KR20210072844A - Method and program for calculating preference ranking using two-alternative forced-choice - Google Patents

Abstract

Provided is a method and a program for calculating preference ranking using alternatives. The method of calculating the preference ranking using the above alternative includes the steps of: providing the alternatives to a pre-determined number of data collection subjects by providing two each of a plurality of preference collection targets which require a computer to collect preference rankings; listing the preference rankings for the plurality of preference collection targets based on a principle of a majority vote by the computer for each of alternative outcomes from the subjects of data collection; and storing, by the computer, a preference order as group preference data for the plurality of preference collection subjects.

Description

Method and program for calculating preference ranking using alternatives {METHOD AND PROGRAM FOR CALCULATING PREFERENCE RANKING USING TWO-ALTERNATIVE FORCED-CHOICE}

The present invention relates to a method and program for calculating a preference ranking using alternatives.

When an individual is asked to rank in order of preferred objects after showing them all at once, the individual cannot easily rank them.

For example, if you show several things within the same or different categories, such as the order of preferred food, preferred color order, preferred animal order, etc., and then list them, it takes a long time and does not lead to an accurate order.

Accordingly, although several methods have been developed as a method of listing the rankings for several objects, a method for accurately and quickly listing the rankings for several objects has not yet been derived.

On the other hand, when companies develop products, it is very important to calculate the preference order of the group of consumers because they want to know what consumers' preferences are from design to function, etc.

Korean Patent Publication No. 10-2013-0083058, 2013.07.22.

SUMMARY OF THE INVENTION An object of the present invention is to provide a method and a program for calculating a preference ranking that can determine the user's own preference.

In addition, the problem to be solved by the present invention is to provide a method and a program for calculating a preference ranking that can determine how much a user's preference has a collectiveness.

In addition, the problem to be solved by the present invention is to provide a method and a program for calculating a preference ranking that can grasp the group's preference.

The problems to be solved by the present invention are not limited to the problems mentioned above, and other problems not mentioned will be clearly understood by those skilled in the art from the following description.

A preference ranking calculation method using alternatives according to an embodiment of the present invention for solving the above-described problems is a method for calculating a predetermined number of data collection targets by providing a computer two each of a plurality of preference collection targets requiring preference ranking collection providing an alternative to, the computer receiving each alternative result from the data collection subjects, the computer based on a majority rule for each of the alternative results from the data collection subjects to list the preference rankings for the plurality of preference collection objects, the computer storing the preference rankings as group preference ranking data for the plurality of preference collection objects, wherein the predetermined number is , is more than the minimum number of data collection subjects in which there is no disturbance selector when listing the preference rankings for the plurality of preference collection targets based on the principle of the majority vote.

The method for calculating a preference ranking using alternatives according to an embodiment of the present invention for solving the above-described problems is the step of providing, by the computer, each of the plurality of preference collection objects two by one, to provide the user with an alternative, the computer Receiving each alternative result from the user, the computer representing the alternative result received from the user as a personal preference matrix, and the computer representing the group preference data as a matrix for a group preference matrix The method further includes orthoprojecting the individual preference matrix to derive a group preference coefficient of the user.

A preference ranking calculation method using alternatives according to an embodiment of the present invention for solving the above-described problems is provided by the computer to provide each of a plurality of preference collection objects for confirmation of preference rankings by two, thereby providing the user with an alternative step, the computer receiving each alternative result from the user; representing, by the computer, the alternative result received from the user in a personal preference matrix; and deriving the user's group preference coefficient by orthogonally projecting the individual preference matrix to the group preference matrix representing the group preference ranking data as a matrix, by the computer, wherein the group preference ranking data is a preference ranking by the computer By providing two each of a plurality of preferred collection objects that need to be collected, providing alternatively to a predetermined number of data collection objects, the computer receiving each alternative result from the data collection objects, and the computer for each of the alternative results from the data collection subjects, based on the principle of majority vote, lists the preference rankings for the plurality of preference collection objects, and the computer sets the preference rankings for the plurality of preference collections It is obtained by storing as group preference ranking data for an object, and the predetermined number is a disturbance that prevents listing in preference order when listing preference rankings for the plurality of preference collection objects based on the principle of the majority vote It is more than the minimum number of data collection subjects for which no selectors exist.

In the method for calculating a preference ranking using alternatives according to an embodiment of the present invention for solving the above-mentioned problems, the computer uses the non-group preference of the user based on the individual preference matrix, the group preference coefficient, and the group preference matrix The method further comprising the step of deriving a coefficient and a non-group preference matrix, wherein the product of the non-group preference coefficient and the non-group preference matrix is a value obtained by subtracting the product of the group preference coefficient and the group preference matrix from the individual preference matrix. It corresponds to an approximate value.

In deriving the non-group preference coefficient and the non-group preference matrix, the non-group preference matrix is derived using only an optimal ranking matrix with respect to the individual preference matrix, and the non-group preference coefficient is the non-group preference matrix. It is derived based on

In the method for calculating a preference ranking using alternatives according to an embodiment of the present invention for solving the above-described problem, the group preference coefficient and the absolute value of the derived group preference coefficient and the non-group preference coefficient are equal to 1. and providing, to the user, an aggregation index and a non-grouping index obtained by correcting the non-group preference coefficient.

In the method of calculating a preference ranking using alternatives according to an embodiment of the present invention for solving the above-described problem, the data collection subjects are divided by a predetermined specific group, the group preference ranking is derived for each group, and the The method further includes the step of providing, by the computer, a group preference order derived for each group with respect to a plurality of objects included in the same specific group.

A preference ranking calculation program using alternatives according to another embodiment of the present invention for solving the above-described problems is combined with a computer that is hardware, and is stored in a medium to execute any one of the methods described above.

Other specific details of the invention are included in the detailed description and drawings.

According to the present invention, the user can grasp his or her preference by using the alternative method, and furthermore, the degree of collectivity and non-collectivity can be grasped through the degree to which the user's preference coincides with the group preference.

In addition, according to the present invention, it is possible to derive a group preference order, so that when a company develops a product, it is possible to grasp what the consumer's preference is.

In addition, in the case of infants, children, or depressed patients, the preference is clear, and the number of confounding selectors is close to 0 in the calculation of preference ranking using alternatives. Therefore, the present invention can be utilized for psychiatric diagnosis or psychological diagnosis using this.

Effects of the present invention are not limited to the effects mentioned above, and other effects not mentioned will be clearly understood by those skilled in the art from the following description.

1 is a diagram for explaining a method of calculating a preference ranking using alternatives according to an embodiment of the present invention.
2 is a diagram for explaining a disturbance selector making a loop in the alternative selection.
3 is a graph showing the number of confounding selectors according to the number of data collection subjects.
4 is a diagram for explaining a method of deriving a user's group preference coefficient.
5 is a diagram for explaining a method of deriving a non-group preference coefficient and a non-group preference matrix.
6 is a view for explaining a method of deriving a collective index and a non-aggregation index.

Advantages and features of the present invention and methods of achieving them will become apparent with reference to the embodiments described below in detail in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but may be implemented in various different forms, and only these embodiments allow the disclosure of the present invention to be complete, and those of ordinary skill in the art to which the present invention pertains. It is provided to fully understand the scope of the present invention to those skilled in the art, and the present invention is only defined by the scope of the claims.

The terminology used herein is for the purpose of describing the embodiments and is not intended to limit the present invention. As used herein, the singular also includes the plural unless specifically stated otherwise in the phrase. As used herein, “comprises” and/or “comprising” does not exclude the presence or addition of one or more other components in addition to the stated components. Like reference numerals refer to like elements throughout, and "and/or" includes each and every combination of one or more of the recited elements. Although "first", "second", etc. are used to describe various elements, these elements are not limited by these terms, of course. These terms are only used to distinguish one component from another. Accordingly, it goes without saying that the first component mentioned below may be the second component within the spirit of the present invention.

Unless otherwise defined, all terms (including technical and scientific terms) used herein may have the meaning commonly understood by those of ordinary skill in the art to which this invention belongs. In addition, terms defined in a commonly used dictionary are not to be interpreted ideally or excessively unless specifically defined explicitly.

As used herein, the term 'computer' includes various devices capable of performing arithmetic processing. For example, computers include desktop PCs, notebooks (Note Books) as well as smart phones, tablet PCs, cellular phones, PCS phones (Personal Communication Service phones), synchronous/asynchronous A mobile terminal of International Mobile Telecommunication-2000 (IMT-2000), a Palm Personal Computer (PC), a Personal Digital Assistant (PDA), and the like may also be applicable. Also, the computer may correspond to a server computer that receives information from a client.

In the present specification, a 'data collection target' refers to a target who collects preference rankings in order to derive a group's preferred ranking for a plurality of specific preference collection targets, and the data collection target belongs to the group.

In the present specification, 'a plurality of preference collection objects' refers to objects requiring preference collection, and the type or number of objects may be predetermined or an administrator who needs data collection may determine as needed. Accordingly, the type or number of objects is not limited, and the types of objects may also include random objects not included in the same category.

In the present specification, 'group preference ranking data' is a ranking preferred by a group with respect to a plurality of preference collection targets obtained from data collection targets. The group preference ranking data is a list of clear rankings for a plurality of preference collection targets according to the principle of majority vote, and corresponds to ranking data in which a disturbing selector does not exist.

In the present specification, the term 'trickster' refers to a specific preference collection object that shifts the ranking when selecting alternatives with respect to the preference collection objects, and a specific preference that forms a loop rather than a line when the preference collection objects are listed in order. preferred collection. The 'disturbance selector' will be described in detail with reference to FIG. 2 .

In the present specification, a 'user' is a person who needs to derive his or her preference tendency to see how much difference his or her personal preference ranking is from a group preference ranking.

In the present specification, the 'personal preference matrix' is a preference matrix of a specific user or a specific data collection target, and indicates alternative results for each preference collection target as a matrix. In the matrix representation method, when each preference collection object is equally arranged in a row and column, a specific user or a specific data collection target selects the result of choosing the preference collection object arranged in the row and column corresponding to the position of each element. It indicates the position of the element.

In the present specification, the 'group preference matrix' represents group preference data as a matrix. In the matrix representation method, when each preferred collection object is equally arranged in a row and column, the selection result is determined based on the group preference data for the preferred collection object arranged in the row and column corresponding to the position of each element. It is indicated by location, or by applying the principle of majority vote based on the individual preference matrix of each of the data collection subjects.

In the present specification, the 'group preference coefficient' indicates how much the user's preference rank follows the group preference ranking when the user's preference ranking is compared with the group preference ranking data. A method of deriving the group preference coefficient will be described later.

In the present specification, the 'non-group preference coefficient' indicates how much the user's preference does not follow the group preference when the user's preference ranking is compared with the group preference ranking data. A method of deriving the non-group preference coefficient will be described later.

In the present specification, the 'non-group preference matrix' is a matrix indicating an individual's preference for non-group preference, except for a portion that follows the group preference matrix from the individual preference matrix. That is, the non-group preference matrix is a matrix indicating preferences excluding collectivity from individual preferences. A method of deriving the non-group preference matrix will also be described later.

In the present specification, the 'collectiveness index' is a processed value of the 'group preference coefficient', and represents a value obtained by processing the group preference coefficient to show the degree of aggregation that the user's preference follows based on a certain numerical value.

In this specification, the 'non-group preference index' is a value obtained by processing the 'non-group preference coefficient', and is a value obtained by processing the non-group preference coefficient to show the degree of non-groupness that the user's preference follows based on a certain numerical value. represents a value.

In this specification, 'optimal rank' means a rank of a plurality of preferred collection targets created by changing a downstream link, and the optimal rank may be one or more, and the 'optimal rank matrix' is based on the 'optimal rank'. means a matrix. Here, the downstream link refers to the individual's alternatives to be changed when finding a preference matrix that minimizes the number of individual alternatives that need to be changed in order to arrange the rankings of a plurality of preference collection objects in a line with respect to the personal preference matrix. It means target.

When the 'optimal ranking' is used, when any alternative preference is changed from the optimum ranking to a minimum, it is possible to know whether the individual preference with the loop type ranking can be changed with the individual preference without the loop type ranking.

If the 'optimal ranking matrix' is described in more detail, according to the description of FIG. 2 to be described later, the minimum downstream link that needs to be changed in FIG. 2 is one, changing C selected from both C and D to D will be.

Therefore, the optimal ranking is also one of the following: C, A, B, and D. In this case, C is ranked as the most preferred and D as the least preferred. For example, when the group preference ranks are in the order of A, B, C, and D, the optimal rank matrix corresponding to the optimal ranks C, A, B, and D of FIG. 2 is as follows.

First, when there are a plurality of preferred collection objects, the present invention is for a specific person to arrange the preferred collection objects in a line in order of preference.

The best way to list multiple favorites in a line in order of preference is to show them all at once and then list them in order, but most people take a long time to make a selection. feel difficult

Accordingly, in the present invention, a plurality of preferred collection objects are alternatively provided two by two to receive selections, and thus are arranged in order of preference.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 is a diagram for explaining a method of calculating a preference ranking using alternatives according to an embodiment of the present invention as deriving group preference ranking data.

Referring to FIG. 1 , the preference ranking calculation method using alternatives according to an embodiment of the present invention provides each of a plurality of preference collection objects to the data collection objects to be alternatively provided (S110), and the alternative result is received (S130). ), listing the preference rankings for a plurality of preference collection targets (S150) and storing the preference rankings as group preference ranking data (S170).

In the step of providing (S110) alternatively to each of a plurality of preferred collection objects to data collection subjects, the computer provides two each of a plurality of preferred collection objects that require preference priority collection to a predetermined number of data collection subjects. It is to provide an alternative.

In this case, the predetermined number is greater than or equal to the minimum number of data collection subjects in which a disturbing selector does not exist when a preference order for a plurality of preference collection objects is listed based on the principle of majority vote.

The reason why the predetermined number is as described above is that the preference order for a plurality of preference collection objects can be arranged in a line only through the selection results of the data collection subjects greater than or equal to the minimum number of data collection subjects in which the disturbance selector does not exist. This is because ranking data can be derived.

Then, the computer receives each alternative result from the data collection subjects (S130), and for each alternative result from the data collection subjects, preference for a plurality of preferred collection objects based on the principle of majority vote List the rankings (S150).

Applying the principle of majority to each alternative result means that two of a plurality of preferred collection objects are provided so that they can be alternatively applied, and the principle of majority is applied to each of the received alternative results.

For example, if the number of data collection subjects is 3, and the preferred collection subjects are red, yellow, and blue, provide each data collection subject to choose between red VS yellow, red VS blue, yellow VS blue, The first data collection subject chooses red for red vs yellow, red for red vs blue, yellow for yellow vs blue, and the second data collection subject selects red for red vs yellow, blue for red vs blue, and blue for yellow vs blue. , and the third data collection subject chooses yellow for red VS yellow, blue for red VS blue, and blue for yellow VS blue, as a rule of majority vote, the one with the highest majority of the three choices for each option Red vs. yellow, red vs. blue, red vs. blue, yellow vs. blue, blue.

However, the above example is only an example of a method of listing a preference order based on the principle of majority vote, and the number of data collection subjects is as described above.

In addition, with respect to listing the preference ranking, as well as applying the principle of majority vote that follows the majority opinion, it is also possible to list the preference ranking according to a predetermined percentage criterion. It may also include listing preferences by super-majority or sub-majority, such as following cases of 60-90% or more, or following minority opinions of 50% or less.

When the preference rankings are listed for a plurality of preference collection objects, the preference rankings listed by the computer are stored as group preference data for the plurality of preference collection objects (S170).

In addition, although not shown, since group preferences may be different for each group, data collection subjects are divided into predetermined specific groups, and group preference rankings can be derived for each group, in which case the computer includes multiple objects included in the same specific group. It may further include the step of providing a group preference order derived for each group with respect to the target of.

If the group preference ranking is derived for each group, the group preference ranking among groups can be compared, or in a later step, whether a specific user selects a specific group among several groups and corresponds to the group preference ranking of the selected specific group can be derived. .

2 is an exemplary diagram for explaining a perturbing selector that creates a loop in the alternative selection.

2 shows alternative results with arrows, for example, when a preferred collection object has A corresponding to yellow green, B corresponding to yellow, C corresponding to purple, and D corresponding to green, among A and B Choose A so that A is ranked ahead of B, among A and C, choose C and rank ahead of A, and among A and D, choose A so that A is ahead of D in rank, B and C Among them, C is selected and C is ranked ahead of B. Among B and D, B is selected and B is ranked ahead of D. Among C and D, D is selected and D is ahead of C in rank.

In this case, the ranks of A, B, C, and D are not arranged in a line, but ranks are formed in the form of a loop.

As an example, even if two preferred collection objects are given so that a plurality of preferred collection objects can be selected alternatively, the ranking of the preferred collection objects cannot be arranged in a line as a result of the alternative in many cases.

The case where the rankings of the preferred collection objects cannot be arranged in a line using the alternative is even more so as the number of preferred collection objects increases.

At this time, the specific selection of the alternative that prevents the ranking of the preferred collection objects from being arranged in a line is called a disturbance selector. However, the disturbance selector cannot designate a specific selection, but can be expressed as a number. This is because, according to which selection is designated as a disturbance selector in the formed loop, the ranking of preferred collection objects arranged in a row may be changed.

In this example, the perturb selector is a result of the selection that led to the formation of a loop. In this example, if C is removed or D is removed, the loop shape is removed, and the remaining preferred collection objects A, B, D or A, B , since the ranks of C can be arranged in a row, the perturbation selector becomes one C or D.

That is, as in the example, what the disturbance selector is may be changed depending on how the preference collection objects are ranked, but the number of the disturbance selectors is derived the same.

3 is a graph showing the number of confounding selectors according to the number of data collection subjects.

Referring to FIG. 3 , the x-axis is the number of data collection subjects, and the y-axis is the number of perturbation selectors. As described above in FIG. 2 , the number of disturbance selectors according to the alternative results is the same in any case regardless of how the preference collection objects are ranked.

In FIG. 3 , the number of disturbance selectors becomes 0 to indicate the number of data collection subjects capable of arranging the ranks of preferred collection objects in a line.

Looking at the group preference rankings of FIG. 3 , various types of group preference rankings when the data collection subjects were 119 people are listed. Preference ranking for 8 pictures, preference ranking for 8 words, preference ranking for 8 colors, preference ranking for 8 female faces, preference ranking for 8 male faces, preference ranking for 9 faces Ranks are listed, and ranks are listed in a line, not in a loop.

Looking at the graph again, in the case of words and pictures, the number of confounding selectors becomes 0 when the number of data collection subjects is about 30 or more, but in other fields, the number of confounding selectors starts when the number of data collection subjects is about 115 or more. It can be seen that 0

Through the results of the graph, it can be confirmed that the number of confounding selectors can vary depending on which field the preferred collection targets are, and it can be seen that the random 8 and 9 types that are not limited to any field at the top of the graph are The selection result confirms that no matter how large the number of data collection subjects is, the number of confounding selectors does not decrease.

Accordingly, the present invention limits the preferred collection targets to a specific field in the same field, and since the number of data collection subjects for which the number of confounding selectors can be 0 may vary in each field, the number of data collection subjects according to each field can be applied otherwise.

4 is a diagram for explaining a method of deriving a user's group preference coefficient.

Referring to FIG. 4 , the method of deriving the user's group preference coefficient is provided to the user as an alternative to each of the plurality of preference collection objects in FIG. 1 ( S190 ), and receiving the alternative result from the user ( S210 ), received The method further includes a step of representing the selection result in a personal preference matrix (S230) and a step of deriving a user's group preference coefficient (S250).

In providing each of the plurality of preferred collection objects to the user alternatively ( S190 ), the computer provides each of the plurality of preferred collection objects by two, thereby providing the user with an alternative.

Then, the computer receives the alternative result from the user (S210), and represents the alternative result received from the user as a personal preference matrix (S230).

In the method of representing the alternative results received from the user as a personal preference matrix, when each preference collection object is equally arranged in a row and column, the preference collection object arranged in the row and column corresponding to the position of each element is selected by a specific user or a specific user. The result of the data collection subject is displayed at the position of each element.

In more detail, for example, if there are two preferred collection objects, a 2X2 matrix is created, and if there are three preferred collection objects, a 3X3 matrix is created. Each element in the matrix represents the result of a comparison between favorite ingests, where each favorite is listed as a comparison in rows and columns, and each favorite in the same order as that in which they are listed as a comparison in a row. The order of the comparison objects on the column of the object is the same as the order of the comparison object on the row.

As a specific example, _{if there are three preferred collection objects, a 1} , a ₂ , and a 3 , a 3X3 matrix is created as follows.

In row 1, column ₁ , it is a _{choice between a 1 and a 1} , and it is expressed as 0 as a comparison with itself, and the same is true for rows 2, 2, and 3, and columns. _{In row 1} and column 2, a 1 arranged in row _{1 is more preferred than a 2} arranged in column 2, and it is expressed as 1. In row 1 and _{column 3,} a 3 arranged in column 3 rather than _{a 1 arranged in row 1} It is expressed as -1, meaning that it is more preferred.

What kind of number to be expressed in the matrix may be variously determined, and only the indication of preference should be clearly displayed in the matrix, and the expression is not necessarily limited to the above example.

The step of deriving the user's group preference coefficient ( S250 ) is to derive the user's group preference coefficient by orthogonally projecting the personal preference matrix to the group preference matrix representing the group preference ranking data as a matrix by the computer.

The group preference matrix in which the group preference data is expressed as a matrix is a group preference matrix by applying the principle of majority vote to each selection on the matrix after each selection result of the data collection subjects is expressed as a personal preference matrix of the form as above. Alternatively, the group preference matrix may be generated by applying a ranking derived as a result of the group preference ranking data to represent it like an individual preference matrix. Either way, the meaning is the same, but the precedence and precedence of the intermediate process may be different.

For example, in the case of generating a group preference matrix by applying the principle of majority vote to each selection on the matrix after representing the selection results of the data collection subjects as individual preference matrices in the form above, first, each data collection After summing all the individual preference matrices of the subjects, a group preference matrix can be generated as follows. The expression method for the matrix is expressed as 1, 0, -1 as in the example of the above description.

이라 하면, 행렬 내의 3은 1이 2번 초과한 것이고, -3 또한 -1이 2번 초과한 것으로서, 집단 선호 행렬은 각각 1과 -1로 나타낸

가 된다.The sum of the personal preference matrices of the data collection subjects is

, 3 in the matrix means 1 is exceeded twice, -3 and -1 are exceeded twice, and the group preference matrix is represented by 1 and -1, respectively.

becomes

The user's group preference coefficient is derived by orthographically projecting the individual preference matrix with respect to the group preference matrix.

The user's group preference coefficient indicates how much the orthographic value of the personal preference matrix affects the group preference matrix, that is, the user's group preference coefficient is the group preference by orthographically projecting the personal preference matrix to the vector of the group preference matrix. It indicates how much the vector of the projected personal preference matrix compared to the vector of the matrix.

Therefore, when the user's group preference coefficient is α, the individual preference matrix is M _o , and the group preference matrix is M _c , the user's group preference coefficient is expressed by Equation 1 below.

<Equation 1>

5 is a diagram for explaining a method of deriving a non-group preference coefficient and a non-group preference matrix.

Referring to FIG. 5 , the method for calculating a preference ranking using alternatives according to an embodiment of the present invention further includes the step of deriving a non-group preference coefficient and a non-group preference matrix ( S270 ).

In the step of deriving the non-group preference coefficient and the non-group preference matrix ( S270 ), the computer derives the non-group preference coefficient and the non-group preference matrix of the user based on the individual preference matrix, the group preference coefficient, and the group preference matrix. .

In this case, the product of the non-group preference coefficient and the non-group preference matrix corresponds to a matrix that is maximally similar to a matrix obtained by subtracting the product of the group preference coefficient and the group preference matrix from the individual preference matrix.

The most similar matrix is βMp in Equation 2 below.

Here, the non-group preference matrix becomes one of the optimal ranking matrices defined above, and the non-group preference coefficient is determined from the non-group preference matrix, the group preference coefficient, and the group preference matrix. Equation 2 below clearly defines the corresponding content.

That is, when the non-group preference coefficient is β and the non-group preference matrix is Mp, the non-group preference coefficient and the non-group preference matrix satisfy Equation 2 below.

<Equation 2>

That is, looking at Equation 2 above, as a result of subtracting the group preference matrix multiplied by the group preference coefficient indicating the degree of aggregation included in the group preference matrix from the individual preference matrix, it is possible to derive a non-collective individual preference that does not follow the group characteristic at all. And, it is also derived in the form of a non-group preference matrix multiplied by a constant non-group preference coefficient.

Hereinafter, a process of calculating and deriving the non-group preference coefficient and the non-group preference matrix will be described with equations.

을

라고 하고

라고 한다. 또한 선호 수집 대상의 개수가 n이라고 할 때,

을

의 최적 순위 행렬의 집합이라고 할 때,in Equation 2

of

say

It is said Also, when the number of preferred collection targets is n,

of

Assuming that the set of optimal ranking matrices of

here

이다.

to be.

으로 구한다.At this time, the non-group preference coefficient and the non-group preference matrix are

save with

According to the calculation,

에 따라, 결과적으로,

According to, consequently,

이 도출되고,

is derived,

에 따라 비집단 선호 계수

와 비집단 선호 행렬

을 구한다.

according to the subgroup preference coefficient

and the non-group preference matrix

save

는

의 원소이면서

을 최대로 만드는 행렬이고 이런

을 구하면

가 결정된다.here

is

being an element of

is a matrix that maximizes

If you get

is decided

Accordingly, β, which is a non-group preference coefficient satisfying Equation 2, and M _p , a non-group preference matrix, are derived.

At this time, if the number of preferred collection targets is 10, _{the number of Ω tran} is 3,628,800 which is 10!, so it may take a lot of time to calculate.

Therefore, in the present invention, _{a method of reducing the number of Ω tran} and calculating the same result while taking less time was derived.

In deriving the non-group preference coefficient and the non-group preference matrix, the non-group preference matrix is derived using only the optimal ranking preference matrix with respect to the individual preference matrix, and the non-group preference coefficient is derived based on the non-group preference matrix will do

As described above, if the non-group preference coefficient and the non-group preference matrix are derived using only the personal preference matrix when the number of disturbance selectors is less than or equal to the predetermined number of disturbance selectors, the calculation time is reduced by reducing the number of individual preference matrices used for calculation. While reducing, it is possible to derive the same result as calculated based on the entire personal preference matrix.

6 is a view for explaining a method of deriving a collective index and a non-aggregation index.

Referring to FIG. 6 , the method for calculating a preference ranking using alternatives according to an embodiment of the present invention further includes a step ( S290 ) of providing a collective index and a non-collective index.

The step of providing the aggregation index and the non-group preference index ( S290 ) is a collectivity index in which the group preference coefficient and the non-group preference coefficient are corrected so that the sum of the absolute values of the derived group preference coefficient and the non-group preference coefficient becomes 1. and to provide an aggregation index to the user.

The method of correcting the group preference coefficient and the non-group preference coefficient so that the sum of the absolute values of the group preference coefficient and the non-group preference coefficient is 1 is as follows, when the aggregation index is a and the non-group preference coefficient is b. according to Equation 3.

<Equation 3>

,

,

According to Equation 3, the sum of the absolute value a and the absolute value b becomes 1, so that the user can grasp the degree of aggregation and non-collectivity at a glance as an objective index.

However, even if the degree of collectivity and non-collectiveness is not corrected by the collective and non-collective indicators, it is possible to grasp the magnitude of the group preference coefficient and the non-group preference coefficient.

In addition, although not shown, the preference ranking calculation method using alternatives according to another embodiment of the present invention, except for the step of deriving the above-described group preference data, provides each of a plurality of preference collection objects to the user as an alternative Only the step of providing (S190), the step of receiving the selection for the alternative from the user (S210), the step of representing the received selection result as a personal preference matrix (S230), and the step of deriving the user's group preference coefficient (S250) only can be configured.

At this time, the group preference data is provided to alternatively to a predetermined number of data collection subjects by providing two of each of a plurality of preference collection objects that require the computer to collect preference rankings, and the computer collects the data from the data collection subjects. receiving each alternative result, the computer lists the preference rankings for a plurality of preferred collection objects based on the principle of majority vote for each alternative result from the data collection subjects, and the computer sets the preference rankings It is obtained by storing as group preference ranking data for a plurality of preference collection objects, and a predetermined number is at least a minimum in which a disturbing selector does not exist when listing preference rankings for a plurality of preference collection objects based on the principle of majority vote. more than the number of subjects for which data was collected.

In addition, the method may further include the step of deriving the non-group preference coefficient and the non-group preference matrix ( S270 ), and may further include the step of providing the aggregation index and the non-group preference index ( S290 ).

The contents corresponding to this embodiment are applied in the same manner as those described above with reference to FIGS. 1 to 6 .

The method according to an embodiment of the present invention described above, the preference ranking calculation method using the alternative, is implemented as a preference ranking calculation computer program (or application) using the alternative in order to be executed in combination with a computer that is hardware. can be stored in

The steps of a method or algorithm described in relation to an embodiment of the present invention may be implemented directly in hardware, as a software module executed by hardware, or by a combination thereof. A software module may contain random access memory (RAM), read only memory (ROM), erasable programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), flash memory, hard disk, removable disk, CD-ROM, or It may reside in any type of computer-readable recording medium well known in the art to which the present invention pertains.

As mentioned above, although embodiments of the present invention have been described with reference to the accompanying drawings, those skilled in the art to which the present invention pertains can realize that the present invention can be embodied in other specific forms without changing its technical spirit or essential features. you will be able to understand Therefore, it should be understood that the embodiments described above are illustrative in all respects and not restrictive.

Claims (8)

providing, by the computer, two of each of a plurality of preference collection objects requiring preference priority collection to alternatively provide a predetermined number of data collection objects;
receiving, by the computer, each alternative result from the data collection subjects;
listing, by the computer, a preference order for the plurality of preferred collection objects, based on a majority rule, for each of the alternative results from the data collection objects;
and storing, by the computer, the preference ranking as group preference data for the plurality of preference collection objects,
The predetermined number is
Based on the principle of the majority vote, when listing the preference order for the plurality of preference collection objects, there is no disturbance selector or more than the minimum number of data collection subjects,
A method of calculating a preference ranking using alternatives. According to claim 1,
providing, by the computer, two of each of the plurality of preferred collection objects to the user to alternatively provide;
the computer receiving each alternative result from the user;
representing, by the computer, the alternative result received from the user in a personal preference matrix; and
The computer orthogonal projection of the individual preference matrix to the group preference matrix representing the group preference ranking data as a matrix, further comprising the step of deriving the group preference coefficient of the user,
A method of calculating a preference ranking using alternatives. providing, by the computer, two of each of a plurality of preference collection objects for confirmation of preference ranking, to the user to choose between;
the computer receiving each alternative result from the user;
representing, by the computer, the alternative result received from the user in a personal preference matrix; and
and deriving the user's group preference coefficient by orthogonally projecting the personal preference matrix to the group preference matrix representing the group preference ranking data as a matrix, by the computer;
The group preference ranking data is,
By providing two of each of a plurality of preferred collection objects that require the computer to collect preference rankings, a predetermined number of data collection objects are provided to alternatively;
the computer receives each alternative result from the data collection subjects;
the computer lists the preference rankings for the plurality of preferred collection objects, based on the principle of majority vote, for each of the alternative results from the data collection objects;
is obtained by the computer storing the preference ranking as group preference data for the plurality of preference collection objects,
The predetermined number is
Based on the principle of the majority vote, when listing the preference order for the plurality of preferred collection objects, there is no disturbance selector or more than the minimum number of data collection subjects,
A method of calculating a preference ranking using alternatives. 4. The method of claim 2 or 3,
The method further comprising the step of the computer deriving a non-group preference coefficient and an a-group preference matrix of the user based on the individual preference matrix, the group preference coefficient, and the group preference matrix,
The product of the non-group preference coefficient and the non-group preference matrix is,
That corresponds to a matrix that is maximally similar to a matrix obtained by subtracting the product of the group preference coefficient and the group preference matrix from the individual preference matrix,
A method of calculating a preference ranking using alternatives. 5. The method of claim 4,
In deriving the non-group preference coefficient and the non-group preference matrix,
The non-group preference matrix is
It is derived using only the optimal ranking matrix for the personal preference matrix,
The non-group preference coefficient is derived based on the non-group preference matrix, the group preference matrix, and the group preference coefficient,
A method of calculating a preference ranking using alternatives. 6. The method of claim 5,
Providing the user with a collectiveness index and a non-group preference index obtained by correcting the group preference coefficient and the non-group preference coefficient so that the sum of the derived absolute values of the group preference coefficient and the non-group preference coefficient becomes 1. containing,
A method of calculating a preference ranking using alternatives. According to claim 1,
The data collection subjects are divided by a predetermined specific group, and the group preference order is derived for each group,
Further comprising the step of providing, by the computer, a group preference order derived for each group with respect to a plurality of objects included in the same specific group,
A method of calculating a preference ranking using alternatives. Combined with a computer that is hardware, claim 1 to claim 3, claim 5 to claim 7 stored in the medium to execute any one of the method, using alternatives, preference ranking calculation program.

Images