JP7071428B2 - Information processing equipment, information processing methods and computer programs for information processing - Google Patents

Description

The present invention relates to an information processing device, an information processing method, and a computer program for information processing, and particularly to a device / method / program that facilitates analysis of a large amount of information in a group.

It has been practiced to process information, plot the collected information on a graph, and analyze the distribution state.

Japanese Unexamined Patent Publication No. 2016-180599 Japanese Unexamined Patent Publication No. 2009-098072 Japanese Unexamined Patent Publication No. 2016-176811

The present invention has been made to solve the above-mentioned problems, and an object of the present invention is to provide a device / method / program for analyzing the distribution state of collected information in a new style.

In order to achieve the above object, the information processing apparatus of the present invention causes one sample to be plotted on a graph with at least two factors for a large number of samples, and each factor is arranged in a predetermined arrangement on an axis on the graph. The plotting means for scaling, the separation distance means for obtaining the mutual separation distance of each sample, and the sample having the short / short separation distance, the samples related to the separation distance are assigned to one group, and the separation distance is set to the distance. If one of the samples already belongs to one of the groups, the sample that does not belong to the group is attributed to the group of the samples that belong to the group, and both of the samples related to the separation distance are already one of them. If it belongs to a group, it is provided with an attribution means that does not attribute the sample to any group, and a stop termination means that suspends or terminates the attribution processing of the attribution means by having a plurality of the above groups and under predetermined conditions. rice field.

As a result, a large number of samples are assigned to a plurality of groups, and close samples belong to the same group. Depending on the number and size of these groups, the distribution state can be easily analyzed, which can greatly contribute to the analysis by artificial intelligence. The same applies to information processing methods and computer programs for information processing.

The whole circuit of the information processing apparatus is shown. The program / information storage unit 7 (information storage unit 6) is shown. The flowchart of the whole processing (information processing) is shown. The flowchart (first half) of the group processing of step 13 is shown. The flowchart (second half) of the group processing of step 13 is shown. A concrete example of the face shape of information processing is shown. A concrete example of the face shape of information processing is shown. A specific example of the sample S on the information processing graph is shown. A specific example of the sample S on the information processing graph is shown. A specific example of the sample S on the information processing graph is shown. A specific example of the processing of the group G on the information processing graph is shown. A specific example of the processing of the group G on the information processing graph is shown.

(1) Overall circuit Figure 1 shows an information processing device / system that processes information, a device / system that realizes an information processing method, a device / system that executes a computer program for information processing, and an information processing device / system. , Servers, personal computers, smartphones, mobile information terminals, Internet devices, car navigation systems, microcomputers, and the entire system circuit.

This whole circuit is input via the address / information bus line 2 in which the address signal and the information signal are transmitted in both directions, centering on the controller 1 (CPU, ROM, RAM, clock circuit, etc., which is a central processing unit). It is configured to be electrically connected to a unit 3, an output unit 4, an interface unit 5, an information storage unit 6, a program / information storage unit 7, and the like.

The input unit 3 includes a character input device such as a keyboard, a mouse, a touch panel, and a digitizer, or an image input device. From the character input device or image input device, various information of information processing, date information, sample, number of samples, factor, graph, group, number of groups, scale, separation distance, quadrant, number of quadrants, dimension, number of dimensions, The angle of the sample, the angle of the group, and the like may be input, and further, commands for processing various other information, setting / instruction information, and other information may be input.

The output unit 4 is composed of a display, a printer, and the like, and includes various information on the above information processing, date information, samples, number of samples, factors, graphs, groups, groups, scales, separation distances, quadrants, and quadrants. , Dimensions, number of dimensions, sample angle, group angle, and other various information, input screens for various information, control screens, output screens, and the like are displayed / printed on the display screen.

On the display of the output unit 4, the sample, the number of samples, the factor, the graph, the number of groups, the number of groups, the scale, the separation distance, the quadrant, the number of quadrants, the dimension, the number of dimensions, the angle of the sample, the information of the angle of the group individually, and others. Information is displayed. The display magnification value, other values, status, and mode are also displayed. These samples, sample numbers, factors, graphs, groups, groups, scales, distances, quadrants, quadrants, dimensions, dimensions, sample angles, group angles are non-display printers, holograms, and three-dimensional images. It may be displayed on a display or the like.

The interface unit 5 includes a LAN, the Internet, a transmission / reception circuit for connecting an external device, and the like, and is connected to an external device including an external input device and an external output device. Through this interface unit 5, various information of the above information processing, date information, sample, sample number, factor, graph, group, group number, scale, separation distance, quadrant, quadrant number, dimension can be obtained from other personal computers. In addition to various information such as the number of dimensions, sample angle, group angle, etc., other information, commands, setting / instruction information, other information, and programs are sent and received.

The program / information storage unit 7 (internal storage medium / means) includes non-volatile memory such as flash memory, EEPROM, EPROM, and CD-ROM, RAM, cache memory, hard disk, magnetic card, magnetic disk, and CD-RAM. It is composed of two types of storage devices (storage means), which are roughly classified as a temporary storage memory using a DVD (digital video disk) or the like.

The ROM of the program / information storage unit 7 stores general-purpose information used for programs and various processes executed by an information processing device (whole circuit) that processes information. This program corresponds to the flowchart described later. Further, the temporary storage memory of the program / information storage unit 7 is used as a temporary storage of information during program processing, that is, as a working memory or a cache memory, and also stores flags, modes, and the like.

The information storage unit 6 (external storage auxiliary medium / means, auxiliary storage medium) includes a magnetic disk, an optical disk, a flexible disk (propy disk), a magnetic drum, a magnetic tape, a memory card, a CD-RAM, and a CD-ROM. It is composed of a large amount of information storage / storage media such as a DVD and an information writing / reading device (drive device, interface circuit) for those media.

In some cases, the information processing device that processes information is a dedicated device in which a single process is performed by a program stored in advance in the ROM of the information storage unit 6, and the program stored in the information storage unit 6 is used. As appropriate, it is a general-purpose computer device that is called into the RAM of the program / information storage unit 7 and started. In this case, the information storage unit 6 stores each program including the basic software (operating software), the system software, and the application program executed by the controller 1.

Then, when the power of the information processing device for processing information is turned on, or when a new program is started, the information is read from the information storage unit 6 and taken in, and temporarily stored in the RAM or the like of the program / information storage unit 7. It is saved and executed by controller 1. That is, the program executed by the information processing apparatus that processes the information is stored in the information storage unit 6, and is appropriately read and fetched and executed by the controller 1.

In this case, when changing to a new program, the program itself stored in the information storage unit 6 is replaced, or the CD-ROM, magnetic disk, flexible magnetic disk, etc. in which the program is stored are replaced. It is done by (installing / transferring).

Further, in the case of a computer device that exclusively executes the program stored in the EPROM of the program / information storage unit 7, the CD-ROM, magnetic disk, flexible magnetic disk, etc. are replaced (installation / transfer) in the same manner. It is achieved by rewriting the EPROM or the like in response to the above, or by exchanging the ROM itself.

In some cases, the new replacement of this program is sent from an external device such as another computer connected via an Internet line, a telephone line, a public line, or a LAN line through the interface unit 5. That is, the program executed by the computer device (whole circuit) is directly exchanged with the information storage unit 6 or exchanged with the one sent from an external device in Japan or overseas via a communication network such as the Internet. It is done.

Therefore, in some cases, a program executed in the information processing device (whole circuit) that processes information is sent from an external device in Japan or overseas every time it is started. That is, the application program is not stored in the information storage unit 6 of the information processing device that processes information, and only the role of the terminal device that executes the program sent from the external device in Japan or overseas is to store the information of the present invention. It is given to the information processing device (whole circuit) to be processed. Further, the program and the information processed by the program are sent together from the external devices in Japan and overseas, and after the information processing is performed by the program, the result is sent back to the external devices in Japan and abroad.

In addition, the program executed by the information processing device (computer device) that processes information is a program by a general computer system such that an application program is executed based on a basic program such as an OS in some cases. , Is a dedicated computer system in which only one unified program including the OS is executed. Therefore, those unified programs are given from domestic and foreign external devices, and only a part of the application program or the basic program or the program is sent from the domestic and foreign external devices and executed.

In some cases, another operating system, system program (OS), and other programs are stored in advance in the information processing device (computer device) of this information processing, and the above programs are executed together with these OSs and other programs. .. When this program is installed and executed on the device (computer body), the processing described in each claim of the claims (claims) or the effects of other inventions at the end, together with another program or alone. -Execute the function.

In some cases, a part or all of this program is stored and executed in one or more domestic and foreign single or multiple separate devices other than this device, and a communication means is provided between this device and another device. The information to be processed / the information already processed / the program is transmitted / received through the device, and the present invention is executed as a whole of the present device and another device. These communication means are an internet communication system and the like.

(2) Program / information storage unit 7 (information storage unit 6)
FIG. 2 shows the storage contents of the program / information storage unit 7 (information storage unit 6). In this program / information storage unit 7 (information storage unit 6), a plurality / large number of samples, a number of samples, a factor, a graph, a group, a number of groups, a scale, a separation distance, a quadrant, a number of quadrants, a dimension, a number of dimensions, and a sample. The angle of and the angle of the group are memorized.

A large number of samples S1 (X1, Y1), S2 (X2, Y2), S3 (X3, Y3), S4 (X4, Y4), S5 (X5, Y5), ... Factors X1, X2, X3, X4, X5, ..., Second factors Y1, Y2, Y3, Y4, Y5, ... Are stored in association with each other.

Each sample is plotted on a graph with factors such as the first and second factors, and the first and second factors are scaled in a predetermined arrangement on the X-axis and axes on the graph. The scale represents the X-axis and the scale or scale of the axis, and represents an integer axis, a natural number axis, a logarithmic axis, and the like.

The number of samples is the total number of the large number of samples S. The graph is data showing the graph, and is, for example, orthogonal coordinates, polar coordinates, the number of axes (number of dimensions), and the like. The separation distances D1, D2, D3, D4, ... Y5), ... Indicates the distance between each other.

Each sample S1 (X1, Y1), S2 (X2, Y2), S3 (X3, Y3), S4 (X4, Y4), S5 (X5, Y5), ... It is divided into a plurality of groups G and grouped together, and the G1, G2, G3, ..., Refer to the group G. The number of groups is the total number of these plurality of groups.

The number of groups is smaller than the number of samples, and is usually 1/2 to 1/10000, for example, about 1/5 to 1/50, which is the optimum number for analyzing the distribution state. Further, one sample does not belong to a plurality of groups, but usually belongs to only one group.

The quadrant represents each part where the space is divided by the coordinate axes of the above graph. The above factors can also serve as this quadrant. The factors X and Y are plus + or minus-, and the quadrant to which the sample belongs becomes the first quadrant, the second quadrant, the third quadrant, and the fourth quadrant. Is represented. The quadrant number indicates the number of this quadrant.

The dimension represents the degree of expansion of the graph and corresponds to the coordinate axes. The number of dimensions indicates the number of this dimension. For the above X-axis and axis, the number of dimensions is 2, and for the X-axis, Y-axis and Z-axis, the number of dimensions is 3, and for the X-axis only, the number of quadrants is 8. The number of dimensions is 1 and the number of quadrants is 2.

The angles As1, As2, As3, ... Of the samples S indicate the angles formed by each sample S with respect to a predetermined reference, for example, the X-axis, and each takes a value of 0 degrees to 360 degrees. The angles Ag1, Ag2, Ag3, ... Of the group G are angles indicating the spread of each group G1, G2, G3, ... Seen from a predetermined reference, for example, the origin O.

As the number of axes increases, so does the number of dimensions and the number of quadrants. A set of such samples, sample counts, factors, graphs, groups, groups, scales, distances, quadrants, quadrants, dimensions, dimensions, sample angles, group angles can be further different samples, factors, Multiple sets are stored in multiple layers for each graph. The above graph does not need to be actually drawn and may be data for drawing.

(3) Overall processing FIG. 3 shows a flowchart of overall processing (information processing) executed by the controller 1. The process of FIG. 3 is started by turning on the power or a predetermined process of executing the program, first, the initialization process is executed (step 11), the error check of the information storage unit 6 and the program / information storage unit 7, and other initializations are performed. Is done.

Next, the input information such as sample S, number of samples, factor, graph, group G, number of groups, scale, separation distance D, quadrant, number of quadrants, dimension, number of dimensions, sample angle As, group angle Ag, etc. Input processing such as writing or input is executed (step 12), group processing such as grouping or grouping of the above samples is executed (step 13), and output processing such as reading or output of the processed information is executed. (Step 14), and thereafter, this input process and output process are repeated until the power is turned off.

(4) Group processing (step 13)
4 and 5 show flowcharts (first half and second half) of the group processing in step 13. First, a large number of samples S1 (X1, Y1), S2 (X2, Y2), S3 (X3, Y3), S4 (X4, Y4), S5 (X5, Y5), ..., Are taken in (step 21). In this sample S, one sample S is plotted on a graph with at least two factors X and Y, and the factors X and Y are scaled on the X-axis and Y-axis on the graph in a predetermined arrangement. be.

These large numbers of samples S are measured in advance or sent from others. A large number of captured samples S1 (X1, Y1), S2 (X2, Y2), S3 (X3, Y3), S4 (X4, Y4), S5 (X5, Y5), ... , D2, D3, D4, ... Are sequentially obtained and output (step 22), written to the program / information storage unit 7 (information storage unit 6), and input. This separation distance D can be obtained from each factor X, ... Using, for example, the Pythagorean theorem / the three-square theorem.

In this input, the separation distances D1, D2, D3, D4, ... Are the samples S1 (X1, Y1), S2 (X2, Y2), S3 (X3, Y3), S4 (X4, Y4) at both ends. It is associated with S5 (X5, Y5), .... Then, whether the quadrants of the graphs to which the samples at both ends of this separation distance D belong are different, that is, plus +/- minus (positive / negative) of one factor X, Y and the other factor X, Y at both ends. It is determined whether they are different (step 23).

If they are different, both ends of the separation distance D are in different quadrants. Therefore, the separation distance D is increased, for example, 1.5 times, 2 times, 3 times, ... Processing is performed and weighted (step 24). As a result, the separation distance D of the two samples spanning different quadrants is corrected to be larger than the actual distance with respect to the separation distance in which both ends are in the same quadrant, and it becomes difficult to belong to the same group.

If the distance D before correction in steps 23 and 24 is the smallest in the whole, and the samples at both ends of the distance D are in different quadrants, the correction process in steps 23 and 24 should not be performed. You may do it. That is, if the samples at both ends having the smallest corrected distance D in steps 23 and 24 are in different quadrants, the correction process of steps 23 and 24 is not performed, and only in this case, the distance D before correction is performed. You may return to.

If they are the same, since both ends of the separation distance D are in the same quadrant, the separation distance D is not changed and the weighting correction process is not performed (step 25). As a result, the separation distance D of the two samples whose ends belong to the same quadrant is not corrected, and it becomes easy to belong to the same group. If the attribution quadrants in step 24 are different, the separation distance D is not corrected, and the separation distance D is reduced if the attribution quadrants are the same in step 25 with respect to the separation distances in different quadrants at both ends. The correction process may be executed.

Next, from the shortest separation distances D1, D2, D3, D4, ..., Sorting / ordering / ordering (step 27), the sample Si, Si + 1 at both ends having the shortest separation distance Di If they do not belong to any of the groups G1, G2, G3, ... (Step 28), both samples are assigned to one new unused group G1, G2, G3, ... (Step 30).

For the next long separation distance Dj, if one of the samples Sj and Sj + 1 Sj at both ends related to the separation distance Dj already belongs to any of the groups G1, G2, G3, ... (Step 28), the group G1 ,. The sample Sj + 1, which does not belong to G2, G3, ..., Is the same group G1, G2, G3 as the group G1, G2, G3, ... , ..., to be attributed to (step 32).

Further, if both the samples Sk and Sk + 1 related to the next longer separation distance Dk already belong to any of the groups G1, G2, G3, ... (Step 28), the sample Sk and Sk + 1 are included in any of the groups. It is not attributed to G1, G2, G3, ... (Step 34).

The processes of steps 28 to 34 are repeated together with the following processes (steps 38 and 40), and each sample S is sequentially assigned to any of the groups G1, G2, G3, ... In order of decreasing distance D. go. Here, again, if both ends do not belong to any of the groups G1, G2, G3, ... (Step 28), both samples S are transferred to one new unused group G1, G2, G3, .... Attributed (step 30).

Then, the angles A1, A2 of the samples of the large number of samples S1 (X1, Y1), S2 (X2, Y2), S3 (X3, Y3), S4 (X4, Y4), S5 (X5, Y5), ... , A3, A4, ... Are sequentially obtained and output (step 36), written to the program / information storage unit 7 (information storage unit 6), and input.

The angles As1, As2, As3, ... Of the samples S indicate, for example, the angles formed by each sample S with respect to the X-axis, and take a value of 0 degrees to 360 degrees. The angle As of this sample S is obtained, for example, from the values of each factor X and Y of each sample S by using the trigonometric function arctan / arctangent, arctan / arctangent, arcsin / arcsine, arccos / arctangent, etc. May be used as it is.

Next, with respect to the groups G1, G2, G3, ..., Which were changed in steps 30 and 32 above, the angles Ag1, Ag2 of the group G1, G2, G3, ... , Ag3, ..., Are required (step 37). This is obtained by subtracting the minimum value from the maximum value of the above angles As1, As2, As3, ... Of each sample belonging to the group G1, G2, G3, ....

Then, when the angle Ag of the attribution group G becomes equal to or larger than a predetermined angle (step 38), the attribution process to the group in steps 28 to 34 is stopped or terminated, and the sample S of the group G now assigned is used. Exclude from the above group G, create a new group G, and assign this sample S to the group G (step 39).

As a result, the number of samples S belonging to one group G does not increase too much, or even if the number of samples S belonging to one group G is small, the variation spreads and the analysis is hindered. It will not be lost. Further, if the angle Ag of the group G is not affected, the number of samples of the group G is increased.

The sample S belonging to this new group G can be either both ends or one of the separation distances D. The predetermined angle at which the predetermined angle or more is determined is set and input by the operator in advance, or is stored in advance. For example, 45 degrees, 60 degrees, 90 degrees, 135 degrees, 180 degrees, 225 degrees, 270 degrees, 315 degrees, 360 degrees and the like.

If the angle is not equal to or larger than the predetermined angle in step 38, the process of belonging to the group in steps 28 to 34 is continued. The reference points of the angles Ag1, Ag2, Ag3, ... Are not limited to the origin O, and may be any point on the graph. It may be a point.

The reference axis of the angles As1, As2, As3, ... Is not limited to the X axis, but may be the Y axis, and the possible values may be other values such as −180 degrees to +180 degrees, and other values on the graph. It may be any line or axis, and it may be an outside, top, bottom, front, back, left, right line or axis in addition to the inside of the graph.

Next, if the number of the group G becomes a plurality and the number of the group G in which the sample S is put together is made up to a predetermined number, for example, a predetermined ratio of the sample number, for example, 1/10, that is, 1/10 of the sample number. If the number is reduced to 10 (step 40), the attribution process to the group G in steps 28 to 34 is stopped or terminated. The predetermined number 1/10 may be another value as described above, may be set in advance by the operator and input, or may be predetermined and input by the operator or the like and stored.

Then, if a sample S that does not belong to any group G remains (step 41), a new group G is assigned to each of the remaining samples S (step 42). Thus, all samples S belong to any group G. There is also a group G that has only one sample S. However, such a group G having only one sample S may be omitted. When the number of groups is too large or too small than desired, the number of groups allowed in step 40, the angle allowed in step 38, and the magnitude of the weighting of the modification in step 24 may be changed.

In reality, it is almost impossible, but if the number of groups is "1" and is not plural in step 40, the number of groups is increased by returning to steps 28 to 34. In reality, it is almost impossible, but if the number of groups remains "1" and there is no sample S remaining that does not belong to any group G as in step 41 above, the number of groups remains "1". Is terminated.

The graph is two-dimensional, but may be three-dimensional, four-dimensional, five-dimensional, six-dimensional, seven-dimensional, or eight-dimensional. The number of the above quadrants increases accordingly. The group G and the separation distance D may span a plurality of different quadrants such as two, three, and so on. In this way, in steps 40, 38, etc., the group G becomes a plurality, and the attribution process to the group G is stopped or terminated under predetermined conditions.

(5) Specific Examples FIGS. 6 and 7 show specific examples of the above information processing. This information shows the shape of the human face, and the length from the eyebrows to the mouth of this face is called the "vertical length of the face", and the line of the mouth is extended to the left and right between the points at both ends that intersect the contour of the face. Let the length of be "the horizontal length of the face".

The ratio of the "vertical length of the face" to the "horizontal length of the face" is defined as an egg shape, a round shape, a face length, a triangle, a quadrangle, or the like. For example, the ratio is egg type 1: 1, round type 2: 3, surface length 3: 2, triangle 3: 3, and quadrangle 1: 2. As shown in FIG.

In FIG. 7, for these egg-shaped, round-shaped, face-length, triangular, and square face shapes, "reduce the volume of the width", "create volume around the chin", and "moderate the volume of the top" for each hairstyle. "Volume on the back" "Volume on the side" "Curl the hair around the face" "Volume on the top" "Hide the cheeks with the hair on the side" "Create space on the forehead" "Make bangs" Is associated with it. Furthermore, these are associated with satisfaction levels of "quite dissatisfied", "slightly dissatisfied", "slightly dissatisfied", "normal", "slightly satisfied", "slightly satisfied", and "quite satisfied".

The egg-shaped face group (cluster) corresponds to the group (cluster) of "reducing the volume of the width". The round face group (cluster) corresponds to the "hide cheeks with side hair" group (cluster). The face group (cluster) corresponds to the group (cluster) of "moderate the volume of the top", "make the volume of the side", "curl the hair around the face", and "make the bangs". Other triangular faces and quadrangular faces also correspond to some of the above groups (clusters).

8, 9, and 10 show specific examples on the graph of the information processing. In FIG. 8, the separation distance D between the two samples S1 and S2 is not large, but the angles As1 and As2 of the samples S1 and S2 seen from the origin O are considerably different, so that the attribution group G of both samples S1 and S2 is They will be in different groups G1 and G2.

In FIG. 9, the samples S3, S4, S5, and S6 straddle different quadrants, but the angles As3, As4, As5, and As6 of the samples S3, S4, S5, and S6 seen from the origin O are not so different. Since they are close to each other, the attribution group G of each sample S3, S4, S5, S6 is the same group G3.

In the example of FIG. 9, when the weighting correction when the quadrants are different is small, the samples S3, S4, S5, and S6 belong to the same group, but when the weighting correction when the quadrants are different is large, the sample. S3, S4 and samples S5, S6 may belong to different groups.

In FIG. 10, among the samples S7, S8, and S9, the samples S7 and S8 belong to the same quadrant, and only the sample S9 belongs to a different quadrant. If the correction amount of the correction process in step 24 is large and the predetermined angle allowed for one group G4 in step 38 is small with respect to the separation distance D of both ends S8 and S9 having different quadrants, the sample S9 is sample S7, S8. Belongs to a different group G.

If the correction amount of the correction process of the step 24 is small and the predetermined angle allowed for one group G4 of the step 38 is large with respect to the separation distance D of both ends S8 and S9 having different quadrants, the sample S9 is the samples S7 and S8. Belongs to the same group G4 as. As described above, the attribution group G and the number of groups are changed even in the same sample S depending on the correction ratio in step 24 and the size of the predetermined angle allowed for one group G in step 8.

11 and 12 show specific examples of processing of group G on the information processing graph. In the example of FIG. 11, the above "vertical length of the face" and "horizontal length of the face" are scaled on the X-axis and the Y-axis, and the face is oval, round, face length, triangular, and quadrangular. The shapes are plotted on the graph as samples S11 to S15.

In the example of FIG. 11, the weighting correction process when the quadrants are different is increased. Therefore, only the samples S14 and S15 in the same quadrant are set to the same group G11, and the other groups G12, G13 and G14 are formed in each quadrant.

In the example of FIG. 12, the above "vertical length of the face" and "horizontal length of the face" are scaled on the X-axis and the Y-axis, and the hairstyle "reduces the volume of the width" and "volume around the chin". "Put out the volume of the top" "Put the volume on the back" "Put the volume on the side" "Curl the hair around the face" "Put the volume on the top" "Hide the cheeks with the side hair" " "Making a space on the forehead" and "Making bangs" are plotted on the graph as samples S21 to S30.

In the example of FIG. 12, the weighting correction process when the quadrants are different is increased. Therefore, the samples S22, S24, S27, S29 and S23, S25, S26, S30 in the same quadrant are set to the same groups G22 and G21, respectively, and the other groups G23 and G24 are formed in each quadrant. Orthant.

(6) Other Embodiments The present invention is not limited to the above embodiments and can be variously modified. For example, the above step 38 may be changed as follows. For example, if the size of the separation distance D or the group G to which the group G belongs becomes a predetermined value or more other than the predetermined angle (step 38), the process of attribution to the group in steps 28 to 34 is stopped or terminated. The sample S of the group G to which it has just been assigned is excluded from the group G, a new group G is created, and this sample S is assigned to the group G (step 39).

Here, the predetermined value determined in the changed step 38 may be set and input in advance by the operator, or may be stored in advance, and may be stored in each group or the magnitude of the separation distance D. It may be a predetermined ratio with respect to.

This may be a predetermined ratio of the maximum value in the separation distance D, for example, 1/2, 2/3, 1/3, 3/4, 1/4, ..., Etc., or the separation distance D in the group. A predetermined ratio of the maximum value of, for example, 1/2, 2/3, 1/3, 3/4, 1/4, etc. may be used. The maximum value may be the same as the average value, the mode value, the minimum value, 1.5 times, 2 times, 2.5 times, 3 times, and the like.

The positive value part on the X axis is the first quadrant, the negative value part on the X axis is the third quadrant, the positive value part on the Y axis is the second quadrant, and the negative value part on the Y axis is the fourth quadrant, the origin. Assuming that the sample S on O belongs to the first quadrant, the processes of steps 23 to 25 are executed. It should be noted that these attribution quadrants are not limited to the above, and may be other quadrants.

The positive value part on the X-axis, the negative value part on the X-axis, the positive value part on the Y-axis, the negative value part on the Y-axis, and the sample S on the origin O are the fifth quadrant and the fifth quadrant different from the above four quadrants. The correction process of step 24 may be executed as the 6th quadrant, the 7th quadrant, the 8th quadrant, and the 9th quadrant. In this case, the content of the weighting correction may be different for each different quadrant, may be the same as one of the adjacent quadrants, or may be predetermined, for example, a value of 1/2 of the weighting correction between the two adjacent quadrants. It may be the value of, and this is the same for the origin O.

The information processing device, information processing method, and computer program for information processing may be executed by devices and programs distributed in a plurality of places, a plurality of countries, and domestically and internationally, and in this case, each distributed device. The system is connected through an internet line and other communication lines to send and receive information.

In response to this, the information storage unit 6 and the program / information storage unit 7 may be divided and separated into a plurality of parts, and the information / information stored in these storage means may be distributed and stored in a plurality of places. , May be stored in the program according to the above flowchart.

The sample S of the above information was a human face type and hair type, but is not limited to this, the distance between the eyes, the distance between the ears, the distance from the eyes to the mouth, the distance from the eyes to the nose, and the distance from the nose. Distance to mouth, arm / hand length / thickness / width, leg / foot length / thickness / width, bust, waist, hips, head circumference, height, weight, body fat ratio, BMI, Muscle mass, basal metabolism, body age, actual age, non-human animals, insects, their behavior, changes, vehicles such as automobiles, motorcycles, ships, aircraft, writing tools, stationery such as erasers, personal computers, electricity such as TVs Products, paintings, calligraphy, books, figurines, furniture, other objects, structures, real estate, etc., these actions, changes, etc. can be anything.

The predetermined angle at which the predetermined angle or more is determined in step 38 is determined in advance, such as the number of groups G converged in step 40 being calculated from some of the above values and being obtained, or being input in advance by the operator. You may go. For example, 1 degree to 359 degrees, 1 piece to 100 pieces, and the like.

The information processing (input processing / output processing / group processing / writing processing / reading processing) of each step of FIGS. 3 to 5 is not executed in the whole processing, but is performed by an administrator or the like by an interrupt processing. It may be executed when an input, output, write, or read command is input, or it may be executed when it is accessed from a contact or the like via an internet line or the like by an interrupt process.

The processing of steps 41 to 42 may be omitted. As a result, the group G having only one sample S is omitted. Even if this omission is made, the distribution state of a large number of samples S is optimally analyzed / analyzed.

The process of steps 23 to 25 may be omitted. As a result, the distribution state of a large number of samples S is optimally analyzed / analyzed regardless of the quadrant. The processing of steps 38 to 39 may be omitted. As a result, the distribution state of a large number of samples S is optimally analyzed / analyzed regardless of the angle of the group G.

The process of step 40 may be omitted. As a result, the distribution state of a large number of samples S is optimally analyzed / analyzed regardless of the number of groups G. In step 32 above, it may be assigned to a new unused group G. In step 34 above, both ends or one may be assigned to a new unused group G.

The order of the respective processes of steps 11 to 42 may be partially or wholly replaced, some or all may be omitted, or each step may be replaced with another equivalent process. , AI (artificial intelligence), change table, hardware or circuit that exerts the same function can be partially or completely replaced / replaced, and some may be artificially executed within an even range, or divided and separated. The number may be increased or decreased, some or all of these may be omitted, and each process may be singular, plural, three or more, or two or three of these. One or more may be combined or integrated and used together.

The controller 1, the address / information bus line 2, the input unit 3, the output unit 4, the interface unit 5, the information storage unit 6, the program / information storage unit 7, and the steps 11 to 14 and 21 to 42 may be singular or plural. It may be three or more, a part or all of each of them may be omitted, or the number may be increased or decreased by being divided and separated, and the shape, arrangement form, and information form may be arbitrarily changed. It is possible and may be replaced with other equivalents, or two or three or more of these may be combined or integrated and used together.

The controller 1, the information storage unit 6 and the program / information storage unit 7 are combined or separated into three or more, or the input unit 3, the output unit 4 and the interface unit 5 are combined or separated into three or more. Or, some or all of the above-mentioned systems, processes, circuits, programs, software, and hardware may be executed by AI (artificial intelligence).

The processing device of the entire circuit of FIG. 1, the controller 1, the address / data bus line 2, the input unit 3, the output unit 4, the interface unit 5, the information storage unit 6, the program / information storage unit 7, and FIGS. 3 to 5. Each step may be singular, plural, or three or more, and some or all of these configurations, actions, actions, processes or functions may be omitted, or may be divided and separated to increase the number. It may be reduced, its shape may be changed arbitrarily, it may be replaced with another equivalent material, or two or three or more of these may be combined or integrated. In addition to the ones shown in the figure, the orientation may be 90 degrees, 0 degrees to 180 degrees, up / down, left / right, front / back, front / back, front, back, tilted / turned / inverted, and relative to each other. The target size may be switched, may be the same, may be partially or completely omitted, the shape may be flat, curved, uneven, or partially or completely omitted. The material may be resin, metal, wood, bamboo, glass, cotton, cloth, thread, fiber, rubber, paper, ceramic, carbon, a composite of these / It may be made of a mixture / multilayer laminate, a rigid urethane, or a composite / mixture / multilayer laminate thereof.

(7) Effects of other inventions [1] A computer is made to plot one sample on a graph with at least two factors for a large number of samples, and each factor is arranged in a predetermined arrangement on the axis on the graph. Scale,
The distance between the samples is calculated, and the samples related to the distance are assigned to one group from the one with the shortest distance, and one of the samples related to the distance already belongs to one of the groups. If so, the sample that does not belong to the group is attributed to the group related to the sample belonging to the group, and if both of the samples related to the separation distance already belong to any group, the sample is assigned to any group. An information processing method characterized in that the above-mentioned group becomes a plurality of groups and is stopped or terminated under a predetermined condition without being attributed to a group.

[2] For a large number of samples, one sample is plotted on a graph with at least two factors, and each factor is scaled on an axis on the graph in a predetermined arrangement, and the distance between the samples is separated from each other. And if the sample related to the separation distance is assigned to one group from the one with the short separation distance and one of the samples related to the separation distance already belongs to one of the groups, the sample is assigned to the group. A process in which a sample that does not belong is attributed to a group related to a sample belonging to a group, and if both of the samples related to the separation distance already belong to one of the groups, the sample is not attributed to any group. And, a computer program for information processing that causes a computer to execute a process of stopping or terminating this attribution process by having a plurality of the above groups and under predetermined conditions.

[3] For a large number of samples, one sample is plotted on a graph with at least two factors, and each factor is scaled on an axis on the graph in a predetermined arrangement. If the separation distance means for obtaining the distance and the sample related to the separation distance are assigned to one group from the one having the short separation distance, and one of the samples related to the separation distance already belongs to one of the groups. , A sample that does not belong to a group is attributed to the group that belongs to the sample that belongs to the group, and if both of the samples that belong to the separation distance already belong to one of the groups, the sample is assigned to any group. An information processing apparatus characterized in that the attribution means not to be attributed and the stop-end means for stopping or terminating the attribution process of the attribution means in a plurality of groups and under predetermined conditions.

[4] For each of the above separation distances, if the samples related to the separation distance belong to different quadrants in the above graph, the separation distance is made larger than the actual distance or is in the same quadrant for the separation distance belonging to the same quadrant. The information processing apparatus according to claim, wherein if it belongs, a correction process is performed to make the separation distance smaller than the actual distance with respect to the separation distance belonging to a different quadrant. This allows each sample to be grouped along the quadrants of the graph.

[5] A claim characterized in that, when the angle formed by the group is equal to or greater than the predetermined angle when viewed from the origin of the graph or the predetermined position of the graph, the attribution process to the group by the attribution means is stopped or terminated. Item 4. The information processing apparatus according to item 4. As a result, each sample is not concentrated and biased in one group.

[6] The information according to claim 5, wherein a new group is created for the sample for which the attribution process to the group by the attribution means has been stopped or terminated, and the sample is assigned to the group. Processing equipment. As a result, the angle of a certain group becomes large, and even if there is a sample that is out of alignment, it can be attributed to one of the groups.

[7] A claim characterized in that it is determined whether or not the number of the above groups has decreased to a predetermined ratio of the number of samples, and if the number has decreased to a predetermined ratio, the attribution processing of the attribution means is stopped or terminated. Item 6. The information processing apparatus according to Item 6. As a result, the distribution state can be converged to the number of groups that can be optimally analyzed / analyzed.

[8] In the attribution process by the attribution means, when the separation distance becomes larger than a predetermined ratio with respect to the size of each group among the plurality of groups created, or the size of the group becomes larger. The information processing apparatus according to claim 7, wherein when the value becomes larger than a predetermined value, the attribution process of the attribution means is stopped or terminated. This makes it possible to exclude samples in which the separation distance or a specific group is large and the group or distribution state is significantly changed.

[9] The information processing apparatus according to claim 8, wherein the group assigned by the attribution means spans different quadrants of the graph. As a result, samples that cross quadrants, which are necessary for analysis / analysis in some cases, can be grouped in the same group, and the distribution state can be optimally analyzed / analyzed.

[10] The information processing apparatus according to claim 9, wherein the graph is two-dimensional, three-dimensional, four-dimensional, five-dimensional, or six-dimensional, and the number of the quadrants increases accordingly. This makes it possible to analyze / analyze a multidimensional sample with many factors.

The distribution state of a large number of samples can be optimally analyzed / analyzed. Of each separation distance D of each sample S, only the separation distance straddling the quadrant on the graph is largely modified (steps 23 and 24), and the following processing is performed in order from the shortest separation distance.

If the samples at both ends of the distance do not belong to any group G, the samples at both ends belong to a new group, and if only one belongs to one of the groups, the samples at the other end belong to the group. If both belong to some group and both belong to some group, no new group allocation is performed (steps 28 to 34).

If the angle of a certain group becomes large, it is not assigned to the group, but it is assigned to another new group (steps 38 and 39), and if the number of allotted groups exceeds a predetermined value, it is assigned to each of the remaining samples S. Assign groups (steps 40-42).

1 ... Controller 2 ... Address / Information bus line 3 ... Input unit 4 ... Output unit 5 ... Interface unit 6 ... Information storage unit 7 ... Program / Information storage unit

Claims (11)

Against the computer
For a large number of samples, one sample is plotted on a graph with at least two factors, and each factor stores information scaled in a predetermined arrangement on an axis on the graph in a scale storage means .
The stored information of this scale storage means is read out to obtain the mutual separation distance of each sample.
From the one with the shortest separation distance, the samples related to the separation distance are assigned to one group, and if one of the samples related to the separation distance already belongs to one of the groups, the samples not belonging to the group are selected. If it is attributed to the group related to the sample belonging to the group and both of the samples related to the separation distance already belong to one of the groups, the sample is not attributed to any group and is related to the separation distance. If both of the samples do not belong to any group, the sample is attributed to another new group, and the result of these attribution processes is stored in the attribution storage means.
When viewed from the origin of the graph or a predetermined position of the graph, the angle formed by the group related to the result of the attribution processing stored in the attribution storage means is equal to or larger than the predetermined angle, and the result of the attribution processing stored in the attribution storage means. The information processing method is characterized in that, when the group to which the sample is assigned becomes a plurality of groups, the attribution process is stopped or terminated. For a large number of samples, one sample is plotted on a graph with at least two factors, and each factor stores information scaled in a predetermined arrangement on an axis on the graph in a scale storage means .
The process of reading the stored information of this scale storage means to obtain the mutual separation distance of each sample, and
From the one with the shortest separation distance, the samples related to the separation distance are assigned to one group, and if one of the samples related to the separation distance already belongs to one of the groups, the samples not belonging to the group are selected. If it belongs to the group related to the sample belonging to the group and both of the samples related to the separation distance already belong to one of the groups, the sample does not belong to any group and is related to the separation distance. If both of the samples do not belong to any group, the process of assigning the sample to another new group, the process of storing the result of these attribution processes in the attribution storage means, and the process of storing the results of these attribution processes in the attribution storage means.
When viewed from the origin of the graph or a predetermined position of the graph, the angle formed by the group related to the result of the attribution processing stored in the attribution storage means is equal to or larger than the predetermined angle, and the result of the attribution processing stored in the attribution storage means. Therefore, if the group to which the above sample belongs becomes a plurality of groups, a computer program for information processing that causes a computer to execute a process of stopping or terminating the attribution process . For a large number of samples, one sample is plotted on a graph with at least two factors, and each factor is a scale storage means for storing information scaled in a predetermined arrangement on an axis on the graph.
A separation distance means that reads out the storage information of this scale storage means and obtains the mutual separation distance of each sample.
From the one with the shortest separation distance, the samples related to the separation distance are assigned to one group, and if one of the samples related to the separation distance already belongs to one of the groups, the samples not belonging to the group are selected. If it is attributed to the group related to the sample belonging to the group and both of the samples related to the separation distance already belong to one of the groups, the sample is not attributed to any group and is related to the separation distance. An attribution storage means for storing the results of these attribution processes, which assigns the sample to another new group if both of the samples do not belong to any group .
When viewed from the origin of the graph or a predetermined position of the graph, the angle formed by the group related to the result of the attribution processing stored in the attribution storage means is equal to or larger than the predetermined angle, and the result of the attribution processing stored in the attribution storage means. The information processing apparatus is provided with a stop / end means for stopping or terminating the attribution process when the number of groups to which the sample belongs becomes a plurality of groups. Against the computer
For a large number of samples, one sample is plotted on a graph with at least two factors, and each factor stores information scaled in a predetermined arrangement on an axis on the graph in a scale storage means .
The stored information of this scale storage means is read out to obtain the mutual separation distance of each sample.
From the one with the shortest separation distance, the samples related to the separation distance are assigned to one group, and if one of the samples related to the separation distance already belongs to one of the groups, the samples not belonging to the group are selected. If it is attributed to the group related to the sample belonging to the group and both of the samples related to the separation distance already belong to one of the groups, the sample is not attributed to any group and is related to the separation distance. If both of the samples do not belong to any group, the sample is attributed to another new group, and the result of these attribution processes is stored in the attribution storage means.
When the group to which the sample belongs becomes a plurality of groups, and the separation distance becomes larger than a predetermined ratio with respect to the size of each group among the plurality of groups, or the size of the group is predetermined. An information processing method characterized by stopping or terminating the attribution process when the value becomes larger than the value . For a large number of samples, one sample is plotted on a graph with at least two factors, and each factor stores information scaled in a predetermined arrangement on an axis on the graph in a scale storage means .
The process of reading the stored information of this scale storage means to obtain the mutual separation distance of each sample, and
From the one with the shortest separation distance, the samples related to the separation distance are assigned to one group, and if one of the samples related to the separation distance already belongs to one of the groups, the samples not belonging to the group are selected. If it belongs to the group related to the sample belonging to the group and both of the samples related to the separation distance already belong to one of the groups, the sample does not belong to any group and is related to the separation distance. If both of the samples do not belong to any group, the process of assigning the sample to another new group, the process of storing the result of these attribution processes in the attribution storage means, and the process of storing the results of these attribution processes in the attribution storage means.
When the group to which the sample belongs becomes a plurality of groups, and the separation distance becomes larger than a predetermined ratio with respect to the size of each group among the plurality of groups, or the size of the group is predetermined. A computer program for information processing that causes a computer to execute a process of stopping or terminating the above attribution process when the value becomes larger than the value . For a large number of samples, one sample is plotted on a graph with at least two factors, and each factor is a scale storage means for storing information scaled in a predetermined arrangement on an axis on the graph.
A separation distance means that reads out the storage information of this scale storage means and obtains the mutual separation distance of each sample.
From the one with the shortest separation distance, the samples related to the separation distance are assigned to one group, and if one of the samples related to the separation distance already belongs to one of the groups, the samples not belonging to the group are selected. If it is attributed to the group related to the sample belonging to the group and both of the samples related to the separation distance already belong to one of the groups, the sample is not attributed to any group and is related to the separation distance. An attribution storage means for storing the results of these attribution processes, which assigns the sample to another new group if both of the samples do not belong to any group .
When the group to which the sample belongs becomes a plurality of groups, and the separation distance becomes larger than a predetermined ratio with respect to the size of each group among the plurality of groups, or the size of the group is predetermined. An information processing apparatus including a stop / end means for stopping or terminating the attribution process when the value becomes larger than the value . Against the computer
For a large number of samples, one sample is plotted on a graph with at least two factors, and each factor stores information scaled in a predetermined arrangement on an axis on the graph in a scale storage means .
The stored information of this scale storage means is read out to obtain the mutual separation distance of each sample.
From the one with the shortest separation distance, the samples related to the separation distance are assigned to one group, and if one of the samples related to the separation distance already belongs to one of the groups, the samples not belonging to the group are selected. If it is attributed to the group related to the sample belonging to the group and both of the samples related to the separation distance already belong to one of the groups, the sample is not attributed to any group and is related to the separation distance. If both of the samples do not belong to any group, the sample is attributed to another new group, and the result of these attribution processes is stored in the attribution storage means.
It is determined whether or not the number of groups to which the sample is attributed has decreased to a predetermined ratio, and if the number of the groups has decreased to a predetermined ratio, the attribution process is stopped or the attribution process is stopped. An information processing method characterized by termination. For a large number of samples, one sample is plotted on a graph with at least two factors, and each factor stores information scaled in a predetermined arrangement on an axis on the graph in a scale storage means .
The process of reading the stored information of this scale storage means to obtain the mutual separation distance of each sample, and
From the one with the shortest separation distance, the samples related to the separation distance are assigned to one group, and if one of the samples related to the separation distance already belongs to one of the groups, the samples not belonging to the group are selected. If it belongs to the group related to the sample belonging to the group and both of the samples related to the separation distance already belong to one of the groups, the sample does not belong to any group and is related to the separation distance. If both of the samples do not belong to any group, the process of assigning the sample to another new group, the process of storing the result of these attribution processes in the attribution storage means, and the process of storing the results of these attribution processes in the attribution storage means.
It is determined whether or not the number of groups to which the sample is assigned becomes a plurality of groups, and the number of the groups is reduced to a predetermined ratio, and if the number is reduced to a predetermined ratio, the attribution process is stopped or the attribution process is stopped. A computer program for information processing that causes a computer to perform processing to be terminated. For a large number of samples, one sample is plotted on a graph with at least two factors, and each factor is a scale storage means for storing information scaled in a predetermined arrangement on an axis on the graph.
A separation distance means that reads out the storage information of this scale storage means and obtains the mutual separation distance of each sample.
From the one with the shortest separation distance, the samples related to the separation distance are assigned to one group, and if one of the samples related to the separation distance already belongs to one of the groups, the samples not belonging to the group are selected. If it is attributed to the group related to the sample belonging to the group and both of the samples related to the separation distance already belong to one of the groups, the sample is not attributed to any group and is related to the separation distance. An attribution storage means for storing the results of these attribution processes, which assigns the sample to another new group if both of the samples do not belong to any group .
It is determined whether or not the number of groups to which the sample is attributed has decreased to a predetermined ratio, and if the number of the groups has decreased to a predetermined ratio, the attribution process is stopped or the attribution process is stopped. An information processing device provided with a stop / end means for terminating. For each of the above separation distances, if the samples related to the separation distance belong to different quadrants in the above graph, the separation distance is made larger than the actual distance with respect to the separation distance belonging to the same quadrant, or if the separation distance belongs to the same quadrant. The information processing apparatus according to claim 3, 6 or 9 , wherein a correction process is performed to make the separation distance smaller than the actual distance with respect to the separation distance belonging to different quadrants. The information processing apparatus according to claim 10 , wherein a new group is created for the sample whose attribution process to the group has been stopped or terminated, and the sample is assigned to the group.

Images