CN109325115A - A kind of role analysis method and analysis system - Google Patents
A kind of role analysis method and analysis system Download PDFInfo
- Publication number
- CN109325115A CN109325115A CN201810934896.9A CN201810934896A CN109325115A CN 109325115 A CN109325115 A CN 109325115A CN 201810934896 A CN201810934896 A CN 201810934896A CN 109325115 A CN109325115 A CN 109325115A
- Authority
- CN
- China
- Prior art keywords
- label
- numerical value
- information
- word
- drama
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/284—Lexical analysis, e.g. tokenisation or collocates
Abstract
This application discloses a kind of role analysis method and analysis systems, and wherein role analysis method includes: S2: the first label word is extracted in pre-set text；S4: the first label numerical value is determined according to the first label word and preset rules；S6: it is worth calculated result according to the second number of tags corresponding to the first label word numerical value and the second label word；S8: according to calculated result obtain the first label numerical value corresponding to the degree of correlation between the second information corresponding to the first information and the second label numerical value.Compared with prior art, role analysis method provided herein and analysis system can extract herein the first label word by that will preset, and by the label word assignment, the second label value corresponding to the assignment of the label word and the second label word is passed through to the matching degree that the two is calculated.Therefore, director can find out the performer for being most suitable for this drama automatically by the above method or system, while performer can also find most suitable drama by above system.
Description
Technical field
This application involves field of information processing, especially a kind of role analysis method and analysis system.
Background technique
Currently, there are many systems to the digitization analysis for star and drama in media industry market, but lack
One can cause to have had in somebody's hand by realizing the assessment of drama the matched system of performer of role in drama
The drama that drama can not still film it or TV play has either had still cannot select suitable performer to clap
At movie or television play.
Summary of the invention
The application's aims to overcome that the above problem or at least is partially solved or alleviates the above problem.
According to the one aspect of the application, a kind of role analysis method is provided, includes the following steps: S2: in default text
The first label word is extracted in this；S4: the first label numerical value is determined according to the first label word and preset rules；S6: according to institute
Stating the second number of tags corresponding to the first label word numerical value and the second label word is worth calculated result；S8: according to the calculating
As a result obtain the second information corresponding to the first information corresponding to the first label numerical value and the second label numerical value it
Between the degree of correlation.
Optionally, the step S6 includes: S61: according in the first label numerical value at least one with it is described
At least one of second label numerical value obtains the first calculated result；S62: according in the first label numerical value at least its
One of obtain the second calculated result at least one of the second label numerical value；The step S8 are as follows: according to described
First calculated result and second calculated result obtain the degree of correlation of the first information Yu second information.
Optionally, the step S61 is to pass through formula: COSA=(X1*Y1+X2*Y2+ ...+Xn*Yn)/((X1^2+X2^
2+ ...+Xn^2) ^1/2) * (y1^2+Y2^2+...+Yn^2) ^1/2 obtains first calculated result, the step S62 is logical
Cross formula: COSA=(X1*Y1+X2*Y2+ ...+Xn*Yn)/((X1^2+X2^2+ ...+Xn^2) ^1/2) * (y1^2+Y2^2+...
+ Yn^2) ^1/2 obtains second calculated result；Wherein A is angle, and A is more than or equal to 0 degree, is less than or equal to 90 degree, X1-Xn
For the first label numerical value, Y1-Yn is the second label numerical value, and A is bigger, then the first information is got over second information correlation
Low, included angle A is smaller, then the first information and second information correlation are higher.
Optionally, the step S8 includes: S81: multiple first calculated results are sorted according to preset rules, and
To matrix A 1；S82: multiple second calculated results are sorted according to preset rules, and obtain matrix B 1；S83: according to described
Matrix A 1 and the matrix B 1 obtain the degree of correlation of the first information Yu second information.
Optionally, the role analysis method further include: S10: pass through formula: COSA=(X1*Y1+X2*Y2+ ...+Xn*
Yn)/((X1^2+X2^2+ ...+Xn^2) ^1/2) * (y1^2+Y2^2+...+Yn^2) ^1/2, according in the first label numerical value
At least one at least one with third label numerical value corresponding to third label word obtain third calculated result,
Pass through formula: COSA=(X1*Y1+X2*Y2+ ...+Xn*Yn)/((X1^2+X2^2+ ...+Xn^2) ^1/2) * (y1^2+Y2^2
+ ...+Yn^2) ^1/2, according at least one in the first label numerical value with the third label numerical value at least its
One of obtain the 4th calculated result；Wherein A is angle, and A is more than or equal to 0 degree, is less than or equal to 90 degree, X1-Xn is the first mark
Numerical value is signed, Y1-Yn is third label numerical value, and A is bigger, then the first information is believed with third corresponding to the third label value
The degree of correlation of breath is lower, and included angle A is smaller, then the first information and the third information correlation are higher；S12: by multiple institutes
It states third calculated result to sort according to preset rules, and obtains matrix a1；S14: by multiple 4th calculated results according to pre-
If rule compositor, and obtain matrix b1；S16: calculating first goodness of fit according to formula A1=[M-a1]/M*100, and wherein A1 is the
One goodness of fit, M are the quantity of the third information；Second goodness of fit is calculated according to formula B1=[M-b1]/M*100, wherein B1
For second goodness of fit, M is the quantity of the third information；S18: comprehensive coincide is calculated according to formula S=0.35A1+0.65B1
Degree, wherein S is the comprehensive goodness of fit.
Optionally, the step S2 includes: S22: being segmented according to word frequency to the pre-set text is analyzed；S24: root
Participle database is established according to the participle；S26: the participle in the participle database is classified according to preset rules；S28:
The first label word is extracted in sorted participle.
According to the another aspect of the application, a kind of role analysis system is provided, comprising: the first client, for receiving
Pre-set text；Server connects first client, the pre-set text sent for receiving first client,
The first label word is extracted in the pre-set text, and the first label numerical value is determined according to the first label word and preset rules,
And it is worth the first information and institute corresponding to the first label numerical value according to the first label numerical value and the second number of tags
State the degree of correlation between the second information corresponding to the second label numerical value.
Optionally, the role analysis system further include: the second client connects the server, for receiving third
Label；The server passes through formula: COSA=(X1*Y1+X2*Y2+ ...+Xn*Yn)/((X1^2+X2^2+ ...+Xn^2) ^1/
2) * (y1^2+Y2^2+...+Yn^2) ^1/2, according at least one and the third mark in the first label numerical value
At least one of the corresponding third label numerical value of label obtains third calculated result, passes through formula: COSA=(X1*Y1+
X2*Y2+ ...+Xn*Yn)/((X1^2+X2^2+ ...+Xn^2) ^1/2) * (y1^2+Y2^2+...+Yn^2) ^1/2, according to described
At least one of at least one in one label numerical value with the third label numerical value obtains the 4th calculated result；Its
Middle A is angle, and A is more than or equal to 0 degree, is less than or equal to 90 degree, and X1-Xn is the first label numerical value, and Y1-Yn is third number of tags
Value, A is bigger, then the first information is lower with the degree of correlation of third information corresponding to the third label value, and included angle A is got over
Small, then the first information and the third information correlation are higher；The server presses multiple third calculated results
It sorts according to preset rules, and obtains matrix a1, multiple 4th calculated results are sorted according to preset rules, and obtain matrix
b1；First goodness of fit is calculated according to formula A1=[M-a1]/M*100, wherein A1 is first goodness of fit, and M is the third information
Quantity；Second goodness of fit is calculated according to formula B1=[M-b1]/M*100, wherein B1 is second goodness of fit, and M is the third
The quantity of information calculates the comprehensive goodness of fit according to formula S=0.35A1+0.65B1, and wherein S is the comprehensive goodness of fit.
According to the another aspect of the application, a kind of computer equipment is provided, including memory, processor and be stored in institute
State the computer program that can be run in memory and by the processor, wherein the processor executes the computer program
The described in any item methods of Shi Shixian such as preceding claim.
According to the another aspect of the application, a kind of computer readable storage medium, preferably non-volatile readable are provided
Storage medium, is stored with computer program, and the computer program is realized when executed by the processor as in claim
State described in any item methods.
Compared with prior art, role analysis method provided herein and analysis system can be by by default this paper
The first label word (such as leading role's label word, personality label word, Appearance tab word etc.) is extracted in (by taking drama as an example), and should
Label word assignment, by the second label value corresponding to the assignment of the label word and the second label word (personality, appearance of performer etc.)
By the matching degree that the two is calculated.Therefore, director can be found out automatically by the above method or system and be most suitable for this
The performer of drama, while performer can also find most suitable drama by above system.
Further, after working as the later predetermined performer of director, by the personality label word of predetermined performer, Appearance tab word etc.
After (third label word) input system, by drama and predetermined performer by the way that matching degree is calculated, and then can be intuitive
Show whether the predetermined performer meets the drama.
According to the accompanying drawings to the detailed description of the specific embodiment of the application, those skilled in the art will be more
Above-mentioned and other purposes, the advantages and features of the application are illustrated.
Detailed description of the invention
Some specific embodiments of the application are described in detail by way of example and not limitation with reference to the accompanying drawings hereinafter.
Identical appended drawing reference denotes same or similar part or part in attached drawing.It should be appreciated by those skilled in the art that these
What attached drawing was not necessarily drawn to scale.In attached drawing:
Fig. 1 is the flow chart according to the role analysis method of the application one embodiment；
Fig. 2 is the flow chart according to the role analysis method of the application one embodiment；
Fig. 3 is the flow chart according to the role analysis method of the application one embodiment；
Fig. 4 is the flow chart according to the role analysis method of the application one embodiment；
Fig. 5 is the flow chart according to the role analysis method of the application one embodiment；
Fig. 6 is the flow chart according to the role analysis method of the application one embodiment；
Fig. 7 is the flow chart according to the role analysis method of the application one embodiment；
Fig. 8 is the schematic diagram according to the role analysis system of the application one embodiment；
Fig. 9 is the schematic diagram according to the role analysis system of the application one embodiment；
Figure 10 is the schematic diagram according to the computer program product of the application one embodiment；
Figure 11 is the schematic diagram according to the computer readable storage medium of the application one embodiment.
Specific embodiment
According to the accompanying drawings to the detailed description of the specific embodiment of the application, those skilled in the art will be more
Above-mentioned and other purposes, the advantages and features of the application are illustrated.
In this application, pre-set text is drama, and the first information is the various roles in drama, and the first label word is drama
In various roles corresponding to feature (such as appearance, age, personality etc.), the second label word is corresponding to the performer that prestores
Feature (such as appearance, age, personality etc.), the second information are the performer prestored, and third label word is corresponding to the performer of input
The characteristics of (such as appearance, age, personality etc.), third information be input performer.But above-mentioned setting is only to be convenient for explaining
It only uses, is not used to limit the use environment and the scope of application of the application.In other use environments, mentioned as long as using the application
The analysis method and analysis system of confession, belong to the application it is claimed within the scope of.
It is an object of the present invention to have the group of conductive shadow demand, in face of student enrollment and network writer etc. for him
A drama that can upload oneself is provided, then system is role match performer in drama by the analysis to drama
APP.First, the films and television programs for completing them for the professional student in part provide convenience；Second, help the director for realizing groups of people
Dream.Third, for profession, director provides new reference model and method
To achieve the above object, the technical solution adopted by the present invention is that:
One drama assessment and role match system include: terminal access and Back Administration Module, terminal AM access module master
If user's drama uploads or the interface of actor information registration, Back Administration Module is in the form of establishing database, according to me
The various data of collected performer and its quantitative analysis, keep and maintenance actor data update, then by user upload play
This plot carries out intelligent, labeling processing, finally obtains for drama assessment as a result, system recommendation is present in data in turn
Matched performer comes out in library, to complete drama assessment and role match.
Drama role match system the following steps are included:
One, data are collected
(1) crawler software: (1) ForeSpider is grabbed from website；(2) octopus.
(2) it is introduced directly into existing text file
Two, natural language processing and participle
Step S1: it is segmented according to frequency.The decomposition of a word may have many methods, but the knot of each method gone out
Fruit centainly alternates betwwen good and bad.It needs to be analyzed according to the result of different methods gone out at this time.On this system proposed adoption network
Numerous segmenting methods carry out first step analysis.Assuming that 2 two kinds of segmenting methods of existing method one and method.Such as method one segments
After obtain { a1, b1, c1 } three parts, method two is segmented { a1, b2, c2 } ... next just according to probability, i.e. behind a1
The probability height for meeting b1 still connects the probability height of b2.This probability just depends on network or existing participle database etc..Frequency is high
It indicates usually to be commonly used, such as " school bag installs ", both may be considered " book " " packaging " " good ", and be also possible to " school bag "
" dress " " good " just uses latter segmenting method as one can imagine being that we are common for latter.
Step S2: establish participle database.The participle database mentioned in step S1, model are as follows.Wherein before a expression
One word section, b indicate the latter word section, and f indicates frequency, it is believed that are advisable with the number for remaining three after decimal point.This is sentenced
For sentence in step S1.
When a variety of situations occurs in word segmentation result, participle database is needed to refer to.Assuming that end value is k, the value of k
Equal to a kind of frequency summation of all connections in segmenting method.As word number of segment n=3, need to carry out database twice altogether
It searches.I.e. in the first segmenting method, need the lookup of primary " book " and " packaging ", and " packaging " and " good " is looked into
It looks for.F can be obtained by searching twice, and add up to obtain k1.Similarly second of segmenting method obtains end value k2.K1 and k2 are compared,
Bigger person indicates that participle is more reasonable, more meets expression way.
Notice that this participle database is needed by a large amount of drama training.
Step S3: the word segmentation result of processing step S1.It is certain there are many analyzing drama useless word section in word segmentation result,
Therefore it is screened and carrys out Simplified analysis.This system is quasi- to use for reference Skip-gram model, and the effect of the model is to skip certain words
To reduce to sentence.Such as in the present system, unnecessary auxiliary words such as " " " " is deleted.
By taking a lines in the small epoch as an example, " no, he is stayed on top of Mountain Tai.But, just it is female spy, also loses
When hand is overboard, moreover, also others is also dragged through the mud." segmented first according to step S1.
The result of the first segmenting method are as follows: no, he stays in, Mount Taishan, and on top, but, just, at last, female, spy also has,
It accidentally drops, it is overboard, when, moreover, going back handle, not, people is dragged, and is descended, water.
The result of second of segmenting method are as follows: no, he, firmly, safe, mountain top, on, but, just, at last, female, spy,
, have, accidentally drops, it is overboard, when, moreover, also, others, it drags, under, water.
By the result of step S2 frequency segmenting method are as follows: no, he, it stays in, Mount Taishan, on top, but, just, at last, female,
Spy, have, accidentally drops, when, moreover, handle is gone back, and others, it drags, under, water.
Using the processing of step S3, should finally result be obtained: no, he, it stays in, Mount Taishan, on top, but, just, at last,
Female, spy have, and accidentally drop, when, handle is gone back, others, it drags, under, water.
Step S4: part of speech analysis.Point hereafter or how many word sections example sentence is, if not carried out parts of speech classification, not
It can facilitate subsequent operation well.In this system, part of speech can be divided into following several: noun, verb, preposition, pronoun, number,
Conjunction, auxiliary word, interjection, quantifier, adverbial word, modal particle, onomatopoeia, character string, adjective, punctuation mark and custom words.It can
Come according to network method for the classification of word section.
Three, intelligent Matching and recommender system
According to drama role label and performer's label, according to certain algorithm, intelligent is carried out to performer and drama role
With simultaneously sort recommendations.
Intelligent Matching and recommended method step:
1, system typing performer data, and it is tagged.
2, director enters system upload drama, and screening system goes out drama dominant role, and tagged for role.
3, system is that drama role selects performer using certain matching way according to the label of drama role and the label of performer,
And sort, obtain most suitable top ten performer.
4, system, which allows to direct, checks all data of performer, and all or part for allowing to direct selection system recommendation is drilled
Member.
5, system informs that performer is selected in the certain role of the drama by way of short message or mail.
6, system allows performer to enter system and check the drama, and chooses whether interested in the drama, system after confirmation
It will notify to direct by way of short message or mail.
It further illustrates:
1.1 performer's intelligent tagging systems
One, labeling:
Classified first by the character trait to existing drama or practical personage, the description of personage in drama is roughly divided into
With Types Below: age, gender, nationality, macroscopic features, language, positive villain.
Secondly in system database performer and its performing art experience carry out labeling to it: the age, gender, nationality,
Macroscopic features, language, performer's bean vermicelli influence power, scandal coefficient, works quantity of taking part in a performance, are obtained and are drilled the positive villain's analysis of the role that takes part in a performance
Skill awards.
Two, labeling is illustrated:
1, the age (a)
Age-based section is matched: being for example a section according to 10 years old by the age, 10 sections were divided into from 1 years old to 100 years old.1-
10 years old are 1 point, and 11-20 years old is 2 points.And so on, 91-100 years old is 10 points.Very likely occur the fuzzy of age in drama to retouch
State and the otherness of the appearance at age performance, therefore the above classification carried out to the age according to the age bracket, and with performer year
Age bracket where age is matched.Such as there is the big situation of drama personage's age range, is occurred in drama according to all ages and classes
Number calculated, it is perhaps last to determine using the average value at age or by frequency or by other means
Age score.
Gender (b)
It is divided into male, women, male denaturation person, female denaturation person, corresponding score is 1,2,3,4 point.
3, nationality (c)
It is accurate to country, according to the sequencing of the national ranking in the publication civitas gentium maxima list of newest the United Nations, row
First it is national be 1 point, come the 2nd it is national be 2 points, and so on.
4, macroscopic features (d)
1), face
It include eyes, eyebrow, nose, mouth, ear in face.Setting rule can be carried out to these organs respectively.This
In by taking eyes as an example.Such as eyes have oxeye, pigsney, single-edge eyelid, double-edged eyelid etc. factor.Here only with it is aforementioned four because
For element, aforementioned four factor is combined.The combination one of 4 factors shares 15 kinds, by this combination assignment 1- respectively in 15
15 points.
2), stature feature
Height: being distinguished by height section, and is matched according to gender and description and height section: 105cm or less, 150-
160cm, 160-170cm, 170-180cm, 180-190cm, 190cm or more.According to assignment 1-6 points of above-mentioned 6 sections difference.
3), impediment
It is identical that principle is described with face, the combination of various defects is enumerated, then assignment both may be used.
Language (e)
Language can use method identical with international assignment, can also use other methods assignment.
6, positive villain (f)
The role positioning of drama personage is possible to be divided into: upright, villain, upright is reversed in neutral, villain's reversion neutrality
It is vertical, upright be reversed to villain, villain is reversed to upright 7 seed type (neutral reversion is according to the role positioning after reversion), according to
Above-mentioned assignment 1-7 points of 7 seed type difference.
By the database drama role analysis of the works of taking part in a performance of performer, such role positioning mode is equally taken, according to
It takes part in a performance the localization ratio of role, gives a mark to the role positioning of performer.For example, villain's score of certain performer=performer takes part in a performance
Villain's number * (role win a prize the situation)/performer for crossing works takes part in a performance all role's numbers；Prize-winning then * 2, nomination *
1.5, * 1 is normally performed, by dispute * 0.5.
Performer's bean vermicelli influence power (g)
The factors such as number are counted and commented on as evaluation factor according to its microblogging bean vermicelli quantity ranking and thumbing up for single microblogging, comment
Sentencing factor weight is respectively α, beta, gamma etc., and 50 first item factor weights are α before bean vermicelli quantity ranking, and ranking declines 50, then
Specific gravity decline 0.1, is at least 0.Or be combined according to above-mentioned factor, assignment is then carried out according to each case.
Scandal coefficient (h)
100 bad news and its party performer before the annual ranking of amount of reading are searched between taking nearly 10 years according to crawler technology,
A scandal exposure rate successively is made to every performer.In 1 year before scandal amount of reading ranking ten performers this year scandal system
Number is 0, and then coefficient increases by 0.1 for ranking decline 10, and exposure rate adds up to 1,0 exposure rate then h=1 during the decade.Or according to above-mentioned
Factor is combined, and then carries out assignment according to each case.
It takes part in a performance works quantity (i)
Works quantity of taking part in a performance includes take part in a performance TV play and film appearances quantity, and leading role, supporting role, group drill, and coefficient is respectively α,
Beta, gamma carries out ranking to the works quantity of taking part in a performance of performers all in database, and preceding 50 i=1, successively declining 50, then i is successively
0.1 is reduced, minimum value 0.Or be combined according to above-mentioned factor, assignment is then carried out according to each case.
Obtained performing art awards (j)
Obtained performing art awards include winning a prize, nominating, being shortlisted for, and coefficient is respectively α, beta, gamma, wherein win a prize, nominate, being shortlisted for
It further include a kind of awards, two class awards, three classes awards, a kind of awards established according to backstage, two class awards, three in awards type
Class awards database respectively to it is all kinds of win a prize, nominate, coefficient of being shortlisted for is divided into α 1, α 2, α 3, β 1, β 2, β 3, γ 1, γ 2, γ 3.
Respectively according to its win a prize number and coefficient carry out achievement make and, > 10 j=1, then j reduces 0.1, j minimum 0 for every decline 1.Or
Person is combined according to above-mentioned factor, then carries out assignment according to each case.
In the present embodiment, note: A, B, C, D, E are as external condition judgment criteria, and F, G, H, I, J are as artistic skills condition
Judgment criteria, but be not limited thereto.
2.1 drama role's intelligent tagging systems
One, drama bridge section analyzed, handled
1, entire drama is first cut into bridge section one by one (can directly divide or according to it according to the chapters and sections directory
His plot standard is divided)
2, each bridge section obtained to the first step carries out Chinese word segmentation, filters out unrelated word, obtains the pass in bridge section
Key word, the part of speech and weight of analysis keyword: weight is measured TF-IDF by science and is calculated, and analysis text is bridge section text
This, traverses each word, a number occurs in each bridge section using each word to obtain divided by total word number of place bridge section
TF1, TF2 ..., such as "Yes", "and", " in " etc. stop-words set its weight as 0, then calculate the inverse document frequency of this word,
Formula is IDF=log (D/Dw), and D is the total word number of bridge section, and the numerical value of Dw is this word appeared in all bridge sections of entire drama
Dw bridge section, that is to say Dw bridge sections containing this word, weight would not be carried out when not occurring this word in drama
Calculate, that is to say Dw there is no be 0 the case where；The weight of this final word is TF*IDF, as the weight of keyword.
3, label is added according to the part of speech of keyword and weight: is labeled using various dimensions label, by drama bridge section
It is divided into personage's level (capture sentence in object factory), people information level (by extraction event, describes etc. and to match the age,
Gender, nationality, macroscopic features), emotion character layer face (specific related to heart activity to mood swing) etc. passes through syntactic analysis
The correspondence text of emotion word combination corpus adds label；
Noun present in parsing sentence first as unit of each complete words, found from noun name and
His appellation of some you I has the sentence of name noun as analysis object, extracts entire sentence with basis using each
Information, emotion information.Each sentence subject is found by syntactic analysis.Such as: I met Kitty when interview
Once.She is the sootiness adornment for drawing exquisiteness, wears sexy skirt, carries the woman that brand B packet wraps class, I and Kitty；
And there is the only Kitty that the third person can occur, thereby confirm that the second word personage matching is Kitty.
Second is that crawl is using personage's subject as personage's essential information in the sentence of core: age, gender, nationality, appearance are special
Sign, language, positive villain, by the well-established character attribute tag library of the keyword match in sentence, such as: I is interviewing
When to meet Kitty primary.She is to draw the sexy skirt of an exquisite sootiness adornment, dress, carry brand B packet and wrap class
Woman, the noun keyword obtained in: woman brand B, skirt, verb keyword: working, interview, adjective keyword: property
Sense, it is exquisite；Matching: gender female, age 20-30 or so, macroscopic features: sexy exquisiteness (appearance are speculated thus according to tag library
Parameter setting is higher), other information not obtained are by the maximum automatic filling of personage's template probabilities known in tag library.
Leading man A be exactly that walk on the Week extending table of Milan, the lifeless handsome unmatched man of face, just
As we open those ineffable cloudy stubborn and unruly but beauty on the brand B that Fashion Magazines can all be seen or brand C advertisement every time
Obtain impeccable plane model.Analysis obtains: gender male, the age 30 is hereinafter, mixed-blood face, appearance parameter are higher.
Second is that going out emotion vocabulary in parsing sentence, character personality in drama is speculated with this, stop-word is first removed to sentence,
Then front is carried out to sentence, neutral, negative analysis is syncopated as the phrase that sentence has emotion later, and realizes structure
The front built up, the data of the positive and negative word of negative emotion word Chinese language shelves matching primitives (happiness, sorrow, hardship, is feared anger), by entire drama
The emotion score of middle each section is added, and obtains emotion behavior of the personage in drama, corresponding personage's affective tag；Such as male
Leading role A has only been drunk flatly, is just raised the head, and has been looked me up and down one minute with that double long and narrow eye, has then been shaken, not any
It says to expression: " repurchasing one glass again." after he just raise the head never again it is any if.Emotion vocabulary 4, be only respectively
It is, is long and narrow, shaking, without any expression, negation words quantity is 4, and negative accounting is 100%, and positive accounting is 0%, always
Body is positive to be divided into -0.01, the sentence emotion that leading man A is analyzed in entire drama is added to get the people about leading man A is arrived
Object affection data.
Two, it is mapped and is calculated according to the various dimensions label marked
1, by the character personality label of the level of emotion label mapping of drama bridge section to performer, personage's layer of drama bridge section
Face is mapped to character physical's label of performer etc., and the plot level of drama bridge section is mapped to the individual speciality level of personage, originally
The label weight and data of drama bridge section are also mapped onto performer's label, obtain performer's template label
2, mapping comes in obtained performer's template, and there are the weight of each label and numerical value, with drilling in actor data library
The weight and numerical value of member's label are calculated: because being the label of various dimensions here using calculating multidimensional characteristic vectors distance
Method,
About each tag parameter of personage's template be x1, x2, x3 ..., as leading man A personage's template label in by drama
Label storehouse matching obtains, and personage's appearance coefficient etc. of final mapping quantization is exactly x1, x2, x3 ..., and performer to be matched is each
A tag parameter is y1, y2, y3 ..., and parameter is the parameter values uploaded by user oneself here.
Here it using the method for calculating multidimensional characteristic vectors distance, is calculated using the cosine law: cosA=(x1*y1+x2*
Y2+x3*y3+ ...+xn*yn)/((x1^2+x2^2+ ...) ^1/2) * (y1^2+y2*2+ ...) ^1/2, calculate performer and bridge section
The angle for the performer's template come is mapped out, more related if angle is smaller, angle is bigger (to orthogonal 90 degree), then more not phase
It closes.
Such as it the mentioned-above age, gender, nationality, face, stature, impediment, language, positive villain, scandal, takes part in a performance
Works quantity obtains performing art awards etc. parameter.The above-mentioned parameter for wherein having middle role is x1-xn, and the performer's prestored is upper
Stating parameter is y1-yn.The corresponding drama role of x and y institute can be mapped with performer is prestored by the cosine law.It will
Each of one performer role and a drama bridge section map out the performer's template come and calculate, to obtained cosine value
It sums, the performer for finally obtaining the smallest cosine value of summation is the performer for being best suitable for this drama.Technology in the application
What scheme was obtained is the correlation between role and performer.Therefore the value of x and y is not to be the bigger the better, but closer to angle
Color and the correlation of performer are better.
In above-mentioned parameter, have plenty of about external condition, such as face, stature etc.；Have plenty of about artistic skills condition
, such as works quantity of taking part in a performance, acquisition performing art awards etc..Therefore, not Tongfang can be obtained by selecting different parameters to carry out calculating
The correlation in face.Such as it only selects in drama role about the parameter external condition corresponding with performer is prestored of external condition
Parameter is calculated, then the result drawn is exactly the correlation for prestoring performer with drama role in terms of external condition.Such as
Fruit only selects the parameter of the artistic skills condition corresponding with performer is prestored of the parameter in drama role about artistic skills part to calculate, that
The result drawn is exactly the correlation for prestoring performer with drama role in terms of artistic skills condition.Certainly, it character parameters and drills
Member's parameter is not limited in above-mentioned cited parameters, those skilled in the art can arbitrarily modify according to actual needs or
Increase and decrease above-mentioned parameter.So correspondingly, the correlation obtained is also not limited to above-mentioned external condition correlation and artistic skills item
Part correlation, other correlations obtained according to other parameters belong in the application range claimed.
3.1 intelligent Matchings and recommender system
External condition sequence A1 is carried out to all performers prestored in database
Artistic skills condition sequence B1 is carried out to all performers prestored in database
External condition is carried out to the performer being matched to sort to obtain a1
Artistic skills condition is carried out to the performer being matched to sort to obtain b1
Wherein, the performer being matched is the performer being temporarily added.Such as there is a drama to want to look for performer, but be not desired to use
The performer that prestores does not need to see whether the performer prestored is suitable, merely desires to look at whether some specific performer is suitable.This
When, this specific performer is exactly above-mentioned matched performer.Using the parameters such as the age-sex above-mentioned of the specific performer as
Y1-yn, it is corresponding by the cosine law with the x1-xn in drama.It just may know that the correlation for the performer and role being matched.If
The performer being equipped with has multiple, then is ranked up.
Then, goodness of fit calculating is carried out according to A1, B1, a1, b1, if the performer's sum being matched is M, external condition coincide
Degree calculates Aa=[M-a1]/M*100；The artistic skills condition goodness of fit calculates Bb=[M-b1]/M * 100；Comprehensive goodness of fit S=
0.35Aa+0.65Bb.Role match sequence is finally provided according to the comprehensive goodness of fit.
Finally being provided according to the comprehensive goodness of fit for director is suitble to the performer of the role to sort.
APP uses (terminal AM access module)
User is allowed to carry out the user's registration of APP, setting " I will work as director ", " I wants going on the stage " and " I will be as playwright, screenwriter " three
The interface of a mode allows user to select to enter on demand, to carry out information input and output.
User is allowed to carry out the user's registration of APP, setting " I will work as director ", " I wants going on the stage " and " I will ought write a play ",
The interface of " masses " four modes, allows user to select to enter on demand, to carry out information input and output.After each mode logs in
Different databases is corresponded to be checked and operated.
Direct end input (terminal AM access module)
The plate includes two databases, the actor data library that a backstage provides, the personal information (year including performer
Age, height, weight, photo (including certificate photo and whole body photograph), contact method), educational background (graduated school, achievement at school list,
Works and the role played in works in school), influence power and situation of winning a prize.The other is director data library, chooses for performer
Choosing director, the database include the personal information (age, height, weight, photo, contact method) of director, complete works situation
(scoring) and prize-winning situation.System registry director is audited.Every director is when being registered, need real name verification and
The certificate (credentials, prize-winning certificate etc.) for providing proof of identification, according to the prize-winning situation of the director, carries out backstage to the director
The scoring of prestige, the credit value initially write a play are X0, comment on, thumb up any drama credit value increase x, finally obtain final prestige
Value X.
The calculation method of credit value are as follows:
The achievable function of the plate has: checking and handles message and (checks performer's self-recommendation message, performer's refusal or receiving are invited
About, the information such as playwright, screenwriter refuses or receiving is invited)；It uploads drama and performer is selected according to drama；Issue wanted advertisement (recruitment performer
It is then shown in performer's connection interface, recruitment playwright, screenwriter is then shown in playwright, screenwriter's connection interface)；Search is write a play according to drama label lookup
Drama is uploaded；Drama (specifying certain playwright, screenwriter that him is allowed to create drama according to demand) is invited to playwright, screenwriter；It can read and write a play in this APP
The drama of upload simultaneously provides scoring or comment (can have collection drama function)；The drama or the play oneself uploaded that selection playwright, screenwriter uploads
This, carries out drama analysis and selects performer (thus issue and invite to performer and playwright, screenwriter)
Performer end inputs (terminal AM access module)
Performer fills in personal information (age, height, weight, photo (including certificate photo and whole body photograph), connection when registering
Mode), educational background (graduated school, achievement at school list, in school works and the role played in works), influence power, feelings of winning a prize
Condition establishes actor data library.System registry performer audits.When every performer registers, need real name verification and
The certificate (credentials, prize-winning certificate etc.) for providing proof of identification, according to the prize-winning situation of the performer, carries out backstage to the performer
The scoring of prestige, the credit value of initial performer are Y0, comment on, thumb up any drama credit value increase y, finally obtain final prestige
Value Y.
Following functions can be achieved after registering and logging in performer: checking and handling director's message request (according to drama or can lead
It drills directly refusal or receives)；Check director information and the scenario information that director provides；Check director's wanted advertisement；It is sent out Xiang director
Send message (performer can be shown on trial self-recommendation by wanted advertisement to this director) can read upload of writing a play in this APP drama and to
It scores or comments on out；
Playwright, screenwriter's end input (terminal AM access module)
User uploads the personal information of oneself, once had those works, what awards is obtained, plate user can look into
It sees and handles message (director invites information, the comment and ranking of uploaded drama)；Upload drama (can also be divided according to drama
It analyses and checks the performer role met with this drama role but message request can not be sent to performer)；Message is sent (certainly Xiang director
Recommend drama)；It reads and writes a play uploaded drama and provide comment and scoring.Increase in the drama in the software systems script library
Hold audit.The legal judgement of copyright and content is carried out to the contents of the section.Copyright judgement user can be inputted/upload drama
Duplicate checking detection is carried out, value of coincideing is greater than some threshold value and is then determined as invalid drama, cannot be loaded into database.Legal judgement:
Sensibility judgement is carried out to the participle of the drama content after carrying out natural language processing to drama, then needs to establish a sensitive word
Library, if the frequency of occurrences of sensitive vocabulary is greater than some threshold in drama, which is determined as invalid drama, cannot be loaded into number
According in library, and the high drama of the frequency of occurrences converged to sensitive word and carries out manual examination and verification, it is ensured that the legitimacy and safety of drama.Initially
The credit value of playwright, screenwriter is Z0, comments on, thumbs up any drama credit value increase z, finally obtains final credit value Z.
The masses' still browsable system homepage content in no registration or login.User can after registration logs in
Open drama is read in hall；There is the collection of itself, can be used to collect scenario information and drama is commented
By with thumb up；Resume can be delivered to play staff utility man channel；Issuing advertisement information, while server-side pair can be contacted with server-side
Advertising information carries out reasonable charge.And the masses need real name verification in registration, such as upload own identification card, student's identity card, mobile phone
Short-message verification etc..The credit value of the initial masses is K0, comments on, thumbs up any drama credit value increase k, finally obtains final prestige
Value K.
Drama quantifies (Back Administration Module)
Intelligent fractionation is carried out to the drama that director uploads, garbage is weeded out, grabs core information and the pass of drama
Key node carries out intelligent label processing, realizes for the analysis of the plot bridge section of drama and drama type itself and element
Positioning, the digitization in this, as drama assessment are analyzed, to direct the drama assessment for showing a labeling, digitization.
User is also required to drama comment content to carry out content legality judgement, and content legality can be to comment content uploading
Unite and according to the comment number of every user, thumb up scoring of the several and authentication situation to user's progress backstage prestige, with
When family scores to drama, the prestige scoring according to every user carries out weight addition, letter to the scoring situation of every user
It is bigger to praise higher user's scoring weight, finally obtains the final score of portion's drama.Specific calculation is as follows: in formula
In Weighted rank (WR)=[v ÷ (v+m) ÷ R+ (m ÷ (v+m)] ÷ C, R is averaged with what common calculation method obtained
Value, v is voter turnout, and including the talent conference only often voted is counted, m is to enter U.S. IMDb, the scoring such as bean cotyledon website
Required minimum poll, even the film full marks of only two single peoples ballot are not also used, c is the flat of current all films
Respectively, and only the talent conference that credit value is higher than some threshold value is calculated within effectively scoring, i.e., the range of marking people
The main method being limited within the scope of the ballot of senior movie-buff as far as possible, and the requirement Standard General for marking person can be limited to sight
See how many films or more, the film for participating in marking is more than how many or more and mays be eligible to vote.And not for credit value
Together, user's ballot quantity is the practical quantity of voting of the user multiplied by the absolute credit value of the user, the i.e. higher user of credit value
Share of voting is more, and credit value is lower, and user ballot share is fewer.By using the algorithm of such complexity, film can be allowed to comment
Divide deviation average, film is voted, and number is more, its scoring is just closer to true average mark, otherwise with regard to closer institute
There is the average mark of film.
Database (Back Administration Module)
Four databases are established, realize different functions respectively.
Seven databases are established, realize different functions respectively.(newly-increased sensitivity dictionary, is drilled at director's awards taxonomy database
Member's awards taxonomy database)
Both data matches (Back Administration Module)
According to the organic connections of the digitization analysis of the plot element of drama and label and drama role, using to both sides'
Label carries out linking matched method, different further according to the matching degree of each label, provides each label and sets different power
Reassignment designs the algorithm of weighting to be matched, is the intelligent label and analysis that drama adds with us, to match me
Before establish lane database storage performer personal information data and label, the drama transmitted with this side of director that sorts out
The suitable character list of the role playing of middle personage.
Direct end output (terminal AM access module)
The available performer's list by being obtained after Data Matching in director end, then chained in list and obtain performer's
The related datas such as personal information and label, are referred to this for director side, so as to allow director oneself come confirm performer whether with
Uploaded drama has more other meeting points, while if director wants to select this performer, can thus obtain the performer's
Contact method.So far, which just completes assessment and the role match of drama.
Please refer to Fig. 1, in one embodiment of the application, role analysis method includes the following steps:
S2: the first label word is extracted in pre-set text；
S4: the first label numerical value is determined according to the first label word and preset rules；
S6: it is worth to calculate knot according to the second number of tags corresponding to the first label word numerical value and the second label word
Fruit；
S8: according to the calculated result obtain the first label numerical value corresponding to the first information and second label
The degree of correlation between second information corresponding to numerical value.
Referring to figure 2., in one embodiment of the application, the step S6 includes:
S61: according at least one in the first label numerical value with the second label numerical value at least within
One of obtain the first calculated result；
S62: according at least one in the first label numerical value with the second label numerical value at least within
One of obtain the second calculated result；
The step S8 are as follows: according to first calculated result and second calculated result obtain the first information with
The degree of correlation of second information.
In the present embodiment, the step S61 is to pass through formula:
COSA=(X1*Y1+X2*Y2+ ...+Xn*Yn)/((X1^2+X2^2+ ...+Xn^2) ^1/2) * (y1^2+ Y2^2+ ...
+ Yn^2) ^1/2 obtains first calculated result, the step S62 is to pass through formula:
COSA=(X1*Y1+X2*Y2+ ...+Xn*Yn)/((X1^2+X2^2+ ...+Xn^2) ^1/2) * (y1^2+ Y2^2+ ...
+ Yn^2) ^1/2 obtains second calculated result；Wherein A is angle, and A is more than or equal to 0 degree, is less than or equal to 90 degree, X1-Xn
For the first label numerical value, Y1-Yn is the second label numerical value, and A is bigger, then the first information is got over second information correlation
Low, included angle A is smaller, then the first information and second information correlation are higher, but are not limited thereto.
Referring to figure 3., in one embodiment of the application, the step S8 includes:
S81: multiple first calculated results are sorted according to preset rules, and obtain matrix A 1；
S82: multiple second calculated results are sorted according to preset rules, and obtain matrix B 1；
S83: the degree of correlation of the first information Yu second information is obtained with the matrix B 1 according to the matrix A 1.
Referring to figure 4., in one embodiment of the application, the role analysis method further include:
S10: pass through formula:
COSA=(X1*Y1+X2*Y2+ ...+Xn*Yn)/((X1^2+X2^2+ ...+Xn^2) ^1/2) * (y1^2+ Y2^2+ ...
+ Yn^2) ^1/2, according at least one and the third number of tags corresponding to third label word in the first label numerical value
At least one of value obtains third calculated result, passes through formula:
COSA=(X1*Y1+X2*Y2+ ...+Xn*Yn)/((X1^2+X2^2+ ...+Xn^2) ^1/2) * (y1^2+ Y2^2+ ...
+ Yn^2) ^1/2, according at least one in the first label numerical value with the third label numerical value at least within it
One obtains the 4th calculated result；Wherein A is angle, and A is more than or equal to 0 degree, is less than or equal to 90 degree, X1-Xn is the first number of tags
Value, Y1-Yn are third label numerical value, and A is bigger, then the first information and third information corresponding to the third label value
The degree of correlation is lower, and included angle A is smaller, then the first information and the third information correlation are higher；
S12: multiple third calculated results are sorted according to preset rules, and obtain matrix a1；
S14: multiple 4th calculated results are sorted according to preset rules, and obtain matrix b1；
S16: first goodness of fit is calculated according to formula A1=[M-a1]/M*100, wherein A1 is first goodness of fit, and M is described
The quantity of third information；Second goodness of fit is calculated according to formula B1=[M-b1]/M*100, wherein B1 is second goodness of fit, and M is
The quantity of the third information；
S18: calculating the comprehensive goodness of fit according to formula S=0.35A1+0.65B1, and wherein S is the comprehensive goodness of fit.
Referring to figure 5., in one embodiment of the application, the step S2 includes:
S22: it is segmented according to word frequency to the pre-set text is analyzed；
S24: participle database is established according to the participle；
S26: the participle in the participle database is classified according to preset rules；
S28: the first label word is extracted in participle after sorting.
Fig. 6 is please referred to, in one embodiment of the application, further includes: S25: simplifying the participle database.
Fig. 7 is please referred to, in one embodiment of the application, the first label word is extracted in pre-set text includes:
S21: crawler software is obtained
S22: the first label word is extracted in pre-set text by the crawler software
Fig. 8-Fig. 9 is please referred to, in one embodiment of the application, provides a kind of role analysis system, which is characterized in that packet
It includes: the first client, for receiving pre-set text；Server connects first client, for receiving first client
The pre-set text sent is held, the first label word is extracted in the pre-set text, according to the first label word and is preset
Rule determines the first label numerical value, and is worth the first label numerical value according to the first label numerical value and the second number of tags
The degree of correlation between second information corresponding to the corresponding first information and the second label numerical value.
In one embodiment of the application, the role analysis system further include: the second client connects the server,
For receiving third label；The server passes through formula: COSA=(X1*Y1+X2*Y2+ ...+Xn*Yn)/((X1^2+X2^2
+ ...+Xn^2) ^1/2) * (y1^2+Y2^2+...+Yn^2) ^1/2, according at least one in the first label numerical value
Third calculated result is obtained at least one of third label numerical value corresponding to the third label, passes through formula:
COSA=(X1*Y1+X2*Y2+ ...+Xn*Yn)/((X1^2+X2^2+ ...+Xn^2) ^1/2) * (y1^2+Y2^2+...+Yn^2) ^
1/2, at least one according at least one in the first label numerical value with the third label numerical value obtains
4th calculated result；Wherein A is angle, and A is more than or equal to 0 degree, is less than or equal to 90 degree, X1-Xn is the first label numerical value, Y1-
Yn is third label numerical value, and A is bigger, then the degree of correlation of the first information and third information corresponding to the third label value
Lower, included angle A is smaller, then the first information and the third information correlation are higher；The server is by multiple described
Three calculated results sort according to preset rules, and obtain matrix a1, and multiple 4th calculated results are arranged according to preset rules
Sequence, and obtain matrix b1；First goodness of fit is calculated according to formula A1=[M-a1]/M*100, wherein A1 is first goodness of fit, M
For the quantity of the third information；Second goodness of fit is calculated according to formula B1=[M-b1]/M*100, wherein B1 is second identical
Degree, M are the quantity of the third information, calculate the comprehensive goodness of fit according to formula S=0.35A1+0.65B1, wherein S is synthesis
The goodness of fit.
Please refer to Figure 10, in one embodiment of the application, computer equipment, including memory, processor and be stored in described
In memory and the computer program that can be run by the processor, wherein when the processor execution computer program
Realize such as the described in any item methods of preceding claim.
Please refer to Figure 11, in one embodiment of the application, the storage of computer readable storage medium, preferably non-volatile readable
Medium, is stored with computer program, and the computer program realizes that preceding claim such as is appointed when executed by the processor
Method described in one.
Compared with prior art, role analysis method provided herein and analysis system can be by by default this paper
The first label word (such as leading role's label word, personality label word, Appearance tab word etc.) is extracted in (by taking drama as an example), and should
Label word assignment, by the second label value corresponding to the assignment of the label word and the second label word (personality, appearance of performer etc.)
By the matching degree that the two is calculated.Therefore, director can be found out automatically by the above method or system and be most suitable for this
The performer of drama, while performer can also find most suitable drama by above system.
Further, after working as the later predetermined performer of director, by the personality label word of predetermined performer, Appearance tab word etc.
After (third label word) input system, by drama and predetermined performer by the way that matching degree is calculated, and then can be intuitive
Show whether the predetermined performer meets the drama.
In the above-described embodiments, can come wholly or partly by software, hardware, firmware or any combination thereof real
It is existing.When implemented in software, it can entirely or partly realize in the form of a computer program product.The computer program
Product includes one or more computer instructions.When computer loads and executes the computer program instructions, whole or portion
Ground is divided to generate according to process or function described in the embodiment of the present application.The computer can be general purpose computer, dedicated computing
Machine, computer network obtain other programmable devices.The computer instruction can store in computer readable storage medium
In, or from a computer readable storage medium to the transmission of another computer readable storage medium, for example, the computer
Instruction can pass through wired (such as coaxial cable, optical fiber, number from a web-site, computer, server or data center
User's line (DSL)) or wireless (such as infrared, wireless, microwave etc.) mode to another web-site, computer, server or
Data center is transmitted.The computer readable storage medium can be any usable medium that computer can access or
It is comprising data storage devices such as one or more usable mediums integrated server, data centers.The usable medium can be with
It is magnetic medium, (for example, floppy disk, hard disk, tape), optical medium (for example, DVD) or semiconductor medium (such as solid state hard disk
Solid State Disk (SSD)) etc..
Professional should further appreciate that, described in conjunction with the examples disclosed in the embodiments of the present disclosure
Unit and algorithm steps, can be realized with electronic hardware, computer software, or a combination of the two, hard in order to clearly demonstrate
The interchangeability of part and software generally describes each exemplary composition and step according to function in the above description.
These functions are implemented in hardware or software actually, the specific application and design constraint depending on technical solution.
Professional technician can use different methods to achieve the described function each specific application, but this realization
It is not considered that exceeding scope of the present application.
In addition, the system database legalizes, secret.After customer data base content is always ensured that obtain user's license
The data obtained, and promise to undertake to the data base encryption and carry out safety control, not external leak data content of trying one's best.
Those of ordinary skill in the art will appreciate that implement the method for the above embodiments be can be with
By program come instruction processing unit completion, the program be can store in computer readable storage medium, and the storage is situated between
Matter is non-transitory (English: non-transitory) medium, such as random access memory, read-only memory, flash
Device, hard disk, solid state hard disk, tape (English: magnetic tape), floppy disk (English: floppy disk), CD (English:
Optical disc) and any combination thereof.
The preferable specific embodiment of the above, only the application, but the protection scope of the application is not limited thereto,
Within the technical scope of the present application, any changes or substitutions that can be easily thought of by anyone skilled in the art,
Should all it cover within the scope of protection of this application.Therefore, the protection scope of the application should be with scope of protection of the claims
Subject to.
Claims (10)
1. a kind of role analysis method, which comprises the steps of:
S2: the first label word is extracted in pre-set text；
S4: the first label numerical value is determined according to the first label word and preset rules；
S6: it is worth calculated result according to the second number of tags corresponding to the first label word numerical value and the second label word；
S8: according to the calculated result obtain the first label numerical value corresponding to the first information and the second label numerical value
The degree of correlation between the second corresponding information.
2. role analysis method according to claim 1, which is characterized in that the step S6 includes:
S61: at least one according at least one in the first label numerical value with the second label numerical value
Obtain the first calculated result；
S62: at least one according at least one in the first label numerical value with the second label numerical value
Obtain the second calculated result；
The step S8 are as follows: according to first calculated result and second calculated result obtain the first information with it is described
The degree of correlation of second information.
3. role analysis method according to claim 2, which is characterized in that the step S61 is to pass through formula:
COSA=(X1*Y1+X2*Y2+ ...+Xn*Yn)/((X1^2+X2^2+ ...+Xn^2) ^1/2) * (y1^2+Y2^2+ ...+Yn^
2) ^1/2 obtains first calculated result, and the step S62 is to pass through formula:
COSA=(X1*Y1+X2*Y2+ ...+Xn*Yn)/((X1^2+X2^2+ ...+Xn^2) ^1/2) * (y1^2+Y2^2+ ...+Yn^
2) ^1/2 obtains second calculated result；Wherein A is angle, and A is more than or equal to 0 degree, less than or equal to 90 degree, X1-Xn the
One label numerical value, Y1-Yn are the second label numerical value, and A is bigger, then the first information is lower with second information correlation,
Included angle A is smaller, then the first information and second information correlation are higher.
4. role analysis method according to claim 3, which is characterized in that the step S8 includes:
S81: multiple first calculated results are sorted according to preset rules, and obtain matrix A 1；
S82: multiple second calculated results are sorted according to preset rules, and obtain matrix B 1；
S83: the degree of correlation of the first information Yu second information is obtained with the matrix B 1 according to the matrix A 1.
5. role analysis method according to claim 4, which is characterized in that the role analysis method further include:
S10: pass through formula:
COSA=(X1*Y1+X2*Y2+ ...+Xn*Yn)/((X1^2+X2^2+ ...+Xn^2) ^1/2) * (y1^2+Y2^2+ ...+Yn^
2) ^1/2, according at least one and the third label numerical value corresponding to third label word in the first label numerical value
At least one obtains third calculated result, passes through formula:
COSA=(X1*Y1+X2*Y2+ ...+Xn*Yn)/((X1^2+X2^2+ ...+Xn^2) ^1/2) * (y1^2+Y2^2+ ...+Yn^
2) ^1/2, at least one according at least one in the first label numerical value with the third label numerical value obtain
4th calculated result out；Wherein A is angle, and A is more than or equal to 0 degree, is less than or equal to 90 degree, and X1-Xn is the first label numerical value,
Y1-Yn is third label numerical value, and A is bigger, then the phase of the first information and third information corresponding to the third label value
Guan Du is lower, and included angle A is smaller, then the first information and the third information correlation are higher；
S12: multiple third calculated results are sorted according to preset rules, and obtain matrix a1；
S14: multiple 4th calculated results are sorted according to preset rules, and obtain matrix b1；
S16: calculating first goodness of fit according to formula A1=[M-a1]/M*100, and wherein A1 is first goodness of fit, and M is the third
The quantity of information；Second goodness of fit is calculated according to formula B1=[M-b1]/M*100, wherein B1 is second goodness of fit, and M is described
The quantity of third information；
S18: calculating the comprehensive goodness of fit according to formula S=0.35A1+0.65B1, and wherein S is the comprehensive goodness of fit.
6. role analysis method according to claim 1, which is characterized in that the step S2 includes:
S22: it is segmented according to word frequency to the pre-set text is analyzed；
S24: participle database is established according to the participle；
S26: the participle in the participle database is classified according to preset rules；
S28: the first label word is extracted in participle after sorting.
7. a kind of role analysis system characterized by comprising
First client, for receiving pre-set text；
Server connects first client, the pre-set text sent for receiving first client, described
The first label word is extracted in pre-set text, and the first label numerical value is determined according to the first label word and preset rules, and according to
The first label numerical value and the second number of tags are worth the first information and described second corresponding to the first label numerical value
The degree of correlation between second information corresponding to label numerical value.
8. role analysis system according to claim 7, which is characterized in that the role analysis system further include:
Second client connects the server, for receiving third label；
The server passes through formula:
COSA=(X1*Y1+X2*Y2+ ...+Xn*Yn)/((X1^2+X2^2+ ...+Xn^2) ^1/2) * (y1^2+Y2^2+ ...+Yn^
2) ^1/2, according at least one and the third label numerical value corresponding to the third label in the first label numerical value
At least one obtain third calculated result, pass through formula:
COSA=(X1*Y1+X2*Y2+ ...+Xn*Yn)/((X1^2+X2^2+ ...+Xn^2) ^1/2) * (y1^2+Y2^2+ ...+Yn^
2) ^1/2, at least one according at least one in the first label numerical value with the third label numerical value obtain
4th calculated result out；Wherein A is angle, and A is more than or equal to 0 degree, is less than or equal to 90 degree, and X1-Xn is the first label numerical value,
Y1-Yn is third label numerical value, and A is bigger, then the phase of the first information and third information corresponding to the third label value
Guan Du is lower, and included angle A is smaller, then the first information and the third information correlation are higher；The server is by multiple institutes
It states third calculated result to sort according to preset rules, and obtains matrix a1, by multiple 4th calculated results according to default rule
It then sorts, and obtains matrix b1；First goodness of fit is calculated according to formula A1=[M-a1]/M*100, wherein A1 is first identical
Degree, M are the quantity of the third information；Second goodness of fit is calculated according to formula B1=[M-b1]/M*100, wherein B1 is second
The goodness of fit, M are the quantity of the third information, calculate the comprehensive goodness of fit according to formula S=0.35A1+0.65B1, wherein S is comprehensive
Close the goodness of fit.
9. a kind of computer equipment, which is characterized in that in the memory and can be by institute including memory, processor and storage
State the computer program of processor operation, wherein the processor realizes such as claim 1-6 when executing the computer program
Any one of described in method.
10. a kind of computer readable storage medium, which is characterized in that preferably non-volatile readable storage medium, interior storage
There is computer program, the computer program realizes such as side of any of claims 1-6 when executed by the processor
Method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810934896.9A CN109325115B (en) | 2018-08-16 | 2018-08-16 | Role analysis method and analysis system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810934896.9A CN109325115B (en) | 2018-08-16 | 2018-08-16 | Role analysis method and analysis system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109325115A true CN109325115A (en) | 2019-02-12 |
| CN109325115B CN109325115B (en) | 2021-03-09 |
Family
ID=65264181
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810934896.9A Active CN109325115B (en) | 2018-08-16 | 2018-08-16 | Role analysis method and analysis system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109325115B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110990709A (en) * | 2019-12-17 | 2020-04-10 | 北京奇艺世纪科技有限公司 | Role automatic recommendation method and device and electronic equipment |
| CN111062435A (en) * | 2019-12-13 | 2020-04-24 | 北京奇艺世纪科技有限公司 | Image analysis method and device and electronic equipment |
| CN112035703A (en) * | 2020-08-31 | 2020-12-04 | 西安君悦网络科技有限公司 | Method and system for searching actors in short video |
| CN113748439A (en) * | 2019-05-20 | 2021-12-03 | 索尼集团公司 | Prediction of successful quotient for motion pictures |
| CN113792547A (en) * | 2021-07-28 | 2021-12-14 | 中国科学院自动化研究所 | Corner selection method and system for film and television works |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101923689A (en) * | 2009-06-15 | 2010-12-22 | 中国移动通信集团公司 | Method for determining advertising information launched audience and related device thereof |
| US20150088911A1 (en) * | 2013-09-25 | 2015-03-26 | Alibaba Group Holding Limited | Method and system for extracting user behavior features to personalize recommendations |
| CN105631025A (en) * | 2015-12-29 | 2016-06-01 | 腾讯科技（深圳）有限公司 | Normalization processing method and device for query tags |
| CN106033444A (en) * | 2015-03-16 | 2016-10-19 | 北京国双科技有限公司 | Method and device for clustering text content |
| US20170132237A1 (en) * | 2015-11-09 | 2017-05-11 | Institute For Information Industry | Display system, method and computer readable recording media for an issue |
-
2018
- 2018-08-16 CN CN201810934896.9A patent/CN109325115B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101923689A (en) * | 2009-06-15 | 2010-12-22 | 中国移动通信集团公司 | Method for determining advertising information launched audience and related device thereof |
| US20150088911A1 (en) * | 2013-09-25 | 2015-03-26 | Alibaba Group Holding Limited | Method and system for extracting user behavior features to personalize recommendations |
| CN106033444A (en) * | 2015-03-16 | 2016-10-19 | 北京国双科技有限公司 | Method and device for clustering text content |
| US20170132237A1 (en) * | 2015-11-09 | 2017-05-11 | Institute For Information Industry | Display system, method and computer readable recording media for an issue |
| CN105631025A (en) * | 2015-12-29 | 2016-06-01 | 腾讯科技（深圳）有限公司 | Normalization processing method and device for query tags |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113748439A (en) * | 2019-05-20 | 2021-12-03 | 索尼集团公司 | Prediction of successful quotient for motion pictures |
| CN113748439B (en) * | 2019-05-20 | 2024-03-12 | 索尼集团公司 | Prediction of successful quotient of movies |
| CN111062435A (en) * | 2019-12-13 | 2020-04-24 | 北京奇艺世纪科技有限公司 | Image analysis method and device and electronic equipment |
| CN110990709A (en) * | 2019-12-17 | 2020-04-10 | 北京奇艺世纪科技有限公司 | Role automatic recommendation method and device and electronic equipment |
| CN110990709B (en) * | 2019-12-17 | 2023-07-21 | 北京奇艺世纪科技有限公司 | Role automatic recommendation method and device and electronic equipment |
| CN112035703A (en) * | 2020-08-31 | 2020-12-04 | 西安君悦网络科技有限公司 | Method and system for searching actors in short video |
| CN113792547A (en) * | 2021-07-28 | 2021-12-14 | 中国科学院自动化研究所 | Corner selection method and system for film and television works |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109325115B (en) | 2021-03-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Lucy et al. | Content analysis of textbooks via natural language processing: Findings on gender, race, and ethnicity in Texas US history textbooks | |
| CN109325115A (en) | A kind of role analysis method and analysis system | |
| CN111177569B (en) | Recommendation processing method, device and equipment based on artificial intelligence | |
| WO2021174890A1 (en) | Data recommendation method and apparatus, and computer device and storage medium | |
| CN107491531B (en) | Chinese network comment sensibility classification method based on integrated study frame | |
| Kaushik et al. | A comprehensive study of text mining approach | |
| CN104281699B (en) | Method and device is recommended in search | |
| Yangarber | Counter-training in discovery of semantic patterns | |
| US6856986B1 (en) | Answer collection and retrieval system governed by a pay-off meter | |
| Munk et al. | The Thick Machine: Anthropological AI between explanation and explication | |
| JP2012027845A (en) | Information processor, relevant sentence providing method, and program | |
| CN108874783A (en) | Power information O&M knowledge model construction method | |
| CN109408811A (en) | A kind of data processing method and server | |
| US20170236095A1 (en) | System and method for generating a career path | |
| WO2016147276A1 (en) | Data analysis system, data analysis method, and data analysis program | |
| CN102955848A (en) | Semantic-based three-dimensional model retrieval system and method | |
| CN107908653A (en) | A kind of data processing method and device | |
| CN109597990A (en) | A kind of matching process of social hotspots and commodity category | |
| Spasić et al. | Idiom-based features in sentiment analysis: Cutting the Gordian knot | |
| CN108920521A (en) | User's portrait-item recommendation system and method based on pseudo- ontology | |
| CN106528676A (en) | Entity semantic retrieval processing method and device based on artificial intelligence | |
| CN110413767A (en) | System and method based on spatial term rendering content | |
| Feldman | The answer machine | |
| Wang et al. | Harshness-aware sentiment mining framework for product review | |
| Klein et al. | The usual and the unusual: Solving remote associates test tasks using simple statistical natural language processing based on language use |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |