparison of intelligent systems in detecting a child’s mathematical gift | |
| Fischetti et al. | R: Data Analysis and Visualization | |
| Tandon et al. | Introduction to Artificial Intelligence using Python | |
| Vachkova et al. | The application of text mining algorithms to discover one topic objects in digital learning repositories |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20160824 |
|
| WD01 | Invention patent application deemed withdrawn after publication |