CN104915742A - User preference-based resource allocation method and apparatus - Google Patents

Abstract

The invention relates to a user preference-based resource allocation method and apparatus. The method includes the following steps that: the attribute information of resources to be allocated and the preference information of a plurality of users are acquired; as for an arbitrary resource allocation scheme, the user's satisfaction degree of each user for allocated resources is evaluated based on the preference information of each user and the attribute information of the resources allocated for the user; the overall satisfaction degree of the resource allocation scheme is evaluated based on the user's satisfaction degree of each user for the allocated resources; the resource allocation scheme is adjusted based on the user's satisfaction degree of each user, and the overall satisfaction degree of the adjusted resource allocation scheme is re-evaluated, and an adjustment process is repeated until an adjustment termination condition is satisfied, so that optimal comprehensive satisfaction degree can be obtained; and the resources are allocated to the plurality of users through utilizing the resource allocation scheme with the optimal comprehensive satisfaction degree.

Description

Based on the method and apparatus of user preference Resources allocation

Technical field

The application relates to Resourse Distribute, more specifically, relates to the method and apparatus based on user preference Resources allocation.

Background technology

At present for the service trade providing resource to user, the satisfaction of user to provided resource be weigh the key factor of the quality of service is provided.In public service occasion, for user provides the attribute of the such as position of resource etc. often to have decisive role to users satisfaction degree.Such as, at the cinema, the Public place of restaurant and hotel and various public transport such as train, aircraft etc., people can rely on the preference of their uniqueness separately when selecting their seat or room.

Existing service industry adopts the Service Principle of " First come first served " all the time, and the client namely arrived first preferentially can obtain service, namely can resource needed for its preference demand prioritizing selection, such as position etc.But the Service Principle of this " First come first served " easily causes the user satisfaction of suboptimum, cause primarily of following two reasons:

(1) satisfaction between the client arrived first and the client arrived afterwards is seriously unbalanced.Under the condition that resource quantity is limited, the client arrived first may occupy all good resources, and leave the negligible amounts of resource that can select for client afterwards for and second-rate, this will cause the satisfaction of client afterwards to reduce widely, thus it is seriously unbalanced to cause between different customer satisfaction, easily causes unnecessary contradiction and conflict;

(2) client arrived first may due to rear to client interference and reduce satisfaction.Although the client arrived first has preoption, still have the final satisfaction that many unforeseen factors can affect him.Than if any client like quiet environment, sit down so come have selected comparatively quietly position early, but get more and more along with the client arrived, his periphery has occupied people at leisure, and so his satisfaction will reduce greatly.

Due to these reasons above, some service providers have started " First come first served " Service Principle discarded tradition, but turn to the advanced services system of global optimization, uniform service.But current technology can not fast searching optimum Resource Allocation Formula and effectively realize balanced user satisfaction.Below some existing technology of carrying out unified resource distribution based on global optimization method.

Patent documentation 1 is the U.S. Patent application submitted on June 26th, 2012 to United States Patent Office (USPO) by ConUS airline, and publication number is US2012/0022901A1.It discloses the airline reservation system that a kind of airline is supplied to passenger, different from traditional airline reservation system, this airline reservation system is display machines seating module distribution map not, but allow the seat classification of passenger's selection preference and feature, like this after all passengers all have input individual preference information, the background program of this airline reservation system carries out unified allocation of resources, to meeting the personal needs of each passenger as far as possible by according to the individual preference of each passenger and the situation of actual cabin seats.But the concrete methods of realizing of the seating allocation of this patented claim its airline reservation system unexposed.

Prior art document 2 is U.S. Patent applications that Toshiyuki Omiya submitted on March 28th, 2013 to United States Patent Office (USPO), and publication number is US2013/0076089A1.It discloses the seat arrangement of a kind of office and the warning system of employee individual preference when clashing.This system also needs the personalization preferences information of known each employee and the attribute information at each office seat, but this system is only responsible for prediction when the seat of certain employee changes, the preference of which employee is had to be affected, and these results predicted are aggregated into a report, submit to relevant managerial personnel, be responsible for coordinating rearranging of seat further by managerial personnel again, until the generation that do not conflict.That is, be only that a kind of seat of semi-automation arranges means disclosed in the document, give the alarm when its automated system is only responsible for prediction and conflict, and the rearranging of not responsible seat, so often cause artificial arrangement repeatedly, reduce the efficiency that seat arranges widely.

Summary of the invention

Consider above problem, present applicant proposes the resource allocation system needing to carry out expeditiously Resourse Distribute according to the preference of people.

According to an aspect of the present invention, the method to multiple user resource allocation is provided.The method can comprise: obtain the attribute information of resource to be allocated and the preference information of described multiple user; For arbitrary Resource Allocation Formula, preference information and the attribute information of the resource distributing to this user based on each user assess the user satisfaction of each user to the resource of distributing; Based on the user satisfaction of each user to the resource of distributing, assess the comprehensive satisfaction of this Resource Allocation Formula; Based on the user satisfaction adjustresources allocative decision of each user, and the Resource Allocation Formula after adjustment is reappraised its comprehensive satisfaction, repeat adjustment process until reach adjustment end condition, to obtain best comprehensive satisfaction; And utilize the Resource Allocation Formula with this best comprehensive satisfaction to described multiple user resource allocation.

Alternatively, adjustresources allocative decision can comprise the resource changing randomly and distribute to the user of the predetermined quantity with minimum user satisfaction.

Alternatively, the method can also comprise: to the controlled number of times of distribution of each its resource of user number counting, adjust counting as user; And based on each user to the user satisfaction of the resource of current distribution and current user adjust counting all users are sorted.Adjustresources allocative decision can comprise the resource changing the user distributing to the minimum predetermined quantity of sequence randomly.

Alternatively, assess the user satisfaction of each user to the resource of distributing based on the preference information of each user and the attribute information of resource of distributing to this user can comprise: assess the initial satisfaction of each user to the resource of distributing based on the preference information of each user and the attribute information of the resource of distributing to this user; And based on this initial satisfaction of relativeness correction between the relation preference information between user and the resource of distributing, to obtain the described user satisfaction of each user to the resource of distributing.

Alternatively, can comprise the comprehensive satisfaction of this Resource Allocation Formula of satisfaction assessment of the resource of distributing based on each user: based on the user satisfaction of each user to the resource of distributing, assess the initial synthetic satisfaction of this Resource Allocation Formula; And based on each user to the difference between the user satisfaction of the resource of distributing, revise this initial synthetic satisfaction to obtain the described comprehensive satisfaction of described Resource Allocation Formula.

Alternatively, described user can be utilized to the minimum value of the user satisfaction of the resource of distributing to revise described initial synthetic satisfaction to obtain the described comprehensive satisfaction of described Resource Allocation Formula.

Alternatively, can comprise the comprehensive satisfaction of this Resource Allocation Formula of satisfaction assessment of the resource of distributing based on each user: based on the user satisfaction of each user to the resource of distributing, assess the initial synthetic satisfaction of this Resource Allocation Formula; And based on energy resource consumption total amount corresponding to this Resource Allocation Formula, revise this initial synthetic satisfaction to obtain the described comprehensive satisfaction of described Resource Allocation Formula.

Alternatively, can comprise based on the initial synthetic satisfaction of each user to this Resource Allocation Formula of satisfaction assessment of the user resources of distributing and calculate described initial synthetic satisfaction to the user satisfaction weighting of each user to the resource of distributing, wherein the weighting coefficient of each user depends on the priority of this user in resource allocation process.

Alternatively, stop regularization condition and can comprise following any one: the adjustment number of times adjusting described Resource Allocation Formula reaches the first pre-determined number, adjust the regulation time of described Resource Allocation Formula more than the first predetermined amount of time, assess Resource Allocation Formula comprehensive satisfaction is more than or equal to predetermined threshold, in the second predetermined amount of time or in the second predetermined adjustment number of times, described best comprehensive satisfaction does not improve.This best comprehensive satisfaction can be the maximal value of current obtained comprehensive satisfaction.

According to a further aspect in the invention, the device to multiple user resource allocation is provided.This device can comprise: obtaining widget, obtains the attribute information of resource to be allocated and the preference information of described multiple user; User satisfaction evaluation means, for arbitrary initial resource allocation scheme, preference information and the attribute information of the resource distributing to this user based on each user assess the user satisfaction of each user to the resource of distributing; Comprehensive satisfaction evaluation means, based on the user satisfaction of each user to the resource of distributing, assesses the comprehensive satisfaction of this Resource Allocation Formula; Adjustment component, based on each user user satisfaction adjustresources allocative decision and make described user satisfaction evaluation means and described comprehensive satisfaction evaluation means to the Resource Allocation Formula assessment user satisfaction after adjustment and comprehensive satisfaction, repeat adjustment process until reach adjustment end condition, to obtain best comprehensive satisfaction; And distribution member, utilize the Resource Allocation Formula with this best comprehensive satisfaction to described multiple user resource allocation.

According to the present invention, carry out the global optimization of Resourse Distribute based on the preference information of total user and the attribute information of allowable resource, unified allocation of resources resource, can indistinguishably treat each user.

Accompanying drawing explanation

Fig. 1 illustrates the method to multiple user resource allocation according to an embodiment of the invention.

Fig. 2 illustrates the process flow diagram of each user of assessment according to an embodiment of the invention to the method for the user satisfaction of the resource of distributing.

Fig. 3 illustrates the process flow diagram of Resource Allocation Formula method of adjustment according to an embodiment of the invention.

Fig. 4 illustrates the process flow diagram of Resource Allocation Formula method of adjustment according to another embodiment of the present invention.

Fig. 5 illustrates the schematic diagram of an embody rule example of resource allocation methods of the present invention.

Fig. 6 illustrates the device to multiple user resource allocation according to an embodiment of the invention.

Embodiment

Present by detail with reference to specific embodiments of the invention, in the accompanying drawings exemplified with example of the present invention.Although the present invention will be described in conjunction with specific embodiments, will understand, and be not intended to limit the invention to disclosed specific embodiment.It should be noted that method step described herein can be arranged by any functional block or function realize, and any functional block or function are arranged and can be implemented as physical entity or logic entity or both combinations.

Herein if no special instructions, so-called Resourse Distribute to user and change Resource Allocation Formula all refer to " virtual assigned " process, " virtual assigned " process should can carry out on the computing equipment of such as computing machine, but not resource is actually allocated to user.The result of Resourse Distribute can inform user by oral or written form, and to the explanation that the result of the program gives to a certain degree, such as consider certain aspect of this user needs or and other users between relation etc.

According to the present invention, for the Resource Allocation Formula of arbitrary initial, assess the comprehensive satisfaction of this Resource Allocation Formula, repeatedly adjustresources allocative decision, and the comprehensive satisfaction of the Resource Allocation Formula after each adjustment of assessment, until obtain best comprehensive satisfaction, in order to there is the Resource Allocation Formula of this best comprehensive satisfaction to user resource allocation.Thereby, it is possible to carry out with best comprehensive satisfaction to user resource allocation expeditiously.

Fig. 1 shows according to one embodiment of present invention to the process flow diagram of the method 100 of multiple user resource allocation.As shown in Figure 1, method 100 comprises the following steps:

S110, obtains the attribute information of resource to be allocated and the preference information of described multiple user;

S120, for arbitrary Resource Allocation Formula, preference information and the attribute information of the resource distributing to this user based on each user calculate the user satisfaction of each user to the resource of distributing;

S130, based on the user satisfaction of each user to the resource of distributing, calculates the comprehensive satisfaction of this Resource Allocation Formula;

S140, based on the user satisfaction adjustresources allocative decision of each user, and to reappraise its comprehensive satisfaction to the Resource Allocation Formula after adjustment, repeats adjustment process until reach adjustment end condition, to obtain best comprehensive satisfaction;

S150, utilizes the Resource Allocation Formula with this best comprehensive satisfaction to described multiple user resource allocation.

Below describe the specific embodiment of the method in detail.

In step S110, obtain the attribute information of resource to be allocated.As an example to multiple user resource allocation, such as, under seat distribution condition, there are some attributes corresponding with the preference of user at each seat, such as position, towards, humiture and brightness etc.The value of these attributes can utilize specific surveying instrument to carry out measuring and record, and the list of attribute values at three seats such as, provided in table 1 for simplicity, only gives the value of position and illumination two attributes here.But when practical application, the specific object can choosing resource according to concrete demand carries out measuring to obtain the Resource Properties list being similar to table, and this Resource Properties list easily can expand to resource and the attribute of greater number.

Table 1

Seat is numbered	Position feature	Illumination (lx)
			1	Near window	800
2	Near window	500

3

Away from window

300

In step S110, also obtain the preference information of each user to resource.This preference information represents the preference of user to each attribute of resource.Usually, there is the method for multiple description user preference, preferably utilize in this embodiment to the expectation value of certain Resource Properties and tolerance as preference information, so that can based on distributing to the attribute assessment user of any cost of user for the satisfaction of this resource.Still for the situation of seat allocation, such as, in table 2, give the position of certain user to arranged seat and the preference information of illumination.Similarly, this table also easily can expand to the situation of multiple user and multiple Resource Properties.

Table 2

Attribute	Expectation value	Tolerance
			Position	Near window	Within distance window 2 meters
Illumination	500lx	±100lx
			Relation	Near user B	Within distance 5 meters

There is the method for multiple acquisition user preference, such as, by acquisitions such as talk, survey, online registration, PUSH message face to face.Suitable mode can be selected when practical application as the case may be to obtain user preference information, and note the privacy of protection user.If certain user's refusal provides relevant information, then process according to default user preference.

In the step s 120, for arbitrary initial resource allocation scheme, preference information and the attribute information of the resource distributing to this user based on each user assess the user satisfaction of each user to the resource of distributing.Such as, when M Resourse Distribute being given N number of user, wherein M >=N, can in M resource, random choose be N number of distributes to N number of user, as initial resource allocation scheme.Or, the existing Resource Allocation Formula determined by other means can be utilized as the initial scheme of this method.Or, Resourse Distribute can be carried out according to predetermined order and user preference information per user, namely simulate the allocation strategy of " First come first served ".No matter select any initial resource allocation scheme, according to simulation result, the Resource Allocation Formula finally determined does not rely on the selection of initial scheme substantially.

For the Resource Allocation Formula that this is initial, based on each user obtained preference information and distribute to the attribute information of resource of this user, assess the user satisfaction of each user to the resource of distributing.As an example, if for the user with the user preference information in table 2 is assigned with the seat 1 shown in table 1, seat feature due to this position near window and in the preference information of this user is also near window to the preference of position, then the position (near window) of this user to this seat 1 is satisfied, but it is just not too satisfied to the illumination (100lx) at this seat 1, because the illumination that he expects is at 500lx, tolerance is within 100lx, see so both integrate, the satisfaction at first seat described in user's his-and-hers watches 1 corresponding in table 2 is in medium level.

Fig. 2 shows according to the process flow diagram of each user of the assessment of an embodiment to the exemplary method 200 of the user satisfaction of the resource of distributing.As shown in Figure 2, the method 200 can comprise the following steps:

Step S210, obtain user for distributed resource each attribute be satisfied with angle value;

Step S220, based on user for distributed resource each attribute be satisfied with angle value, obtain user to the user satisfaction of distributed resource.

Particularly, in one embodiment, the matching degree that can quantize between the attribute of resource and the preference information of user with concrete numerical value, as user angle value is satisfied with to the attribute of resource, and set up in advance and be satisfied with the look-up table of angle value, search for the specific object information according to each preference information and resource the user with any preference information for distributed resource any attribute be satisfied with angle value.Certainly, obtain the mode being satisfied with angle value and be not limited thereto, the method that those skilled in the art can adopt any other suitable completely as required.

In one embodiment, can based on user for the resource be assigned with each attribute be satisfied with angle value, obtain user to the user satisfaction of distributed resource.Such as, user can be calculated to distributing to its evaluation of estimate ES of resource according to following formula (1):

$\mathrm{ES}=\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{w}_i{p}_i---\left(1\right)$

Wherein p _ifor this resource i-th attribute be satisfied with angle value, w _ifor the weight that this attribute is corresponding.

The satisfaction SD of user is generally into following linear relationship with between this evaluation of estimate ES:

SD=c*ES (2)

Wherein c is normal scale-up factor.Under normal conditions, the normal scale-up factor in formula (2) can be set to c=1.But, on the other hand, in order to make, between the satisfaction of different user, there is comparability, also the value of satisfaction SD can be normalized, namely

$c=\frac{1}{\underset{i}{\max }{\mathrm{ES}}_i}---\left(3\right)$

Wherein ES _ifor user is to the evaluation of estimate of i-th resource.

For under extreme case , we can suppose that the satisfaction of user to any cost is SD=100%.As an example, table 3 shows the initial satisfaction of three users be distributed on three seats.

Table 3

Seat is numbered	Initial satisfaction
		1	50%
2	80%
		3	30%

Thus, based on each user preference information and distribute to the attribute information of resource of this user, assess the user satisfaction of each user to the resource of distributing.Certainly, assess the method for user to the satisfaction of resource to be not limited thereto.The method that those skilled in the art can take other suitable completely as required.

In some cases, consider the social relationships between different user, user not only depends on the attribute of this resource itself for the satisfaction of distributed resource, but also may depend on the relative position relation between this user and other users.Social relationships described herein can comprise relatives, friend, colleague and between the superior and the subordinate etc.In resource allocation process, especially when arranging seat, these social relationships may be more important than the preference of user to resource itself in some cases.

Consider this, the attribute information of Gains resources and the preference information of user can comprise the attribute information of the relativeness obtaining related resources in step s 110, relative position or the distance at such as seat obtain, and the relation preference information that can obtain between user, the preference of the relative position between such as user couple and other users or distance etc.

Therefore, when assessing user to the satisfaction of resource of distributing, alternatively, the user satisfaction of above obtained each user to the resource of distributing can be revised based on the relativeness between the relation preference information between user and the resource of distributing.For convenience of description, below above-mentioned user satisfaction is called initial user satisfaction.

For example, when distributing social resources and being arrangement seats, can be reduced to the consideration of social relationships and get close to and become estranged two kinds of situations.Such as the relation preference information of user A, user A tends to get close to user B, so he to the satisfaction at distributed seat except depending on the attribute at seat itself, also depend on the relativeness between his seat and the seat of user B, the satisfaction modified value for user A in different distance situation between user B such as, has been shown in table 4.

Table 4

Seat is numbered	And the distance (rice) between user B	Modified value
			1	0～5	+10%

2	5～10	0%
			3	>10	-10%

In this example, the user's initial satisfaction shown in modified value correction chart 3 of table 4 can be utilized, thus the final satisfaction list of user can be obtained, as shown in table 5.

Table 5

Seat is numbered	Final satisfaction
		1	50%+10%=60%
2	80%+0%=80%
		3	30%+10%=40%

Certainly, to correction just a kind of optional embodiment of initial satisfaction, when Resources allocation does not need the social relationships considering user, can not be carried out this and revise.

After step S120, carry out step S130, based on the user satisfaction of each user to the resource of distributing, assess the comprehensive satisfaction of this Resource Allocation Formula.

According to an embodiment, based on each user assessed in the step s 120 to the user satisfaction of institute's Resources allocation, the comprehensive satisfaction of whole Resource Allocation Formula can be assessed to evaluate the quality of this Resource Allocation Formula.The simplest a kind of integrated evaluating method is averaged to the satisfaction of all users, shown in following formula (4):

$f_0=\frac{1}{n}\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{\mathrm{SD}}_i---\left(4\right)$

Wherein f ₀represent the comprehensive satisfaction of Resource Allocation Formula, SD _irepresent the user satisfaction of i-th user.

Note, user satisfaction SD used herein can be initial user satisfaction assessed in the step s 120 as mentioned above, or alternatively, when needing the relation preference considering user, this SD also can be revised final user's satisfaction described above.

Alternatively, if need to consider that different user has different priority levels when Resources allocation, then by arranging different weights to realize to the satisfaction of different user, can assess comprehensive satisfaction shown in following formula (5) in the case:

$f_i=\frac{1}{n}\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{w}_i{\mathrm{SD}}_i---\left(5\right)$

Wherein f ₁represent the comprehensive satisfaction of Resource Allocation Formula, w _irepresent the weight of i-th user.

Similarly, also can consider to be normalized to facilitate to this weight to compare between different resource allocative decision, not repeat them here.

According to another embodiment, in some cases, in order to embody fairness further, may also need to consider the difference size between different user satisfaction, because the Resource Allocation Formula that the difference typically, between user satisfaction is less more can embody the fairness of Resourse Distribute.In the case, can revise based on the difference between user satisfaction the comprehensive satisfaction f that assesses ₀or f ₁, below, for convenience of description, by above assessed comprehensive satisfaction f ₀or f ₁be called initial synthetic satisfaction.

Usually, the method for calculated difference has a variety of, the method for such as conventional variance and standard deviation etc.In the present embodiment, calculate to simplify thus carry out resource allocation process fast, adopting the method Ru shown in following formula (6) to revise initial synthetic satisfaction:

$f_2=\frac{1}{n}\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{w}_i{\mathrm{SD}}_i+\mathrm{\α}\underset{i}{\min }{\mathrm{SD}}_i---\left(6\right)$

Wherein f ₂represent the comprehensive satisfaction revised, α represents the relative weight of the minimum satisfaction of user, represent SD _iminimum value.This f ₂value larger, represent that the result of Resourse Distribute is more inclined to fairness.

In the above-described assessment of the comprehensive satisfaction to Resource Allocation Formula, only consider the impact of satisfaction on Resource Allocation Formula of user.But, in addition, may also need in some cases to consider other constraint condition, the consumption etc. of the such as energy.Such as, in the office of an opening, if only have a few individual in office and their difference that is dispersed in office is local, so in order to provide necessary illumination and air-conditioning, then need to expend more electricity.If but they are arranged in work together, then can reduce the electric energy of consumption widely, because need the illumination of opening and air-conditioning equipment to greatly reduce.

Consider this point, alternatively, when evaluating Resource Allocation Formula, the restriction to these constraint conditions can be increased, namely revise above-mentioned initial synthetic satisfaction f based on the energy resource consumption total amount corresponding to Resource Allocation Formula ₀or f ₁, or revise the comprehensive satisfaction f revised based on satisfaction difference further ₂.

Certainly, also have a variety of to the restriction of constraint condition, some is rigid, namely not violable, and some can be flexible.Such as, to the total quantitative limitation of energy resource consumption, what generally speaking should not limit is too strict, so it is flexible, the energy namely consumed is more few better.To revise comprehensive satisfaction f ₂for example, following formula (7) can be adopted to revise:

$f_3=\frac{1}{n}\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{w}_i{\mathrm{SD}}_i+\mathrm{\α}\underset{i}{\min {\mathrm{SD}}_i-\mathrm{\βEC}---\left(7\right)}$

Wherein f ₃represent the comprehensive satisfaction revised based on energy ezpenditure, EC is total energy ezpenditure (comprising illumination, air-conditioning equipment etc.), and β is corresponding energy ezpenditure weight.

In above formula (7), subtraction form is adopted to punish energy ezpenditure.In addition, also can select other forms of evaluation function, such as adopt inverse etc.

Similarly, if there is other constraint condition except energy resource consumption when Resourse Distribute, also can to carrying out similar correcting process.

After the comprehensive satisfaction that have evaluated Resource Allocation Formula, in step S140, based on the user satisfaction adjustresources allocative decision of each user, and the Resource Allocation Formula after adjustment is reappraised its comprehensive satisfaction, repeat adjustment process until reach adjustment end condition, to obtain best comprehensive satisfaction.

In order to promote its comprehensive satisfaction by changing existing Resource Allocation Formula, can repeatedly adjustresources allocative decision, and the Resource Allocation Formula after each adjustment is reappraised comprehensive satisfaction, until reach adjustment end condition.The mode changing Resource Allocation Formula has a lot, and in one embodiment, we have employed comparatively simple and practical heuristic randomized optimization process, and its main thought is to search the more excellent scheme of performance by change Resource Allocation Formula random on a large scale.But when selectable Resource Allocation Formula is too much (such as because when the quantity of resource and user is more), the efficiency of random search is lower on a large scale.

Consider this, in order to the scope reducing to search for also promotes the efficiency of search, present inventors have proposed the method utilizing the random search improved to carry out adjustresources allocative decision, describe this Resource Allocation Formula method of adjustment in detail below with reference to Fig. 3.

As shown in Figure 3, this method of adjustment 300 can comprise the following steps:

Step S310, obtains the user satisfaction of each user for the resource of distributing according to Current resource allocative decision; And

Step S320, changes the resource distributing to the user of the predetermined quantity with minimum user satisfaction randomly.

Certainly, depend on embody rule situation, the user satisfaction obtained in step S310 can be initial user satisfaction, also can be the user satisfaction revised.Then, the resource distributing to n the user with minimum user satisfaction is changed randomly.Such as, can sort to all users according to user satisfaction, and the resource of distribution is changed randomly to sequence n user the most rearward.

At this, the selection of digital n is general relevant with the size of resource and userbase, if n choosing is excessive, then hunting zone is very large, is difficult to ensure speed of convergence; And if n choosing is too small, then performance boost is comparatively slow, and is easy to reach local minimum.Such as, in one implementation, can choose n be total number of users 1/1 to four/3rd between integer.Certainly, depend on embody rule situation, those skilled in the art can take other suitable n values completely.

But, lower and that its satisfaction is still lower after repeatedly changing their the Resourse Distribute situation of the satisfaction of certain user may be there is, the Resourse Distribute of other users will be caused like this to can not get all the time changing, and the performance of overall Resource Allocation Formula can not get significantly promoting.

Consider this, the restriction user repeatedly redistributing resource being redistributed again to resource can be increased, such as when the number of times that certain user is reallocated resource is increased to a certain degree, be not reallocated resource by the chance redistributing resource being left for other or distribute number of times less user.An optional implementation of Resource Allocation Formula method of adjustment is in the case described in below with reference to Fig. 4.

As shown in Figure 4, method of adjustment 400 can comprise the following steps:

Step S410, obtains the user satisfaction of each user for the resource of distributing according to Current resource allocative decision;

Step S420, to the controlled number of times of distribution of each its resource of user number counting, adjusts counting as user; And

Step S430, adjusts counting based on each user to the user satisfaction of the resource of current distribution and current user and sorts to all users,

Step S440, changes the resource of the user distributing to the minimum predetermined quantity of sequence randomly.

As above-mentioned, the user satisfaction obtained in step S410 can be initial user satisfaction, also can be the user satisfaction revised.Such as, can arrange counter for each user, to count the number of times that he is reallocated resource, after each its resource of distributing of change, this counter increases by one automatically.Then, adjust counting based on each user to the user satisfaction of the resource of current distribution and current user to sort to all users.

Such as, can sort to all users according to the weighted sum of satisfaction and code reassignment number of times.At this, weight choose outbalance, the weights of code reassignment number of times are excessive, and hunting zone can be caused excessive, rate of convergence reduce; The weights of code reassignment number of times are too small, are easily absorbed in the process of iterating, and rate of convergence is still lower.Therefore, suitable weight can be selected according to embody rule situation, to promote rate of convergence.Such as, in one implementation, the weight of code reassignment number of times can be chosen for one of percentage of maximum satisfaction left and right.

After count all user's sequences based on user satisfaction and adjustment, change the resource distributing to n minimum user of sequence randomly.As mentioned above, depend on embody rule situation, those skilled in the art can take suitable n value completely.

Thus, repeatedly based on the user satisfaction adjustresources allocative decision of each user, namely after carrying out above-mentioned Resource Allocation Formula method of adjustment 300 or method 400 at every turn, the user satisfaction of the Resource Allocation Formula of adjustment is assessed according to above-mentioned step S120, and the comprehensive satisfaction of the Resource Allocation Formula of this adjustment is assessed according to above-mentioned step S130, based on the user satisfaction of evaluation, adjustresources allocative decision assess user satisfaction and comprehensive satisfaction (such as the resource of the user of the predetermined quantity that Random assignment user satisfaction is again minimum) again, until reach predetermined adjustment end condition.

In one embodiment, after each adjustresources allocative decision, judge whether to meet this adjustment end condition.If met, then stop adjustment, if do not met, then continue to change Resource Allocation Formula, stop regularization condition until meet.

This predetermined end condition can pre-set.Such as, this termination regularization condition can include but not limited to following: the adjustment number of times of adjustresources allocative decision reach the first pre-determined number, adjustresources allocative decision regulation time more than the first predetermined amount of time, assess Resource Allocation Formula comprehensive satisfaction is more than or equal to predetermined threshold, in the second predetermined amount of time or in the second predetermined adjustment number of times, comprehensive satisfaction does not improve.Those skilled in the art can pre-set suitable adjustment end condition according to embody rule situation.

Such as, when adjusting end condition and being the restriction to adjustment number of times, adjustment number of times is counted, when reaching the first predetermined adjustment number of times, stop adjustment, choose the maximal value to the comprehensive satisfaction that each Resource Allocation Formula has been assessed, as best comprehensive satisfaction.As an object lesson, such as, after adjustresources allocative decision, the comprehensive satisfaction value of current Resource Allocation Formula can be compared with the comprehensive satisfaction value of initial resource allocation scheme, preserve the greater, and afterwards after each adjustresources allocative decision, by current comprehensive satisfaction compared with the comprehensive satisfaction preserved, preserve the greater, iteration like this until adjustment end condition, using the comprehensive satisfaction finally preserved as optimal satisfaction.

Or the target comprehensive satisfaction can expected according to actual conditions definition is as adjustment end condition.When the comprehensive satisfaction of current Resource Allocation Formula does not reach this target comprehensive satisfaction, carry out Resource Allocation Formula adjustment, until this comprehensive satisfaction reaches or close to this target comprehensive satisfaction, stop adjustment process.In the case, this best comprehensive satisfaction is the maximal value of the comprehensive satisfaction assessed, and namely reaches or close to the comprehensive satisfaction of this target comprehensive satisfaction.Owing to being provided with target comprehensive satisfaction, close to or stop iteration when reaching target comprehensive satisfaction, the whole efficiency of Resourse Distribute can be promoted.Also should judge whether this target comprehensive satisfaction may reach according to actual conditions during Offered target comprehensive satisfaction, if arrange too high, then Adjustable calculation process can be caused long.

Or, in order to avoid this comprehensive satisfaction does not improve within a period of time or in certain adjustment number of times, namely there is not more excellent comprehensive satisfaction, can consider to stop adjustment in the case.

Certainly, the above adjustresources allocative decision that only provides is to obtain some sample implementation of best comprehensive satisfaction, and those skilled in the art can conceive other implementations according to embody rule environment completely.

After the best comprehensive satisfaction of acquisition, in step S150, utilize the Resource Allocation Formula with this best comprehensive satisfaction to described multiple user resource allocation.

Thus, according to the resource allocation methods of the application, the global optimization of Resourse Distribute can be carried out based on the attribute of the preference information of total user and allowable resource, unified allocation of resources resource, can indistinguishably treat each user, thus overcome the unfair distribution of traditional " First come first served " distribution principle and there is the problems such as control unknown risks.

Fig. 5 shows the schematic diagram of an object lesson of this resource allocation methods.As shown in Figure 5, still with to three user A, B, C Resources allocation is example, choose arbitrary Resource Allocation Formula as initial resource allocation scheme, assess user satisfaction and the comprehensive satisfaction of this Resource Allocation Formula, judge whether to reach adjustment end condition, if do not reach this condition, carry out Resource Allocation Formula adjustment, and with the Resource Allocation Formula of adjustment again to three user resource allocations, iteration like this is until reach adjustment end condition, obtain current best comprehensive satisfaction, thus the Resource Allocation Formula with this comprehensive satisfaction can be utilized to user resource allocation.

Below provide an application example simplified of the resource allocation methods of the application.Such as, in an office, there are three idle seats, numbering 1,2,3, as shown in Figure 5.Their attribute list is as shown in table 1.Existing three user A, B and C prepare to enter this office and carry out work, and wherein the preference information of user A is as shown in table 2.

Suppose that initial seat arrangement scheme is:

User A---seat 1

User B---seat 2

User C---seat 3

For this scheme, calculate the user satisfaction of each user.

For user A, the preference information (as shown in table 2) of contrast user A and the attribute (as shown in table 1) at seat 1, can obtain user A satisfaction is:

To position satisfaction: 100%(because seat 1 near window and distance is less than 2 meters)

To illumination satisfaction: 0%(because the illumination at seat 1 exceeds tolerance 400 ~ 600lx)

Then the initial satisfaction of user A is:

SD _a=[(position) 100%+ (illumination) 0%]/2=50%

Satisfaction for the relation between other users :+10%(, because user A and user B distance are 2 meters, is less than 5 meters, tables look-up 5 known, satisfaction modified value+10%)

After considering relation satisfaction, the final satisfaction of user A is:

SD _A=50%+10%=60%

Similarly, the initial satisfaction of user B and user C and final satisfaction can be calculated.Suppose that the final satisfaction through calculating user B is 80%, the final satisfaction of user C is 40%, as shown in Table 3 and Table 4.

So, this seat is arranged to comprehensive satisfaction=(the final satisfaction of the final satisfaction+user C of the final satisfaction+user B of user A)/3=(60%+80%+40% of scheme)/3=60%

Arrange scheme to optimize seat, suppose to choose lower user A and C of satisfaction and carry out seat and redistribute, then new seat arrangement scheme is:

User C---seat 1

User B---seat 2

User A---seat 3

Now, recalculate the satisfaction of each user, still for user A, now user A

Satisfaction to position: 0%(is because seat 3 is away from window)

Satisfaction to illumination: 0%(is because the illumination at seat 3 is lower than tolerance 400 ~ 600lx)

The then initial satisfaction of user A: SD _a=[(position) 0%+ (illumination) 0%]/2=0%

Satisfaction for the relation between other users :+10%(is because the air line distance between user A and user B is less than 5 meters)

So the final satisfaction of user A is: SD _a=0%+10%=10%

Visible, now the satisfaction of user A greatly reduces.After the satisfaction of further other users of assessment, select two minimum users of satisfaction again to carry out seat and redistribute, until comprehensive satisfaction no longer raises in a long time or reaches certain iterations.In fact, as mentioned above, it is also conceivable to when selecting the lower user of satisfaction the number of times that each user is reallocated seat, repeatedly being redistributed to avoid certain user as far as possible.

Certainly, above-mentioned simple numerical example is only schematic, and the complicacy of practical problems may considerably beyond this example, and the object describing this example is only used to illustrate how to apply the present invention.

On the other hand, the invention provides the device to multiple user resource allocation.Fig. 6 shows the schematic block diagram of the device 600 to multiple user resource allocation according to an embodiment.

As shown in Figure 6, this device 600 can comprise with lower component:

Obtaining widget 610, obtains the attribute information of resource to be allocated and the preference information of described multiple user;

User satisfaction evaluation means 620, for arbitrary Resource Allocation Formula, preference information and the attribute information of the resource distributing to this user based on each user assess the user satisfaction of each user to the resource of distributing;

Comprehensive satisfaction evaluation means 630, each user assessed based on user satisfaction evaluation means 620, to the user satisfaction of resource of distributing, assesses the comprehensive satisfaction of this Resource Allocation Formula;

Adjustment component 640, based on each user user satisfaction adjustresources allocative decision and make described user satisfaction evaluation means 620 and described comprehensive satisfaction evaluation means 630 to the Resource Allocation Formula assessment user satisfaction after adjustment and comprehensive satisfaction, repeat adjustment process until reach adjustment end condition, to obtain best comprehensive satisfaction; And

Distribution member 650, utilizes the Resource Allocation Formula with this best comprehensive satisfaction to described multiple user resource allocation.

Thus, the global optimization of Resourse Distribute can be carried out based on the attribute of the preference information of total user and allowable resource, unified allocation of resources resource, can indistinguishably treat each user, thus overcomes the unfair distribution of traditional " First come first served " distribution principle and there is the problems such as control unknown risks.

Alternatively, this adjustment component 640 can change the resource of the user distributing to the predetermined quantity with minimum user satisfaction randomly.

Alternatively, this device 600 can also comprise with lower component (not shown):

Counter block 660, to the controlled number of times of distribution of each its resource of user number counting, adjusts counting as user; And

Ordering element 670, each user based on user satisfaction evaluation means 620 assessment adjusts counting to the current user that the user satisfaction of the resource of current distribution and counter block 660 count and sorts to all users.

This adjustment component 640 can carry out adjustresources allocative decision by the resource changing the user distributing to the minimum predetermined quantity of sequence randomly.

In addition, as mentioned above, according to the present invention, in resource allocation process, except considering user preference, by the difference between consideration user satisfaction and the assessment result to Resource Allocation Formula can also be revised to the constraint of the aspects such as energy resource consumption.

And the present invention can provide full automatic most optimum distribution of resources scheme, can according to the user preference information of preserving in advance and Resource Properties information automatic Prediction user satisfaction, and to Resource Allocation Formula correct, whole process prosthetic is intervened.

The example of the block scheme of the device related in the disclosure, device, equipment, system only illustratively property, and be not intended to require or hint must carry out connecting according to the mode shown in block scheme, arranges, configure.As the skilled person will recognize, can connect by any-mode, arrange, configure these devices, device, equipment, system.Such as " comprise ", " comprising ", " having " etc. word be open vocabulary, refer to " including but not limited to ", and can use with its exchange.Here used vocabulary "or" and " with " refer to vocabulary "and/or", and can to use with its exchange, unless it is not like this that context clearly indicates.Here used vocabulary " such as " refer to phrase " such as, but not limited to ", and can to use with its exchange.

Flow chart of steps in the disclosure and above method only describe the example of illustratively property, and are not intended to require or imply the step must carrying out each embodiment according to the order provided.As the skilled person will recognize, the order of the step in above embodiment can be carried out in any order.Such as the word of " thereafter ", " then ", " next " etc. is not intended to limit the order of step; The description of these words only for guiding reader to read over these methods.In addition, such as use article " ", " one " or " being somebody's turn to do " be not interpreted as this key element to be restricted to odd number for any quoting of the key element of odd number.

The above description of disclosed aspect is provided to make to enable any technician of this area or use the present invention.Be very apparent to those skilled in the art to the various amendments of these aspects, and can be applied in other in General Principle of this definition and do not depart from the scope of the present invention.Therefore, the present invention be not intended to be limited to shown in this in, but according to consistent with principle disclosed herein and novel feature most wide region.

Claims (10)

1., to a method for multiple user resource allocation, comprising:

Obtain the attribute information of resource to be allocated and the preference information of described multiple user;

For arbitrary Resource Allocation Formula, preference information and the attribute information of the resource distributing to this user based on each user assess the user satisfaction of each user to the resource of distributing;

Based on the user satisfaction of each user to the resource of distributing, assess the comprehensive satisfaction of this Resource Allocation Formula;

Based on the user satisfaction adjustresources allocative decision of each user, and the Resource Allocation Formula after adjustment is reappraised its comprehensive satisfaction, repeat adjustment process until reach adjustment end condition, to obtain best comprehensive satisfaction; And

Utilize the Resource Allocation Formula with this best comprehensive satisfaction to described multiple user resource allocation.

2. the method for claim 1, wherein said adjustresources allocative decision comprises the resource changing randomly and distribute to the user of the predetermined quantity with minimum user satisfaction.

3. the method for claim 1, also comprises:

To the controlled number of times of distribution of each its resource of user number counting, adjust counting as user; And

Adjust counting based on each user to the user satisfaction of the resource of current distribution and current user to sort to all users,

Wherein said adjustresources allocative decision comprises the resource changing the user distributing to the minimum predetermined quantity of sequence randomly.

4. the method for claim 1, the wherein said preference information based on each user and the attribute information of the resource distributing to this user are assessed the user satisfaction of each user to the resource of distributing and are comprised:

Preference information based on each user assesses the initial satisfaction of each user to the resource of distributing with the attribute information of the resource distributing to this user; And

Based on this initial satisfaction of relativeness correction between the relation preference information between user and the resource of distributing, to obtain the described user satisfaction of each user to the resource of distributing.

5. the method for claim 1, wherein saidly comprises the comprehensive satisfaction of this Resource Allocation Formula of satisfaction assessment of the resource of distributing based on each user:

Based on the user satisfaction of each user to the resource of distributing, assess the initial synthetic satisfaction of this Resource Allocation Formula; And

Based on each user to the difference between the user satisfaction of the resource of distributing, revise this initial synthetic satisfaction to obtain the described comprehensive satisfaction of described Resource Allocation Formula.

6. method as claimed in claim 5, the minimum value wherein based on each user, this initial synthetic satisfaction of difference correction between the user satisfaction of the resource of distributing being comprised to the user satisfaction of the resource utilizing described user to distribution revises this initial synthetic satisfaction.

7. the method for claim 1, wherein saidly comprises the comprehensive satisfaction of this Resource Allocation Formula of satisfaction assessment of the resource of distributing based on each user:

Based on the user satisfaction of each user to the resource of distributing, assess the initial synthetic satisfaction of this Resource Allocation Formula; And

Based on the energy resource consumption total amount that this Resource Allocation Formula is corresponding, revise this initial synthetic satisfaction to obtain the described comprehensive satisfaction of described Resource Allocation Formula.

8. the method as described in claim 5 or 7, wherein said comprising based on the initial synthetic satisfaction of each user to this Resource Allocation Formula of satisfaction assessment of the user resources of distributing calculates described initial synthetic satisfaction to the user satisfaction weighting of each user to the resource of distributing, and wherein the weighting coefficient of each user depends on the priority of this user in resource allocation process.

9. the method for claim 1, wherein said termination regularization condition comprises following any one: the adjustment number of times adjusting described Resource Allocation Formula reaches the first pre-determined number, adjust the regulation time of described Resource Allocation Formula more than the first predetermined amount of time, assess Resource Allocation Formula comprehensive satisfaction is more than or equal to predetermined threshold, in the second predetermined amount of time or in the second predetermined adjustment number of times, described best comprehensive satisfaction does not improve

Wherein said best comprehensive satisfaction is the maximal value of current obtained comprehensive satisfaction.

10., to a device for multiple user resource allocation, comprising:

Obtaining widget, obtains the attribute information of resource to be allocated and the preference information of described multiple user;

User satisfaction evaluation means, for arbitrary initial resource allocation scheme, preference information and the attribute information of the resource distributing to this user based on each user assess the user satisfaction of each user to the resource of distributing;

Comprehensive satisfaction evaluation means, based on the user satisfaction of each user to the resource of distributing, assesses the comprehensive satisfaction of this Resource Allocation Formula;

Adjustment component, based on each user user satisfaction adjustresources allocative decision and make described user satisfaction evaluation means and described comprehensive satisfaction evaluation means to the Resource Allocation Formula assessment user satisfaction after adjustment and comprehensive satisfaction, repeat adjustment process until reach adjustment end condition, to obtain best comprehensive satisfaction; And

Distribution member, utilizes the Resource Allocation Formula with this best comprehensive satisfaction to described multiple user resource allocation.