Summary of the invention
To overcome the problems in correlation technique, present description provides electronic ticket granting, resource allocation methods, devices
And computer equipment.
According to this specification embodiment in a first aspect, providing a kind of electronic ticket distribution method, comprising:
The history service data of user are obtained, the user is distinguished by multiple dimensions, the history service data
Indicate the portfolio variation after electronic ticket is provided;
The data distribution of history service data is corresponded to according to the dimension and difference that target is evenly distributed, determine the dimension
Spend the influence degree changed to portfolio;
According to definitive result, adjustment is issued to the electronic ticket for belonging to the dimension user.
Optionally, the difference is determined by calculating the cross entropy that the data distribution is evenly distributed with the target.
Optionally, the dimension has multiple enumerated values, and the target is evenly distributed through the enumerated value to the dimension
It is determined after carrying out following beta pruning processing: deleting the enumerated value for not meeting power law distribution.
Optionally, the data distribution is determined by calculating the following accounting parameter of each enumerated value of corresponding dimension:
Enumerated value corresponds to the ratio of the increased portfolio of entirety of increased portfolio and the dimension.
Optionally, the cross entropy is calculated in the following way:
Wherein, C_K is the cross entropy, P_KiFor the accounting parameter of the enumerated value i of the dimension, Q_K is the dimension
Target be evenly distributed, Q_K=1/m, m are the dimension beta pruning treated enumerated value number.
Optionally, the definitive result includes: to influence maximum first dimension collection to portfolio variation and influence minimum
The second dimension collection;
The adjustment for belonging to the first electronic ticket of the first dimension collection user, and belongs to the second dimension collection user
The second electronic ticket adjustment have negative correlativing relation.
Optionally, the coefficient of the first electronic ticket adjustment and the coefficient of the second electronic ticket adjustment meet following relationship: ax
=by, wherein a is the number parameter for belonging to the first dimension collection user, and b belongs to the second dimension collection to be described
The number parameter at family is shared, x is the coefficient of first electronic ticket adjustment, and y is the coefficient of second electronic ticket adjustment.
According to the second aspect of this specification embodiment, a kind of resource allocation methods are provided, comprising:
The history service data of user are obtained, the user is distinguished by multiple dimensions, the history service
Data indicate the portfolio variation after resource allocation;
The data distribution of history service data is corresponded to according to the dimension and difference that target is evenly distributed, determine institute
State the influence degree that dimension changes portfolio;
According to definitive result, adjustment is issued to the resource for belonging to the dimension user.
Optionally, the difference is determined by calculating the cross entropy that the data distribution is evenly distributed with the target.
Optionally, the dimension has multiple enumerated values, and the target is evenly distributed through the enumerated value to the dimension
It is determined after being handled as follows: deleting the enumerated value for not meeting power law distribution.
Optionally, the data distribution is determined by calculating the following accounting parameter of each enumerated value of corresponding dimension:
Enumerated value corresponds to the ratio of the increased portfolio of entirety of increased portfolio and the dimension.
Optionally, the cross entropy is calculated in the following way:
Wherein, C_K is the cross entropy, P_KiFor the accounting parameter of the enumerated value i of the dimension, Q_K is the dimension
Target be evenly distributed, Q_K=1/m, m are the dimension beta pruning treated enumerated value number.
Optionally, the definitive result includes: to influence maximum first dimension collection to portfolio variation and influence minimum
The second dimension collection;
The adjustment for belonging to the first resource of the first dimension collection user is used with second dimension collection is belonged to
The adjustment of the resource of side has negative correlativing relation.
Optionally, the coefficient of the first resource adjustment and the coefficient of Secondary resource adjustment meet following relationship: ax=by,
Wherein, a is the number parameter for belonging to the first dimension collection user, and b belongs to second dimension collection to be described
The number parameter of user, x are the coefficient of first resource adjustment, and y is the coefficient of Secondary resource adjustment.
According to the third aspect of this specification embodiment, a kind of electronic ticket dispensing apparatus is provided, comprising:
Module is obtained, is used for: obtaining the history service data of user, the user is distinguished by multiple dimensions, institute
State the portfolio variation after history service data instruction electronic ticket is provided;
Determining module is used for: according to the dimension corresponding to the data distribution of history service data and target is evenly distributed
Difference determines the influence degree that the dimension changes portfolio;
Adjust module, be used for: according to definitive result, adjustment is issued to the electronic ticket for belonging to the dimension user.
Optionally, the difference is determined by calculating the cross entropy that the data distribution is evenly distributed with the target.
Optionally, the dimension has multiple enumerated values, and the target is evenly distributed through the enumerated value to the dimension
It is determined after carrying out following beta pruning processing: deleting the enumerated value for not meeting power law distribution.
Optionally, the data distribution is determined by calculating the following accounting parameter of each enumerated value of corresponding dimension:
Enumerated value corresponds to the ratio of the increased portfolio of entirety of increased portfolio and the dimension.
Optionally, the cross entropy is calculated in the following way:
Wherein, C_K is the cross entropy, P_KiFor the accounting parameter of the enumerated value i of the dimension, Q_K is the dimension
Target be evenly distributed, Q_K=1/m, m are the dimension beta pruning treated enumerated value number.
Optionally, the definitive result includes: to influence maximum first dimension collection to portfolio variation and influence minimum
The second dimension collection;
The adjustment for belonging to the first electronic ticket of the first dimension collection user makes with second dimension collection is belonged to
There is negative correlativing relation with the adjustment of the second electronic ticket of side.
Optionally, the coefficient of the first electronic ticket adjustment and the coefficient of the second electronic ticket adjustment meet following relationship: ax
=by, wherein a is the number parameter for belonging to the first dimension collection user, and b belongs to second dimension to be described
Gather the number parameter of user, x is the coefficient of first electronic ticket adjustment, and y is the coefficient of second electronic ticket adjustment.
According to the fourth aspect of this specification embodiment, a kind of resource allocation device is provided, comprising:
Module is obtained, is used for: obtaining the history service data of user, the user carries out area by multiple dimensions
Point, the portfolio variation after the history service data instruction resource allocation;
Determining module is used for: according to the dimension corresponding to the data distribution of history service data and target is evenly distributed
Difference determines the influence degree that the dimension changes portfolio;
Adjust module, be used for: according to definitive result, adjustment is issued to the resource for belonging to the dimension user.
Optionally, the difference is determined by calculating the cross entropy that the data distribution is evenly distributed with the target.
Optionally, the dimension has multiple enumerated values, and the target is evenly distributed through the enumerated value to the dimension
It is determined after being handled as follows: deleting the enumerated value for not meeting power law distribution.
Optionally, the data distribution is determined by calculating the following accounting parameter of each enumerated value of corresponding dimension:
Enumerated value corresponds to the ratio of the increased portfolio of entirety of increased portfolio and the dimension.
Optionally, the cross entropy is calculated in the following way:
Wherein, C_K is the cross entropy, P_KiFor the accounting parameter of the enumerated value i of the dimension, Q_K is the dimension
Target be evenly distributed, Q_K=1/m, m are the dimension beta pruning treated enumerated value number.
Optionally, the definitive result includes: to influence maximum first dimension collection to portfolio variation and influence minimum
The second dimension collection;
The adjustment for belonging to the first resource of the first dimension collection user is used with second dimension collection is belonged to
The adjustment of the Secondary resource of side has negative correlativing relation.
Optionally, the coefficient of the first resource adjustment and the coefficient of Secondary resource adjustment meet following relationship: ax=by,
Wherein, a is the number parameter for belonging to the first dimension collection user, and b belongs to second dimension collection to be described
The number parameter of user, x are the coefficient of first resource adjustment, and y is the coefficient of Secondary resource adjustment.
According to the 5th of this specification embodiment aspect, a kind of computer equipment is provided, including memory, processor and deposit
Store up the computer program that can be run on a memory and on a processor, wherein the processor is realized when executing described program
Method as the aforementioned.
The technical solution that the embodiment of this specification provides can include the following benefits:
In this specification embodiment, by obtaining the history service data of user, which indicates electronic ticket
Portfolio variation after granting, and user is distinguished by multiple dimensions, it is based on this, history industry is corresponded to according to the dimension
The difference that the data distribution and target for data of being engaged in are evenly distributed, can determine the influence degree that the dimension changes portfolio,
And then belong to the electronic ticket of dimension user according to influence degree adjustment, so that the available effective use of electronic ticket.
It should be understood that above general description and following detailed description be only it is exemplary and explanatory, not
This specification can be limited.
Specific embodiment
Example embodiments are described in detail here, and the example is illustrated in the accompanying drawings.Following description is related to
When attached drawing, unless otherwise indicated, the same numbers in different drawings indicate the same or similar elements.Following exemplary embodiment
Described in embodiment do not represent all embodiments consistent with this specification.On the contrary, they are only and such as institute
The example of the consistent device and method of some aspects be described in detail in attached claims, this specification.
It is only to be not intended to be limiting this explanation merely for for the purpose of describing particular embodiments in the term that this specification uses
Book.The "an" of used singular, " described " and "the" are also intended to packet in this specification and in the appended claims
Most forms are included, unless the context clearly indicates other meaning.It is also understood that term "and/or" used herein is
Refer to and includes that one or more associated any or all of project listed may combine.
It will be appreciated that though various information may be described using term first, second, third, etc. in this specification, but
These information should not necessarily be limited by these terms.These terms are only used to for same type of information being distinguished from each other out.For example, not taking off
In the case where this specification range, the first information can also be referred to as the second information, and similarly, the second information can also be claimed
For the first information.Depending on context, word as used in this " if " can be construed to " ... when " or
" when ... " or " in response to determination ".
By taking electronic ticket provides scene as an example, electronic ticket refers to resource provider granting, the system in the form of various electronic medias
The electronic certificate made, propagate and used.Electronic ticket includes: cash certificate, minimum price certificate, deduction certificate, coupons, such as red packet, discount coupon
Deng;And various types of stock trader's product, such as game article.In the business initial stage, electronic ticket granting may follow certain plan
Slightly, such as it can be and each user is provided in a manner of relative equilibrium, be issued to the amount or number of the electronic ticket of each user
Amount etc. remains basically stable.After shopping platform side provides electronic ticket, Data Analyst can pay close attention to the influence after electronic ticket is provided to business.
For example, whether portfolio has the increase of increase and decrease, portfolio mainly to be contributed by which user group after electronic ticket granting, and
Which user group to the electronic ticket of granting and insensitive, contribution it is less etc..
Business datum would generally be there are many index, by way of example, the index of data may include: order volume, light exposure, wide
Accuse clicking rate, channel conversion ratio, the number of users that increases day by day, active users, user conversation number or user's value index nember etc..Data
Analyst can pay close attention to the fluctuation of these indexs, and the fluctuation of index is characterized using the change rate of index, such as year-on-year or ring ratio etc.;
It indicates to compare current statistic period and history with the period on year-on-year basis, ring is than indicating current statistic period and a upper measurement period
Compare.
The fluctuation of index may be generated by many reasons, by way of example, if index is user's quantity on order, change rate is same
Than influencing the dimension of quantity on order, city, age or gender etc., different dimensions pair where may include user's occupation, user
It may be different in the influence degree of user's quantity on order fluctuation.If can identify that the correlation for causing index to fluctuate is former in time
Cause, it will be able to which corresponding analysis and solution are made in the variation for business datum in time, for example, analysis electronic ticket granting
Which user group brings biggish positive influences to the increase of business afterwards, which user group does not bring the increase of business
It influences, so as to make service platform more stably provide service for user, improves the utilization rate of electronic ticket.
Based on this, the present embodiment provides a kind of electronic ticket distribution methods, and by obtaining the history service data of user, this is gone through
History business datum indicates the portfolio variation after electronic ticket granting, and user is distinguished by multiple dimensions, therefore, according to
The difference that the dimension corresponds to the data distribution of history service data and target is evenly distributed can determine the dimension to business
The influence degree of variation is measured, and then belongs to the electronic ticket of dimension user according to influence degree adjustment, electronic ticket is obtained
To effective use.
As shown in Figure 1, being a kind of this specification process of electronic ticket distribution method shown according to an exemplary embodiment
Figure, includes the following steps:
In a step 102, the history service data of user are obtained, the user is distinguished by multiple dimensions, described
History service data indicate the portfolio variation after electronic ticket granting.
At step 104, the data distribution of history service data is corresponded to according to the dimension and difference that target is evenly distributed
It is different, determine the influence degree that the dimension changes portfolio.
In step 106, according to definitive result, adjustment is issued to the electronic ticket for belonging to the dimension user.
Some words that the present embodiment is related to are illustrated first.Dimension, for characterizing certain spy of things or phenomenon
Sign, such as gender, area, time are all dimensions.Wherein the time is a kind of common, special dimension, passes through pair before and after the time
Than, so that it may know that the development of things improves or degenerates, if order volume ring is than increasing by 10% last month, increasing compared with the same period of last year together
Long 20%, here it is temporal comparisons, also referred to as vertical ratio;Another kind is relatively horizontal ratio, such as country variant population, GDP
Compare, different province incomes, the comparison of number of users, different company, comparison between different department etc., referred to as horizontal ratio.Dimension can
It to be divided into qualitative dimension and quantitation dimension, that is, is divided according to data type, data type is character type (text-type) number
According to, be exactly qualitative dimension, such as area, gender be all qualitative dimension;Data type is numeric type data, is just quantitation dimension,
Such as income, age, consumption.All values of dimension are known as enumerated value, for example, the value of dimension gender includes male and female, then
The enumerated value of the dimension is male and female, and there are two the numbers of enumerated value.
The granting object of electronic ticket is user, distinguishes user by multiple dimensions in the present embodiment, which to be analyzed
Which kind of influence a little users have to business.Specific dimension, city, occupation, age, gender or education degree etc. where may include
Deng can according to need flexible configuration in practical application, such as can divide according to the registration information of user or basis
Practical business scene and the customized dimension divided, are also possible to be provided with configuration interface, such as the configuration interface can be one
Visualization interface can wrap the function containing interaction, by the Data Detection configuration information of the available input of configuration interface,
Such as the change rate of dimension and index.In practical application, can for user divide in advance there are many dimension, for what is divided
Dimension can choose the dimension that its expectation is distinguished for Data Analyst.
In the present embodiment, the concern target of history service data is portfolio variation, therefore using portfolio as index,
Index can also include population, GDP, receipts in other application scenarios for measuring the unit or method of things development degree
Enter, number of users, profit margin, retention ratio, coverage rate etc..Index can summarize calculation by adduction, average etc. and obtain, and
And can carry out summarizing calculating in certain precondition, such as time, place, range, that is, Statistical Criteria and range.Index
Can be divided into absolute number index and relative number index, the reflection of absolute number index be scale index, as population, GDP,
Income, number of users, and relative number index is mainly used to reflect the index of quality, such as profit margin, retention ratio, coverage rate.
It can analyze the influence that the user of different dimensions changes portfolio in the present embodiment, by way of example, have a use
The history service data at family, the index of this part of data are portfolio (specifically being characterized using order volume), and index change rate is using same
Than as an example, user's dimension may include: occupation, city, gender and age bracket in this part of data with ring ratio.In practical application
Also it can according to need and other a variety of dimensions be set, the present embodiment is not construed as limiting this.For example, the enumerated value of occupation includes as follows
University student, teaching and administrative staff or blue collar etc. 11, the corresponding portfolio variation of each enumerated value is as follows:
Occupation |
Order volume is year-on-year |
Order volume ring ratio |
It can not identify |
1.21% |
0.13% |
University student |
3.29% |
0.35% |
Famous enterprise employee |
21.87% |
6.1% |
Teaching and administrative staff |
1.44% |
0.62% |
Self-employed worker |
2.03% |
0.27% |
Blue collar |
0.69% |
0.07% |
Civil servant |
0.81% |
0.05% |
Medical worker |
1.88% |
0.19% |
White collar |
0.13% |
- 0.02% |
Listed company employee |
- 1.04% |
- 0.35% |
Listed Company |
0.43% |
0.03% |
In this part of data of the present embodiment, the ring ratio of order volume quantity be 0.18% (indicate the situation of change of number of users,
The average rate of change for the order volume quantity being calculated with whole part data), the target of data analysis is, from occupation, city, property
And in 4 dimensions of age bracket influence of the enumerated value of each dimension to user's ring ratio is not analyzed, that is to say, that detection causes user
Influences of the main reason for ring is than changing, each dimension of detection have it is much, detect which dimension influences maximum and the dimension
Influence degree etc..
Above-mentioned example may divide more coarse, to business by taking 4 occupation, city, gender and age bracket dimensions as an example
The effect of analysis is not obvious.Optionally, dimension can also be that each enumerated value in above-mentioned occupation is " university student, blue collar, white
Neck, medical worker etc. " distinguishes user, and such dimension can also further have one or more enumerated values, such as
The enumerated value of " medical worker " may include " director, deputy director, attending physician, resident or nurse etc. ".
In the present embodiment, the data distribution of history service data is corresponded to using dimension and difference that target is evenly distributed determines
Influence of the dimension to portfolio;It in information theory, works to something, often a few factor, other are big
Partial factor is all unimportant, and in the present embodiment, the influence for portfolio variation often concentrates on least a portion of use
Family.Therefore, it can use the data distribution that each dimension corresponds to history service data in the present embodiment, to analyze portfolio variation
Which user group concentrated on.Specifically, data distribution can be determined using the variation degree of the enumerated value of each dimension;Mesh
Mark, which is evenly distributed, can according to need flexible configuration, such as can be normal distribution, power-law distribution or Poisson distribution etc..Pass through ratio
The data distribution of history service data is corresponded to compared with dimension and difference that target is evenly distributed, so that it is determined that dimension changes portfolio
Influence degree.
Optionally, in the present embodiment, what which can be evenly distributed by calculating the data distribution with the target
Cross entropy and determine.Cross entropy (Cross Entropy) is a key concept in Shannon information theory, is mainly used for measurement two
Otherness information between probability distribution.In information theory, cross entropy is to indicate two probability distribution p, q, and wherein p indicates true point
Cloth, q indicates non-genuine distribution, in identical one group of event, wherein indicates that institute occurs for some event with non-genuine distribution q
The average needed.
As an example, cross entropy can be calculated by following formula:
Wherein, C_K is the cross entropy, P_KiFor the accounting parameter of the enumerated value i of the dimension, Q_K is the dimension
Target be evenly distributed, Q_K=1/m, m are the dimension beta pruning treated enumerated value number.Wherein, cross entropy is bigger, table
The influence for showing that the dimension changes portfolio is bigger.
In the present embodiment, target is evenly distributed by determining after the following beta pruning processing of enumerated value progress to the dimension,
Specific prune approach includes: to delete the enumerated value for not meeting power law distribution.Wherein, power law distribution (pow law
Distribution), probability density function form is as follows:
Y=et
Power-law distribution shows as the straight line for the negative that a slope is power exponent, and power law distribution indicates, rises to something
Effect, often a few factor, other most factors are all unimportant;Field is provided in the electronic ticket of the present embodiment
Jing Zhong, it can be understood as, some dimension often focuses on wherein a small number of enumerated values to the increased influence of portfolio, other are big
Partial enumerated value influences smaller.Based on this, the present embodiment deletes the enumerated value for being lower than power law distribution, it is possible to reduce noise is got rid of
Abandon the lower three kinds of situations of data information amount.For example, the enumerated value corresponding data missing values when the dimension are fewer, it is unknown only to account for
When than sub-fraction, this certain customers can be precisely kicked off;If the missing values accounting of certain dimension is bigger, and to there is information
The data of content are also relatively low, and the enumerated value of the dimension can be also removed.If itself distributional difference of data is bigger, such as
90% user is the user of Hangzhou, then the variation of itself just will affect whole unusual fluctuation analyses, then the dimension
Enumerated value can also be deleted.
By above-mentioned processing, the influence degree that each dimension changes portfolio can be determined, in the present embodiment, the business
Amount variation can refer to increased portfolio, may also mean that reduced portfolio.The influence that each dimension changes portfolio
Degree may vary.Wherein, according to definitive result, adjustment is issued to the process for the electronic ticket for belonging to the dimension user,
Also it can according to need flexible configuration, such as according to the influence degree of all dimensions, correspondingly to the electronics of all dimension users
Certificate is adjusted;It can also be adjusted for the electronic ticket of partial dimensional user, the adjustment amplitude of influence degree and electronic ticket
It can be positive correlation, influence degree is biggish, and also corresponding larger for the adjustment amplitude of electronic ticket, influence degree is smaller
, it is also corresponding smaller for the adjustment amplitude of electronic ticket.
Optionally, which user group couple is the influence degree changed by the determination dimension to portfolio, can know
The increased contribution of portfolio is larger, which user group contribution increased to portfolio is smaller, for example, according to calculated result, it is right
Portfolio, which increases, contributes maximum first dimension collection to may is that Beijing, Shanghai, Hangzhou, Shenzhen and Guangzhou;Portfolio is increased
Contribution minimum enumerates value set and may is that Kunming.Based on this, it can be improved to be issued to and belong to the first dimension collection user's
Electronic ticket, or reduce and be issued to the electronic ticket for belonging to the second dimension collection user.
Optionally, the quantity that can be for electronic ticket is increased or decreased, and the preferential amount of electronic ticket is also possible to
Etc. being increased or decreased, it is specific improve mode and reduce mode can be with flexible configuration, the present embodiment is not construed as limiting this.Make
For example, the definitive result includes: to influence maximum first dimension collection to portfolio variation and influence the smallest second dimension
Degree set;In the case where needing to adjust the first electronic ticket and the second electronic ticket, in order to make raising mode and reduction mode can be with
It more balances, in the present embodiment, belongs to the adjustment of the first electronic ticket of the first dimension collection user, and belong to described second
The adjustment of the second electronic ticket of dimension collection user has negative correlativing relation, that is to say, that the biggish user group's tune of contribution
High electronic ticket, then contribute lesser user group to turn down electronic ticket other, so that amplitude and electronics that electronic ticket improves
The amplitude that certificate reduces can match.
Optionally, the quantity or total value of possible electronic ticket have certain limitations in practical application, or for the electronics of adjustment
The quantity or total value of certificate have certain limitations, and in the present embodiment, the coefficient and the second electronic ticket that first electronic ticket improves are reduced
Coefficient can also meet following relationship: ax=by, wherein a is the quantity ginseng for belonging to the first dimension collection user
Number, b are the number parameter for belonging to the second dimension collection user, and x is the coefficient that first electronic ticket improves, and y is
The coefficient that second electronic ticket reduces.Above-mentioned formula is characterized in during the adjustment of electronic ticket granting, belongs to first dimension
The electronic ticket of degree aggregate user improves how many, can correspondingly reduce the electronic ticket for belonging to the second dimension collection user, make
The adjustment for obtaining electronic ticket can satisfy demand, be also possible to prevent the excessive adjustment of partial mass user, thus will not be to other use
Family impacts.
Next it is illustrated again by an embodiment.After electronic ticket granting, the present embodiment can going through by user
History business datum searches out most sensitive crowd and least sensitive crowd.Wherein, sensitive group refers to, because electronic ticket is sent out
After putting, the crowd of transaction stroke count amplification maximum or the biggest drop;Insensitive crowd refers to, after electronic ticket is provided, stroke count of trading
The crowd that amplification is kept approximately constant.
Using stroke count of trading on year-on-year basis as the index for measuring business datum variation, first according to each dimension for determining the index
(such as gender, city or occupation etc.);The each of each dimension is calculated for each dimension according to the selection of Data Analyst
The following accounting parameter of enumerated value: enumerated value corresponds to the ratio of the increased portfolio of entirety of increased portfolio and the dimension
P_Ki(1 > P_K > 0).
For the P_K of each enumerated value of the dimensioniIt is ranked up, and is fitted based on power law distribution, deleted not
The enumerated value for meeting power law distribution, by the enumerated value number of dimension from n beta pruning to m.
According to the enumerated value number after beta pruning, the target for calculating a hypothesis is evenly distributed:
The difference for calculating original data distribution and being evenly distributed:
Wherein, C_K is the cross entropy, P_KiFor the accounting parameter of the enumerated value i of the dimension, Q_K is the dimension
Target be evenly distributed, Q_K=1/m, m are the dimension beta pruning treated enumerated value number.
The C_K of the enumerated value of each dimension is ranked up, the maximum i.e. sensitive group of C_K, changing on portfolio influences
It is maximum;The smallest i.e. insensitive crowd of C_K, changing on portfolio influences minimum.
Finally according to the number ratio of sensitive group and insensitive crowd, dynamic regulation coefficient, for example, can be for sensitivity
The coefficient of crowd is turned up, and reduces for the coefficient of insensitive crowd, by way of example, is adjusted by following formula
It is whole: ax=by, wherein a is the number parameter for belonging to the first dimension collection user, and b belongs to described second to be described
The number parameter of dimension collection user, x are the coefficient of first electronic ticket adjustment, and y is for second electronic ticket adjustment
Number.
Optionally, can be using x=0.1 as initial value and step-length, dynamic increment adjustment;Optionally, it excessively adjusts in order to prevent
It is whole, the upper limit or lower limit of coefficient can also be set, for example, sensitive group amount is no more than 0.1% or coefficient does not exceed more than s, s
It can be set as needed.
Above-described embodiment is illustrated so that electronic ticket provides scene as an example, and the present embodiment additionally provides resource allocation methods,
It is applicable to multiple resources allocation scenarios.As shown in Fig. 2, Fig. 2 is this specification one kind shown according to an exemplary embodiment
The flow chart of resource allocation methods, comprising the following steps:
In step 202, the history service data of acquisition user, the user is distinguished by multiple dimensions, institute
Portfolio variation after stating history service data instruction resource allocation;
In step 204, correspond to according to the dimension data distribution of history service data and difference that target is evenly distributed,
Determine the influence degree that the dimension changes portfolio;
In step 206, according to definitive result, adjustment is issued to the resource for belonging to the dimension user.
In the present embodiment, after resource allocation, in order to determine resource user upon receipt after resource to the shadow of business
It rings, the portfolio variation after resource allocation can be indicated by the history service data of acquisition user, the history service data,
And distinguish user by multiple dimensions, it is based on this, the data distribution of history service data is corresponded to according to the dimension
The difference being evenly distributed with target can determine the influence degree that the dimension changes portfolio, and then according to influence degree
Adjustment belongs to the resource of dimension user, so that the available effective use of resource.
Resource in the present embodiment may include electronic ticket, processing thread, hardware resource or bandwidth resources etc., and resource makes
It may include user, trade company, computer equipment or server cluster etc. with side;The specific dimension of resource user, according to reality
Border business scenario and flexible configuration.
In the present embodiment, the concern target of history service data is portfolio variation, therefore using portfolio as index,
Index is used to measure the unit or method of things development degree, in other examples, can also include portfolio, population,
GDP, income, number of users, profit margin, retention ratio, coverage rate etc..Index can summarize calculation by adduction, average etc. and obtain
It arrives, and can carry out summarizing calculating in certain precondition, such as time, place, range, that is, Statistical Criteria and model
It encloses.
In the present embodiment, the data distribution of history service data is corresponded to using dimension and difference that target is evenly distributed determines
Influence of the dimension to portfolio;It in information theory, works to something, often a few factor, other are big
Partial factor is all unimportant, and in the present embodiment, the influence for portfolio variation often concentrates on least a portion of use
Family.Therefore, it can use the data distribution that each dimension corresponds to history service data in the present embodiment, to analyze portfolio variation
Which concentrate in user group.Specifically, data distribution can be using the variation degree of the enumerated value of each dimension come really
It is fixed;Target, which is evenly distributed, can according to need flexible configuration, such as can be normal distribution, power-law distribution or Poisson distribution etc..
The data distribution of history service data is corresponded to by comparing dimension and difference that target is evenly distributed, so that it is determined that dimension is to business
Measure the influence degree of variation.
Optionally, the difference is determined by calculating the cross entropy that the data distribution is evenly distributed with the target.
Optionally, the dimension has multiple enumerated values, and the target is evenly distributed through the enumerated value to the dimension
It is determined after carrying out following beta pruning processing: deleting the enumerated value for not meeting power law distribution.
Power-law distribution shows as the straight line for the negative that a slope is power exponent, and power law distribution indicates, rises to something
Effect, often a few factor, other most factors are all unimportant;Scene is provided in the resource of the present embodiment
In, it can be understood as, some dimension often focuses on wherein a small number of enumerated values, other big portions to the increased influence of portfolio
The enumerated value divided influences smaller.Based on this, the present embodiment deletes the enumerated value for being lower than power law distribution, it is possible to reduce noise is abandoned
The lower three kinds of situations of data information amount.For example, the enumerated value corresponding data missing values when the dimension are fewer, unknown only accounting
When sub-fraction, this part resource user can be precisely kicked off;If the missing values accounting of certain dimension is bigger, and to having
The data of information content are also relatively low, and the enumerated value of the dimension can be also removed.If itself distributional difference of data is bigger,
Such as 90% user be all Hangzhou resource user, then the variation of itself just will affect whole unusual fluctuation analyses,
Then the enumerated value of the dimension can also be deleted.
Optionally, the data distribution is determined by calculating the following accounting parameter of each enumerated value of corresponding dimension:
Enumerated value corresponds to the ratio of the increased portfolio of entirety of increased portfolio and the dimension.
Optionally, the cross entropy is calculated in the following way:
Wherein, C_K is the cross entropy, P_KiFor the accounting parameter of the enumerated value i of the dimension, Q_K is the dimension
Target be evenly distributed, Q_K=1/m, m are the dimension beta pruning treated enumerated value number.
By above-mentioned processing, the influence degree that each dimension changes portfolio can be determined, in the present embodiment, the business
Amount variation can refer to increased portfolio, may also mean that reduced portfolio.The influence that each dimension changes portfolio
Degree may vary, and according to definitive result, adjustment is issued to the process for the resource for belonging to the dimension user, can also
With flexible configuration as needed, such as according to the influence degree of all dimensions, correspondingly to the resource of all dimension users into
Row adjustment;It can also be adjusted for the resource of partial dimensional user.The adjustment amplitude of influence degree and resource can be
Positive correlation, influence degree is biggish, also corresponding larger for the adjustment amplitude of resource, and influence degree is lesser, for money
The adjustment amplitude in source is also corresponding smaller.
Optionally, the definitive result includes: to influence maximum first dimension collection to portfolio variation and influence minimum
The second dimension collection;
The adjustment for belonging to the first resource of the first dimension collection user is used with second dimension collection is belonged to
The adjustment of the Secondary resource of side has negative correlativing relation.
In order to balance raising mode and reduction mode more, in the present embodiment, belong to first dimension collection
The adjustment of the first resource of user, the adjustment with the Secondary resource for belonging to the second dimension collection user have negative correlation
Relationship, that is to say, that contributing biggish user group that resource has been turned up, then lesser user group is contributed to other
Resource is turned down, the amplitude that the amplitude that resource is improved is reduced with resource matches.
Optionally, the quantity of possible resource or total value have certain limitations in practical application, or the resource for adjustment
Quantity or total value have certain limitations, and the coefficient of the first electronic ticket adjustment and the coefficient of the second electronic ticket adjustment meet such as ShiShimonoseki
System: ax=by, wherein a is the number parameter for belonging to the first dimension collection user, and b is described to belong to described the
The number parameter of two-dimensions set user, x are the coefficient of first resource adjustment, and y is for Secondary resource adjustment
Number.
In this specification embodiment, by obtaining the history service data of user, which indicates resource
Portfolio variation after granting, and user is distinguished by multiple dimensions, it is based on this, history is corresponded to according to the dimension
The difference that the data distribution and target of business datum are evenly distributed, can determine the influence journey that the dimension changes portfolio
Degree, and then the resource for belonging to dimension user is adjusted according to influence degree, so that the available effective use of resource.
Corresponding with aforementioned electronic certificate granting/embodiment of resource allocation methods, this specification additionally provides electronic ticket hair
Put/the embodiment of resource allocation device and its applied computer.
This specification electronic ticket granting/resource allocation device embodiment can be applied on a computing device, such as take
Business device or terminal device.Installation practice can be by software realization, can also be by way of hardware or software and hardware combining
It realizes.It taking software implementation as an example, is by electronic ticket granting/resource allocation where it as the device on a logical meaning
Processor by computer program instructions corresponding in nonvolatile memory be read into memory operation formed.From hardware layer
For face, as shown in figure 3, for one kind of computer equipment where this specification embodiment electronic ticket granting/resource allocation device
Hardware structure diagram, in addition to processor 310 shown in Fig. 3, memory 330, network interface 320 and nonvolatile memory 340 it
Outside, the server or electronic equipment in embodiment where device 331 may be used also generally according to the actual functional capability of the computer equipment
To include other hardware, this is repeated no more.
As shown in figure 4, Fig. 4 is a kind of this specification frame of electronic ticket dispensing apparatus shown according to an exemplary embodiment
Figure, described device include:
Module 41 is obtained, is used for: obtaining the history service data of user, the user is distinguished by multiple dimensions,
Portfolio after the history service data instruction electronic ticket is provided changes;
Determining module 42, is used for: corresponding to the data distribution of history service data according to the dimension and target is evenly distributed
Difference, determine the influence degree that the dimension changes portfolio;
Adjust module 43, be used for: according to definitive result, adjustment is issued to the electronic ticket for belonging to the dimension user.
Optionally, the difference is determined by calculating the cross entropy that the data distribution is evenly distributed with the target.
Optionally, the dimension has multiple enumerated values, and the target is evenly distributed through the enumerated value to the dimension
It is determined after carrying out following beta pruning processing: deleting the enumerated value for not meeting power law distribution.
Optionally, the data distribution is determined by calculating the following accounting parameter of each enumerated value of corresponding dimension:
Enumerated value corresponds to the ratio of the increased portfolio of entirety of increased portfolio and the dimension.
Optionally, the cross entropy is calculated in the following way:
Wherein, C_K is the cross entropy, P_KiFor the accounting parameter of the enumerated value i of the dimension, Q_K is the dimension
Target be evenly distributed, Q_K=1/m, m are the dimension beta pruning treated enumerated value number.
Optionally, the definitive result includes: to influence maximum first dimension collection to portfolio variation and influence minimum
The second dimension collection;
The adjustment for belonging to the first electronic ticket of the first dimension collection user, and belongs to the second dimension collection user
The second electronic ticket adjustment have negative correlativing relation.
Optionally, the coefficient of the first electronic ticket adjustment and the coefficient of the second electronic ticket adjustment meet following relationship: ax
=by, wherein a is the number parameter for belonging to the first dimension collection user, and b belongs to the second dimension collection to be described
The number parameter at family is shared, x is the coefficient of first electronic ticket adjustment, and y is the coefficient of second electronic ticket adjustment.
As shown in figure 5, Fig. 5 is a kind of this specification frame of resource allocation device shown according to an exemplary embodiment
Figure, comprising:
Module 51 is obtained, is used for: obtaining the history service data of user, the user carries out area by multiple dimensions
Point, the portfolio variation after the history service data instruction resource allocation;
Determining module 52, is used for: corresponding to the data distribution of history service data according to the dimension and target is evenly distributed
Difference, determine the influence degree that the dimension changes portfolio;
Adjust module 53, be used for: according to definitive result, adjustment is issued to the resource for belonging to the dimension user.
Optionally, the difference is determined by calculating the cross entropy that the data distribution is evenly distributed with the target.
Optionally, the dimension has multiple enumerated values, and the target is evenly distributed through the enumerated value to the dimension
It is determined after being handled as follows: deleting the enumerated value for not meeting power law distribution.
Optionally, the data distribution is determined by calculating the following accounting parameter of each enumerated value of corresponding dimension:
Enumerated value corresponds to the ratio of the increased portfolio of entirety of increased portfolio and the dimension.
Optionally, the cross entropy is calculated in the following way:
Wherein, C_K is the cross entropy, P_KiFor the accounting parameter of the enumerated value i of the dimension, Q_K is the dimension
Target be evenly distributed, Q_K=1/m, m are the dimension beta pruning treated enumerated value number.
Optionally, the definitive result includes: to influence maximum first dimension collection to portfolio variation and influence minimum
The second dimension collection;
The adjustment for belonging to the first resource of the first dimension collection user is used with second dimension collection is belonged to
The adjustment of the Secondary resource of side has negative correlativing relation.
Optionally, the coefficient of the first resource adjustment and the coefficient of Secondary resource adjustment meet following relationship: ax=by,
Wherein, a is the number parameter for belonging to the first dimension collection user, and b belongs to second dimension collection to be described
The number parameter of user, x are the coefficient of first resource adjustment, and y is the coefficient of Secondary resource adjustment.
According to the 5th of this specification embodiment aspect, a kind of computer equipment is provided, including memory, processor and deposit
Store up the computer program that can be run on a memory and on a processor, wherein the processor is realized when executing described program
Such as aforementioned electronic certificate distribution method/resource allocation methods.
The function of modules and the realization process of effect are specifically detailed in the above method and correspond to step in above-mentioned apparatus
Realization process, details are not described herein.
For device embodiment, since it corresponds essentially to embodiment of the method, so related place is referring to method reality
Apply the part explanation of example.The apparatus embodiments described above are merely exemplary, wherein described be used as separation unit
The module of explanation may or may not be physically separated, and the component shown as module can be or can also be with
It is not physical module, it can it is in one place, or may be distributed on multiple network modules.It can be according to actual
The purpose for needing to select some or all of the modules therein to realize this specification scheme.Those of ordinary skill in the art are not
In the case where making the creative labor, it can understand and implement.
It is above-mentioned that this specification specific embodiment is described.Other embodiments are in the scope of the appended claims
It is interior.In some cases, the movement recorded in detail in the claims or step can be come according to the sequence being different from embodiment
It executes and desired result still may be implemented.In addition, process depicted in the drawing not necessarily require show it is specific suitable
Sequence or consecutive order are just able to achieve desired result.In some embodiments, multitasking and parallel processing be also can
With or may be advantageous.
Those skilled in the art will readily occur to this specification after considering specification and practicing the invention applied here
Other embodiments.This specification is intended to cover any variations, uses, or adaptations of this specification, these modifications,
Purposes or adaptive change follow the general principle of this specification and do not apply in the art including this specification
Common knowledge or conventional techniques.The description and examples are only to be considered as illustrative, the true scope of this specification and
Spirit is indicated by the following claims.
It should be understood that this specification is not limited to the precise structure that has been described above and shown in the drawings,
And various modifications and changes may be made without departing from the scope thereof.The range of this specification is only limited by the attached claims
System.
The foregoing is merely the preferred embodiments of this specification, all in this explanation not to limit this specification
Within the spirit and principle of book, any modification, equivalent substitution, improvement and etc. done should be included in the model of this specification protection
Within enclosing.