CN106708939A - Target person scoring and pushing methods, apparatuses and systems - Google Patents
Target person scoring and pushing methods, apparatuses and systems Download PDFInfo
- Publication number
- CN106708939A CN106708939A CN201611041080.0A CN201611041080A CN106708939A CN 106708939 A CN106708939 A CN 106708939A CN 201611041080 A CN201611041080 A CN 201611041080A CN 106708939 A CN106708939 A CN 106708939A
- Authority
- CN
- China
- Prior art keywords
- target person
- index
- score
- score index
- scoring
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Theoretical Computer Science (AREA)
- Databases & Information Systems (AREA)
- Educational Administration (AREA)
- Entrepreneurship & Innovation (AREA)
- Strategic Management (AREA)
- Physics & Mathematics (AREA)
- Development Economics (AREA)
- General Physics & Mathematics (AREA)
- Economics (AREA)
- Data Mining & Analysis (AREA)
- Game Theory and Decision Science (AREA)
- General Engineering & Computer Science (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The technical scheme of the invention provides target person scoring and pushing methods, apparatuses and systems. The target person scoring method comprises the steps of obtaining a plurality of first scoring indexes and distribution of dimensions under the first scoring indexes; performing statistics on probability distribution of the dimensions under the first scoring indexes for a target person; calculating the fit degrees of the target person for the first scoring indexes based on state distribution of the dimensions under the first scoring indexes; taking an average value for the fit degrees of the first scoring indexes belonging to the same index type to obtain index values; and outputting a score of the target person based on a weighted average result of all the index values and corresponding index weights. According to the technical scheme, automatic evaluation, selection and recommendation of the target person can be realized.
Description
Technical field
The invention belongs to field of computer technology, in particular it relates to a kind of method scored target person, one kind
Method, a kind of device scored target person, a kind of system pushed to personage pushed to personage.
Background technology
With the popularization of internet, same website can be used by panoramic people.For example, a navigation website meeting
Used by the people of more than 60 years old by age range, meanwhile, theirs is professional different, and the demand of online is also with big phase footpath
Front yard.The information of magnanimity is emerged in large numbers in face of everybody like the tidal wave, and we already enter an epoch for information explosion, in this background
Under, one side user is increasingly not easy therefrom to find the content that oneself is interested, on the other hand be also substantial amounts of information without
People makes inquiries, it is impossible to acquired in user.
The pairing between information and corresponding user is how realized, is that user is obtained in that more, more practical information, just needed
Judge the attribute of targeted customer, based on the attribute, understand the situation of targeted customer, need and like, could be to targeted customer
Relevant information is provided.
At this stage, the extraordinary method of neither one solves the problems, such as mentioned above.
The content of the invention
Technical solution of the present invention solve technical problem be：How on the basis processed targeted customer's information data
On, scored by algorithm, then realize the recommendation of corresponding informance.
In order to solve the above-mentioned technical problem, technical solution of the present invention provides a kind of side scored target person
Method, including：
Obtain the state distribution of dimension under some first Score indexes and first Score index；
The state distribution of each dimension of the first Score index is counted for the target person；
State distribution based on each dimension of the first Score index calculates the target person for the described first scoring
The compatible degree of index；
The first Score index compatible degree to belonging to identical pointer type averages to obtain desired value；
Result of weighted average based on the corresponding index weights of all desired values exports the scoring of the target person.
Optionally, the pointer type includes the first pointer type and the second pointer type, described that target person is carried out
The method of scoring also includes：
Obtain the attention rate of the target person；
Attention rate based on the target person calculates the attention rate index multiplier of the target person, and the attention rate refers to
Scalar multiplication number is attention rate of the target person relative to other personages；
If first Score index belongs to the first pointer type, described pair of the first scoring for belonging to identical pointer type refers to
Mark compatible degree is averaged and is included with obtaining desired value：By the first Score index compatible degree average value and the target person
Attention rate index multiplier be multiplied to obtain the desired value；
If first Score index belongs to the second pointer type, described pair of the first scoring for belonging to identical pointer type refers to
Mark compatible degree is averaged and is included with obtaining desired value：Using the first Score index compatible degree average value as the index
Value.
Optionally, also include：
Obtain some second Score indexes；
The statistical value of second Score index is counted for the target person；
Statistical value based on second Score index calculates the relative liveness of second Score index；
Second Score index is taken with respect to the average value of liveness to obtain the desired value.
Optionally, for the second Score index, the desired value is obtained based on following steps：
Obtain the liveness of the target person；
Liveness based on the target person calculates the liveness multiplier of the target person；
Second Score index is multiplied to respect to the average value of liveness with the liveness multiplier of the target person
Obtain the desired value.
Optionally, if the compatible degree is cos θi, have：
Wherein, i is the numbering of index, RiIt is the state distribution matrix of each dimension of the first Score index,It is the target
The state distribution matrix of each index of personage, (Ri)TIt is RiTransposed matrix.
Optionally, if attention rate index multiplier is Ti, have：
Wherein, i is the numbering of the target person, is the attention rate of the target person, t1, t2..., tnFor it is described its
His personage and the attention rate of target person, n are personage's number of described other personages and target person.
Optionally, if the relative liveness of second Score index is Fi, have：
Wherein, i is the numbering of the target person, is the statistical value of the Score index of the target person second, is described
The statistical value of the second Score index of other personages and target person, n is personage's number of described other personages and target person.
Optionally, if liveness multiplier is Si, have：
Wherein, i is the numbering of the target person, giIt is the liveness of the target person, g1, g2..., gnFor described
Other personages and the liveness of target person, n are personage's number of described other personages and target person.
Optionally, the multiplied result based on the corresponding index weights of all desired values exports the target person
Scoring includes：
The multiplied result is added up to obtain the scoring of the target person.
Optionally, the multiplied result based on the corresponding index weights of all desired values exports the target person
Scoring includes：
Obtain the 3rd Score index；
The 3rd Score index numerical value is counted for the target person；
The multiplied result is added up and is multiplied to the accumulated value and the 3rd Score index numerical value to obtain described
The scoring of target person.
In order to solve the above-mentioned technical problem, technical solution of the present invention additionally provides a kind of method pushed to personage,
Including：
All target persons are scored based on foregoing method；
Scoring based on the target person carries out personage's push.
Optionally, also include：
Receive user personage and push request；
Described pair of all target persons are carried out scoring and are performed based on the request.
In order to solve the above-mentioned technical problem, technical solution of the present invention additionally provides a kind of dress scored target person
Put, including：
Acquiring unit, is suitable to the distribution of dimension under some first Score indexes of acquisition and first Score index；
Statistic unit, is suitable to be counted for the target person state distribution of each dimension of the first Score index；
Computing unit, be suitable to the state distribution based on each dimension of the first Score index calculate the target person for
The compatible degree of first Score index；
Averaging unit, is suitable to the first Score index compatible degree for belonging to identical pointer type is averaged to obtain index
Value；
Output unit, is suitable to the multiplied result based on the corresponding index weights of all desired values and exports the target person
Scoring.
In order to solve the above-mentioned technical problem, technical solution of the present invention provides a kind of system pushed to personage, bag
Include：
Device as described above, is suitable to output and all target persons is scored；
Push unit, being suitable to the scoring based on the target person carries out personage's push.
The beneficial effect of technical solution of the present invention at least includes：
Technical solution of the present invention is based on default evaluation system, has tag set in evaluation system, and tag set can be with
It is relevant to classification of assessment (game, shopping, economic dispatch), row label evaluation is entered based on the web data that browser client is browsed, lead to
The data for crossing big data/historical viewings device user formulate preference, judge the attribute of gender and are pushed away preferably to carry out webpage
Send.So contribute to realize the demand that pushed information meets browser client.
On the other hand, technical solution of the present invention can be evaluated all kinds of public figures, and judges public figure's
Attribute, to determine the various objective evaluation information of comprehensive public figure, and carries out commenting for public figure based on above-mentioned evaluation information
Estimate, so as to adaptively carry out personage's recommendation and selection, realize the synthesis of various people informations under network and big data background
Evaluate, that realizes target person automatically selects recommendation.
Brief description of the drawings
The detailed description made to non-limiting example with reference to the following drawings by reading, other features of the invention,
Objects and advantages will become more apparent upon：
Fig. 1 shows the first specific embodiment of the invention, a kind of method flow scored target person
Figure；
Fig. 2 shows the second specific embodiment of the invention, a kind of method flow scored target person
Figure；
Fig. 3 shows the 3rd specific embodiment of the invention, a kind of method flow scored target person
Figure；
Fig. 4 shows a change case of the 3rd specific embodiment of the invention, a kind of acquisition side to desired value
Method flow chart；
Fig. 5 shows the 4th specific embodiment of the invention, a kind of method flow diagram pushed to personage；
Fig. 6 shows the 5th specific embodiment of the invention, a kind of apparatus structure scored target person
Figure.
Specific embodiment
In order to preferably make technical scheme clearly show, the present invention is made into one below in conjunction with the accompanying drawings
Step explanation.
It is to be appreciated that the present invention is main and is applicable but is not limited to such a scene, skilled artisan understands that this
The control process of sample has great importance, and browser client may correspond to multiple users；One user is also likely to be multiple clear
Look at device user.Preference is formulated by the data of big data/historical viewings device user, judges the attribute of gender so as to more preferable
Carry out Web page push.Technical solution of the present invention sets evaluation system, has tag set in evaluation system, and tag set is to evaluating
Classification (game, shopping, economic dispatch) is relevant, enters row label evaluation based on the web data that browser client is browsed, in evaluation then
The calculating on basic score value (0) is carried out to correspondence classification of assessment, such as then (+1), inclined women is then (- 1) for inclined male；Commented all
The final calculated value of valency classification is added, so as to obtain the negative value on basic score value or on the occasion of being pushed away so as to carry out webpage/product
Recommend.
Fig. 1 shows the first specific embodiment of the invention, a kind of method flow scored target person
Figure.Comprise the following steps that：
Into step S101, the state distribution of dimension under some first Score indexes and first Score index is obtained.
Specifically, whether described acquisition on the i.e. page of detection user terminal has user input instruction and judges the class of user input instruction
Type.User terminal can be the browser in computer, mobile phone etc..The page of user terminal is accessed typically by browser
Response message is beamed back in a page in portal management service, operation of the server according to user on the page for being accessed；Institute
The tag set during the first scoring refers to evaluation system is stated, tag set is relevant to classification of assessment (game, shopping, economic dispatch).
Step S102 is performed, specifically, the state of each dimension of the first Score index is counted for the target person
Distribution.
Into step S103, the state distribution based on each dimension of the first Score index calculate the target person for
The compatible degree of first Score index.Specifically, the compatible degree, i.e. system judge that the target person is commented with described first
Divide the similarity of index.More specifically, if the compatible degree is cos θi, have：
Wherein, i is the numbering of index, RiIt is the state distribution matrix of each dimension of the first Score index,It is the target
The state distribution matrix of each index of personage, (Ri)TIt is RiTransposed matrix.
Step S104 is performed, the first Score index compatible degree to belonging to identical pointer type averages to obtain index
Value.Specifically, the desired value is the judgement to the target person attribute, for example, can be used to indicate the target person
Inclined male/female.
Into step S105, the result of weighted average based on the corresponding index weights of all desired values exports the target
The scoring of personage.Specifically, the scoring is comprehensive scoring, that is, refer to that the target person indices are integrated and reflected
The target person attribute.
Further, the multiplied result based on the corresponding index weights of all desired values exports the target person
Scoring include：The multiplied result is added up to obtain the scoring of the target person.
Fig. 2 shows the second specific embodiment of the invention, a kind of method flow scored target person
Figure.The pointer type includes the first pointer type and the second pointer type, wherein second pointer type belongs to index class
Type, is here that pointer type is divided, and a portion pointer type (correspondence first pointer type) makees average value
Be multiplied treatment with attention rate, and another part pointer type (correspondence second pointer type) makees average value treatment.Specific steps
It is as follows：
Into step S201, the attention rate of the target person is obtained.Specifically, the target person is not real
People, and refer to the Internet portals such as browser；The attention rate can be clicking rate or light exposure.
Step S202 is performed, the attention rate index that the attention rate based on the target person calculates the target person multiplies
Number.Specifically, the attention rate index multiplier is attention rate of the target person relative to other personages.More specifically,
If attention rate index multiplier is Ti, have：
Wherein, wherein, i is the numbering of the target person, tiIt is the attention rate of the target person, t1, t2..., tn
It is described other personages and the attention rate of target person, n is personage's number of described other personages and target person.
Into step S203, the type of first Score index is judged.
Step S204 is performed, if first Score index belongs to the first pointer type, described pair belongs to identical index class
First Score index compatible degree of type is averaged and is included with obtaining desired value：By the first Score index compatible degree average value
Attention rate index multiplier with the target person is multiplied to obtain the desired value；If first Score index belongs to second
Pointer type, described pair of the first Score index compatible degree for belonging to identical pointer type is averaged and is included with obtaining desired value：
Using the first Score index compatible degree average value as the desired value.
Fig. 3 shows the 3rd specific embodiment of the invention, a kind of method flow scored target person
Figure.
Into step S301, some second Score indexes are obtained.Specifically, it is described second scoring and the first Score index all
It is a certain class of the Score index, including sex, occupation etc., is simply distinguish between in statement.
Step S302 is performed, the statistical value of second Score index is counted for the target person.
Into step S303, the statistical value based on second Score index calculates living relatively for second Score index
Jerk.Specifically, the relative liveness of second Score index is Fi, have：
Wherein, i is the numbering of the target person, is the statistical value of the Score index of the target person second, is described
The statistical value of the second Score index of other personages and target person, n is personage's number of described other personages and target person.
Step S304 is performed, second Score index is taken with respect to the average value of liveness to obtain the desired value.
Further, the desired value is obtained based on the method shown in Fig. 4.Comprise the following steps that；
Into step S401, the liveness of the target person is obtained.
Step S402 is performed, the liveness based on the target person calculates the liveness multiplier of the target person.Tool
Body ground, if liveness multiplier is Si, have：
Wherein, i is the numbering of the target person, giIt is the liveness of the target person, g1, g2..., gnFor described
Other personages and the liveness of target person, n are personage's number of described other personages and target person.
Into step S403, by second Score index with respect to the average value of liveness and enlivening for the target person
Degree multiplier is multiplied to obtain the desired value.
Further, the multiplied result based on the corresponding index weights of all desired values exports the target person
Scoring include：Obtain the 3rd Score index；The 3rd Score index numerical value is counted for the target person；By the phase
Multiply the scoring that result adds up and the accumulated value is multiplied to obtain the target person with the 3rd Score index numerical value.
Fig. 5 shows the 4th specific embodiment of the invention, a kind of method flow diagram pushed to personage.The
Any one method scores all target persons during four specific embodiments are based on first, second and third specific embodiment, then
The scoring of the target person is carried out into task push.Comprise the following steps that：
Into step S501, receive user personage and push request.Specifically, the user personage is the operation of browser end
Person；Described push asks to refer to the demand that the user personage is sent when browser end is operated that the demand is probably described
User personage directly passes on, and is also probably the demand implied in a certain operation.For example, user searches for pen at night, then hidden
May be copybook, ink etc. containing demand.
Step S502 is performed, described pair of all target persons are carried out scoring and are performed based on the request.
Fig. 6 shows the 5th specific embodiment of the invention, a kind of apparatus structure scored target person
Figure, the device includes acquiring unit 61, statistic unit 62, computing unit 63, averaging unit 64 and output unit 65.The acquisition
Unit 61 is suitable to the distribution of dimension under some first Score indexes of acquisition and first Score index；The statistic unit 62 is fitted
In the state distribution that each dimension of the first Score index is counted for the target person；The computing unit 63 is suitable to be based on
The state distribution of each dimension of the first Score index calculates compatible degree of the target person for first Score index；
The averaging unit 64 is suitable to the first Score index compatible degree for belonging to identical pointer type is averaged to obtain desired value；
The output unit 65 is suitable to the multiplied result based on the corresponding index weights of all desired values and exports the target person
Scoring.
It should be noted that this embodiment is not limited to for the structure of the above-mentioned device scored target person,
Such as, above-mentioned acquiring unit 61, statistic unit 62, computing unit 63, averaging unit 64 can synthesize in same module, such as one
Processing module, the processing module completes above-mentioned scoring function with output unit 65；For another example, above-mentioned averaging unit 64 can synthesize in
In computing unit 63, or, statistic unit 62, averaging unit 64 can also together synthesize in above-mentioned computing unit 63.The present invention
Technical scheme is not construed as limiting to said structure form.
Further, the device scored target person collectively constitutes with push unit and task is pushed
System.The push unit is suitable to the scoring based on the target person carries out personage's push.
According to the above-mentioned technical characteristic of technical solution of the present invention, the present embodiment additionally provides first application examples, even if using
State the method for scoring target person to score performer, its main thought is to comment target person based on above-mentioned
The method divided sets up a kind of performer's Rating Model.Performer's scoring mould is being set up using the above-mentioned method scored target person
During type, its application target includes：Influence of performer's factor to movie and television play benefit is many, is embodied in popularity, artistic skills, work
The aspects such as jerk；Selection performer need to carry out the consideration of various dimensions, to realize the maximization of benefit.Because performer's selection is involved
Dimension it is more, and otherness is big between dimension, therefore is difficult to form complete considering by subjective judgement.Set up performer's scoring
Model is meant to ensure that the comprehensive and diversity of evaluation index, and different indexs is closed by certain statistics aspect
The synthesis of reason, the readability and practicality of enhancing result.Furthermore, it is contemplated that the factor of influence movie and television play benefit is many,
In addition to performer's factor, also including all many factors such as drama, shooting, publicity；Therefore, performer's Rating Model is from it
In a link set out, given drama and in the case of not considering to shoot and publicize, selection maximizing the benefits is drilled
Member.In these conditions, given drama is most important premise, you can fully to ask for theme, the class of shooting to movie and television play
Type and the story of a play or opera are understood.
Based on above-mentioned thinking, and the method scored target person with reference to described in the present embodiment, by above-mentioned technology
Scheme is applied to performer's scoring, first, including the step of set up following index system, refers to performer's Rating Model of the setting of table one
Index.
Table one：Performer's Rating Model index
Based on table one, above-mentioned performer's Rating Model is divided into following pointer type by the application example, including：Drama-bean vermicelli contract
Right (C), bean vermicelli multiplier, drama-ability compatible degree (R), popularity (F) and liveness multiplier.Wherein：
Signified drama-bean vermicelli the compatible degree (C) of table one includes following index dimension (i.e. described index name)：
(1) age bracket (C1)
Age bracket is divided into 6 sections, 0-12 represents boyhood crowd, 12-18 represents puberty crowd (junior-senior high school life),
18-24 represents university student crowd, and 24-35 represents the Young Patients for just entering workplace, and 35-55 represents middle-aged population, and more than 55 represent
Elderly population.
The age of performer's bean vermicelli is a frequency distribution, and the correspondence bean vermicelli age compares for the performer of rejuvenation, its bean vermicelli
Based on student crowd, a kind of typical distribution can be：C1=(0,0.3,0.5,0.2,0,0), certainly, the year of performer's bean vermicelli
Age distribution may be different according to the difference that bean vermicelli crowd is distributed in years, for the performer of bean vermicelli age comparative maturity,
Its distribution situation can be C1=(0,0.2,0.3,0.4,0.1,0).
(2) sex (C2)
Based on the female ratio of bean vermicelli, it is divided into 6 grades, i.e. 0-15%, 15%-30%, 30%-50%, 50%-
70%th, 70%-85% and more than 85%.
Sex index is a state, if the female ratio of performer's bean vermicelli is 60%, distribution is：C2=(0,
0,0,1,0,0), if the female ratio of performer's bean vermicelli is 44%, distribution is：C2=(0,0,1,0,0,0)
(3) flow state (C3)
Flow state reflects attention rate of performer's bean vermicelli to amusement circles.Investigation 50-100 enlivens bean vermicelli, and its nearest 20
The average accounting related to amusement circles in bar microblogging；This ratio is divided into 6 grades, i.e. 0-15%, 15%-30%, 30%-
50%th, 50%-70%, 70%-85% and more than 85%.
Flow state index is equally a state, if the average accounting of amusement circles microblogging is 34%,：C3=(0,0,
1,0,0,0)
(4) hobby (C4)
The theme and plot of drama tend to correspond to some animations, for example residence, nature, motion, reading, capable and experienced
Deng different people can select different plays.The state of this life is divided into quiet (residence, literature and art, workplace) by hobby index,
Dynamic (motion, natural), and thought (illusion), totally 6 class.
Hobby index is equally a frequency distribution, and investigation 50-100 enlivens bean vermicelli, obtains all kinds of animations
Accounting.If accounting is respectively 35%, 20%, 10%, 5%, 5%, 25%,：C4=(0.35,0.2,0.1,0.05,
0.05,0.25)
Table two gives drama-bean vermicelli compatible degree (C) and is distributed in each index and its dimension.
Table two：Drama-bean vermicelli agrees with (C) degree index and dimension
Based on above-mentioned table one and table two, the computational methods for calculating drama-bean vermicelli compatible degree (C) given below.
According to target group's feature of above-mentioned drama-bean vermicelli compatible degree (C), if given drama, the spy of its target group
Age-based section, sex, flow state and hobby this 4 dimensions are levied to can be expressed as：
Wherein,WithAge and hobby are represented, is a probability distribution, andWithRepresent sex and trend
State, is a state.
When calculating compatible degree, using cosine similarity, a certain index (age, sex, flow state and interest are calculated
Hobby) compatible degree be：
Then the drama of performer-bean vermicelli compatible degree (C) is 4 averages of index compatible degree：
Signified drama-ability the compatible degree (R) of table one includes following index dimension (i.e. described index name)：
(1) role takes on power (R1)
Performer will influence acute total quality to the annotation ability of role, and the experience of performer can reflect to role in play
Competence level.According to character types, index is divided into compassion, happiness, feelings, cold, steady, power, 7 dimensions of property.
The index that role takes on power (R1) is a frequency distribution, counts all character types performed of performer, can be counted
Calculate this 7 accountings of dimension.
(2) story of a play or opera takes on power (R2)
It is similar that the equally influence of movie and television play type that actor features cross takes on power to acute annotation, with role, according to movie and television play
Type, ancient costume, history, city, idol, science fiction, 7 dimensions of suspense and fantasy are divided into by index.
Index is equally a frequency distribution, counts all acute types performed of performer, calculates accounting for for each dimension
Than.Table three gives drama-ability compatible degree (R) and is distributed in each index and its dimension.
Table three：Drama-ability compatible degree (R) index and dimension
Based on above-mentioned table one and table two, calculating given below：The calculating side of drama-ability compatible degree (R) index and dimension
Method.
According to given drama, then role takes on power and the story of a play or opera takes on power and can be according to the dimension distribution and expression of table three：
WhereinDrama is performed before according to statistics performer, is performer in drama role class
Type compassion, happiness, feelings, cold, steady, power, the probability distribution in property,It is performer in drama story of a play or opera type ancient costume, history, city, idol
Probability distribution on picture, science fiction, suspense and fantasy.
It should be noted thatWithThe type of character types and play is represented in the application example, is a probability distribution,
But closer to state distribution.
Be with the calculating of drama-bean vermicelli compatible degree (C) it is similar, drama-ability compatible degree (R) based role take on power and
The story of a play or opera takes on the two indices of power, and drama-ability compatible degree (R) calculating formula is calculated including cosine：
The overall target of drama-ability compatible degree (R) is：
For the signified popularity (F) of table one, its calculating process includes：
Setting relative score system, with reference to the index name of table one, popularity index includes：
F1：Engage in this profession the time limit-from perform first play to scoring when time, in units of year；
F2：Perform number of times-by the end of scoring when, the movie and television play quantity performed altogether；
F3：Prize-winning number of times-and when scoring, obtain the number of times of international awards；
F4：Public's scoring-and when scoring, scoring average of all movie and television plays performed in video website.
If Y11, Y12..., Y1nThe time limit of engaging in this profession of N number of performer, then i-th F of performer1It is scored at：
I-th F of performer is calculated with same method2、F3And F4Score, is designated as F2i, F3iAnd F4i。
Then popularity (the relative liveness i.e. in above-described embodiment) comprehensive grading of performer i is：
Wherein, j is 1~4.
The calculating process of signified liveness multiplier (S) of table one includes：Popularity represents the performing art achievement of performer, is one quiet
The history index of state.However, the rate of decay of performer's attention rate is exceedingly fast, therefore, the influence of recent liveness to popularity compared with
Greatly.
If gi be i-th performer during certain in (such as, nearly 1 year or several years) searchable index, the search mesh
Mark can be based on the average value of a kind of statistics or various statisticses in all search engines such as www.baidu.com, the system
The summation that result can be i-th performer searching times on the search engine, or the search ratio based on total calculating are counted,
Because the statistics can carry out objective evaluation according to the algorithm of searching statistical or statistic, the application example is not to the statistics knot
Being specifically defined for fruit is limited.
Based on above-mentioned giStatistics, the liveness multiplier (S) of its i-th performer is：
The liveness multiplier (S) of performer is the popularity for adjusting the performer in performer's Rating Model of the application example
(F), although popularity (F) algorithm that the application example is used alone can also set up above-mentioned performer's Rating Model, but the degree of accuracy
It is barely satisfactory.The application example also adjusts the popularity (F) of performer using the liveness multiplier (S) of performer, comments final performer
Index in sub-model is more accurate, based on above-mentioned elaboration, there is the popularity F of the performer after adjustmenti'=Fi·Si。
Continuing with the index parameter in table one, drama-bean vermicelli compatible degree (C) has weighed the relative of performer from correlation
Benefit, but bean vermicelli scale can also embody scale and benefit.Real active bean vermicelli can be having embodied in the practical action, therefore can
Produced with money commodity or other works related to performer etc. with by the performer shown in network selling business or other statistics platforms
At least one product accounting amount in sales volume, sale quantity, restocking quantity of product etc. is used as index basis.
If using performer in Taobao with the quantity of money commodity as index basis, setting ti, to be i-th performer washing in a pan
With the quantity of money commodity on treasured, if 1...n is the numbering (n is the natural number more than 1) of all performers, its bean vermicelli multiplier (T) is (i.e.
Attention rate multiplier in above-described embodiment) be：
Bean vermicelli multiplier (T) is for adjusting drama-bean vermicelli compatible degree (C), after making adjustment in performer's Rating Model
Drama-bean vermicelli compatible degree (C) can more be used to evaluate the objectivity of performer's bean vermicelli effect.I-th performer adjusts by bean vermicelli multiplier (T)
Whole drama-bean vermicelli compatible degree C 'i=Ci·Ti。
Certainly, it is necessary to explanation be in other embodiments, directly export the drama-bean vermicelli compatible degree (C) and without
It is also feasible to cross the adjustment of bean vermicelli multiplier (T).
After the preliminary foundation of model above parameter, drama-bean vermicelli compatible degree, drama-ability compatible degree and tune after adjustment
Popularity after whole constitutes the evaluation model of performer, there is evaluation model P=(C ', R, F ').That is, in the application example
Evaluation model includes three parameters, that is, well-known after the drama-bean vermicelli compatible degree, drama-ability compatible degree and adjustment after adjusting
Degree.
Parameters in above-mentioned evaluation model, that is, adjust after drama-bean vermicelli compatible degree, drama-ability compatible degree and
Also there is popularity after adjustment respective output to drive weight.According to the difference of input, the output of movie and television play drives weight to deposit
In difference, that is to say, that driving force according to performer is different, performer towards the public when publicity can be subject to drive, itself performance it is real
The differences such as power, the parameters of performer's evaluation model can drive compared to above-mentioned publicity driving, performing quality and set each ginseng
Weighted value between number.
For example, publicity is driven in the case where cost space is larger, and income is realized by popularity and channel, therefore
Shadow different to the balance between bean vermicelli compatible degree, ability compatible degree and popularity, popularity can generally speaking being caused
Ringing power can be bigger than normal, can such as set the weighted value of drama-bean vermicelli compatible degree, drama-between ability compatible degree and popularity as (0.2,
0.2,0.6).In addition, quality drives the accreditation that the public is obtained by good performance, the expansion of income is realized by public praise.
In this case, the quality of performer is relatively more important, therefore the balance of bean vermicelli compatible degree, ability compatible degree and popularity may
It is (0.3,0.5,0.2).Only as an example, the driving factors according to setting (drive by such as described publicity for the division of above-mentioned weighted value
Dynamic, performer's example etc.) more proper parameters weighting is set to the parameters in above-mentioned evaluation model, infer mesh for accurate
Balance in last stage model between parameters, is that those of ordinary skill in the art can be public according to technical solution of the present invention institute
The technical characteristic flow opened makes confirmation by oneself.
According to above-mentioned evaluation model, based on the parameters weight in driving factors setting evaluation model, and set drama-
Weight W=(the W of popularity after bean vermicelli compatible degree, drama-ability compatible degree and adjustmentc’, WR, WF’), wherein, Wc’、WRAnd WF’Point
The weight of the drama-bean vermicelli compatible degree after Wei not adjusting, drama-ability compatible degree weight and after adjustment popularity weight.
Final Rating Model for performer described in the application example is：Score=PT·W。
In one change case of the application example, the wind of comprehensive assessment performer can also be included in above-mentioned performer's assessment models
Danger value.Performer is used as public figure, and the influence of its public image and spin to input and output is larger.Front, good shape
As if the guarantee of income is realized, conversely, negative news is mostly important to the potential impact of income.In above-mentioned Rating Model Score
=PTOn the basis of W, negative news can be brought into Rating Model, specific practice (ratio interior during being search agreement
Such as 6 months) news disclosed in network or media, can specifically determine that some network platform does news search, such as Baidu
Or Sina, judge the value-at-risk Z of the performer.The concrete numerical value of value-at-risk Z can be carried out any one of by the following method
It is determined that：
(1) there is the serious negative news such as violence, illegal, crime, then Z=0；
(2) general negative news (life, emotion, conflict and opposition etc.), if news accounting is more than 50%, Z=0.5；
(3) general negative news accounting is less than 50%, then Z=0.8；
(4) there is no negative news, then Z=1.
Therefore, the performer's assessment models after value-at-risk adjustment is added, specially：
Adjusted Score=PT·W·Z。
Above-mentioned application examples be specifically described how the scoring that the method scored target person is applied to performer,
Certainly, the scoring object of the above method is not limited to performer, in fact, the evaluation for all kinds of public figures is all feasible.This
Inventive technique scheme can be evaluated all kinds of public figures, and judges the attribute of public figure, to determine comprehensive public people
The various objective evaluation information of thing, and the assessment of public figure is carried out based on above-mentioned evaluation information, so as to adaptively enter pedestrian
Thing is recommended and is selected, and realizes the overall merit of various people informations under network and big data background, realizes target person oneself
Dynamic selection is recommended.
On specific embodiment of the invention is described.It is to be appreciated that the invention is not limited in above-mentioned spy
Determine implementation method, those skilled in the art can within the scope of the claims make various deformations or amendments, and this has no effect on
Substance of the invention.
Claims (14)
1. a kind of method scored target person, it is characterised in that including：
Obtain the state distribution of dimension under some first Score indexes and first Score index；
The state distribution of each dimension of the first Score index is counted for the target person；
State distribution based on each dimension of the first Score index calculates the target person for first Score index
Compatible degree；
The first Score index compatible degree to belonging to identical pointer type averages to obtain desired value；
Result of weighted average based on the corresponding index weights of all desired values exports the scoring of the target person.
2. the method for being scored target person as claimed in claim 1, it is characterised in that the pointer type includes first
Pointer type and the second pointer type, the method scored target person also include：
Obtain the attention rate of the target person；
Attention rate based on the target person calculates the attention rate index multiplier of the target person, and the attention rate index multiplies
Attention rate of the number for the target person relative to other personages；
If first Score index belongs to the first pointer type, described pair of the first Score index contract for belonging to identical pointer type
Right averaging is included with obtaining desired value：By the pass of the first Score index compatible degree average value and the target person
Note degree index multiplier is multiplied to obtain the desired value；
If first Score index belongs to the second pointer type, described pair of the first Score index contract for belonging to identical pointer type
Right averaging is included with obtaining desired value：Using the first Score index compatible degree average value as the desired value.
3. the method for being scored target person as claimed in claim 1, it is characterised in that also include：
Obtain some second Score indexes；
The statistical value of second Score index is counted for the target person；
Statistical value based on second Score index calculates the relative liveness of second Score index；
Second Score index is taken with respect to the average value of liveness to obtain the desired value.
4. the method for being scored target person as claimed in claim 3, it is characterised in that for the second Score index, institute
Desired value is stated to be obtained based on following steps：
Obtain the liveness of the target person；
Liveness based on the target person calculates the liveness multiplier of the target person；
Second Score index is multiplied to obtain with respect to the average value of liveness with the liveness multiplier of the target person
The desired value.
5. the method for being scored target person as claimed in claim 1, it is characterised in that it is cos θ to set the compatible degreei,
Have：
Wherein, i is the numbering of index, RiIt is the state distribution matrix of each dimension of the first Score index,It is the target person
The state distribution matrix of each index, (Ri)TIt is RiTransposed matrix.
6. the method for being scored target person as claimed in claim 2, it is characterised in that it is T to set attention rate index multiplieri,
Have：
Wherein, i is the numbering of the target person, is the attention rate of the target person, t1, t2..., tnFor it is described other people
The attention rate of thing and target person, n is personage's number of described other personages and target person.
7. the method for being scored target person as claimed in claim 3, it is characterised in that set second Score index
It is F with respect to livenessi, have：
Wherein, i is the numbering of the target person, is the statistical value of the Score index of the target person second, be it is described other
The statistical value of the second Score index of personage and target person, n is personage's number of described other personages and target person.
8. the method for being scored target person as claimed in claim 4, it is characterised in that it is S to set liveness multiplieri, have：
Wherein, i is the numbering of the target person, giIt is the liveness of the target person, g1, g2..., gnFor it is described other
Personage and the liveness of target person, n are personage's number of described other personages and target person.
9. the method that is scored target person as claimed in claim 1, it is characterised in that it is described based on all desired values with
The scoring that the multiplied result of its correspondence index weights exports the target person includes：
The multiplied result is added up to obtain the scoring of the target person.
10. the method for being scored target person as claimed in claim 1, it is characterised in that described based on all desired values
The scoring that the multiplied result of corresponding index weights exports the target person includes：
Obtain the 3rd Score index；
The 3rd Score index numerical value is counted for the target person；
The multiplied result is added up and is multiplied to obtain the target with the 3rd Score index numerical value by the accumulated value
The scoring of personage.
A kind of 11. methods pushed to personage, it is characterised in that including：
All target persons are scored based on the method described in any one of claim 1 to 10；
Scoring based on the target person carries out personage's push.
12. methods for being pushed to personage as claimed in claim 11, it is characterised in that also include：
Receive user personage and push request；
Described pair of all target persons are carried out scoring and are performed based on the request.
A kind of 13. devices scored target person, it is characterised in that including：
Acquiring unit, is suitable to the distribution of dimension under some first Score indexes of acquisition and first Score index；
Statistic unit, is suitable to be counted for the target person state distribution of each dimension of the first Score index；
Computing unit, is suitable to the state distribution based on each dimension of the first Score index and calculates the target person for described
The compatible degree of the first Score index；
Averaging unit, is suitable to the first Score index compatible degree for belonging to identical pointer type is averaged to obtain desired value；
Output unit, is suitable to the multiplied result based on the corresponding index weights of all desired values and exports commenting for the target person
Point.
A kind of 14. systems pushed to personage, it is characterised in that including：
Device as claimed in claim 13, is suitable to output and all target persons is scored；
Push unit, being suitable to the scoring based on the target person carries out personage's push.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611041080.0A CN106708939A (en) | 2016-11-24 | 2016-11-24 | Target person scoring and pushing methods, apparatuses and systems |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611041080.0A CN106708939A (en) | 2016-11-24 | 2016-11-24 | Target person scoring and pushing methods, apparatuses and systems |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106708939A true CN106708939A (en) | 2017-05-24 |
Family
ID=58933778
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611041080.0A Pending CN106708939A (en) | 2016-11-24 | 2016-11-24 | Target person scoring and pushing methods, apparatuses and systems |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106708939A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108599168A (en) * | 2018-03-30 | 2018-09-28 | 中国电力科学研究院有限公司 | A kind of method and system for carrying out reasonable evaluation to bulk power grid plan trend |
| CN110162545A (en) * | 2019-04-18 | 2019-08-23 | 平安城市建设科技(深圳)有限公司 | Information-pushing method, equipment, storage medium and device based on big data |
| CN114444987A (en) * | 2022-04-11 | 2022-05-06 | 深圳小库科技有限公司 | Automatic analysis method and device for house type graph |
-
2016
- 2016-11-24 CN CN201611041080.0A patent/CN106708939A/en active Pending
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108599168A (en) * | 2018-03-30 | 2018-09-28 | 中国电力科学研究院有限公司 | A kind of method and system for carrying out reasonable evaluation to bulk power grid plan trend |
| CN108599168B (en) * | 2018-03-30 | 2020-12-04 | 中国电力科学研究院有限公司 | Method and system for carrying out rationality evaluation on planned power flow of large power grid |
| CN110162545A (en) * | 2019-04-18 | 2019-08-23 | 平安城市建设科技(深圳)有限公司 | Information-pushing method, equipment, storage medium and device based on big data |
| CN114444987A (en) * | 2022-04-11 | 2022-05-06 | 深圳小库科技有限公司 | Automatic analysis method and device for house type graph |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Li et al. | Project success prediction in crowdfunding environments | |
| Krauss et al. | Predicting movie success and academy awards through sentiment and social network analysis | |
| CN106372249B (en) | A kind of clicking rate predictor method, device and electronic equipment | |
| Jin et al. | The wisdom of social multimedia: using flickr for prediction and forecast | |
| CN103038769B (en) | System and method for content to be directed into social network engine user | |
| CN110196946A (en) | A kind of personalized recommendation method based on deep learning | |
| US8893012B1 (en) | Visual indicator based on relative rating of content item | |
| US20090327120A1 (en) | Tagged Credit Profile System for Credit Applicants | |
| JP2008537817A (en) | Systems, methods, and computer program products for scoring items and determining predictor proficiency based on user sentiment | |
| CN103377250A (en) | Top-k recommendation method based on neighborhood | |
| US20150287051A1 (en) | System and method for identifying growing companies and monitoring growth using non-obvious parameters | |
| CN105677767A (en) | Equipment configuration recommending method and device | |
| WO2012161740A2 (en) | System and method for generating recommendations | |
| CN110737859A (en) | UP main matching method and device | |
| CN108322317A (en) | A kind of account identification correlating method and server | |
| US20150134380A1 (en) | Computerized system and method for determining an action person's value to parties to a transaction | |
| Choo et al. | Understanding and promoting micro-finance activities in kiva. org | |
| CN106708939A (en) | Target person scoring and pushing methods, apparatuses and systems | |
| CN104008204A (en) | Dynamic multi-dimensional context awareness film recommending system and achieving method thereof | |
| WO2019242453A1 (en) | Information processing method and device, storage medium, and electronic device | |
| Burtch et al. | Peer awards increase user content generation but reduce content novelty | |
| US7716209B1 (en) | Automated advertisement publisher identification and selection | |
| Zhang et al. | Research on movie box office forecasting based on internet data | |
| Banhawi et al. | User engagement attributes and levels in facebook | |
| CN116010696A (en) | News recommendation method, system and medium integrating knowledge graph and long-term interest of user |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20170524 |
|
| WD01 | Invention patent application deemed withdrawn after publication |