JP6808599B2 - Classification support device, classification support method and program - Google Patents

Description

The present invention relates to a classification support device, a classification support method and a program.

For example, when conducting a workshop, the participants may be grouped by referring to the attributes of the participants for the subsequent group work. In such cases, as conditions for grouping, for example, "I want to have the same number of genders in all groups as much as possible", "I want to gather participants of the same affiliation in the same group as much as possible", "Participation in the same group of a wide range of ages" I want to include people. " As described above, grouping based on a plurality of conditions has many considerations and is complicated, so that proper grouping takes a long time.

Conventionally, as a method of classifying users into two or more groups based on user attributes, a method of quantifying the contents of user attribute information and forming a multidimensional vector to take a clustering method has been proposed. In this method, by performing clustering, it is possible to divide similar users into the same group after adding attribute information about a plurality of attributes.

Japanese Unexamined Patent Publication No. 2011-128677

However, in the above method, although it is possible to add multiple attribute information and collect similar users in the same group, for example, based on a condition such as "only people with similar attributes cannot be included in the same group". Cannot be grouped together. In addition, grouping based on conditions such as "include at least one woman in the group" is not possible. Furthermore, it is not possible to make adjustments such as creating a group of a specified number of people.

The present invention has been made in view of the above points, and an object of the present invention is to support grouping to which a plurality of conditions are applied.

Therefore, in order to solve the above problem, the classification support device is
First information indicating the respective values of the plurality of classification targets for each of the plurality of items constituting the attributes relating to the classification targets to the plurality of groups, and
Second information showing a plurality of classification results when the plurality of classification targets are classified into the plurality of groups, and
A third piece of information, which is set for one or more of the above items and indicates two or more conditions for classification into groups, and a priority for the degree to which each of the above conditions is satisfied.
For each of the classification results, a first calculation unit that calculates a first index indicating the degree to which each condition is met, and
Each of the classification results has a second calculation unit for calculating a second index indicating the degree to which the two or more conditions are met, based on the first index calculated for the classification result.
The above two or more conditions are
The first condition regarding the bias of values between groups for the item subject to the condition,
The second condition regarding the number of values included in each group for the item subject to the condition,
The third condition regarding whether or not the number of types of values included in each group for the item subject to the condition is equal to or greater than the specified value,
Fourth condition regarding whether or not each group includes all possible values for the item subject to the condition,
Fifth condition regarding whether each group contains one or more values specified for the item subject to the condition,
The sixth condition regarding the number of groups including each value for the item subject to the condition,
Any two or more of these conditions.

In addition, the classification support device is
First information indicating the respective values of the plurality of classification targets for each of the plurality of items constituting the attributes relating to the classification targets to the plurality of groups, and
When the plurality of classification targets are classified into the plurality of groups, the fourth information indicating a plurality of classification results including a predetermined group and
Fifth information including a plurality of combinations of two or more conditions related to group classification set for one or more of the above items, and
See,
For each of the combinations, a first calculation unit that calculates a first index indicating the degree to which each of the above conditions is met for each of the classification results,
For each of the combinations, a second calculation unit that calculates a second index indicating the degree to which the two or more conditions are met, based on the first index calculated for the classification result. When,
Among the combinations, the second index indicating that the degree of satisfying the two or more conditions is the highest is provided with an output unit that outputs information based on the calculated combination.
The above two or more conditions are
The first condition regarding the bias of values between groups for the item subject to the condition,
The second condition regarding the number of values included in each group for the item subject to the condition,
The third condition regarding whether or not the number of types of values included in each group for the item subject to the condition is equal to or greater than the specified value,
Fourth condition regarding whether or not each group includes all possible values for the item subject to the condition,
Fifth condition regarding whether each group contains one or more values specified for the item subject to the condition,
The sixth condition regarding the number of groups including each value for the item subject to the condition,
Any two or more of these conditions.

It can support grouping where multiple conditions apply.

It is a figure which shows the hardware configuration example of the classification support apparatus 10 in 1st Embodiment. It is a figure which shows the functional structure example of the classification support apparatus 10 in 1st Embodiment. It is a flowchart for demonstrating an example of the processing procedure executed by the classification support apparatus 10 in 1st Embodiment. It is a figure which shows the structure example of the attribute item information. It is a figure which shows an example of the display component for inputting a group requirement. It is a figure which shows the structure example of the participant information. It is a figure which shows the structural example of the condition information. It is a figure which shows an example of the classification result of a participant. It is a figure for demonstrating an example of the calculation method of a suitable group point. It is a figure which shows an example of the calculation result of the appropriate group point for each classification result. It is a figure which shows the functional structure example of the classification support apparatus 10 in the 2nd Embodiment. It is a flowchart for demonstrating an example of the processing procedure executed by the classification support apparatus 10 in the 2nd Embodiment.

[First Embodiment]
Hereinafter, embodiments of the present invention will be described with reference to the drawings. In this embodiment, an example of classifying a plurality of participants in a workshop into a plurality of groups will be described. That is, in the present embodiment, the target of classification into groups is a plurality of participants.

FIG. 1 is a diagram showing a hardware configuration example of the classification support device 10 according to the first embodiment. The classification support device 10 of FIG. 1 includes a drive device 100, an auxiliary storage device 102, a memory device 103, a CPU 104, an interface device 105, a display device 106, an input device 107, and the like, which are connected to each other by a bus B, respectively.

The program that realizes the processing in the classification support device 10 is provided by a recording medium 101 such as a CD-ROM. When the recording medium 101 storing the program is set in the drive device 100, the program is installed in the auxiliary storage device 102 from the recording medium 101 via the drive device 100. However, the program does not necessarily have to be installed from the recording medium 101, and may be downloaded from another computer via the network. The auxiliary storage device 102 stores the installed program and also stores necessary files, data, and the like.

The memory device 103 reads and stores the program from the auxiliary storage device 102 when the program is instructed to start. The CPU 104 realizes the function related to the classification support device 10 according to the program stored in the memory device 103. The interface device 105 is used as an interface for connecting to a network. The display device 106 displays a programmatic GUI (Graphical User Interface) or the like. The input device 107 is composed of a keyboard, a mouse, and the like, and is used for inputting various operation instructions.

FIG. 2 is a diagram showing a functional configuration example of the classification support device 10 according to the first embodiment. In FIG. 2, the classification support device 10 includes an attribute item input unit 11, a group requirement input unit 12, a participant information input unit 13, a condition information input unit 14, a start instruction input unit 15, a classification unit 16, and an appropriate group point calculation unit. It has 17 and a result output unit 18 and the like. Each of these parts is realized by a process of causing the CPU 104 to execute one or more programs installed in the classification support device 10.

FIG. 3 is a flowchart for explaining an example of the processing procedure executed by the classification support device 10 in the first embodiment.

In step S110, the attribute item input unit 11 inputs information related to each item constituting the attribute information of the participant in the workshop (hereinafter referred to as "attribute item information"), and the input attribute item information is input to the participant. Output to the information input unit 13 and the condition information input unit 14.

FIG. 4 is a diagram showing a configuration example of attribute item information. As shown in FIG. 4, the attribute item information includes an "item name" and a "type" for each item constituting the participant's attribute information. The "item name" is the name of the item managed as the item of the participant's attribute. "Type" is the type of item value. In the present embodiment, "category" or "continuous" is set as the value of "type". "Continuous" is a continuous numerical value (continuous value) such as age. "Category" is a name for a value other than a continuous value, or a concept classified by category such as gender ("male" or "female"). However, a type other than "continuous" and "category" may be specified as "type". Further, with respect to the attribute item information, information indicating an exclusive relationship between the categories may be given to the same participant, such that the category A and the category B are contradictory to each other. Items that are not used for grouping may not be included in the attribute item information.

In the present embodiment, the "attribute information" refers to information indicating the entire attributes of the participants, which is represented by a set of values of the items shown in FIG. The value of each item indicates the characteristics of each participant regarding the item.

The attribute item information may be manually input by the user, or may be input by the attribute item input unit 11 acquiring from a predetermined database. The predetermined database may be a database in which the attribute item information is registered in advance, or a database or the like generated by the attribute item input unit 11 registering the attribute item information input for the first time. That is, if the attribute item information input once can be reused in, for example, the next workshop, step S101 is omitted in the grouping related to the next workshop, and the database is concerned. The attribute item information registered in may be used.

Subsequently, the group requirement input unit 12 inputs the grouping requirement (hereinafter, referred to as “group requirement”), and outputs the input group requirement to the classification unit 16 (S120). The group requirement is, for example, information indicating the number of groups to be created or the number of members per group. For example, the group requirement input unit 12 may accept the input of the group requirement from the user by displaying the screen including the display component as shown in FIG. 5 on the display device 106.

FIG. 5 is a diagram showing an example of display components for inputting group requirements. FIG. 5 includes two columns, one for accepting the group requirement by specifying the number of groups and the other for accepting the specification of the number of members per group, and the column with the check button selected (if in FIG. 5). An example of a display component in which the “Specify by the number of people in the group” column) with a check mark is valid is shown.

However, a display component including only one of the columns may be displayed. In addition, group requirements may be set in advance and stored in the auxiliary storage device 102 or the like. In this case, the group requirement input unit 12 may read the group requirement from the auxiliary storage device 102 or the like.

Subsequently, the participant information input unit 13 inputs a list of attribute information of each participant in the workshop (hereinafter referred to as “participant information”), and the input participant information is input to the classification unit 16 and an appropriate group. Output to the point calculation unit 17 (S130). That is, the participant information input unit 13 inputs the value of each item included in the attribute item information input from the attribute item input unit 11 for each participant.

FIG. 6 is a diagram showing a configuration example of participant information. As shown in FIG. 6, the participant information includes the value of each item included in the attribute item information and the “ID” for each participant. The "ID" is identification information of each participant.

The participant information may also be manually input by the user at the time of step S130, or may be acquired by the participant information input unit 13 from a database or the like registered in advance. Step S130 may be executed in association with step S110. That is, step S130 may be executed before step S120. In this case, the attribute item input unit 11 may automatically extract the attribute item information from the database, and the participant information input unit 13 may register the participant information based on the attribute item information.

In addition, when the value of a certain item of a certain participant is an invalid value (for example, a blank or a value outside the range), the participant information input unit 13 is, for example, each participant regarding the item. The value of the item of the participant may be complemented by the average value of the values of, or the participant may be excluded from the grouping.

Subsequently, the condition information input unit 14 can be set when classifying (grouping) a plurality of participants into a plurality of groups for each item included in the attribute item information input from the attribute item input unit 11. Information including any of the conditions of the above and the priority for the condition (hereinafter referred to as "condition information") is input, and the input condition information is output to the appropriate group point calculation unit 17 (S140). ).

FIG. 7 is a diagram showing a configuration example of condition information. As shown in FIG. 7, the condition information includes "condition" and "priority" in addition to the "item name" and "type" included in the attribute information for each item included in the attribute item information. In FIG. 7, the parentheses number of any of the following conditions (1) to (10) is specified in the "condition". That is, in the present embodiment, the candidate conditions that can be set for each item are the following 10 conditions (1) to (10).
Condition (1): Not considered as a condition for grouping.
Condition (2): There is no bias in the value of the target item between groups (however, it can be set only for the item whose "type" is "category").
Condition (3): Minimize the bias of values between groups for the target item (however, it can be set only for items whose "type" is "category").
Condition (4): Each group contains more types of values for the target item (however, it can be set only for items whose "type" is "category").
Condition (5): The number of types of values included in each group for the target item is equal to or greater than the specified number (however, it can be set only for items whose "type" is "category").
Condition (6): Each group includes all types of values for the target item (however, it can be set only for items whose "type" is "category"). Here, all kinds of values mean all kinds of values that all participants have for the item.
Condition (7): Each group includes one or more participants having the specified value for the target item (however, it can be set only for the item whose "type" is "category").
Condition (8): Participants who share the same value for the target item are gathered in the same group as much as possible. Participants who have become fractions collect the fractions to form a group (however, it can be set only for items whose "type" is "category").
Condition (9): Minimize the bias of values between groups for the target item (however, it can be set only for items whose "type" is "continuous").
Condition (10): Collect participants having similar values for the target item in the same group as much as possible (however, it can be set only for the item whose "type" is "continuous").

In the above, the condition (2), the condition (3), the condition (9), and the condition (10) are examples of "the first condition regarding the bias of the value among the groups for the item subject to the condition". The condition (4) is an example of the second condition regarding the number of values included in each group for the item subject to the condition. The condition (5) is an example of a third condition relating to whether or not the number of types of values included in each group for the item subject to the condition is equal to or greater than the specified value. Condition (6) is an example of a fourth condition relating to whether or not all the values that each group can take for the item subject to the condition are included. The condition (7) is an example of a fifth condition relating to whether or not each group includes one or more values specified for the item subject to the condition. Condition (8) is an example of the sixth condition regarding the number of groups including each value for the item subject to the condition.

The "priority" is a priority regarding the degree to which the conditions set for each item are satisfied. That is, the "priority" indicates that the condition with the higher priority is more likely to be satisfied, and does not necessarily mean that the condition with the higher priority is completely satisfied. .. Further, although FIG. 7 shows an example in which the “priority” is unique, the same priority may be given to a plurality of conditions. Moreover, the "priority" does not have to be set.

Further, although FIG. 7 shows an example in which one condition is set for one item, a combination of a plurality of conditions may be set for one item. For example, a condition such as "satisfying condition (2) and condition (7) at the same time" may be set. In this case, the priority may be set for each condition (condition (2) and condition (7)) included in the combination. The priority has the meaning that, for example, the condition (7) is absolutely satisfied and then the condition (2) is taken into consideration. Alternatively, the weighting may be set such that the condition (7) is 70% and the condition (2) is 30%. In this case, the match evaluation value described later may be calculated by, for example, a weighted sum in consideration of weighting. When setting a combination of a plurality of conditions for one item, the following restrictions may be considered.
-Condition (1) cannot be combined with conditions (2) to (10).
-Condition (2) cannot be combined with condition (3) condition (8).
-Condition (4) cannot be combined with condition (8).
-Condition (5) cannot be combined with condition (6).
-Condition (6) cannot be combined with condition (7).
-Condition (9) cannot be combined with condition (10).

The above conditions (1) to (10) are examples. Therefore, any of the conditions (1) to (10) may not be set, or conditions other than the conditions (1) to (10) may be set. Regardless of which condition is set, it is desirable that two or more conditions can be set, but one or more conditions may be set.

Further, when the information indicating the exclusive relationship of the category is given to the attribute item information, the condition information input unit 14 warns that an impossible condition is not set at the time of step S140, or warns that the impossible condition is not set. The information may be used to hide impossible conditions.

In step S140, among the condition information, information that cannot be acquired from the attribute item information (that is, "condition" and "priority") may be manually input by the user, or "condition" for each item. The condition information input unit 14 may read from a database or the like in which "priority" and "priority" are set.

Further, steps S110 and S140 may be executed first, and steps S120 and S130 may be executed in no particular order.

Subsequently, when the start instruction input unit 15 receives the input of the grouping start instruction from the user via the input device 107, the start instruction input unit 15 outputs the grouping start instruction to the classification unit 16 (S150). At this time, the start instruction input unit 15 may check whether all necessary information has been input, check for missing values, and the like. When there is a missing value, the start instruction input unit 15 may output, for example, an error message or the like to the display device 106 so that step S160 and subsequent steps are not executed until complete data input is performed.

Subsequently, the classification unit 16 starts processing in response to the input of the grouping start instruction from the start instruction input unit 15. Specifically, the classification unit 16 sets each participant related to the participant information (FIG. 6) input from the participant information input unit 13 so as to satisfy the group requirement input from the group requirement input unit 12. The classification result is output to the appropriate group point calculation unit 17 (S160). At this time, the classification unit 16 generates a plurality of classification results. For example, the classification unit 16 may generate all kinds of classification results satisfying the group requirements, or may refer to the condition information (FIG. 7) and prevent the generation of impossible classification results. In this case, the condition information input unit 14 may output the condition information (FIG. 7) to the classification unit 16. Alternatively, the classification unit 16 may repeatedly calculate the appropriate group points described later and classify the participant information by using a genetic algorithm or the like. Alternatively, the classification unit 16 may calculate a random classification result. In this case, different classification results will be obtained each time the start instruction is received in step S150. Alternatively, the classification unit 16 may calculate a plurality of classification results before the step S160 is executed, and store the classification results in the auxiliary storage device 102 or the like. In this case, the classification unit 16 may acquire the classification result from the auxiliary storage device 102 or the like in step S160.

FIG. 8 is a diagram showing an example of the classification results of the participants. In FIG. 8, the group name of the classification destination is shown for each participant for each of the classification results of the plurality of ways (K ways). Each of the K-style classification results shows the results grouped to at least meet the group requirements. The classification unit 16 outputs the classification result as shown in FIG. 8 to the appropriate group point calculation unit 17.

In the following steps S170 to S220, the appropriate group point calculation unit 17 includes the participant information input from the participant information input unit 13, the condition information input from the condition information input unit 14, and the participation input from the classification unit 16. Based on the classification result of the person, the appropriate group point described later is calculated for each classification result, and the appropriate group point for each classification result is output to the result output unit 18.

First, the appropriate group point calculation unit 17 substitutes 1 for the variable i (S170). The variable i is a variable for identifying the item to be processed (hereinafter, referred to as “item i”) among the items included in the condition information (FIG. 7). In the flowchart of FIG. 3 and the like, “←” indicates that the value on the right side is substituted on the left side.

Subsequently, the appropriate group point calculation unit 17 refers to the condition information (FIG. 7) input from the condition information input unit 14 and determines whether or not the condition is set for the item i (S180). When the condition is not set in the item i (No in S180), the appropriate group point calculation unit 17 proceeds to step S200 without executing step S190. When the condition is set in the item i (Yes in S180), the appropriate group point calculation unit 17 executes step S190 and proceeds to step S200.

In step S190, the appropriate group point calculation unit 17 sets the matching evaluation value for the condition of item i (hereinafter, referred to as “target condition”) for each of the plurality of classification results (FIG. 8) input from the classification unit 16. p is calculated based on the participant information input from the participant information input unit 13. The matching evaluation value p for the condition of item i is an example of a first index indicating the degree to which the classification result matches the condition, and can take a value of 0 to 100. As for the match evaluation value p, the smaller the value, the more the condition is matched. Hereinafter, the matching evaluation value p for the condition of item i relating to the kth classification result is referred to as a matching evaluation value _pki .

In the present embodiment, the above-mentioned conditions (1) to (10) can be set as the conditions for the items, but the method of calculating the matching evaluation value p differs for each condition. Therefore, in the following, a method of calculating the matching evaluation value _pki for each condition will be described.

[When the target condition is condition (1)]
Condition (1) is "not considered as a condition for grouping." Therefore, in this case, the appropriate group point calculation unit 17 sets the match evaluation value p _ki = 0 for all the classification results k. However, if the value is common to all the classification results k, a value other than 0 may be set as the match evaluation value p _ki .

[When the target condition is condition (2)]
Condition (2) is "there is no bias in the value of the target item between groups (however, it can be set only for the item whose" type "is" category "). ". Therefore, based on the idea that the condition (2) is satisfied if the number of types of values (categories) included in each group for item i is the same or the error is within one person, the appropriate group point calculation unit 17 determines the classification result. The match evaluation value p _ki of k is calculated as follows.

First, the appropriate group point calculation unit 17 counts the number of participants having the value for each value (category) related to the item i in the classification result k, and obtains the maximum value and the minimum value thereof.

The appropriate group point calculation unit 17 sets p _ki = 0 when the maximum value − minimum value ≦ = 1 is satisfied for all the values (categories) related to the item i, and sets p _ki = 100 in other cases.

For example, in the case of 3 people x 5 groups and item i is "gender (category)" (category = "male", "female"), classification result 1 = (male male and female), (male male and female), (male male and female) ), (Men and women), (Men and women), and classification result 2 = (Men and men), (Men and women), (Men and women), (Men and women), (Men and women).

In the classification result 1, the maximum value and the minimum value for "male" are as follows.
Maximum number of men in each group = 2
Minimum number of men included in each group = 1
Therefore, the maximum value-minimum value = 1 of the category "male".

Further, in the classification result 1, the maximum value and the minimum value for "woman" are as follows.
Maximum number of women in each group = 2
Maximum number of men in each group = 1
Therefore, the maximum value-minimum value = 1 of the category "woman".

From the above, the classification result 1 satisfies the maximum value − minimum value ≦ = 1 for all the values (categories) related to the item i. Therefore, the match evaluation value p _1i = 0.

On the other hand, in the classification result 2, the maximum value and the minimum value for "male" are as follows.

Maximum number of men in each group = 3
Minimum number of men included in each group = 1
Therefore, the maximum value-minimum value = 2 of the category "male".

In addition, in the classification result 2, the maximum value and the minimum value for "woman" are as follows.

Maximum number of women in each group = 2
Maximum number of men included in each group = 0
Therefore, the maximum value-minimum value of the category "woman" = 2
From the above, the classification result 2 does not satisfy the maximum value − minimum value ≦ = 1 for all the values (categories) related to the item i. Therefore, the match evaluation value p _2i = 100.

Regarding the condition (2), the p _ki does not have to be 0 or 100. p _ki may be a relatively small constant value when the maximum value −minimum value ≦ = 1 is satisfied, and may be a relatively large constant value when the maximum value −minimum value ≦ = 1 is not satisfied.

In short, the calculation of the match evaluation value p when the target condition is the condition (2) is the item i which is the target of the condition among the two or more conditions related to the classification into the group set for one or more items. Is to calculate a first index indicating the degree to which the first condition regarding the bias of the value among the groups is met.

[When the target condition is condition (3)]
Condition (3) is "to minimize the bias of values between groups for the target item (however, it can be set only for the item whose" type "is" category "). ". Therefore, based on the idea that the smaller the difference in the number of people by type (category, gender, etc.) of the value (category) included in each group for item i, the degree to which condition (3) is met increases. The calculation unit 17 calculates the match evaluation value _pki of the classification result k as follows.

但し、
ｃ＝項目ｉの値（カテゴリ）の種類数
ＭＡＸ_Ｃ＝項目ｉに関して値（カテゴリ）ｃの参加者のグループ別の人数の最大値
ＭＩＮ_Ｃ＝項目ｉに関して値（カテゴリ）ｃの参加者のグループ別の人数の最小値
ｎ＝参加者人数
例えば、３名×５グループで、項目ｉが「性別（カテゴリ）」（カテゴリ＝「男」、「女」）の場合、参加者人数ｎ＝１５である。ここで、分類結果１＝（男男女）（男男女），（男男女），（男女女），（男女女）であり、分類結果２＝（男男男），（男男女），（男男女），（男女女），（男女女）であるとする。

However,
c = Number of types of value (category) of item i MAX _C = Maximum number of participants by group of value (category) c with respect to item i MIN _C = Group of participants with value (category) c with respect to item i Minimum value of another number n = number of participants For example, in the case of 3 people x 5 groups and item i is "gender (category)" (category = "male", "female"), the number of participants n = 15. is there. Here, the classification result 1 = (male and female) (male and female), (male and female), (male and female), (male and female), and the classification result 2 = (male and male), (male and female), (male). Men and women), (men and women), (men and women).

In this case, for the item i, matching evaluation value p _2i regarding matching evaluation value p _1i and classification result 2 on classification result 1 is as follows.
p _1i = (1 + 1) / 15 × 100 = 13.3
p _2i = (2 + 2) / 15 × 100 = 26.7
The above equation is an example, and the matching evaluation value p regarding the condition (3) may be calculated by another equation.

In short, the calculation of the match evaluation value p when the target condition is the condition (3) is the item that is the target of the condition among the two or more conditions related to the classification into the group set for one or more items. For i, a first index indicating the degree to which the first condition regarding the bias of the value among the groups is satisfied is calculated.

[When the target condition is condition (4)]
Condition (4) includes "more types of values for each group for the target item (however, it can be set only for items whose" type "is" category "). ". Therefore, based on the idea that the greater the number of types of values (categories) included in each group, the higher the degree of matching with the condition (4), the appropriate group point calculation unit 17 performs a matching evaluation of the classification result k. The value p _ki is calculated as follows.

但し、
ｇ＝グループ数
ｃ＝項目ｉの値（カテゴリ）の種類数
ｃ_ｘ＝グループｘにおける項目ｉに関する値（カテゴリ）の種類数
例えば、３名×５グループで、項目ｉが「性別（カテゴリ）」（カテゴリ＝「男」、「女」）の場合、グループ数ｇ＝５である。更に、項目ｉの値（カテゴリ）の種類数ｃ＝２（「男」、「女」）である場合に、分類結果１＝（男男女），（男男女），（男男女），（男女女），（男女女）、分類結果２＝（男男男），（男男女），（男男女），（男女女），（男女女）であるとする。

However,
g = number of groups c = number of types of value (category) of item i c _x = number of types of value (category) related to item i in group x For example, in 3 people x 5 groups, item i is "gender (category)" In the case of (category = "male", "female"), the number of groups g = 5. Furthermore, when the number of types of the value (category) of item i is c = 2 (“male”, “female”), the classification result 1 = (male male / female), (male male / female), (male male / female), (male / female). It is assumed that (female), (male and female), classification result 2 = (male and male), (male and female), (male and female), (male and female), (male and female).

In this case, for the item i, matching evaluation value p _2i regarding matching evaluation value p _1i and classification result 2 on classification result 1 is as follows.
p _1i = (1- (2 + 2 + 2 + 2 + 2) / 5 × 2) × 100 = 0
p _2i = (1- (1 + 2 + 2 + 2 + 2) / 5 × 2) × 100 = 10
The above equation is an example, and the matching evaluation value p regarding the condition (4) may be calculated by another equation.

In short, the calculation of the match evaluation value p when the target condition is the condition (4) is the item that is the target of the condition among the two or more conditions related to the classification into the group set for one or more items. For i, a first index indicating the degree to which the second condition regarding the number of values included in each group is met is calculated.

[When the target condition is condition (5)]
The condition (5) is "the number of types of values included in each group for the target item is equal to or greater than the specified number (however, it can be set only for the item whose" type "is" category "). ". Therefore, in this case, the appropriate group point calculation unit 17 counts the number of types of values (categories) for each group with respect to the item i, and if all the count results are equal to or greater than the specified number,
p _ki = 0
And otherwise
p _ki = 100
And.

Regarding the condition (5), the p _ki does not have to be 0 or 100. p _ki may be a relatively small constant value if all the count results are equal to or more than a specified number, and may be a relatively large constant value otherwise.

The "specified number" may be preset and stored in the auxiliary storage device 102 or the like, or may be input by the user at the timing of step S190 or earlier. Further, the above is an example, and the matching evaluation value p regarding the condition (5) may be calculated by another method.

In short, the calculation of the match evaluation value p when the target condition is the condition (5) is the item that is the target of the condition among the two or more conditions related to the classification into the group set for one or more items. For i, a first index indicating the degree to which the third condition regarding whether or not the number of types of values included in each group is equal to or greater than the specified value is satisfied is calculated.

[When the target condition is condition (6)]
The condition (6) is "each group includes all kinds of values for the target item (however, it can be set only for the item whose" type "is" category "). ". Therefore, in this case, when the appropriate group point calculation unit 17 includes all kinds of values (categories) in each group with respect to the item i,
p _ki = 0
And otherwise
p _ki = 100
And.

Regarding the condition (5), the p _ki does not have to be 0 or 100. The p _ki may be a relatively small constant value when all kinds of values (categories) are included in each group with respect to the item i, and may be a relatively large constant value otherwise. Further, the above is an example, and the matching evaluation value p regarding the condition (5) may be calculated by another method.

In short, the calculation of the match evaluation value p when the target condition is the condition (6) is the item that is the target of the condition among the two or more conditions related to the classification into the group set for one or more items. It is to calculate a first index indicating the degree to which the fourth condition regarding whether or not each group includes all possible values for i is satisfied.

[When the target condition is condition (7)]
Condition (7) includes "one or more participants in each group having a value specified for the target item (can be set only for the item whose" type "is" category "). ". Therefore, in this case, when the appropriate group point calculation unit 17 includes one or more participants having a specified value (category) in each group with respect to the item i,
p _ki = 0
And otherwise
p _ki = 100
And.

The "designated value (category)" may be preset and stored in the auxiliary storage device 102 or the like, or may be input by the user at the timing of step S190 or earlier.

Regarding the condition (7), the p _ki does not have to be 0 or 100. p _ki may be a relatively small constant value when one or more participants having a specified value (category) are included in each group, and may be a relatively large constant value otherwise. .. Further, the above is an example, and the matching evaluation value p regarding the condition (7) may be calculated by another method. In addition, the number of participants in each group including the participants having the specified values for the target items is not limited to one. That is, the number of people may be r (r ≧ 1) or more.

In short, the calculation of the match evaluation value p when the target condition is the condition (7) is the item that is the target of the condition among the two or more conditions related to the classification into the group set for one or more items. It is to calculate the first index which shows the degree of meeting the fifth condition regarding whether or not each group contains 1 or more values specified for i.

[When the target condition is condition (8)]
Condition (8) states, "Participants who share the same value for the target item are collected in the same group as much as possible. Participants who are fractions collect the fractions to form a group ("type "is" category ". Can be set only for items that are). ". Therefore, for each value (each category) related to the item i, the appropriate group point calculation unit 17 is based on the idea that the smaller the number of groups including the value (the category), the higher the degree of meeting the condition (8). , The matching evaluation value p _ki of the classification result k is calculated as follows.

但し、
ｇ＝グループ数
ｃ＝項目ｉの値（カテゴリ）の種類数
例えば、３名×５グループで、項目ｉが「性別（カテゴリ）」（カテゴリ＝「男」、「女」）の場合、グループ数ｇ＝５、値（カテゴリ）の種類数＝２（「男」、「女」）である。ここで、分類結果１＝（男男女），（男男女），（男男女），（男女女），（男女女）であり、分類結果２＝（男男男），（男男女），（男男女），（男女女），（男女女）であるとする。

However,
g = number of groups c = number of types of value (category) of item i For example, if there are 3 people x 5 groups and item i is "gender (category)" (category = "male", "female"), the number of groups g = 5, the number of types of values (categories) = 2 (“male”, “female”). Here, the classification result 1 = (male and female), (male and female), (male and female), (male and female), (male and female), and the classification result 2 = (male and male), (male and female), ( Men and women), (men and women), and (men and women).

In this case, for the item i, matching evaluation value p _2i regarding matching evaluation value p _1i and classification result 2 on classification result 1 is as follows.
p1i = ((5 + 5) / 5 × 2) × 100 = 100
p2i = ((5 + 4) / 5 × 2) × 100 = 90
The above equation is an example, and the matching evaluation value p regarding the condition (8) may be calculated by another equation.

In short, the calculation of the match evaluation value p when the target condition is the condition (8) is the item that is the target of the condition among the two or more conditions related to the classification into the group set for one or more items. For i, a first index indicating the degree to which the sixth condition regarding the number of groups including each value is satisfied is calculated.

[When the target condition is condition (9)]
Condition (9) is "to minimize the bias of values between groups for the target item (can be set only for items whose" type "is" continuous "). ". Therefore, based on the idea that if the average value and the standard deviation of the values related to the item i of each group are about the same, the degree of matching with the condition (9) is high, the appropriate group point calculation unit 17 matches the classification result k. The evaluation value p _ki is calculated as follows.

但し、
ｍａｘ＝項目ｉの値の中での最大値
ｍｉｎ＝項目ｉの値の中での最小値
ＧＭｍａｘ＝グループ別での項目ｉの値の平均の最大値
ＧＭｍｉｎ＝グループ別での項目ｉの値の平均の最小値
ＧＳＤｍａｘ＝グループ別での項目ｉの値の標準偏差の最大値
ＧＳＤｍｉｎ＝グループ別での項目ｉの値の標準偏差の最小値
Ｍ＝参加者全体での項目ｉの値の平均値
Ｇｎｍａｘ＝グループ別の参加者人数の最大値
例えば、３名×５グループで、項目ｉが「年齢」であるとする。また、分類結果１＝，（２０，３０，４０），（２５，３５，４５），（３０，４０，５０），（２０，３５，５０）、分類結果２＝（２０，２０，２５），（３０，３５，３５），（３５，４０，４０），（４５，５０，５０）であるとする。

However,
max = maximum value among the values of item i min = minimum value among the values of item i GMmax = maximum value of the average value of item i by group GMmin = value of item i by group Minimum average value GSDmax = Maximum standard deviation of item i value by group GSDmin = Minimum standard deviation of item i value by group M = Average value of item i value for all participants Gnmax = maximum number of participants by group For example, in a group of 3 people × 5, item i is “age”. In addition, classification result 1 =, (20,30,40), (25,35,45), (30,40,50), (20,35,50), classification result 2 = (20,20,25) , (30,35,35), (35,40,40), (45,50,50).

In this case, for classification result 1, the average value of each group for item i is 30,35,40,35, the standard deviation of each group is 10,10,10,15, the overall average is 35, and each group. The maximum number of participants is 3. Therefore, the matching evaluation value _p1i regarding the classification result 1 is as follows.
p _1i = ((40-30) / (50-20)) x 50+ ((15-10) / 35 x √2) x 50 = 16.7 + 5.05 = 21.7
On the other hand, regarding the classification result 2, the average value of each group regarding item i is 21.7, 33.3, 38.3, 46.7, and the standard deviation of each group is 2.89, 2.89, 2. 89, 5.77, the overall average is 35, and the maximum number of participants by group is 3. Therefore, the matching evaluation value p _2i regarding the classification result 2 is as follows.
p _2i = ((46.7-21.7) / (50-20)) × 50 + ((5.77-2.89) / 35 × √2) × 50 = 41.7 + 2.91 = 44.6
The above equation is an example, and the matching evaluation value p regarding the condition (9) may be calculated by another equation.

[When the target condition is condition (10)]
Condition (10) is "collect participants having similar values for the target item in the same group as much as possible (can be set only for items whose" type "is" continuous "). ". Therefore, based on the idea that the smaller the standard deviation in each group of the value of item i, the higher the degree of matching with the condition (10), the appropriate group point calculation unit 17 determines the matching evaluation value of the classification result k. Calculate p _ki as follows (use the value obtained by adding all the standard deviations of the values of item i in each group).

但し、
ｇ＝グループ数
ＧｘＳＤ＝グループｘ内での項目ｉの値の標準偏差
Ｍ＝参加者全体での項目ｉの値の平均値
Ｇｎｍａｘ＝グループ別の参加者人数の最大値
例えば、３名×５グループで、項目ｉが「年齢」であるとする。また、分類結果１＝（２０，３０，４０），（２５，３５，４５），（３０，４０，５０），（２０，３５，５０）、分類結果２＝（２０，２０，２５），（３０，３５，３５），（３５，４０，４０），（４５，５０，５０）であるとする。

However,
g = Number of groups GxSD = Standard deviation of the value of item i within group x M = Mean value of the value of item i for all participants Gnmax = Maximum number of participants by group For example, 3 people x 5 groups Then, it is assumed that the item i is "age". In addition, classification result 1 = (20,30,40), (25,35,45), (30,40,50), (20,35,50), classification result 2 = (20,20,25), It is assumed that it is (30, 35, 35), (35, 40, 40), (45, 50, 50).

In this case, for classification result 1, the number of groups is 4, the standard deviation of the value of item i in each group is 10,10,10,15, the overall average is 35, and the maximum number of participants in each group. Is 3. Therefore, the matching evaluation value _p1i regarding the classification result 1 is as follows.
p _1i = (10 + 10 + 10 + 15) / (35 × √2 × 4) × 100 = 22.7
On the other hand, for classification result 2, the number of groups is 4, the standard deviation of the value of item i in each group is 2.89, 2.89, 2.89, 5.77, the overall average is 35, and by group. The maximum number of participants in is 3. Therefore, the matching evaluation value p _2i regarding the classification result 2 is as follows.
p _2i = (2.89 + 2.89 + 2.89 + 5.77) / (35 × √2 × 4) × 100 = 7.29
The above formula is an example, and the matching evaluation value p regarding the condition (10) may be calculated by another formula.

In the above, the match evaluation value p takes a numerical value between 0 and 100, and the smaller the value, the more the condition is met, but the present invention is not limited to this.

In step S200, the appropriate group point calculation unit 17 determines whether or not the processing has been completed for all the items included in the condition information (FIG. 7). That is, the appropriate group point calculation unit 17 determines whether or not the value of the variable i has reached the number of all the items.

When an unprocessed item remains (No in S200), the appropriate group point calculation unit 17 adds 1 to the variable i (S210), and repeats steps S180 and subsequent steps. When the processing is completed for all items (Yes in S200), the appropriate group point calculation unit 17 calculates and calculates the appropriate group points for each classification result based on the match evaluation value p calculated for each item. The appropriate group point is output to the result output unit 18 (S220).

FIG. 9 is a diagram for explaining an example of a method of calculating suitable group points. FIG. 9 shows an example of calculating the appropriate group points for the classification result k. As shown in FIG. 9, the appropriate group points of the classification result k correspond to the priority of the matching evaluation value _pki for the conditions of each item calculated by the appropriate group point calculation unit 17 with respect to the classification result k in step S190. It is a weighted sum with the weighting as the weight. In FIG. 9, O _i indicates a weighting for the condition related to the item i. For example, priority may be directly used as the value of O _i. In this case, O _i = 1 when the priority is 1, and O _i = 2 when the priority is 2. Also, O _i for item priority is not set, for example, it may be zero.

FIG. 10 shows an example of the calculation result of the appropriate group points for each classification result. The appropriate group point calculation unit 17 may specify a ranking based on the matching evaluation value p for each item and calculate an appropriate group point using the ranking, instead of using the matching evaluation value p as it is. Further, the appropriate group point calculation unit 17 may calculate the appropriate group point by a calculation method other than the above. Moreover, the variable corresponding to the priority is not limited to the above.

Subsequently, the result output unit 18 outputs (displays) the classification result to, for example, the display device 106 based on the appropriate group points (S230). For example, the result output unit 18 may output a classification result having the smallest suitable group point. Alternatively, the result output unit 18 may output the results in which each classification result is sorted in ascending order of appropriate group points. In this case, the result output unit 18 may limit the classification result to be output to a part of the higher rank in the sort order. In addition, the result output unit 18 may also output information indicating the degree of matching of the conditions for each item in each classification result. The information indicating the degree of matching of the conditions for each item may be the matching evaluation value p itself. Alternatively, a threshold value (for example, 50) for the match evaluation value p may be set. In this case, the result output unit 18 outputs information indicating that the condition is satisfied for the item whose match evaluation value p is equal to or less than the threshold value, and the condition is not satisfied for the item whose match evaluation value p exceeds the threshold value. Information indicating that may be output. Further, the output destination is not limited to the display device 106. For example, the result output unit 18 may store the classification result of the output target in the auxiliary storage device 102, or may transmit it to another computer via the network.

In the above, the subject of classification into groups is the participants (people) of the workshop, and an example of grouping based on the attributes of each participant has been described. However, the target of classification into groups is people. Not limited. For example, any thing can be classified as long as it has an attribute composed of a plurality of items.

Further, although the example in which the classification support device 10 is realized by one computer is shown above, the classification support device 10 may be realized by being distributed to a plurality of computers connected via a network. For example, a computer as a server has a classification unit 16, an appropriate group point calculation unit 17, and a result output unit 18, and an attribute item input unit 11, a group requirement input unit 12, a participant information input unit 13, and a condition information input unit 14. And the computer as a client may have the start instruction input unit 15. Each part shown in FIG. 2 may be dispersed by other forms.

As described above, according to the first embodiment, it is possible to support grouping to which a plurality of conditions are applied. In addition, in grouping, a matching evaluation value is calculated for each condition for each of a plurality of classification results (results of grouping), and a plurality of conditions are obtained for each classification result by an appropriate group point obtained by summing the matching evaluation values. The degree of matching with is evaluated. Therefore, it is possible to present a grouping plan that has a high degree of comprehensive matching with a plurality of conditions.

[Second Embodiment]
Next, the second embodiment will be described. The second embodiment will explain the differences from the first embodiment. The points not particularly mentioned in the second embodiment may be the same as those in the first embodiment.

In the second embodiment, the configuration information of a certain group (that is, the information indicating the participants constituting the group) is input to the classification support device 10, and the group (hereinafter referred to as "target group") is referred to. ) Is an example in which the classification support device 10 outputs the evaluation result as to what kind of condition the group corresponds to the grouping.

FIG. 11 is a diagram showing a functional configuration example of the classification support device 10 according to the second embodiment. In FIG. 11, the same parts as those in FIG. 2 are designated by the same reference numerals.

In FIG. 11, the classification support device 10 further has a target group input unit 19. The target group input unit 19 inputs the configuration information of the target group (hereinafter, referred to as “target group information”). The classification support device 10 of the second embodiment does not have to have the condition information input unit 14. However, when the classification support device 10 realizes the functions of both the first embodiment and the second embodiment, the classification support device 10 also has the condition information input unit 14.

FIG. 12 is a flowchart for explaining an example of the processing procedure executed by the classification support device 10 in the second embodiment. In FIG. 12, the same steps as those in FIG. 3 are assigned the same step numbers, and the description thereof will be omitted as appropriate.

Following step S130, the target group input unit 19 inputs the target group information and outputs the input target group information to the classification unit 16, the appropriate group point calculation unit 17, and the result output unit 18 (S135). For example, the target group information may be a set of IDs of each participant constituting the target group. As an example, when the IDs of all participants are 1 to 30 (that is, when all participants are 30 people), even if "ID = 1,5,7,10" is input as the target group information. Good. The target group information may be registered in the database or the like in advance, or may be input by the user at the timing of step S135.

In the second embodiment, step S140 does not have to be executed.

Following step S160, the classification unit 16 removes the classification result that does not include the target group related to the target group information input from the target group input unit 19 from the classification result generated in step S150 (S161). Therefore, the classification result in the following means the classification result including the target group. In addition, the classification unit 16 may generate only the classification result including the target group in step S150. In this case, step S161 does not have to be executed. Alternatively, the classification result from which the target group has been removed is registered in the database or the like in advance, and the classification unit 16 may acquire the classification result from the database.

Subsequently, the appropriate group point calculation unit 17 generates a plurality of combinations of conditions for each item constituting the attribute of the participant (S162). That is, the appropriate group point calculation unit 17 generates a plurality of types of condition information shown in FIG. 7. For example, the appropriate group point calculation unit 17 may generate condition information corresponding to each of all combinations, or may generate a random number of condition information. Here, it is assumed that N (N ≧ 2) types of condition information are generated. However, N kinds of condition information are generated in advance and registered in a database or the like, and the appropriate group point calculation unit 17 may acquire the condition information from the database. The condition information generated or acquired in step S162 does not have to include the priority.

Subsequently, the appropriate group point calculation unit 17 assigns 1 to the variable j, 100 to the variable min_p, and makes the variable min_c empty (for example, assigns NULL to the variable min_c) (S163). .. That is, in step S163 of FIG. 12, [] indicates NULL. The variable j is a variable for identifying the condition information to be processed in the following. Hereinafter, the j-th conditional information is referred to as "condition information j". The variable min_p is a list-type variable for storing the minimum value of the appropriate group points calculated for each classification result for each condition information below. The variable min_c is a list-type variable for storing the order of condition information (the value of the variable j at that time) when the appropriate group point of the minimum value is calculated.

Following step S170, the appropriate group point calculation unit 17 determines whether or not a condition is set for item i in the condition information j (S180a). When the condition is not set in the item i (No in S180), the appropriate group point calculation unit 17 proceeds to step S200 without executing step S190a. When the condition is set in the item i (Yes in S180), the appropriate group point calculation unit 17 executes step S190a and proceeds to step S200.

In step S190a, the appropriate group point calculation unit 17 calculates the matching evaluation value p for the condition of item i for each of the plurality of classification results. At this time, the appropriate group point calculation unit 17 stores the calculated match evaluation value p in association with j and i, for example, in the memory device 103. That is, the match evaluation value p is stored in the memory device 103 for each condition information and for each item.

When the condition information j is applied to each classification result including the target group related to the target group information input from the target group input unit 19 by repeating steps S180a to S210 for all the items regarding the condition information j. The match evaluation value p for each item of is calculated.

When the processing for all items is completed (Yes in S200), the appropriate group point calculation unit 17 calculates the appropriate group points for each classification result including the target group based on the match evaluation value p calculated for each item. (S220). At this time, since the condition information j does not include the priority, weighting corresponding to the priority is not performed. For example, the appropriate group point of the classification result k may be the sum of the matching evaluation values _pki . Or a weighted sum of p _ki of carbonochloridate with an equal weight to each p _ki may be a suitable group points.

Subsequently, the appropriate group point calculation unit 17 determines whether or not the calculated appropriate group point is min_p or less (S240). When the appropriate group point exceeds min_p (No in S240), the appropriate group point calculation unit 17 proceeds to step S260 without executing step S250. On the other hand, when the appropriate group point is min_p or less (Yes in S240), the appropriate group point calculation unit 17 substitutes the appropriate group point into min_p and adds the value of the variable j as an element of the list of min_c. (S250), the process proceeds to step S260. However, when the value of min_p becomes smaller than before, in step S250, the appropriate group point calculation unit 17 adds the value of the variable j as an element of the list of min_c after emptying min_c.

In step S260, the appropriate group point calculation unit 17 determines whether or not the value of the variable j is N or more. That is, the appropriate group point calculation unit 17 determines whether or not step S170 or later has been executed for all N kinds of condition information. When there is unexecuted condition information (No in S260), the appropriate group point calculation unit 17 adds 1 to the variable j and repeats step S170 and subsequent steps. When step S170 and subsequent steps are executed for all N kinds of condition information (Yes in S260), the appropriate group point calculation unit 17 outputs the condition information corresponding to the value of each element of min_c at this point in the result output unit 18. Output to. Therefore, the appropriate group point calculation unit 17 outputs a plurality of condition information to the result output unit 18 when there is a plurality of order information from which the minimum appropriate group points are obtained.

Subsequently, the result output unit 18 determines what kind of group the target group related to the target group information input from the target group input unit 19 is, based on the condition information input from the appropriate group point calculation unit 17. The indicated information is output (displayed) to, for example, the display device 106 (S280). For example, the result output unit 18 has the condition that the match evaluation value p stored in the memory device 103 is less than the threshold value among the conditions set for each item in the condition information as follows (that is, the condition). By enumerating the conditions with a high degree of matching), information indicating what kind of group the target group is may be output.
"The target group is
(2) There is no bias in the values of item a among the groups.
(9) Regarding item b, is it a group created as a result of minimizing the bias of values (continuous values) between groups?
(3) For item c, reduce the bias of values (categories) between groups as much as possible.
(9) Item b is a group created as a result of biasing the values (continuous values) among the groups as much as possible. 』\
As described above, according to the second embodiment, by inputting the configuration information of a certain group, what kind of condition the group satisfies (what kind of condition is satisfied). It is possible to provide the user with information indicating (is it a high group?).

In each of the above embodiments, the appropriate group point calculation unit 17 is an example of the first calculation unit and the second calculation unit. The result output unit 18 is an example of an output unit. Participant information is an example of the first information. The plurality of classification results are an example of the second information. The condition information is an example of the third information. Appropriate group points are an example of the second index. The classification result including the target group is an example of the fourth information. The plurality of types of condition information is an example of the fifth information.

Although the embodiments of the present invention have been described in detail above, the present invention is not limited to such specific embodiments, and various aspects are within the scope of the gist of the present invention described in the claims. It can be transformed and changed.

10 Classification support device 11 Attribute item input unit 12 Group requirement input unit 13 Participant information input unit 14 Condition information input unit 15 Start instruction input unit 16 Classification unit 17 Appropriate group point calculation unit 18 Result output unit 19 Target group input unit 100 Drive Device 101 Recording medium 102 Auxiliary storage device 103 Memory device 104 CPU
105 Interface device 106 Display device 107 Input device B Bus

Claims (11)

First information indicating the respective values of the plurality of classification targets for each of the plurality of items constituting the attributes relating to the classification targets to the plurality of groups, and
Second information showing a plurality of classification results when the plurality of classification targets are classified into the plurality of groups, and
A third piece of information, which is set for one or more of the above items and indicates two or more conditions for classification into groups, and a priority for the degree to which each of the above conditions is satisfied.
For each of the classification results, a first calculation unit that calculates a first index indicating the degree to which each condition is met, and
Each of the classification results has a second calculation unit for calculating a second index indicating the degree to which the two or more conditions are met, based on the first index calculated for the classification result.
The above two or more conditions are
The first condition regarding the bias of values between groups for the item subject to the condition,
The second condition regarding the number of values included in each group for the item subject to the condition,
The third condition regarding whether or not the number of types of values included in each group for the item subject to the condition is equal to or greater than the specified value,
Fourth condition regarding whether or not each group includes all possible values for the item subject to the condition,
Fifth condition regarding whether each group contains one or more values specified for the item subject to the condition,
The sixth condition regarding the number of groups including each value for the item subject to the condition,
Any two or more of the conditions,
A classification support device characterized by this. First information indicating the respective values of the plurality of classification targets for each of the plurality of items constituting the attributes relating to the classification targets to the plurality of groups, and
When the plurality of classification targets are classified into the plurality of groups, the fourth information indicating a plurality of classification results including a predetermined group and
Fifth information including a plurality of combinations of two or more conditions related to group classification set for one or more of the above items, and
See,
For each of the combinations, a first calculation unit that calculates a first index indicating the degree to which each of the above conditions is met for each of the classification results,
For each of the combinations, a second calculation unit that calculates a second index indicating the degree to which the two or more conditions are met, based on the first index calculated for the classification result. When,
Among the combinations, the second index indicating that the degree of satisfying the two or more conditions is the highest is provided with an output unit that outputs information based on the calculated combination.
The above two or more conditions are
The first condition regarding the bias of values between groups for the item subject to the condition,
The second condition regarding the number of values included in each group for the item subject to the condition,
The third condition regarding whether or not the number of types of values included in each group for the item subject to the condition is equal to or greater than the specified value,
Fourth condition regarding whether or not each group includes all possible values for the item subject to the condition,
Fifth condition regarding whether each group contains one or more values specified for the item subject to the condition,
The sixth condition regarding the number of groups including each value for the item subject to the condition,
Any two or more of the conditions,
A classification support device characterized by this. The first calculation unit is based on the difference between the maximum value and the minimum value of the number of the classification targets having the same value in each group for the items subject to the first condition. Calculate the index of
The classification support device according to claim 1 or 2, characterized in that. The first calculation unit uses the first index for the item subject to the second condition.

However,
p _ki = the first index c = the number of types of the value of the item MAX _C = the maximum value of the number of the value c for the item by the group of the classification target MIN _C = the value c of the classification target for the item Minimum value of number for each group n = Calculated based on the number of classification targets,
The classification support device according to claim 1 or 2, characterized in that. The first calculation unit uses the first index for the item subject to the third condition.

However,
p _ki = the first index g = the number of groups c = the number of types of values of the item c _g = calculated based on the number of types of values related to the item in the group g.
The classification support device according to claim 1 or 2, characterized in that. The first calculation unit relatives the degree of matching indicated by the first index when all kinds of values are included in each group for the item subject to the fourth condition. To raise the target
The classification support device according to claim 1 or 2, characterized in that. The first calculation unit is a match indicated by the first index when one or more of the classification targets having a specified value are included in each group for the item subject to the fifth condition. To increase the degree relatively,
The classification support device according to claim 1 or 2, characterized in that. The first calculation unit uses the first index for the item subject to the sixth condition.

However,
p _ki = the first index g = number of groups c = calculated based on the number of types of values of the item,
The classification support device according to claim 1 or 2, characterized in that. First information indicating the respective values of the plurality of classification targets for each of the plurality of items constituting the attributes relating to the classification targets to the plurality of groups, and
Second information showing a plurality of classification results when the plurality of classification targets are classified into the plurality of groups, and
A third piece of information, which is set for one or more of the above items and indicates two or more conditions for classification into groups, and a priority for the degree to which each of the above conditions is satisfied.
With reference to the first calculation procedure for calculating the first index indicating the degree of matching with each of the above conditions for each of the classification results.
For each of the classification results, the computer executes a second calculation procedure for calculating a second index indicating the degree to which the two or more conditions are met, based on the first index calculated for the classification result. And
The above two or more conditions are
The first condition regarding the bias of values between groups for the item subject to the condition,
The second condition regarding the number of values included in each group for the item subject to the condition,
The third condition regarding whether or not the number of types of values included in each group for the item subject to the condition is equal to or greater than the specified value,
Fourth condition regarding whether or not each group includes all possible values for the item subject to the condition,
Fifth condition regarding whether each group contains one or more values specified for the item subject to the condition,
The sixth condition regarding the number of groups including each value for the item subject to the condition,
Any two or more of the conditions,
A classification support method characterized by this. First information indicating the respective values of the plurality of classification targets for each of the plurality of items constituting the attributes relating to the classification targets to the plurality of groups, and
When the plurality of classification targets are classified into the plurality of groups, the fourth information indicating a plurality of classification results including a predetermined group and
Fifth information including a plurality of combinations of two or more conditions related to group classification set for one or more of the above items, and
See,
For each of the combinations, a first calculation procedure for calculating a first index indicating the degree to which each of the conditions is met for each of the classification results, and
For each of the combinations, a second calculation procedure for calculating a second index indicating the degree to which the two or more conditions are met, based on the first index calculated for the classification result. When,
Among the combinations, the computer executes an output procedure for outputting information based on the combination in which the second index indicating that the degree of meeting the two or more conditions is the highest is calculated.
The above two or more conditions are
The first condition regarding the bias of values between groups for the item subject to the condition,
The second condition regarding the number of values included in each group for the item subject to the condition,
The third condition regarding whether or not the number of types of values included in each group for the item subject to the condition is equal to or greater than the specified value,
Fourth condition regarding whether or not each group includes all possible values for the item subject to the condition,
Fifth condition regarding whether each group contains one or more values specified for the item subject to the condition,
The sixth condition regarding the number of groups including each value for the item subject to the condition,
Any two or more of the conditions,
A classification support method characterized by this. A program characterized in that a computer functions as each part according to any one of claims 1 to 8.

Images