Invention title: Analyzer, control method, and program
Technical field
[0001]
　The present invention relates to data analysis.
Background technology
[0002]
　Techniques for grasping the relationship between multiple data have been developed. For example, Patent Document 1 discloses a technique for expressing the relationship between data by using a characteristic factor diagram, a pie chart, or the like. Here, the Ishikawa diagram visually displays the characteristics and the factors by showing the factors in which the trunk is extended from the characteristics (results) and grouped into the branches extended from the trunk.
Prior art literature
Patent documents
[0003]
Patent Document 1: Japanese Patent Application Laid-Open No. 2019-36061
Patent Document 2: US Patent Application Publication No. 2014/0222741A1
Outline of the invention
Problems to be solved by the invention
[0004]
　The magnitude of the effect of a factor on its characteristics can vary from factor to factor. However, the characteristic factor diagram of Patent Document 1 does not show the magnitude of the influence of each of such factors.
[0005]
　The present invention has been made in view of the above problems, and an object of the present invention is to provide a technique capable of easily grasping both the characteristics, the factors thereof, and the influence of each factor on the characteristics.
Means to solve problems
[0006]
　The analyzer of the present invention 1) acquires a relational information indicating the degree of influence of each of a plurality of explanatory variables on the objective variable, and 2) expresses the relationship between the objective variable and the explanatory variable using the relational information. It has a generation unit that generates a characteristic factor diagram.
　The generation unit determines the display mode of the display for each explanatory variable in the characteristic factor diagram or the presence or absence of the display based on the degree of influence of the explanatory variable.
[0007]
　The control method of the present invention is executed by a computer. The control method includes 1) an acquisition step of acquiring relational information indicating the degree of influence of each of a plurality of explanatory variables on the objective variable, and 2) a characteristic factor representing the relationship between the objective variable and the explanatory variable using the relational information. It has a generation step to generate a diagram.
　In the generation step, the display mode of the display for each explanatory variable in the characteristic factor diagram or the presence or absence of the display is determined based on the degree of influence of the explanatory variable.
[0008]
　The program of the present invention causes a computer to execute the control method of the present invention.
Effect of the invention
[0009]
　A technique is provided that can easily grasp both the characteristics, the factors thereof, and the influence of each factor on the characteristics.
A brief description of the drawing
[0010]
[Fig. 1] Fig. 1 is a diagram for explaining an outline of the analyzer of the present embodiment.
FIG. 2 is a diagram illustrating a functional configuration of the analyzer of the first embodiment.
[Fig. 3] Fig. 3 is a diagram illustrating a computer for realizing an analyzer.
[Fig. 4] Fig. 4 is a diagram illustrating a usage environment of an analyzer.
FIG. 5 is a flowchart illustrating a flow of processing executed by the analyzer of the first embodiment.
[Fig. 6] Fig. 6 is a diagram illustrating the structure of related information.
FIG. 7 is a diagram illustrating relationship information 50 when the relationship between an objective variable and an explanatory variable is represented by a plurality of linear models.
FIG. 8 is a diagram illustrating a characteristic factor diagram 10 in which a display mode of a factor display is determined by using the degree of influence of an explanatory variable.
FIG. 9 is a diagram illustrating a characteristic factor diagram 10 in which the presence or absence of factor display is determined by using the degree of influence of an explanatory variable.
FIG. 10 is a diagram for explaining an outline of the analyzer of the second embodiment.
FIG. 11 is a flowchart illustrating a flow of processing executed by the analyzer of the second embodiment.
[Fig. 12] Fig. 12 is a diagram illustrating a screen including both a characteristic factor diagram and a graph.
[Fig. 13] Fig. 13 is a diagram illustrating a graph showing data for each of a designated explanatory variable and an objective variable.
[Fig. 14] Fig. 14 is a diagram illustrating a graph when a plurality of factor displays are specified.
Mode for carrying out the invention
[0011]
　Hereinafter, embodiments of the present invention will be described with reference to the drawings. In all drawings, similar components are designated by the same reference numerals, and description thereof will be omitted as appropriate. Further, in each block diagram, unless otherwise specified, each block represents a functional unit configuration rather than a hardware unit configuration.
[0012]
[Embodiment 1]

　FIG. 1 is a diagram for explaining an outline of the analyzer 2000 of the present embodiment. Note that FIG. 1 is an example for facilitating understanding of the analyzer 2000, and the function of the analyzer 2000 is not limited to that shown in FIG.
[0013]
　The analyzer 2000 generates a characteristic factor diagram 10 showing the relationship between the objective variable and the plurality of explanatory variables. Here, the characteristic factor diagram is a diagram that visually shows the relationship between the characteristic and the plurality of factors. Characteristic Factors In FIG. 10, the characteristics correspond to the objective variables and the factors correspond to the explanatory variables. Here, the characteristic may be the objective variable itself or an item related to the objective variable. Items related to the objective variable are, for example, items or events of interest that are common to the objective variable. For example, there are cases where the objective variable represents "presence or absence of defective product" and the characteristic represents "defective product". Similarly, the factor may be the explanatory variable itself or an item related to the explanatory variable.
[0014]
　The characteristic factor diagram 10 has one trunk 11 connected to a display representing the characteristic (characteristic display 12), and the trunk 11 and one or more branch branches 13 are connected to each other. Each branch 13 is connected to a display (group display 14) representing a group of factors. Further, one or more twigs 15 are connected to one twig 13. A display (factor display 16) representing one factor is connected to each twig 15.
[0015]
　In the example of FIG. 1, the objective variable is "presence or absence of defective product" and the characteristic is "defective product". Therefore, the characteristic display 12 indicates "defective". Also, the groups of explanatory variables are materials, environment, and quality. Therefore, the characteristic factor diagram 10 shows "material", "environment", and "quality" as the group display 14. The explanatory variables are average temperature, component 1, quality information 1, and the like. Therefore, the characteristic factor diagram 10 shows the factor display 16 corresponding to each of these explanatory variables.
[0016]
　The analyzer 2000 determines the display mode of each factor display 16 or the presence or absence of display of each factor display 16 based on the degree of influence (hereinafter, the degree of influence) of the explanatory variable corresponding to the factor display 16 on the objective variable. do. Therefore, the analyzer 2000 acquires the relationship information 50 representing the relationship between the objective variable and the explanatory variable. The relationship information 50 indicates the degree of influence of each explanatory variable.
[0017]
　For example, in the analyzer 2000, the display mode is different between the factor display 16 of the explanatory variable having a relatively high degree of influence and the factor display 16 of the explanatory variable having a relatively low degree of influence. More specifically, the factor display 16 of the explanatory variable having a relatively high degree of influence is emphasized more. In the example of FIG. 1, a frame is attached to the factor display 16 of the explanatory variable having a relatively high degree of influence, and a frame is not attached to the other factor display 16. In addition, for example, the analyzer 2000 includes only the factor display 16 of the explanatory variables having a relatively high degree of influence in the characteristic factor diagram 10, and does not include the factor display 16 of the other explanatory variables in the characteristic factor diagram 10. To.
[0018]

　As one of the methods for visually expressing the relationship between the objective variable and the explanatory variable, a method using a characteristic factor diagram can be considered. However, it is not possible to grasp the magnitude of the influence of each explanatory variable on the objective variable from the existing characteristic factor diagram.
[0019]
　Therefore, the analyzer 2000 determines the display mode of the factor display 16 based on the degree of influence. By doing so, it is possible to easily grasp the difference in the degree of influence of each explanatory variable on the objective variable in the characteristic factor diagram that visually shows the relationship between the objective variable and the explanatory variable. That is, by viewing the characteristic factor diagram 10, both the relationship between the objective variable and the explanatory variable (relationship between the characteristic and the factor) and the degree of influence of the explanatory variable can be easily grasped.
[0020]
　Hereinafter, the present embodiment will be described in more detail.
[0021]

　FIG. 2 is a diagram illustrating the functional configuration of the analyzer 2000 of the first embodiment. The analyzer 2000 has an acquisition unit 2020 and a generation unit 2040. The acquisition unit 2020 acquires the related information 5050 for the source data 40. The generation unit 2040 uses the relational information 5050 to generate the characteristic factor diagram 10. Here, the generation unit 2040 determines the display mode of each factor display 16 in the characteristic factor diagram 10 or the presence / absence of the display of the factor display 16 based on the degree of influence of the explanatory variable corresponding to the factor display 16.
[0022]
Each functional component of the analyzer 2000 may be realized by hardware (eg, hard-wired electronic circuit, etc.) that realizes each functional component, or may be hardware. It may be realized by a combination of hardware and software (eg, a combination of an electronic circuit and a program that controls it). Hereinafter, a case where each functional component of the analyzer 2000 is realized by a combination of hardware and software will be further described.
[0023]
　FIG. 3 is a diagram illustrating a computer 1000 for realizing the analyzer 2000. The computer 1000 is an arbitrary computer. For example, the calculator 1000 is a stationary computer such as a PC (Personal Computer) or a server machine. In addition, for example, the computer 1000 is a portable computer such as a smartphone or a tablet terminal.
[0024]
　The computer 1000 may be a dedicated computer designed to realize the analyzer 2000, or may be a general-purpose computer. In the latter case, for example, by installing a predetermined application on the computer 1000, each function of the analyzer 2000 is realized on the computer 1000. The above application is composed of a program for realizing the functional component of the analyzer 2000.
[0025]
　The computer 1000 includes a bus 1020, a processor 1040, a memory 1060, a storage device 1080, an input / output interface 1100, and a network interface 1120. The bus 1020 is a data transmission path for the processor 1040, the memory 1060, the storage device 1080, the input / output interface 1100, and the network interface 1120 to send and receive data to and from each other. However, the method of connecting the processors 1040 and the like to each other is not limited to the bus connection.
[0026]
　The processor 1040 is various processors such as a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), and an FPGA (Field-Programmable Gate Array). The memory 1060 is a main storage device realized by using RAM (Random Access Memory) or the like. The storage device 1080 is an auxiliary storage device realized by using a hard disk, an SSD (Solid State Drive), a memory card, a ROM (Read Only Memory), or the like.
[0027]
　The input / output interface 1100 is an interface for connecting the computer 1000 and the input / output device. For example, an input device such as a keyboard and an output device such as a display device are connected to the input / output interface 1100.
[0028]
　The network interface 1120 is an interface for connecting the computer 1000 to the communication network. This communication network is, for example, a LAN (Local Area Network) or a WAN (Wide Area Network).
[0029]
　The storage device 1080 stores a program (a program that realizes the above-mentioned application) that realizes each functional component of the analyzer 2000. The processor 1040 reads this program into the memory 1060 and executes it to realize each functional component of the analyzer 2000.
[0030]
In order to facilitate understanding of the analyzer 2000, an example of the usage environment of the analyzer 2000 will be described. FIG. 4 is a diagram illustrating a usage environment of the analyzer 2000.
[0031]
　In FIG. 4, the analyzer 2000 is connected to the user terminal 60 via a network. The user operates the user terminal 60 to send a request indicating the conditions related to the relational information 50 to the analyzer 2000 so as to provide the characteristic factor diagram 10 for the specific relational information 50. In response to this request, the analyzer 2000 acquires the relational information 50 corresponding to the condition from the storage device, and uses the acquired relational information 50 to generate the characteristic factor diagram 10. Then, the analyzer 2000 transmits the generated characteristic factor diagram 10 to the user terminal 60.
[0032]
　For example, the analyzer 2000 provides the user terminal 60 with screen data (for example, a Web page) including the image of the characteristic factor diagram 10. In this case, the user terminal 60 displays her received Web page on the browser. By doing so, the user can browse the characteristic factor diagram 10.
[0033]
　The usage environment of the analyzer 2000 is not limited to that shown in FIG. For example, the analyzer 2000 may be operated directly by the user instead of being used via the user terminal 60. Further, the method of specifying the relational information 50 to be generated in the characteristic factor diagram 10 is not limited to the method of specifying the condition regarding the relational information 50 by the user (details will be described later).
[0034]

　FIG. 5 is a flowchart illustrating a processing flow executed by the analyzer 2000 of the first embodiment. The acquisition unit 2020 acquires the relationship information 50 (S102). The generation unit 2040 generates the characteristic factor diagram 10 using the relational information 50 (S104). The generation unit 2040 outputs the characteristic factor diagram 10 (S106).
[0035]

　The relationship information 50 indicates the relationship between the objective variable and a plurality of explanatory variables. FIG. 6 is a diagram illustrating the configuration of the relationship information 50. In FIG. 6, the relational information 50 includes information of the objective variable 52 and the explanatory variable 54. The objective variable 52 indicates identification information (name, etc.) of the objective variable. The explanatory variable 54 indicates the identification information 56 (name, etc.) and the degree of influence 58 for each explanatory variable.
[0036]
　The relationship shown by the relationship information 50 is estimated, for example, by analyzing the data (hereinafter, source data) in which the value of the objective variable and the value of each explanatory variable are associated with each other. The estimated relationship is represented by, for example, a linear model (linear regression model or linear discriminative model) that estimates the value of the objective variable from the value of each explanatory variable. In this case, the relationship information 50 can also be regarded as information representing a linear model.
[0037]
　For example, the degree of influence of an explanatory variable is represented by a coefficient (coefficient multiplied by the value of the explanatory variable) corresponding to the explanatory variable in a linear model. This is because the larger the corresponding coefficient in the linear model, the greater the influence on the value of the objective variable obtained by using the model. Therefore, for example, the relational information 50 indicates the coefficient corresponding to the explanatory variable in the trained linear model as the degree of influence of the explanatory variable.
[0038]
　For example, it is assumed that the presence or absence of defective products is treated as an objective variable, and each index (for example, material components and environment such as temperature) representing the manufacturing conditions of the product is treated as an explanatory variable. In this case, the source data shows the value of each index (content of each component, temperature, etc.) indicating the manufacturing conditions and the presence or absence of defects in the product manufactured under the manufacturing conditions. By training the linear model with this source data, it is possible to obtain a trained linear model showing the relationship between the presence or absence of defects and the manufacturing conditions.
[0039]
　For example, the analyzer 2000 treats the information representing this linear model as the relational information 50. In this case, the relationship information 50 shows the identification information of the objective variable in the generated linear model in the objective variable 52, the identification information of each explanatory variable in the generated linear model in the identification information 56, and corresponds to each explanatory variable. The coefficient to be used is shown in the degree of influence 58.
[0040]
　The relationship between the objective variable and the explanatory variable may be represented by a plurality of linear models. As a method of expressing the relationship between the objective variable and the explanatory variable with a plurality of linear models, there is a method of generating an estimation model by using heterogeneous blended learning (see Patent Document 2). In heterogeneous blended learning, an estimation model defined by a set of a tree consisting of nodes representing conditional branches and a plurality of linear models is generated. One linear model is assigned to each leaf of the tree. When using this estimation model, first, the tree is traced from the root to the leaf by using the data to be estimated (combination of the values ​​of each explanatory variable). Then, the value of the objective variable is obtained by inputting the data to be estimated into the linear model corresponding to the arrived leaf.
[0041]
　When the relationship between the objective variable and the explanatory variable is represented by a plurality of linear models, for example, the relationship information 50 indicates the explanatory variable 54 (combination of the identification information 56 and the degree of influence 58) for each of the plurality of linear models. FIG. 7 is a diagram illustrating the relationship information 50 when the relationship between the objective variable and the explanatory variable is represented by a plurality of linear models. The relationship information 50 in FIG. 7 shows the explanatory variables 54 for each of the plurality of linear models.
[0042]
The
　acquisition unit 2020 acquires the relationship information 50 used for generating the characteristic factor diagram 10 (S102). For example, the acquisition unit 2020 acquires the relational information 50 corresponding to the condition specified by the user from the plurality of relational information 50s stored in the storage device in advance.
[0043]
　For example, the relationship information 50 is specified by the conditions relating to the source used for the relationship between the objective variable and the explanatory variable indicated by the relationship information 50. For example, when the source data is data related to the manufacture of a product, the source data can be specified by conditions such as the name of the product, the place of manufacture, and the date and time of manufacture.
[0044]
　Therefore, the user gives the analyzer 2000 a condition regarding the source data for which the characteristic factor diagram 10 is to be generated. The analyzer 2000 acquires the relational information 50 corresponding to the condition by searching the storage device under the given condition.
[0045]
　When there are a plurality of relational information 50s that meet the conditions specified by the user, the analyzer 2000 may generate a characteristic factor diagram 10 for each of the relational information 50s, or a part of the relational information 50s. The characteristic factor diagram 10 may be generated for only. In the latter case, the acquisition unit 2020 may provide the user with information about each relationship information 50 corresponding to the specified condition, and may allow the user to select one or more relationship information 50. In this case, the analyzer 2000 generates the characteristic factor diagram 10 only for the relationship information 50 selected by the user.
[0046]
　In addition, for example, the acquisition unit 2020 may receive the relationship information 50 transmitted from another device (for example, the user terminal 60). For example, in this case, the user terminal 60 transmits the related information 50 to the analyzer 2000.
[0047]
　In addition, for example, the acquisition unit 2020 may acquire the information necessary for generating the relational information 50 and generate the relational information 50 by using the acquired information. For example, the user provides the acquisition unit 2020 with information indicating the source data, the identification information of the objective variable, the type of the model, and the like. The acquisition unit 2020 generates an estimation model using the provided information, and generates the relational information 50 represented by the generated estimation model.
[0048]
　The process of generating the relational information 50 using the information provided by the user may be performed by an apparatus other than the analyzer 2000. In this case, the acquisition unit 2020 acquires the relationship information 50 from the device that generated the relationship information 50.
[0049]
The
　generation unit 2040 generates the Ishikawa diagram 10 by using the relational information 50. Here, in order to generate the characteristic factor diagram 10, in addition to the identification information of the objective variable and the explanatory variable, information for defining a group of the explanatory variables (hereinafter, group definition information) is required. The group definition information indicates, for example, the identification information (name, etc.) of the group and the identification information of each explanatory variable included in the group. The group definition information may be included in the relational information 50, or may be prepared separately from the relational information 50. If the characteristic is not the objective variable itself (for example, the objective variable is "presence or absence of defect" while the characteristic is "defective"), the identification information (name, etc.) of the characteristic is also identified as a group. Prepare as well as information. The same applies when the factor is not the explanatory variable itself.
[0050]
　For example, the generation unit 2040 generates the characteristic display 12, the group display 14, and the factor display 16, respectively, using the identification information of the objective variable, the group identification information indicated by the group definition information, and the identification information of the explanatory variables. Further, the generation unit 2040 determines the positional relationship between the group display 14 and the factor display 16 by using the group definition information. Then, the generation unit 2040 generates the characteristic factor diagram 10 by connecting the generated displays with the trunk 11, the branch 13, and the twig 14 based on the determined positional relationship.
[0051]
　However, the generation unit 2040 generates the factor display 16 based on the degree of influence of the explanatory variable indicated by the relational information 50. Hereinafter, a method for generating the factor display 16 based on the degree of influence will be specifically illustrated.
[0052]
<< Case of determining the display mode according to the degree of influence >>
　For example, the generation unit 2040 determines the display mode of the factor display 16 of the explanatory variable by comparing the degree of influence of the explanatory variable with a predetermined threshold value. For example, the generation unit 2040 makes the display mode of the factor display 16 of the explanatory variable having an influence degree equal to or higher than the threshold value different from the display mode of the factor display 16 of the explanatory variable having an influence degree lower than the threshold value. As a method of making the display mode of the factor display 16 different, a method of making the background of the factor display 16 different (presence / absence of fill, color, pattern, etc.) and a method of making the frame of the factor display 16 different (presence / absence of frame, color, shape). , Thickness, etc.), a method of making the size of the factor display 16 different, and the like.
[0053]
　The information in which the display mode is determined for each of the case where the influence degree is equal to or more than the threshold value and the case where the influence degree is less than the threshold value is stored in advance in a storage device accessible from the generation unit 2040. However, this information may be changeable by the user.
[0054]
　Here, it is preferable that the factor display 16 of the explanatory variable is made more conspicuous (emphasized) as the degree of influence of the explanatory variable is higher. Therefore, for example, the display mode of the factor display 16 is "to make the background of the factor display 16 of the explanatory variable having an influence degree equal to or higher than the threshold value more conspicuous than the background of the other factor display 16" and "to make the influence degree equal to or higher than the threshold value stand out". Make the frame of the factor display 16 of the explanatory variable to stand out more than the frame of the other factor display 16 "," The size of the factor display 16 of the explanatory variable having an influence degree equal to or higher than the threshold value of the other factor display 16. It is preferable to set the standard such as "make it larger than the frame".
[0055]
　FIG. 8 is a diagram illustrating the characteristic factor diagram 10 in which the display mode of the factor display 16 is determined by using the influence degree of the explanatory variable. In FIG. 8, only the factor display 16 for the explanatory variables (minimum temperature, component 2, quality information 2, and quality information 3) whose degree of influence is equal to or higher than the threshold value is framed. Therefore, the explanatory variables whose influence degree is equal to or higher than the threshold value and the other explanatory variables can be easily distinguished.
[0056]
　There may be a plurality of influence thresholds. In other words, a plurality of numerical ranges may be defined for the degree of influence, and the display mode of the factor display 16 may be different for each numerical range. For example, provide three numerical ranges R1 to R3, "less than Th1", "more than Th1 and less than Th2", and "more than Th2" (Th1 and Th2 are real numbers that satisfy Th1 >
　For example, the generation unit 2040 includes the factor display 16 of the explanatory variable in the characteristic factor diagram 10 by comparing the degree of influence of the explanatory variable with the threshold value. Decide whether or not. More specifically, the generation unit 2040 includes the factor display 16 of the explanatory variable having an influence degree equal to or higher than the threshold value in the characteristic factor diagram 10, and the factor display 16 of the explanatory variable having an influence degree less than the threshold value is included in the characteristic factor diagram 10. Do not include in.
[0063]
　FIG. 9 is a diagram illustrating a characteristic factor diagram 10 in which the presence or absence of the factor display 16 is determined by using the degree of influence of the explanatory variable. In FIG. 9, the explanatory variables in which the degree of influence is equal to or higher than the threshold value are the same as in the example of FIG. However, in FIG. 9, only the factor display 16 of the explanatory variable whose influence degree is equal to or higher than the threshold value is included in the characteristic factor diagram 10.
[0064]
　In addition, for example, the generation unit 2040 may determine the presence or absence of the factor display 16 in the order of the degree of influence. For example, the generation unit 2040 specifies each explanatory variable included within a predetermined order in descending order of influence from all the explanatory variables, and includes only the factor display 16 of the identified explanatory variable in the characteristic factor diagram 10. In addition, for example, the generation unit 2040 specifies each explanatory variable included within a predetermined order in descending order of the degree of influence in each group, and only the factor display 16 of the identified explanatory variable is shown in the characteristic factor diagram 10. Include in.
[0065]
　Here, when the presence or absence of the factor display 16 is determined by the method described above, there may be a group display 14 in which no factor display 16 is included. Therefore, the generation unit 2040 may determine the display mode of the group display 14 depending on whether or not the factor display 16 is included. By doing so, the group display 14 including the factor display 16 is emphasized more than the group display 14 not including the factor display 16. Examples of the method of making the display mode of the group display 14 different include a method of making the background, the frame, and the like different, as in the method of making the display mode of the factor display 16 different. Further, the generation unit 2040 may not display the group display 14 that does not include the factor display 16. In other words, the generation unit 2040 determines whether or not the group display 14 is displayed only for the group display 14 including at least one factor display 16.
[0066]
<< Case where the relationship information 50 includes information about a plurality of linear models >>
　As illustrated with reference to FIG. 7, the relationship information 50 may indicate information about each of the plurality of linear models. In this case, the generation unit 2040 may generate the characteristic factor diagram 10 for each of the plurality of linear models, or accepts the designation of the linear model from the user and generates the characteristic factor diagram 10 for the specified linear model. May be good.
[0067]
　In addition, for example, the generation unit 2040 calculates the statistical value (total value, average value, maximum value, minimum value, etc.) of the influence degree of the explanatory variable for each explanatory variable, and uses the statistical value of the explanatory variable. One characteristic factor diagram 10 may be generated by treating it as the degree of influence. For example, when the total value of the influences shown by the relational information 50 is used to generate the characteristic factor diagram 10, the generation unit 2040 uses the following equation (1).
[Equation 1]

　e [i] [j] is the degree of influence of the explanatory variable i, which the relational information 50 indicates for the j-th linear model. n is the total number of linear models indicated by the relationship information 50. Ei is a value that is treated as the degree of influence of the explanatory variable i when the characteristic factor diagram 10 is generated.
[0068]
The
　analyzer 2000 outputs the Ishikawa diagram 10 generated by the generation unit 2040. There are various specific methods for outputting the characteristic factor diagram 10. For example, the analyzer 2000 stores image data representing the characteristic factor diagram 10 in a storage device, displays it on a display device connected to the analyzer 2000, and transmits it to another device (for example, a user terminal 60). do.
[0069]
[Embodiment 2]

　FIG. 10 is a diagram for explaining an outline of the analyzer 2000 of the second embodiment. FIG. 10 is an example for facilitating the understanding of the analyzer 2000, and the function of the analyzer 2000 is not limited to that shown in FIG. Further, the analyzer 2000 of the second embodiment has the same function as the analyzer 2000 of the first embodiment, except for the points described below.
[0070]
　The analyzer 2000 of the second embodiment receives an explanatory variable (hereinafter, designated) corresponding to the designated factor display 16 in response to an input for designating the factor display 16 to the output characteristic factor diagram 10. The graph 30 for (also referred to as the explanatory variable) is output. In FIG. 10, the factor display 16 of "component 2" is selected by the user. Therefore, the analyzer 2000 generates a line graph representing the data obtained for the component 2 in chronological order as the graph 30.
[0071]
　Here, the source data used for estimating the relationship between the objective variable and the explanatory variable represented by the relationship information 50 includes a plurality of sets of the value of the objective variable and the value of the explanatory variable. For example, the source data shows time series data of a set of values ​​of the objective variable and the value of the explanatory variable. For example, source data relating to the manufacture of a product can be handled, which indicates the presence or absence of defects in the product and a set of manufacturing conditions at different manufacturing time points.
[0072]
　The graph 30 is a graph of a plurality of values ​​indicated by the source data for the specified explanatory variable. For example, the graph 30 is a graph showing the plurality of data in time series and a graph showing the result of statistically processing the plurality of data.
[0073]

　The user can easily grasp the difference in the degree of influence of each explanatory variable on the objective variable by viewing the characteristic factor diagram 10 generated by the analyzer 2000. For example, by emphasizing the factor display 16 for the explanatory variable having a high degree of influence, the explanatory variable having a high degree of influence can be easily grasped.
[0074]
　Then, it can be said that the user who grasps the difference in the influence degree of the explanatory variables in this way has a high probability of wanting to further browse the information related to each explanatory variable. For example, when the factor display 16 for the highly influential explanatory variable is emphasized, it is highly probable that the user wants to browse more detailed information about the highly influential explanatory variable.
[0075]
　In this regard, according to the analyzer 2000 of the present embodiment, when the input for designating the factor display 16 is input to the characteristic factor diagram 10, the graph 30 is generated for the designated explanatory variable. Therefore, the user can easily obtain more detailed information about the specified explanatory variable.
[0076]
　Hereinafter, the analyzer 2000 of the present embodiment will be described in more detail.
[0077]

　The functional configuration of the analyzer 2000 of the second embodiment is represented by FIG. 2 in the same manner as the analyzer 2000 of the first embodiment, for example. However, the generation unit 2040 of the second embodiment generates a graph 30 about the factors corresponding to the designated factor display 16 in response to the input for designating the factor display 16 to the characteristic factor diagram 10. And output.
[0078]

　The hardware configuration of the analyzer 2000 of the second embodiment is represented by FIG. 3 in the same manner as the analyzer 2000 of the first embodiment, for example. However, the storage device 1080 of the second embodiment stores a program that realizes the function of the analyzer 2000 of the second embodiment.
[0079]

　FIG. 11 is a flowchart illustrating a processing flow executed by the analyzer 2000 of the second embodiment. The acquisition unit 2020 acquires the identification information of the designated explanatory variable (S202). The generation unit 2040 acquires a plurality of data indicated by the source data for the specified explanatory variable (S204). The generation unit 2040 generates the graph 30 using the acquired plurality of data (S206). The generation unit 2040 outputs the generated graph 30 (S208).
[0080]
The
　acquisition unit 2020 acquires the identification information of the explanatory variable specified in the characteristic factor diagram 10 (S202). Here, when a specific part is specified in the figure output by a method such as displaying on a display device, various existing techniques are used as a technique for acquiring the identification information of the specified part. be able to.
[0081]
The
　generation unit 2040 acquires a plurality of data indicated by the source data for the specified explanatory variable. The source data is stored in advance in a storage device accessible from the generation unit 2040. By accessing this storage device, the generation unit 2040 acquires a plurality of data for the specified explanatory variable.
[0082]
　For example, as described above, the source data indicates time series data of a set of values ​​of the objective variable and the value of the explanatory variable. In this case, the generation unit 2040 acquires time series data representing the time change of the value of the designated explanatory variable. However, the source data may include a plurality of sets of values ​​of the objective variable and the values ​​of the explanatory variables, and these sets do not have to be data representing the time change.
[0083]
The
　generation unit 2040 generates the graph 30 using a plurality of data acquired for the designated explanatory variables (S206). The type of graph 30 can be any type that can be generated using a plurality of data. When the acquired data is time-series data as described above, for example, the graph 30 is a line graph or the like showing the time change of the value of the designated explanatory variable. In addition, for example, the graph 30 is a histogram or the like showing the result of statistically processing a plurality of values ​​of the explanatory variables.
[0084]
　Here, the generation unit 2040 may generate a plurality of graphs 30 for the designated explanatory variables. For example, the generation unit 2040 generates a graph showing the time-series change of the acquired data and a graph showing the result of statistically processing the data. In addition, for example, the generation unit 2040 statistically processes the acquired data by each of a plurality of methods, and generates a graph 30 for each result.
[0085]
　Here, what kind of graph is generated as the graph 30 may be fixed in advance or may be specified by the user.
[0086]
The
　generation unit 2040 outputs the generated graph 30. The output method of the graph 30 is the same as the output method of the characteristic factor FIG. Further, the generation unit 2040 may output the characteristic factor diagram 10 together with the graph 30. For example, the generation unit 2040 generates and outputs screen data (for example, a Web page) including both the characteristic factor diagram 10 and the graph 30.
[0087]
FIG. 12 is a diagram illustrating a screen including both the characteristic factor diagram 10 and the graph 30. In FIG. 12, two graphs 30 are included, a graph (30-1) representing data in time series and a histogram (30-2).
[0088]
　The graph 30 is not limited to a line graph or a histogram. For example, a boxplot may be generated by performing aggregation for each specific section, or the number of samples in the same section may be displayed as a bar graph. In addition, these aggregation sections may be set in advance, may be selected by the user from a plurality of types of sections (month, week, day, time) prepared in advance, and an arbitrary value may be set by the user. It may be specified.
[0089]
　Further, the entire period (when to display the data) included in the graph 30 may be from the first time point to the last time point included in the source data, or an arbitrary period may be specified by the user. good. In the latter case, for example, a calendar is displayed on the screen so that input for specifying both the first time point and the last time point can be performed on the calendar.
[0090]
The
　generation unit 2040 acquires the objective variable data in addition to the specified explanatory variable from the source data, and generates the graph 30 using the specified explanatory variable and the objective variable data. You may. By doing so, it becomes possible to directly grasp the relationship between the specified explanatory variable and the objective variable.
[0091]
　FIG. 13 is a diagram illustrating graph 30 showing data for each of the designated explanatory variable and the objective variable. In FIG. 13, graph 30 is a graph in which a polygonal line representing data for the objective variable is superimposed on a histogram of the values ​​of the designated explanatory variables. More specifically, for each rank shown on the horizontal axis of the histogram by a broken line, the defective rate of the product corresponding to that rank (the ratio of the number of defective products to the total number of manufactured products corresponding to that rank). It is shown. In FIG. 13, confidence intervals are displayed for the defective rate for each rank.
[0092]
　Further, data of items related to the objective variable may be added to the graph 30 in place of or together with the data of the objective variable. For example, when the objective variable is the presence or absence of defects, the defect rate data is included in the graph 30.
[0093]
The
　user may be able to specify a plurality of factor displays 16 included in the characteristic factor diagram 10. In this case, the generation unit 2040 generates the graph 30 for each of the plurality of designated explanatory variables. At this time, the graph 30 may be individually generated for the data obtained for each of the plurality of designated explanatory variables, or one graph 30 may be generated for these.
[0094]
　FIG. 14 is a diagram illustrating a graph 30 when a plurality of factor indications 16 are designated. In this example, two explanatory variables, component 1 and component 2, are specified, respectively. Therefore, in Graph 30-1, the data for each of the component 1 and the component 2 are shown in chronological order. Further, Graph 30-2 and Graph 30-3 show a histogram for component 1 and a histogram for component 2, respectively.
[0095]
　Although the embodiments of the present invention have been described above with reference to the drawings, these are examples of the present invention, and a combination of the above embodiments or various configurations other than the above can be adopted.
[0096]
　Some or all of the above embodiments may also be described, but not limited to:
　Hereinafter, an example of the reference form will be added.
1. 1. An acquisition unit that acquires relational information indicating the degree of influence of each of the plurality of explanatory variables on the
　objective variable, and a generation unit that generates a characteristic factor diagram showing the relationship between the objective variable and the explanatory variable using the relational information.
　An analyzer that the generator determines the display mode of the display for each explanatory variable in the characteristic factor diagram or the presence or absence of the display based on the degree of influence of the explanatory variable .
2. The relationship between the objective variable and the plurality of explanatory variables is represented by one or more linear models, and the
　degree of influence of the explanatory variables indicated by the relationship information is represented by a coefficient multiplied by the explanatory variables in the linear model. To be done 1. The analyzer described in.
3. 3. The relational information indicates a plurality of influence degrees of the
　explanatory variables for each explanatory variable, and the generation unit calculates statistical values ​​of a plurality of influence degrees indicated by the relational information for the explanatory variables for each explanatory variable. Then, the calculated statistical value is treated as the degree of influence of the explanatory variable to generate the specific factor diagram. Or 2. The analyzer described in.
4. The generation unit
　　relates to an explanatory variable whose display mode is different between the display related to the explanatory variable whose influence degree is equal to or higher than the threshold value and the display related to the other explanatory variables, or
　　whose influence degree is within a predetermined order. Whether the display mode is different between the display and the display related to other explanatory variables,
　　For each group of the explanatory variables, the display mode is different between the display of the explanatory variables having the influence degree within a predetermined order among the explanatory variables belonging to the group and the display of the other explanatory variables. 1. 1. The analyzer according to 3.
5. Whether the generation unit includes
　　only the display related to the explanatory variables whose influence degree is equal to or higher than the threshold value in the characteristic factor diagram, or includes only the display related to
　　the explanatory variables whose influence degree is within a predetermined order in the characteristic factor diagram. For each group of
　　the explanatory variables, only the display relating to the explanatory variables whose influence degree is within a predetermined order among the explanatory variables belonging to the group is included in the characteristic factor diagram. The analyzer according to 3.
6. The generation unit
　　outputs the
　　characteristic factor diagram, and when the explanatory variable is specified in the output characteristic factor diagram, acquires data indicating a plurality of values ​​of the designated explanatory variable, and obtains the
　　data. Generate a graph using 1. 5 The analyzer according to any one of 5.
7. The generation unit
　　acquires time-series data for the designated explanatory variable, and
　　as the graph, the first graph showing the time change of the value of the explanatory variable or the second graph showing the result of statistically processing the time-series data. Generate a graph, 6. The analyzer described in.
8. 6. The generation unit generates screen data including both the first graph and the second graph. The analyzer described in.
9. The generator includes the data of the objective variable in the graph. 8 The analyzer according to any one of 8.
10. It is a control method executed by a computer, 　and represents a relationship between the objective variable and the explanatory variable by using
　the acquisition step of acquiring the relational information indicating the degree of influence of each of the plurality of explanatory variables on the objective variable and the relational information.
It has a generation step for generating a characteristic factor diagram, and in the
　generation step, the display mode of the display for each explanatory variable in the characteristic factor diagram or the presence or absence of the display is determined based on the influence degree of the explanatory variable. Control method.
11. The relationship between the objective variable and the plurality of explanatory variables is represented by one or more linear models, and the
　degree of influence of the explanatory variables indicated by the relationship information is represented by a coefficient multiplied by the explanatory variables in the linear model. 10. The control method described in.
12. The relational information indicates a plurality of influence degrees of the
　explanatory variables for each of the explanatory variables, and in the generation step, for each of the explanatory variables, statistical values ​​of a plurality of influence degrees indicated by the relational information are calculated for the explanatory variables. Then, the calculated statistical value is treated as the degree of influence of the explanatory variable to generate the specific factor diagram. Or 11. The control method described in.
13. In the generation step,
　　the display mode is different between the display related to the explanatory variables whose influence degree is equal to or higher than the threshold value and the display related to the other explanatory variables.
　　The display mode may be different between the display for the explanatory variables whose influence degree is within a predetermined order and the display for the other explanatory variables, or
　　for each group of the explanatory variables, among the explanatory variables belonging to the group. In 10., the display mode is different between the display related to the explanatory variables whose influence degree is within the predetermined order and the display related to the other explanatory variables. 12. The control method according to 12.
14. In the generation step,
　　whether to include only the display related to the explanatory variables whose influence degree is equal to or higher than the threshold value in the characteristic factor diagram, or to include only the display related to
　　the explanatory variables whose influence degree is within a predetermined order in the characteristic factor diagram. For each group of
　　the explanatory variables, only the display relating to the explanatory variables whose influence degree is within a predetermined order among the explanatory variables belonging to the group is included in the characteristic factor diagram. 12. The control method according to 12.
15. In the generation step, the
　　characteristic factor diagram is output, and when
　　the explanatory variable is specified in the output characteristic factor diagram, data indicating a plurality of values ​​of the specified explanatory variable is acquired, and
　　the data is obtained. Use to generate a graph 10. To 14 The control method according to any one of 14.
16. In the generation step, time series data is acquired for the
　　specified explanatory variable, and the time series data is acquired.
　　As the graph, a first graph showing the time change of the value of the explanatory variable or a second graph showing the result of statistically processing the time series data is generated. The control method described in.
17. In the generation step, screen data including both the first graph and the second graph is generated. The control method described in.
18. In the generation step, the data of the objective variable is included in the graph. To 17 The control method according to any one of 17.
19. 10. A program that causes a computer to execute the control method according to any one of 18 to 18.
[0097]
　This application claims priority on the basis of Japanese Application Japanese Patent Application No. 2019-193810 filed on October 24, 2019, the entire disclosure of which is incorporated herein by reference.
Code description
[0098]
10 Characteristic factor Figure
11 Trunk
12 Characteristic display
13 Branch
14 Group display
15 Twig
16 Factor display
30 Graph
50 Relationship information
52 Objective variable
54 Explanatory variable
56 Impact
60 User terminal
1000 Computer
1020 Bus
1040 Processor
1060 Memory
1080 Storage device
1100 Input / output Interface
1120 Network Interface
2000 Analyzer
2020 Acquisition Unit
2040 Generation Unit
The scope of the claims
[Claim 1]
　An acquisition unit that acquires relational information indicating the degree of influence of each of the plurality of explanatory variables on the
　objective variable, and a generation unit that generates a characteristic factor diagram showing the relationship between the objective variable and the explanatory variable using the relational information.
　An analyzer that the generator determines the display mode of the display for each explanatory variable in the characteristic factor diagram or the presence or absence of the display based on the degree of influence of the explanatory variable .
[Claim 2]
　The relationship between the objective variable and the plurality of explanatory variables is represented by one or more linear models, and the
　degree of influence of the explanatory variables indicated by the relationship information is represented by a coefficient multiplied by the explanatory variables in the linear model. The analyzer according to claim 1.
[Claim 3]
　The relational information indicates a plurality of influence degrees of the
　explanatory variables for each explanatory variable, and the generation unit calculates statistical values ​​of a plurality of influence degrees indicated by the relational information for the explanatory variables for each explanatory variable. The analyzer according to claim 1 or 2, wherein the calculated statistical value is treated as the degree of influence of the explanatory variable to generate the specific factor diagram.
[Claim 4]
　The generation unit
　　relates to an explanatory variable whose display mode is different between the display related to the explanatory variable whose influence degree is equal to or higher than the threshold value and the display related to the other explanatory variables, or
　　whose influence degree is within a predetermined order. The display mode is different between the display and the display related to other explanatory variables, or
　　for each group of the explanatory variables, the explanatory variables whose influence degree is within a predetermined order among the explanatory variables belonging to the group. The analyzer according to claim 1 to 3, wherein the display mode differs between the display relating to and the display relating to other explanatory variables.
[Claim 5]
　Whether the generation unit includes
　　only the display related to the explanatory variables whose influence degree is equal to or higher than the threshold value in the characteristic factor diagram, or includes only the display related to
　　the explanatory variables whose influence degree is within a predetermined order in the characteristic factor diagram.
　　The analysis according to claims 1 to 3, wherein for each group of the explanatory variables, only the display relating to the explanatory variables whose influence degree is within a predetermined order among the explanatory variables belonging to the group is included in the characteristic factor diagram . Device.
[Claim 6]
　The generation unit
　　outputs the
　　characteristic factor diagram, and when the explanatory variable is specified in the output characteristic factor diagram, acquires data indicating a plurality of values ​​of the designated explanatory variable, and obtains the
　　data. The analyzer according to any one of claims 1 to 5, which is used to generate a graph.
[Claim 7]
　The generation unit
　　acquires time-series data for the designated explanatory variable, and
　　as the graph, the first graph showing the time change of the value of the explanatory variable or the second graph showing the result of statistically processing the time-series data. The analyzer according to claim 6, which produces a graph.
[Claim 8]
　The analyzer according to claim 7, wherein the generation unit generates screen data including both the first graph and the second graph.
[Claim 9]
　The analyzer according to any one of claims 6 to 8, wherein the generation unit includes data of the objective variable in the graph.
[Claim 10]
　It is a control method executed by a computer, 　and represents a relationship between the objective variable and the explanatory variable by using
　the acquisition step of acquiring the relational information indicating the degree of influence of each of the plurality of explanatory variables on the objective variable and the relational information.
It has a generation step for generating a characteristic factor diagram, and in the
　generation step, the display mode of the display for each explanatory variable in the characteristic factor diagram or the presence or absence of the display is determined based on the influence degree of the explanatory variable. Control method.
[Claim 11]
　The relationship between the objective variable and the plurality of explanatory variables is represented by one or more linear models, and the
　degree of influence of the explanatory variables indicated by the relationship information is represented by a coefficient multiplied by the explanatory variables in the linear model. The control method according to claim 10.
[Claim 12]
　The relational information indicates a plurality of influence degrees of the
　explanatory variables for each of the explanatory variables, and in the generation step, for each of the explanatory variables, statistical values ​​of a plurality of influence degrees indicated by the relational information are calculated for the explanatory variables. The control method according to claim 10 or 11, wherein the calculated statistical value is treated as the degree of influence of the explanatory variable to generate the specific factor diagram.
[Claim 13]
　In the generation step,
　　the display mode is different between the display related to the explanatory variable whose influence degree is equal to or higher than the threshold value and the display related to the other explanatory variables, or the explanatory variable
　　whose influence degree is within a predetermined order. The display mode is different between the display and the display related to other explanatory variables, or
　　for each group of the explanatory variables, the explanatory variables whose influence degree is within a predetermined order among the explanatory variables belonging to the group. The control method according to claim 10 to 12, wherein the display mode is different between the display relating to and the display relating to other explanatory variables.
[Claim 14]
　In the generation step,
　　whether to include only the display related to the explanatory variables whose influence degree is equal to or higher than the threshold value in the characteristic factor diagram, or to include only the display related to
　　the explanatory variables whose influence degree is within a predetermined order in the characteristic factor diagram.
　　The control according to claims 10 to 12, wherein for each group of the explanatory variables, only the display relating to the explanatory variables whose influence degree is within a predetermined order among the explanatory variables belonging to the group is included in the characteristic factor diagram . Method.
[Claim 15]
　In the generation step, the
　　characteristic factor diagram is output, and when
　　the explanatory variable is specified in the output characteristic factor diagram, data indicating a plurality of values ​​of the specified explanatory variable is acquired, and
　　the data is obtained. The control method according to any one of claims 10 to 14, wherein a graph is generated using the method.
[Claim 16]
　In the generation step,
　　time-series data is acquired for the designated explanatory variable, and
　　as the graph, the first graph showing the time change of the value of the explanatory variable or the second graph showing the result of statistically processing the time-series data. The control method according to claim 15, which generates a graph.
[Claim 17]
　The control method according to claim 16, wherein in the generation step, screen data including both the first graph and the second graph is generated.
[Claim 18]
　The control method according to any one of claims 15 to 17, wherein the data of the objective variable is included in the graph in the generation step.
[Claim 19]
　A program that causes a computer to execute the control method according to any one of claims 10 to 18.