JP7576390B2 - Information analysis device, information analysis method, and information analysis program - Google Patents

Description

The present invention relates to an information analysis device, an information analysis method, and an information analysis program. More specifically, the present invention relates to an information analysis device, an information analysis method, and an information analysis program that present different questions to each respondent, for example.

In recent years, analytical devices that can analyze the status of employees from the responses of questionnaires administered to members of an organization have become widespread, and are now being used by companies and other organizations to administer such surveys to their employees. For example, there are known organizational development support systems, methods, and programs that obtain scores from the responses of members to multiple survey questions, enabling on-site and human resources personnel to strengthen and develop the organization themselves (see, for example, Patent Document 1).

JP 2018-18152 A

In order for a company or other organization to understand the condition of its employees, it is necessary to continually diagnose and understand changes in the employee's condition and awareness. However, with the technology described in Patent Document 1, many questionnaires must be conducted for one diagnosis, which places a heavy burden on employees and means that this cannot be done frequently. In addition, the factors that employees look for in a company or other organization vary from employee to employee. However, with the technology described in Patent Document 1, the same questionnaire is uniformly administered to all employees without taking such points into consideration, and there is a problem in that questionnaires are not conducted according to the condition of each employee.

The present invention is based on such social and technological background and achieves the following objectives: The present invention aims to present different questions to each respondent.

In order to achieve the above object, the present invention takes the following measures.
(1) An information analysis device comprising: a storage means for storing respondent information identifying a respondent and question information asking the respondent questions about duties, personal growth, health, human relationships, support, recognition, philosophy strategy, organizational culture, and environment as factors influencing the willingness to contribute to the organization; an answer information receiving means for receiving from the respondent's terminal the values of each factor influencing the willingness to contribute as answer information; an analysis means for analyzing the answer information received by the answer information receiving means and generating an analysis result using factor analysis or Bayesian statistics to determine which factors the respondent considers important and a level of importance for deriving causal relationships from the answer information based on the time-series trends of the values of each factor in the past from the answer information for each respondent; a question selection means for selecting question information to be asked for each respondent based on the level of importance as an analysis result analyzed by the analysis means; and a question presentation means for presenting the question information for each respondent selected by the question selection means to the respondent's terminal for each respondent.
(2) An information analysis device as described in claim 1, characterized in that the question information has a frequency which is a degree to which the question is delivered to the respondent, and the question selection means selects question information to be asked to each respondent according to the analysis results analyzed by the analysis means and the frequency of the question information.
(3) The information analyzing apparatus according to claim 2, further comprising a frequency updating means for updating the frequency of question information in accordance with the importance.
(4) An information analysis device according to any one of claims 1 to 3, wherein the analysis means includes an importance model generation means for generating an estimation model by machine learning, using the response information of all respondents as teacher data, and estimating parameters expressing the relationship between the response information of all respondents and importance using factor analysis or Bayesian statistics , and an importance output means for outputting the importance of one respondent by analyzing the response information of one respondent using the estimation model generated by the importance model generation means.
(5) An information analysis device according to claim 4, wherein the analysis means uses the answer information of all respondents as teacher data, and determines how long ago a question was selected from the answer information of all respondents and the answer information of each respondent , and generates an estimation model by machine learning to estimate parameters expressing the relationship between the amount of remaining information, which is information for determining whether a question to be selected for a respondent is new or old, using factor analysis, Bayesian statistics, or clustering ; and a remaining information amount output means for outputting the amount of remaining information of one respondent by analyzing the answer information of one respondent using the estimation model generated by the remaining information amount model generation means.
(6) An information analysis device according to claim 5, wherein the analysis means comprises: a selective question model generation means for generating an estimation model by machine learning, using the importance of all respondents and/or the amount of remaining information of all respondents as teacher data, to estimate parameters expressing a relationship between the importance of all respondents and/or the amount of remaining information of all respondents and selective question information of all respondents; and a selective question output means for outputting selective question information to one respondent by analyzing the importance of one respondent and/or the amount of remaining information of one respondent using the estimation model generated by the selective question model generation means.
(7) An information analysis method characterized by having a computer realize the following steps: a storage step for storing respondent information identifying a respondent and question information asking the respondent questions about duties, personal growth, health, human relationships, support, recognition, philosophy and strategy, organizational culture, and environment as factors that affect the willingness to contribute; an answer information receiving step for receiving, from the respondent terminal, the values of each factor that affects the willingness to contribute answered by the respondent in response to the questions as answer information; an analysis step for analyzing the answer information received by the answer information receiving step and generating, as an analysis result, an importance level for deriving causal relationships from the answer information of each respondent, based on the time-series trends in the values of each factor in the past; a question selection step for selecting question information to be asked to each respondent based on the importance level that is the analysis result obtained by the analysis step; and a question presentation step for presenting, to the respondent terminal of each respondent, the question information for each respondent selected by the question selection step.
(8) An information analysis program that causes a computer to implement the following: a storage function that stores respondent information identifying a respondent and question information asking the respondent questions about duties, personal growth, health, human relationships, support, recognition, philosophy and strategy, organizational culture, and environment as factors that influence the respondent's willingness to contribute; an answer information receiving function that receives from the respondent's terminal the values of each factor that influences the willingness to contribute as answer information; an analysis function that analyzes the answer information received by the answer information receiving function and generates an analysis result indicating which factors the respondent considers important and a level of importance that derives causal relationships from the answer information based on the time series trends of the values of each element in the past from the answer information for each respondent; a question selection function that selects question information to be asked for each respondent based on the level of importance, which is the analysis result obtained by the analysis function; and a question presentation function that presents the question information for each respondent selected by the question selection function to the respondent's terminal for each respondent.

As described above, the present invention can present different questions to each respondent depending on the analysis results of the answers collected from the respondents.

FIG. 2 is a diagram showing an overview of services provided by the information analysis device 1. FIG. 1 is a block diagram showing a configuration of an information analysis device 1. FIG. 2 is a block diagram showing functions of the information analysis device 1. FIG. 2 is a diagram showing an example of a logical configuration of an employee information database 108. FIG. 2 is a diagram showing an example of a logical configuration of a distribution information database 109.

FIG. 2 is a diagram showing an example of a logical configuration of a question information database 110. FIG. 2 is a diagram showing an example of a logical configuration of a selected question information database 111. FIG. 2 is a diagram showing an example of a logical configuration of a response information database 112. 1 is a diagram showing an example of a logical configuration of an activity plan information database 113. FIG. 1 is a diagram showing an example of a logical configuration of an activity evaluation information database 114. FIG.

11 is a sequence chart for explaining a procedure for a human resources officer H to perform an initial setting process using an information analysis system W. 4 is a flowchart for explaining a procedure for a main system 100 to perform a model generation process. FIG. 2 is a block diagram showing a learning device that executes a model generation process. 10 is a flowchart for explaining a procedure for a main system 100 to perform a question selection process. FIG. 2 is a block diagram showing a learner that performs multiple choice question machine learning processing.

10 is a flowchart for explaining a procedure for main system 100 to carry out a questionnaire process. FIG. 13 is a diagram showing an example of a question input screen. 10 is a flowchart for explaining a procedure for main system 100 to perform analysis result output processing. FIG. 2 is a block diagram showing a learning device that executes feature output processing. 13 is a sequence chart for explaining a procedure in which information analysis system W transmits analysis results to respondents A, B, and C.

FIG. 13 is a diagram showing an analysis result output screen for one respondent. FIG. 13 is a diagram showing a group analysis result output screen. 10 is a flowchart for explaining a procedure for main system 100 to perform analysis result output processing. FIG. 13 is a diagram showing an activity plan information input screen. FIG. 13 is a diagram showing an analysis result output screen.

FIG. 13 is a diagram showing an activity evaluation information output screen. 10 is a flowchart for explaining a procedure for main system 100 to perform second question selection processing. 13 is a flowchart for explaining a procedure for main system 100 to carry out a second questionnaire process. FIG. 13 is a diagram showing an activity evaluation information input screen. 10 is a flowchart for explaining a procedure for main system 100 to perform a tallying process.

An information analysis device according to the present invention will be described in detail below with reference to FIGS. 1 to 30, taking as an example an embodiment in which the device is applied to a company that conducts questionnaires with employees as respondents.
[Overview of the embodiment]
The information analysis device 1 of this embodiment accepts response information, which is responses to a questionnaire, from respondents, analyzes the information, and generates an analysis result for each respondent.
The information analysis device 1 selects question information to be asked to each respondent according to the analysis result for each respondent.
The information analysis device 1 displays the question information selected for each respondent on the respondent terminal of each respondent.
According to this embodiment, the questions change for each respondent based on the analysis results obtained by the information analysis device 1, so that the different psychological states of each respondent can be grasped more specifically.

[Details of the embodiment]
First Embodiment
A first embodiment of the present invention will be described in detail with reference to the drawings. FIG. 1 is a diagram showing an outline of a service provided by an information analysis device 1. The information analysis system W of the first embodiment of the present invention is a system in which the information analysis device 1 analyzes responses to a questionnaire collected from respondents A, B, and C, and changes the questions to be sent to the respondents A, B, and C next based on the analysis results. The information analysis system W is configured, for example, by connecting an information analysis device 1, a human resources manager terminal 2, and respondent terminals 3a, 3b, and 3c via a communication network such as the Internet. The communication network is configured, for example, using the Internet, and can perform communication that is concealed from the outside using an appropriate encryption protocol such as SSL (Secure Sockets Layer). In addition, since these devices and terminals are connected via the Internet, they can be used by people and corporations all over the world across borders without being restricted by location.

[Information analysis device 1]
2 is a block diagram showing the configuration of the information analysis device 1. The information analysis device 1 is a device such as a server that presents questionnaire questions, accepts questionnaire responses, tallys up the responses, and analyzes the responses. The information analysis device 1 has three systems, a main system 100, a tallying system 200, and an analysis system 300, each of which is configured by a separate server. Note that the information analysis device 1 has the three systems configured by separate independent servers, but it can also be realized with a single device configuration, or with a system configuration distributed across multiple devices. In addition, the information analysis device 1 assigns certain functions (processing) to certain devices, but it is possible to freely change which functions (processing) of the three systems are assigned to which devices.

[Main System 100]
The main system 100 is a server device that functions as a survey support server, and provides a function of periodically conducting surveys for respondents A, B, and C who are employees of a company. The main system 100 provides a function of having the survey system 200 collect responses and the analysis system 300 analyze the responses in response to a request from a person having a certain authority in the company. The main system 100 also provides a function of presenting different questions to each of the respondents A, B, and C in response to the collected results of the survey system 200 and the analysis results of the analysis system 300. The main system 100 also provides a function of setting an action plan and managing the progress of the action plan in response to a request from a respondent who has confirmed the collected results and the analysis results. The main system 100 may be installed in an organization that conducts surveys, or may be installed in an ASP (Application Service Provider) or cloud managed by a service company that provides the information analysis system W.

The main system 100 includes an employee information database 108, a distribution information database 109, a question information database 110, a selected question information database 111, an answer information database 112, an activity plan information database 113, and an activity evaluation information database 114. The main system 100 uses the employee information database 108, the distribution information database 109, the question information database 110, etc. to distribute questions to respondents who are company employees at appropriate times.

The main system 100 configured in this manner performs the following processes.
(1) Employee information, which is information about company employees, is received in the employee information database 108 from the personnel manager terminal 2.
(2) The distribution information database 109 receives distribution information such as the timing of distribution of questions from the personnel manager terminal 2.
(3) A question selected from the question information database 110 for each of the respondents A, B, and C is stored in the selected question information database 111.
(4) Questions are presented to the respondent terminals 3a, 3b, and 3c for each of the respondents A, B, and C, and responses are requested.
(5) The answers received from the answerer terminals 3a, 3b, and 3c are accepted in the answer information database 112.
(6) The analysis system 300 analyzes the responses to analyze the importance and amount of remaining information.
(7) The importance analyzed by analysis system 300 is stored in importance database 310, and the analyzed remaining information amount is stored in a remaining information amount database.
(8) In response to a request from the human resources officer terminal 2 or the respondent terminal 3a having the specified authority, the tabulation system 200 tabulates the responses and transmits the tabulated results to the human resources officer terminal 2 or the respondent terminal 3a.

2(a) is a block diagram for explaining the hardware configuration of the main system 100. The main system 100 is configured by connecting a CPU (Central Processing Unit) 101, a RAM (Random Access Memory) 102, a ROM (Read Only Memory) 103, a communication control unit 104, an input unit 105, an output unit 106, and a storage unit 107, etc., via a bus line. The CPU 101 is a central processing unit that operates according to a survey support program (not shown) stored in the storage unit 107, and performs information processing such as registering respondents, providing questions to respondents, collecting responses, making a request for compilation to the compilation system 200, making an analysis request to the analysis system 300, setting action programs, and managing the progress of action plans.

RAM 102 is a readable and writable memory that provides a working memory when CPU 101 performs each process. ROM 103 is a read-only memory that stores basic programs and parameters for operating CPU 101. The communication control unit 104 connects main system 100 to the Internet. Through the communication control unit 104, main system 100 can communicate with HR staff terminal 2, respondent terminals 3a, 3b, 3c, tabulation system 200, and analysis system 300.

The input unit 105 is a device for inputting data from the outside, and includes hard keys for receiving input from a keyboard, an external camera for reading predetermined images, and a touch panel for receiving data input from soft keys. The output unit 106 includes an LCD display for displaying a menu screen, images of the tabulation results, images of the analysis results, images related to printing, and the like, and a speaker for issuing instructions and warning sounds. The storage unit 107 is configured using a large-capacity storage medium such as a hard disk or solid state drive, and stores a survey support program, an employee information database 108, a distribution information database 109, a question information database 110, a selected question information database 111, a response information database 112, an activity plan information database 113, an activity evaluation information database 114, and the like.

[Counting system 200]
The counting system 200 is a server device that functions as a questionnaire counting server, and provides a function of counting answers given by respondents A, B, and C. The counting system 200 provides a function of counting by respondent, by question, by predetermined period, by department in the company, and a combination of these, in response to a request from a person with a certain authority in the company (e.g., the human resources officer terminal 2, respondent A). The counting system 200 transmits the counted results to the main system 100 or the analysis system 300 in response to a request from the main system 100. The counting system 200 may be installed in an organization that conducts the questionnaire, or may be installed in an ASP or cloud managed by a service company that provides the information analysis system W.

The counting system 200 includes a counting information database 209 that stores the counting results. The counting system 200 counts the answer information using the answer information stored in the answer information database 112 stored in the main system 100, and stores the answer information in the counting information database 209.

Figure 2(b) is a block diagram for explaining the hardware configuration of the tallying system 200. The tallying system 200 is configured by connecting a CPU 201, a RAM 202, a ROM 203, a communication control unit 204, a storage unit 207, and the like via a bus line. The CPU 201 is a central processing unit that operates according to a tallying program 208 stored in the storage unit 207, and performs information processing such as tallying up response information.

RAM 202 is a readable and writable memory that provides a working memory when CPU 201 performs each process. ROM 203 is a read-only memory that stores basic programs and parameters for operating CPU 201. Communication control unit 204 connects tallying system 200 to the Internet. Via communication control unit 204, tallying system 200 can communicate with HR person terminal 2, respondent terminals 3a, 3b, 3c, main system 100, and analysis system 300.

The memory unit 207 is configured using a large-capacity storage medium such as a hard disk or solid state drive, and stores a compilation program 208, a compilation information database 209, etc.

[Analysis System 300]
The analysis system 300 is a server device that functions as a questionnaire analysis server, and provides a function of analyzing answers given by respondents A, B, and C. The analysis system 300 provides a function of performing analysis for each respondent, each question, each department within a company, and a combination of these, according to a preset analysis time. The analysis system 300 transmits the results of the analysis to the main system 100 in response to a request from the main system 100. The analysis system 300 may be installed within the organization that conducts the questionnaire, or may be installed in an ASP (Application Service Provider) or cloud managed by a service company that provides the information analysis system W.

The analysis system 300 includes an importance database 310 that stores the importance analyzed from the respondent's responses, a remaining information amount database 311, and the like. The analysis system 300 analyzes the response information using the response information stored in the response information database 112.

Figure 2(c) is a block diagram for explaining the hardware configuration of the analysis system 300. The analysis system 300 is configured by connecting a CPU 301, a RAM 302, a ROM 303, a communication control unit 304, an importance analysis means 305, a remaining information amount analysis means 306, and a memory unit 307, etc., via a bus line. The CPU 301 is a central processing unit, and operates according to an importance analysis program 308 and a remaining information amount analysis program 309 stored in the memory unit 307, and performs information processing such as analysis of response information.

RAM 302 is a readable and writable memory that provides a working memory when CPU 301 performs each process. ROM 303 is a read-only memory that stores basic programs and parameters for operating CPU 301. Communication control unit 304 connects analysis system 300 to the Internet. Analysis system 300 can communicate with HR staff terminal 2, respondent terminals 3a, 3b, 3c, main system 100, and tabulation system 200 via communication control unit 304.

The importance analysis program 308 is a learning device for generating an estimation model, acquiring features, and outputting importance. In this embodiment, the importance analysis program 308 performs machine learning according to a learning model constructed from, for example, factor analysis, Bayesian statistics (Bayesian network and information theory), clustering, etc. The importance is information for determining the time series transition of past answers from the answer information of each respondent, and deriving causal relationships from the answer information as to which items the respondent considers important. For example, by acquiring the importance of respondent A, it is possible to determine the time series transition of past answers (what values were entered into which items) from the answer information of respondent A, and derive which items of the nine key drivers respondent A considers important for the company and/or organization.

The remaining information amount analysis program 309 is a learning device for generating an estimation model, acquiring features, and outputting the amount of remaining information. In this embodiment, the remaining information amount analysis program 309 performs machine learning according to a learning model constructed, for example, from factor analysis, Bayesian statistics (Bayesian network and information amount theory), clustering, etc. The amount of remaining information is information for determining how long ago a question was selected from the answer information of each respondent and for determining whether a question to be selected for the respondent is new or old. For example, by acquiring the amount of remaining information of respondent A, it is possible to determine from the answer information of respondent A whether the question was only asked in the previous questionnaire, or whether it was asked a long time ago and should be asked again, and so on, and derive whether the question should be selected for respondent A or not.

The memory unit 307 is configured using a large-capacity storage medium such as a hard disk or solid state drive, and stores an importance analysis program 308, a remaining information amount analysis program 309, an importance database 310, a remaining information amount database 311, etc.

Figure 3 is a block diagram showing the functions of the information analysis device. Figure 3(a) is a block diagram for explaining the functional configuration of the main system 100. The main system 100 includes, for example, a question presentation means, a question selection means, and an answer information reception means. The question presentation means is a means for presenting a question to the respondent terminal. The question selection means is a means for selecting a question according to the importance. The answer information reception means is a means for receiving answer information from the respondent terminal. Figure 3(b) is a block diagram for explaining the functional configuration of the counting system 200. The counting system 200 includes, for example, a counting means. The counting means is a means for performing counting using the answer information.

Figure 3 (c) is a block diagram for explaining the functional configuration of the analysis system 300. The analysis system 300 includes, for example, an importance model generation means, an importance output means, a remaining information amount model generation means, and a remaining information amount output means. The importance model generation means is a means for generating an importance model using the answer information as training data. The importance output means is a means for outputting importance using the importance model and the importance feature. The remaining information amount model generation means is a means for generating a remaining information amount model using the answer information as training data. The remaining information amount output means is a means for outputting the remaining information amount using the remaining information amount model and the remaining information amount feature.

[Personnel Officer Terminal 2]
The human resources staff terminal 2 is a terminal used by the human resources staff H to register the implementation of a survey, determine the subjects to be surveyed, and set the distribution date and time of the survey. The human resources staff terminal 2 is composed of a personal computer or the like equipped with a browser and application software for managing a survey. The human resources staff H is an employee of the company that is the subject of analysis by the information analysis system W, and is a person who is authorized to manage the information analysis system W in the department in charge of human resources. The human resources staff terminal 2 is given a certain authority and can access the information analysis device 1 to view the tabulation results and analysis results of the survey. In addition, the human resources staff terminal 2 can compare and view the tabulation results and analysis results of the survey conducted in the past and the tabulation results and analysis results of the survey conducted this time. In addition, the human resources staff terminal 2 can compare and view the tabulation results and analysis results of the survey conducted in one department and the tabulation results and analysis results of the survey conducted in another department.

[Respondent terminal 3]
The respondent terminal 3 is a terminal used for inputting responses to the questionnaire, and respondent terminals 3a, 3b, and 3c are provided for each of respondents A, B, and C. The respondent terminal 3 is composed of a personal computer equipped with a browser and application software for responding to questionnaires. Respondents A, B, and C are employees of the company that is the subject of analysis by the information analysis system W. The respondent terminals 3a, 3b, and 3c are also used for viewing, setting, evaluating, and the like of action plans.

The respondent terminal 3a is also a terminal used by respondent A who has certain authority as a company administrator (e.g., manager, etc.). The respondent terminal 3a is granted certain authority and can access the information analysis device 1 to analyze survey results and view the analysis results. The respondent terminal 3a can also compare and view the analysis results of surveys conducted in the past and the analysis results of the current survey. Furthermore, the respondent terminal 3a can compare and view the analysis results of surveys conducted in the department to which the respondent belongs and the analysis results of surveys conducted in other departments.

[Employee Information Database 108]
4 is a diagram showing an example of the logical configuration of employee information database 108. Note that this diagram is an example for the purpose of explanation, and the actual employee information database 108 is configured as a relational database that is logically related to other databases, etc.

The employee information database 108 is composed of items such as "no", "ID", "name", "email address", "gender", "date of birth", "job type", "position", "employment category", "new graduate/mid-career", "date of joining", "group", "department", "management authority", "group authority", etc. "no" is an item for registering the order in which employees were registered in the information analysis device 1, and each item is registered with a number indicating the order in which they were registered. "ID" is an item for uniquely identifying employees of the company registered in the information analysis device 1, and unique ID information assigned to each input name is registered. "Name" is an item for registering the names of employees of the company in the information analysis device 1, and the input name is registered. Both names in Japanese for domestic use and names in foreign languages for overseas use can be registered.

"Email address", "Gender", "Date of birth", "Job type", "Position", "Employment category", "New graduate/mid-career", "Date of joining", "Affiliated group", and "Department" are items for registering personal information of employees of the company registered in the information analysis device 1, and in this example, the personal information of each employee entered by human resources officer H in a specified format is registered. "Administrative authority" and "Group authority" are items that indicate whether or not employees of the company registered in the information analysis device 1 have the authority to use specified functions of the information analysis device 1, and it is registered that only specified employees have been given the authority.

Other employee information registered in the employee information database 108 includes the employee's face photo, self-introduction (profile), credit card information for paying system usage fees to the operator of the information analysis device 1, the URL (Uniform Resource Locator) of the employee's personal website (homepage), the address of the SNS (Social Networking Service), etc.

[Distribution information database 109]
Fig. 5 is a diagram showing an example of the logical configuration of distribution information database 109. Note that this diagram is an example for the purpose of explanation, and the actual distribution information database 109 is configured as a relational database that is logically related to other databases, etc.

The distribution information database 109 is composed of items such as "Distribution ID", "Registration date and time", "Distribution frequency", "Distribution timing", "Response period", "Automatic reminder", and "Initial distribution of all items". "Distribution ID" is an item of distribution information uploaded and registered by the human resources person H in the information analysis device 1, and ID information assigned to the distribution information in the information analysis device 1 is registered. "Registration date and time" is an item of the registration date and time when the human resources person H uploaded and registered the distribution information in the information analysis device 1, and information on the time measured by a clock installed in the information analysis device 1 is registered. "Distribution frequency" is an item indicating the distribution frequency of questions determined by the human resources person H, and information on the distribution frequency entered from the human resources person terminal 2 is registered.

"Delivery timing" is an item that indicates the date and time when the question determined by the human resources person H will be delivered, and information on the delivery timing entered from the human resources person terminal 2 is registered. "Response period" is an item that indicates the response period for the questionnaire determined by the human resources person H, and information on the response period entered from the human resources person terminal 2 is registered. "Automatic reminder" is an item that indicates the timing for sending a reminder to the respondent terminals 3a, 3b, and 3c determined by the human resources person H, and information on the automatic reminder entered from the human resources person terminal 2 is registered. "Initial delivery of all items" is an item that allows the human resources person H to select the question to be sent initially to the respondent terminals 3a, 3b, and 3c determined by the human resources person H, and information to the effect that all items will not be delivered initially entered from the human resources person terminal 2 is registered.

[Question Information Database 110]
6 is a diagram showing an example of the logical configuration of the question information database 110. Note that this diagram is an example for the purpose of explanation, and the actual question information database 110 is configured as a relational database that is logically related to other databases, etc.

The question information database 110 is composed of items such as "no", "frequency", "question", "item", and "answer selection item (weight)". "No" is an item for registering the order in which questions were registered in the information analysis device 1, and each item is registered with a number indicating the order in which they were registered. "Frequency" is a value indicating the frequency at which questions are sent to employees who will answer questions, and a predetermined value is registered according to the content and nature of the "question" or "item". "Question" is an item of a question sent to employees who will answer questions, and registered are questions to ascertain the employee's psychological state regarding work and/or the organization, questions to ascertain the willingness to contribute to work and/or the organization, and questions to ascertain the elements of willingness to contribute that the employee considers important.

"Items" are items that indicate nine elements (key drivers) into which questions are divided according to predetermined criteria, and are registered as "Duties," "Personal Growth," "Health," "Interpersonal Relationships," "Support," "Recognition," "Philosophy Strategy," "Organizational Culture," and "Environment." "Items" are elements that influence motivation to contribute to work and/or the organization, and they measure what the respondent considers important from these nine key drivers, and the measurement result is calculated as a score. "Answer Selection Items (Weights)" are items that the respondent selects based on their answer to the question, and numbers from 1 to 7 are registered.

In this example, the "questions" registered in the question information database 110 present the respondents with a number of questions to gauge the respondent's psychological state regarding their work and/or the organization, the respondent's willingness to contribute to their work and/or the organization, and the factors that each respondent considers important, and answers to the questions are accepted. The questions presented to the respondents are intended to gauge the importance of what motivates the respondent to contribute to an organization such as a company through factors such as "job," "personal growth," "health," "human relationships," "support," "recognition," "philosophy and strategy," "organizational culture," and "environment."

[Selected Question Information Database 111]
7(a) and (b) are diagrams showing an example of the logical configuration of the selected-question information database 111. Note that this diagram is an example for the purpose of explanation, and the actual selected-question information database 111 is configured as a relational database that is logically related to other databases, etc.

The selected question information database 111 is composed of items such as "Question date and time," "Answer date and time," "ID," and "Question 1" to "Question 16." The items "Question 1" to "Question 16" are further divided into items "Question" and "Answer." The "Question date and time" is an item indicating the date and time when the information analysis device 1 distributes a question to the respondent, and each of these items is registered as a date and time linked to the "distribution timing" registered in the distribution information database 109. The "Answer date and time" is an item indicating the date and time when the information analysis device 1 receives an answer from the respondent, and each of these items is registered as information on the time measured by a clock provided in the information analysis device 1. The "ID" is an item indicating the employee who will be the respondent to whom the information analysis device 1 distributes a question, and each of these items is registered as the ID of the employee linked to the "ID" registered in the employee information database 108.

"Question" from "Question 1" to "Question 16" is an item indicating the question that the information analysis device 1 will deliver to the respondent, and each of these is registered with a question associated with a "no" selected from the question information database 110. "Answer" from "Question 1" to "Question 16" is an item indicating the value of the answer to the question delivered to the respondent, and each is registered with a value (1 to 7) input from the respondent terminal 3a, 3b, 3c.

As shown in Figures 7(a) and (b), "Questions 1" to "16" are selected from the question information database 110 according to predetermined conditions, and the questions to be registered are different. For example, when a company is using the information analysis system W for the first time and wants to ask certain predetermined questions to employees (respondents), the questions are selected from the "no" questions in the question information database 110 so that the same questions are delivered, as shown in Figure 7(a).

On the other hand, if the company is already using the information analysis system W and has a certain amount of data for each employee, as shown in FIG. 7(b), the question selection means selects from the "no" questions selected from the question information database 110 so that different questions are delivered to each employee. Also, even if the company is using the information analysis system W for the first time, if the company does not want to ask employees certain predetermined questions, the questions are selected from the "no" questions in the question information database 110 so that random questions are delivered according to the "frequency" of the question information database 110.

[Response information database 112]
8(a) and (b) are diagrams showing an example of the logical configuration of the response information database 112. Note that this diagram is an example for the purpose of explanation, and the actual response information database 112 is configured as a relational database that is logically related to other databases, etc.

The answer information database 112 is composed of items such as "Question date and time", "Answer date and time", "ID", "Question 1" to "Question 16". The items "Question 1" to "Question 16" are further divided into items "Question" and "Answer". "Question date and time" is an item indicating the date and time when the information analysis device 1 distributes a question to the respondent, and each date and time linked to the "distribution timing" registered in the distribution information database 109 is registered. "Answer date and time" is an item indicating the date and time when the information analysis device 1 receives an answer from the respondent, and each time information measured by a clock installed in the information analysis device 1 is registered. "ID" is an item indicating the employee who will be the respondent to whom the information analysis device 1 distributes a question, and each employee ID linked to the "ID" registered in the employee information database 108 is registered.

"Question" from "Question 1" to "Question 16" is an item indicating the question that the information analysis device 1 will send to the respondent, and each of these is registered with a question associated with a "no" selected from the question information database 110. "Answer" from "Question 1" to "Question 16" is an item indicating the value of the answer to the question sent to the respondent, and each is registered with a value (1 to 7) input from the respondent terminals 3a, 3b, and 3c.

In this example, for the sake of convenience, the selected question information database 111 and the answer information database 112 are described as separate databases. This is to clearly explain the situation in which a question is selected for each answerer and the situation in which the answerer answers the question. However, it goes without saying that the selected question information database 111 and the answer information database 112 may be the same database for collecting information.

[Activity plan information database 113]
Fig. 9 is a diagram showing an example of the logical configuration of the activity plan information database 113. Note that this diagram is an example for the purpose of explanation, and the actual activity plan information database 113 is configured as a relational database that is logically related to other databases, etc.

The activity plan information database 113 is composed of items such as "activity ID", "activity plan", "creator ID", "belonging group", "rating (1)" to "rating (3)". The items "rating (1)" to "rating (3)" are further divided into items of "high rating", "low rating" and "improvement plan". The "activity ID" is an item that uniquely identifies the activity plan registered in the information analysis device 1, and ID information assigned to each inputted activity plan is registered. The "activity plan" is an item for registering an activity plan in the information analysis device 1, and activity plans inputted from the respondent terminals 3a, 3b, and 3c are registered. The "creator ID" is an item that indicates the employee who inputted the activity information to the information analysis device 1, and the ID of the employee linked to the "ID" registered in the employee information database 108 is registered. The "belonging group" is an item that indicates the group to which the employee who inputted the activity information belongs, and information linked to the "belonging group" registered in the employee information database 108 is registered.

The "High Rating" in "Rating (1)" to "Rating (3)" is an item that indicates the rating of the answer to the activity information distributed to the respondent, and the ratings entered from the respondent terminals 3a, 3b, and 3c are registered, respectively. The "Low Rating" in "Rating (1)" to "Rating (3)" is an item that indicates the rating of the answer to the question distributed to the respondent, and the ratings entered from the respondent terminals 3a, 3b, and 3c are registered, respectively. The "Improvement Suggestion" in "Rating (1)" to "Rating (3)" is an item that indicates the rating of the answer to the question distributed to the respondent, and the ratings entered from the respondent terminals 3a, 3b, and 3c are registered, respectively.

In this example, for ease of explanation, the activity plan information database 113 is illustrated with different items registered as shown in Figures 9(a) and (b). This is to explain a situation where no activity plan information has been input and a situation where activity plan information has already been input.

[Activity evaluation information database 114]
10 is a diagram showing an example of the logical configuration of the activity evaluation information database 114. Note that this diagram is an example for the purpose of explanation, and the actual activity evaluation information database 114 is configured as a relational database that is logically related to other databases, etc.

The activity evaluation information database 114 is composed of items such as "Activity ID", "Activity Plan", "Creator ID", "Rating (1)" to "Rating (3)". The items "Rating (1)" to "Rating (3)" are further divided into items of "High Rating", "Low Rating" and "Improvement Suggestion". "Activity ID" is an item that uniquely identifies an activity plan registered in the information analysis device 1, and ID information assigned to each input activity plan is registered. "Activity Plan" is an item that registers an activity plan in the information analysis device 1, and activity plans input from respondent terminals 3a, 3b, 3c are registered.

"Creator ID" is an item that indicates the employee who has input activity information by the information analysis device 1, and each employee ID linked to the "ID" registered in the employee information database 108 is registered. "Group" is an item that indicates the group to which the employee who inputted the activity information belongs, and each information linked to the "Group" registered in the employee information database 108 is registered.

The "High Rating" in "Rating (1)" to "Rating (3)" is an item that indicates the rating of the answer to the activity information distributed to the respondent, and the ratings entered from the respondent terminals 3a, 3b, and 3c are registered, respectively. The "Low Rating" in "Rating (1)" to "Rating (3)" is an item that indicates the rating of the answer to the question distributed to the respondent, and the ratings entered from the respondent terminals 3a, 3b, and 3c are registered, respectively. The "Improvement Suggestion" in "Rating (1)" to "Rating (3)" is an item that indicates the rating of the answer to the question distributed to the respondent, and the ratings entered from the respondent terminals 3a, 3b, and 3c are registered, respectively.

In this example, the activity plan information database 113 and the activity evaluation information database 114 are described as separate databases for the sake of convenience. This is to clearly explain the situation in which an activity plan is input for each respondent and the situation in which the respondent evaluates the input activity plan. However, it goes without saying that the activity plan information database 113 and the activity evaluation information database 114 may be the same database to collect information. Also, in this example, for the sake of convenience, the activity evaluation information database 114 is illustrated as having different items registered therein, as shown in Figures 10(a) and (b). This is to explain the situation in which no activity evaluation information has been input and the situation in which activity evaluation information has already been input.

[Total information database 209]
The aggregate information database 209 is configured by dividing the items according to the aggregation results of the answer information in the answer information database 112 (not shown). The aggregate information database 209 stores the aggregation results obtained by aggregating values obtained from answers to questions from respondents. The aggregate information database 209 may store aggregation results according to each respondent, each question, each organization, each department of an organization, each attribute of the respondent, or the like. The aggregate information database 209 may store an average value of the aggregation results of each item for each group indicating the characteristics of the diagnostic area.

[Importance Database 310]
The importance database 310 is configured by dividing items according to the answer information of the answer information database 112 (not shown). The importance database 310 stores the importance of each answerer by analyzing values obtained from the answer information. The importance database 310 may store the importance according to each answerer, each question, each organization, each department of an organization, each attribute of the answerer, etc.

[Residual information amount database 311]
The remaining information amount database 311 is configured by dividing items according to the answer information of the answer information database 112 (not shown). The remaining information amount database 311 stores the remaining information amount for each answerer by analyzing the freshness of information obtained from the answer information. The remaining information amount database 311 may store the remaining information amount according to each answerer, each question, each organization, each department of an organization, each attribute of the answerer, etc.

[Operation of Information Analysis System W]
Next, a description will be given of the operation of the information analysis system W configured as above. The operation of various processes described below is realized by the CPU 101 of the main system 100, the CPU 201 of the tabulation system 200, the CPU 301 of the analysis system 300, etc., executing various programs stored in the respective storage units 107, 207, 307.

[Initial setting process]
11 is a sequence chart for explaining the procedure for the initial setting process performed by the personnel manager H using the information analysis system W. The initial setting process is a process for making initial settings for using the information analysis system W. The following process is performed by the CPU 101 and the CPU of the personnel manager terminal 2 in accordance with the initial setting program.

[System Registration Request SC1]
First, the CPU of the personnel manager terminal 2 accesses the top screen of the portal site of the main system 100 through the browser operation of the personnel manager H. The main system 100 then accepts the access via the communication control unit 104, generates top screen data for displaying the top screen in the RAM 102, and transmits this to the personnel manager terminal 2. When the CPU of the personnel manager terminal 2 receives the top screen data, it stores it in the RAM and uses it to display the top screen on the output unit.

When the registration button on the top screen is selected by HR person H, the CPU of the HR person terminal 2 sends a request for HR person H registration screen data to the main system 100. Furthermore, upon receiving this request, the CPU 101 generates registration screen data for HR person H in the RAM 12 and sends it to the HR person terminal 2. When the CPU of the HR person terminal 2 receives the HR person H registration screen data from the main system 100, it stores it in the RAM and uses it to display the HR person H registration screen on the output section.

When the human resources officer H enters employee information into the human resources officer H registration screen, the CPU of the human resources officer terminal 2 stores the employee information in RAM. When the human resources officer H selects, for example, a registration button, the CPU of the human resources officer terminal 2 transmits the employee information stored in RAM to the main system 100. When the CPU 101 receives the employee information from the human resources officer terminal 2, it registers it in the employee information database 108.

[ID/PW assignment SC2]
The CPU 101 generates a unique ID and password PW for the human resources person H and registers them in the employee information database 108. The CPU 101 sends the generated ID and password PW by email to the human resources person terminal 2. Through the above process, the human resources person H is given a unique ID and password, and becomes able to log in to his/her personal page in the information analysis system W.

[Receiving respondent information SC3]
When the personnel manager H returns to the top screen by operating the browser, enters the ID and password PW of the personnel manager H in the ID and password fields and selects the login button, the CPU of the personnel manager terminal 2 transmits the ID and password PW of the personnel manager H to the main system 100 and requests the personnel manager H to log in to his/her personal page. The CPU 101 stores the ID and password of the personnel manager H in the RAM 12 and performs the login process by comparing these with the ID and password PW of the personnel manager H stored in the employee information database 108.

After authenticating the human resources person H, the CPU 101 generates in the RAM 12 MyPage screen data for displaying a MyPage screen for the human resources person H from the information registered in the employee information database 108. The CPU 101 then transmits the generated MyPage screen data to the human resources person terminal 2.

The CPU of the HR person terminal 2 receives the My Page screen data from the main system 100 and stores it in RAM. The CPU of the HR person terminal 2 then displays the My Page screen 501 on the output section using the stored My Page screen data. When HR person H selects, for example, the employee registration button on the My Page screen 501, the CPU of the HR person terminal 2 sends a request for employee registration screen data to the main system 100. When the CPU 101 receives the request, it generates employee registration screen data and sends it to the HR person terminal 2.

When the CPU of the personnel manager terminal 2 receives employee registration screen data from the main system 100, it stores it in the RAM. Then, it displays the employee registration screen on the output section using the stored employee registration screen data. On the employee registration screen, items such as "Name", "Email address", "Gender", "Date of birth", "Occupation", "Job title", "Employment category", "New graduate/mid-career", "Date of joining the company", "Group", "Department", "Administrative authority", and "Group authority" are displayed so that they can be entered for registration in the employee information database 108. The personnel manager H may enter these items from the input section of the personnel manager terminal 2, or may enter them by exporting data from a spreadsheet software that has been prepared in advance. When the personnel manager H enters employee information from the employee registration screen, the CPU of the personnel manager terminal 2 stores it in the RAM. Then, when the personnel manager H selects the registration button on the employee registration screen, the CPU of the personnel manager terminal 2 transmits the employee information stored in the RAM to the main system 100.

When the CPU 101 receives employee information from the human resources officer terminal 2, it registers it in the employee information database 108 as shown in FIG. 4. The CPU 101 generates a unique ID and password PW for each employee as a respondent and registers them in the employee information database 108. The CPU 101 sends the generated ID and password PW by email to the respondent terminals 3a, 3b, and 3c. Respondents A, B, and C, who are employees, are given a unique ID and password and are able to log in to their personal pages in the information analysis system W. In this way, the main system 100 is equipped with an employee information registration means that receives and registers employee information, which is information about employees who will be respondents, from the human resources officer terminal 2.

[Receiving distribution information SC4]
When the personnel manager H selects, for example, a distribution information registration button on the My Page screen 501, the CPU of the personnel manager terminal 2 transmits a request for distribution information registration screen data to the main system 100. When the CPU 101 receives the request, it generates distribution information registration screen data and transmits it to the personnel manager terminal 2. When the CPU of the personnel manager terminal 2 receives the distribution information registration screen data from the main system 100, it stores it in the RAM. Then, it displays the distribution information registration screen on the output section using the stored distribution information registration screen data.

The distribution information registration screen displays items such as "distribution frequency," "distribution timing," "answer period," "automatic reminder," and "initial distribution of all items" that can be entered to be registered in the distribution information database 109. When personnel H enters distribution information from the distribution information registration screen, the CPU of the personnel terminal 2 stores it in RAM. Then, when personnel H selects the registration button on the distribution information registration screen, the CPU of the personnel terminal 2 transmits the distribution information stored in RAM to the main system 100.

When the main system 100 receives distribution information from the human resources personnel terminal 2, it registers these items in the distribution information database 109 as shown in FIG. 5. Also, when the main system 100 receives distribution information from the human resources personnel terminal 2, it generates a unique distribution ID as shown in FIG. 5 and registers it in the "distribution ID" of the distribution information database 109. Furthermore, when the main system 100 receives distribution information from the human resources personnel terminal 2, it measures the time using a clock installed in it and registers it in the "registration date and time" of the distribution information database 109. In this way, the main system 100 is equipped with a distribution information registration means that receives and registers distribution information, which is information on the distribution of the questionnaire, from the human resources personnel terminal 2.

[Model generation process]
FIG. 12 is a flow chart for explaining the procedure of the main system 100 performing the model generation process. The model generation process is a learning phase in which the response information for each respondent stored in the response information database 112 is input and analyzed to obtain the relationship between the response information and a predetermined feature amount. The CPU 101 monitors whether the current time has reached the model generation time (S1). That is, the CPU 101 obtains the current time from a clock provided therein, and monitors whether the current time has reached the date and time of the "timing of model generation" registered in the storage unit 107. When the current time reaches the date and time of the "timing of model generation" (S101: Y), the CPU 101 requests the analysis system 300 to execute the importance model generation process (S2) and the residual information estimator model process (S3). When the analysis system 300 finishes the model generation process, the CPU 101 again monitors whether the current time has reached the analysis time (S1).

[Importance model generation process]
FIG. 13A is a block diagram showing a learning device that executes the importance model generation process. The importance model generation process is a learning phase that obtains the relationship between the answer information, which is input data, and the importance, which is output data. In response to a request from the CPU 101, the CPU 301 inputs the historical answer information of all respondents stored in the answer information database 112 to the learning device. The learning device obtains the relationship between the answer information and the importance by analyzing the input historical answer information of all respondents. The learning device acquires a model regarding the importance by going through such a learning phase, that is, acquires parameters that express the relationship between the answer information of all respondents and the importance. By acquiring such a model regarding the importance (hereinafter referred to as the "importance model"), the learning device becomes able to predict the importance based on a certain standard from the answer information of one answerer. In this way, the analysis system 300 is provided with an importance model generation means that uses the answer information of all respondents as teacher data and generates an estimation model by machine learning that estimates parameters that express the relationship between the answer information of all respondents and the importance.

[Residual information amount model generation process]
FIG. 13B is a block diagram showing a learning device that executes the remaining information amount model generation process. The remaining information amount model generation process is a learning phase that obtains the relationship between the answer information, which is input data, and the remaining information amount, which is output data. In response to a request from the CPU 101, the CPU 301 inputs the historical answer information of all respondents stored in the answer information database 112 to the learning device. The learning device obtains the relationship between the answer information and the remaining information amount by analyzing the input historical answer information of all respondents. By going through such a learning phase, the learning device obtains a model regarding the remaining information amount, that is, obtains parameters that express the relationship between the answer information of all respondents and the remaining information amount. By obtaining such a model regarding the remaining information amount (hereinafter referred to as the "remaining information amount model"), the learning device becomes able to predict the remaining information amount based on a certain standard from the answer information of one answerer. In this way, the analysis system 300 is provided with a remaining information amount model generation means that uses the answer information of all respondents as teacher data and generates an estimation model by machine learning that estimates parameters that express the relationship between the answer information of all respondents and the remaining information amount.

[Question selection process]
FIG. 14 is a flow chart for explaining the procedure of the question selection process performed by the main system 100. The question selection process is a process for selecting question information to be asked to each respondent. The CPU 101 monitors whether the current time has reached the distribution date and time (S101). That is, the CPU 101 obtains the current time from a built-in clock and monitors whether the current time has reached the date and time of the "distribution timing" registered in the distribution information database 109. When the current time reaches the date and time of the "distribution timing" (S101: Y), the CPU 101 judges whether the question to be presented is the first distribution or not (S102). That is, when a company using the information analysis system W uses this system for the first time, the importance and remaining information amount for each respondent have not been learned, and therefore an appropriate question may not be selected, so it is necessary to select a question prepared in advance.

If it is not the first distribution (S102: N), the CPU 101 requests the analysis system 300 to execute the selective question machine learning process (S103). If it is the first distribution (S102: Y), the CPU 101 judges whether or not "initial all items distribution" in the distribution information database 109 is ON (S104). If "initial all items distribution" is ON (S104: Y), the CPU 101 executes the basic question output process (S105). That is, the CPU 101 reads the question information of 32 basic questions "q1" to "q32" from the "no" items registered in the question information database 110, registers them in the selective question information for each respondent (for example, see FIG. 7(a)), and transmits them for each respondent. In this way, by having a relatively large number of initial questions at 32, the importance and remaining information amount can be obtained reliably as basic data for each respondent.

When "Initial all items delivery" is OFF (S104: N), the CPU 101 executes the frequency question output process (S106). That is, the CPU 101 registers 16 randomly selected question information from the question information database 110 in the selected question information for each respondent according to the "frequency" value registered in the question information database 110, and transmits it to each respondent. Depending on the company, if there are 32 initial questions, employees may not respond proactively. Therefore, by reducing the number of questions to 16, the importance and remaining information amount that are basic data for each respondent can be obtained slowly and steadily. When executing the frequency question output process (S106), the CPU 101 randomly selects the 16 questions from the question information database 110 evenly according to the nine key drivers. In this way, by evenly selecting questions according to the nine key drivers, the average importance and remaining information amount can be obtained.

When the CPU 101 finishes executing either the selective question machine learning process (S103), the basic question output process (S105), or the frequency question output process (S106), it again monitors whether the "distribution timing" date and time has been reached (S101). In this way, the main system 100 is equipped with a question selection means that selects question information to be asked to each respondent according to the analysis results obtained by the analysis means.

[Multiple Choice Question Machine Learning Processing S103]
More specifically, the multiple-choice question machine learning process (S103) shown in FIG. 14 is composed of two processes: a multiple-choice question model generation process shown in FIG. 15(a) and a multiple-choice question output process shown in FIG. 15(b). FIG. 15(a) is a block diagram showing a learning device that executes the multiple-choice question model generation process. The multiple-choice question model generation process is a learning phase that obtains the relationship between the importance and remaining information amount, which are input data, and the multiple-choice question information, which is output data. In response to a request from CPU 101, CPU 301 inputs the importance of all respondents stored in the importance database 310 and the remaining information amount of all respondents stored in the remaining information amount database 311 to the learning device.

The learning device obtains the relationship between the importance and the remaining information amount and the selected question information by analyzing all the input importances and all the remaining information amounts. The learning device obtains a model for the selected question information by going through such a learning phase, that is, obtains parameters that express the relationship between the importance and the remaining information amount of all respondents and the selected question information. By obtaining such a model for the selected question information (hereinafter referred to as the "selected question information model"), the learning device becomes able to select the selected question information based on a certain criterion from the importance and the remaining information amount of each respondent. In this way, the analysis system 300 is provided with a selected question model generation means that uses the importance of all respondents and/or the remaining information amount of all respondents as teacher data and generates an estimation model by machine learning to estimate parameters that express the relationship between the importance of all respondents and/or the remaining information amount of all respondents and the selected question information of all respondents.

Figure 15 (b) is a block diagram showing a learning device that executes the selective question output process. The selective question output process is a prediction phase in which selective question information for one answerer is obtained, predicted through a selective question information model, from the input data, the importance of one answerer and the amount of remaining information of one answerer. In response to a request from CPU 101, CPU 301 inputs the importance of one answerer stored in the importance database 310 and the amount of remaining information of one answerer stored in the remaining information amount database 311 to the learning device. The learning device obtains selective question information for one answerer by analyzing the input importance of one answerer and the amount of remaining information of one answerer. By using the selective question information model obtained in this way in the selective question model generation process, the learning device can select selective question information based on a certain criterion from the importance and remaining information amount of each answerer.

In this way, the analysis system 300 includes a selective question output means for outputting selective question information for one respondent by analyzing the importance of one respondent and/or the amount of remaining information of one respondent using the estimation model generated by the selective question model generation means. The CPU 101 generates the selective question information for the 16 questions obtained by the CPU 301 from the learning device in the RAM 102 for transmission to each respondent. Note that the CPU 101 may register the selective question information for the 16 questions obtained by the CPU 301 from the learning device in the selective question information database 111 (see, for example, FIG. 7(b)), and then generate the selective question information in the RAM 102 for transmission to each respondent.

[Survey processing]
16 is a flow chart for explaining the procedure for the main system 100 to process a survey. The survey process is a process for sending different questions to each respondent and receiving the responses from each respondent. The following process is performed by the CPU 101 and the CPU of each of the respondent terminals 3a, 3b, and 3c according to a predetermined program.

[Question information transmission SC101]
First, the CPU 101 generates question input screen data for each respondent in the RAM 102 from the selected question information generated in the RAM 102 in the selected question machine learning process (S103), the basic question output process (S105), or the frequency question output process (S106) in the question selection process, and transmits a notice that there is a survey to each of the respondent terminals 3a, 3b, and 3c. In this example, the CPU 101 notifies the respondent terminals 3a, 3b, and 3c that there is a survey and includes a URL for transitioning to the question input screen of the survey in an email, thereby notifying the respondent terminals 3a, 3b, and 3c of the presence of the survey. The main system 100 may notify the respondent terminals 3a, 3b, and 3c of the URL for transitioning to the question input screen by a chatbot, SNS, or the like other than email.

When the URL written in the email is selected by the respondent terminal 3a, 3b, 3c, the CPU of the respondent terminal 3a, 3b, 3c sends a request for question input screen data for each respondent A, B, C to the main system 100. Furthermore, upon receiving the request, the CPU 101 sends the question input screen data for each respondent A, B, C to the respondent terminal 3a, 3b, 3c. When the CPU of the respondent terminal 3a, 3b, 3c receives the question input screen data for each respondent A, B, C from the main system 100, it stores it in RAM and uses this to display the question input screen for each respondent A, B, C on the output section (see FIG. 17). In this way, the main system 100 is equipped with a question presentation means for presenting the question information for each respondent selected by the question selection means to the respondent terminal for each respondent.

[Question input screen]
FIG. 17 is a diagram showing an example of a question input screen. The question input screen is a page where the respondents A, B, and C input answers to the questions selected by the analysis system 300 and transmitted by the main system 100 to each of the respondent terminals 3a, 3b, and 3c. The answer information stored in the answer information database 112 is updated according to the content input here (see FIG. 8). The question output field 171 is a field where question information linked to the selected question information selected for each respondent is output. The answer input field 172 is a field where the respondent inputs according to the question output in the question output field 171. The answer input field is a field where data is input to determine for each respondent, such as the state of mind of each respondent toward the work and/or organization, the state of each respondent's willingness to contribute to the work and/or organization, and the elements that each respondent considers important, based on the answer to the selected question.

In this embodiment, seven ranks from "I don't think so at all" to "I think so very much" can be input in the answer input field by selecting buttons 172a-172g, but other ranks (7 or less, 7 or more) may also be set. Respondents A, B, and C select buttons 172a-172g from the input section of respondent terminals 3a, 3b, and 3c. When respondents A, B, and C input answers from the question input screen, the CPUs of respondent terminals 3a, 3b, and 3c store the answers in RAM. The input answers are converted into numerical values according to the selected button 172a-172g, for example, according to conversion data stored in the storage device of respondent terminals 3a, 3b, and 3c (for example, button 172a=1, button 172b=2, button 172c=3, button 172d=4, button 172e=5, button 172f=6, button 172g=7, etc.).

[Response information reception SC102]
When the respondents A, B, and C answer all the questions and select a confirmation button (not shown) to check for input errors, for example, the CPUs of the respondent terminals 3a, 3b, and 3c transmit the answer information stored in the RAM to the main system 100. When the CPU 101 receives the answer information for each of the respondents A, B, and C from the respondent terminals 3a, 3b, and 3c, it registers it in the answer information database 112 as shown in Figure 8(a) or 8(b).

The CPU 101 monitors whether the current time has reached the time for a reminder (not shown). That is, the CPU 101 obtains the current time from a built-in clock and monitors whether the current time has reached the date and time of the "automatic reminder" registered in the distribution information database 109. When the current time reaches the date and time of the "automatic reminder", the CPU 101 generates reminder screen data in the RAM 102 and transmits it to the respondent terminals 3a, 3b, and 3c. In this way, the main system 100 is provided with an answer information receiving means for receiving answer information provided by each of respondents A, B, and C from the respondent terminals 3a, 3b, and 3c.

[Notification of end of response period SC103]
When all respondents have completed their responses or the response period has expired, the CPU 101 generates response period end notification screen data in the RAM 102 to display a response period end notification screen, and transmits this to a person with a certain level of authority (e.g., the personnel manager terminal 2, respondent A). For example, when the CPU of the personnel manager terminal 2 or the respondent terminal 3a receives the response period end notification screen data, it stores it in the RAM and uses it to display the response period end notification screen on the output section.

[Analysis result output processing]
18 is a flowchart for explaining the procedure for analysis result output processing performed by main system 100. The analysis result output processing is processing for outputting analysis results for each respondent in response to a request from each respondent. CPU 101 monitors whether or not there is a request for analysis from the respondent (S201). If there is a request for analysis from the respondent (S201: Y), CPU 101 requests analysis system 300 to execute feature output processing (S202). When main system 100 finishes feature output processing (S202), CPU 101 again monitors whether or not there is a request for analysis from the respondent (S201).

In this example, the analysis result output process outputs the analysis results for each respondent in response to the request of each respondent. However, if a group is created with multiple employees within an organization and it is desired to output the analysis results for that group, the analysis results for the group may be output at a specified time. For example, when the survey is completed, the analysis results for each group may be output to the employees belonging to that group.

[Feature output processing]
19A is a block diagram showing a learning device that executes the importance output process. The importance output process is a prediction phase that obtains the importance of one answerer predicted through an importance model from the answer information of one answerer, which is input data. In response to a request from CPU 101, CPU 301 inputs the answer information of one answerer stored in the answer information database 112 to the learning device. The learning device obtains the importance of one answerer by analyzing the input answer information of one answerer. The learning device can predict the importance of each answerer by using the importance model obtained in this manner by the importance model generation process.

In other words, by predicting the importance of a respondent, it is possible to predict the respondent's psychological state regarding the job and/or organization, willingness to contribute, and factors that he or she considers important. In this way, the analysis system 300 is equipped with an importance output means that outputs the importance of a respondent by analyzing the response information of the respondent using the estimation model generated by the importance model generation means. The CPU 301 registers the importance obtained from the learning device in the importance database 310.

In this way, the analysis system 300 includes an analysis means for analyzing the response information received by the response information receiving means and generating analysis results for each respondent. The analysis system 300 includes a psychological state estimation means for estimating the psychological state of each respondent according to the analysis results obtained by the analysis means. The analysis system 300 also includes a willingness to contribute estimation means for estimating the willingness to contribute to the work and/or organization of each respondent according to the psychological state estimated by the psychological state estimation means. The analysis system 300 also includes an element estimation means for estimating the elements that each respondent considers important according to the contribution estimated by the willingness to contribute estimation means.

Figure 19 (b) is a block diagram showing a learning device that executes the remaining information amount output process. The remaining information amount output process is a prediction phase that obtains the remaining information amount of one respondent predicted through the remaining information amount model from the answer information of one respondent, which is the input data. In response to a request from CPU 101, CPU 301 inputs the answer information of one respondent stored in answer information database 112 to the learning device. The learning device obtains the remaining information amount of one respondent by analyzing the input answer information of one respondent. By using the remaining information amount model obtained in this manner by the remaining information amount model generation process, the learning device can predict the remaining information amount for each respondent.

In other words, by predicting the remaining information amount of a respondent, it is possible to predict the questions that the respondent has answered in the past and to predict the questions that the respondent should answer. In this way, the analysis system 300 is equipped with a remaining information amount output means that outputs the remaining information amount of a respondent by analyzing the answer information of the respondent using the estimation model generated by the remaining information amount model generation means. The CPU 301 registers the remaining information amount obtained from the learning device in the remaining information amount database 311.

[Analysis result transmission process]
20 is a sequence chart for explaining the procedure by which information analysis system W transmits analysis results to respondents A, B, and C. The analysis result transmission process is a process for transmitting the analysis results for each respondent to each respondent. The following process is performed by CPU 101 and the CPU of each of respondent terminals 3a, 3b, and 3c in accordance with a predetermined program.

[Transmit Analysis Result SC201]
When the CPU 101 analyzes the importance of one respondent in the feature output process, it generates analysis result output screen data for the one respondent in RAM 102 according to the importance of the one respondent registered in the importance database 310, and transmits it to the respondent terminal 3a of one respondent A. When the CPU of the respondent terminal 3a receives the analysis result output screen data, it stores it in RAM and uses it to display the analysis result output screen on the output section. The same process is performed for the other respondents, Respondent B and Respondent C (SC202, SC203).

Figure 21 shows the analysis result output screen for the department (sales) to which one respondent A belongs. The analysis result output screen is a screen that organizes and displays the most recent questionnaire in terms of the respondent's psychological state regarding their work and/or organization, their willingness to contribute to their work and/or organization, and the factors they consider important within their willingness to contribute to their work and/or organization. In this way, the main system 100 is equipped with an analysis result presentation means that outputs the analysis results corresponding to the most recent questionnaire to the respondent's terminal. Respondent A accesses the analysis result output screen using a browser, for example, and checks the various analysis results.

The guideline column 21a is a column for inputting guidelines for respondent A and the department (sales) to which respondent A belongs. The engagement column 21b is a column for displaying an overall score indicating the engagement of the department (sales) to which respondent A belongs, calculated according to the importance obtained from the analysis results of the latest questionnaire. In addition to the score indicating the engagement, the engagement column 21b outputs the benchmark score obtained as a reference point or level point, a value indicating the deviation from the benchmark, etc.

The trend column 21c is a column showing the trend of the score indicating the engagement of the department (sales) to which respondent A belongs. In the trend column 21c, a line graph is output that connects points plotted according to the date of distribution to respondent A, showing the score indicating the engagement of the department (sales) to which respondent A belongs. In the trend column 21c, a bar graph showing the score distribution is output. In the trend column 21c, the score indicating the engagement of the department (sales) to which respondent A belongs is output as a line graph according to the trend of the survey responses, making it easy to understand the trends and patterns represented by the scores.

The work driver column 21d is a column that displays the score for each key driver that indicates the engagement that respondent A feels toward the work in the department (sales) to which he or she belongs, obtained from the responses to the latest questionnaire or the latest and past questionnaires. The work driver column 21d outputs scores indicating engagement related to "recognition," "human relationships," "support," "health," "self-growth," and "duties" out of the nine key drivers. In addition to the scores indicating engagement related to the six key drivers, the work driver column 21d also outputs benchmark scores obtained as reference points or standard points for these, values indicating deviations from the benchmark, etc.

The work driver change column 21e is a column that outputs key drivers that have changed among the six key drivers, dividing them into important increases and important decreases. The work driver evaluation column 21f is a column that outputs a written evaluation of key drivers that have changed among the six key drivers.

The organizational driver column 21g is a column that displays the score for each key driver that indicates the engagement felt by respondent A with the organizational philosophy, culture, and environment of the department (sales) to which he/she belongs, obtained from the responses to the latest questionnaire or the latest and past questionnaires. The organizational driver column 21g outputs scores indicating engagement related to "organizational culture," "philosophy strategy," and "environment" out of the nine key drivers. In addition to the scores indicating engagement related to the three key drivers, the organizational driver column 21g also outputs benchmark scores obtained as reference points or standard points for these, values indicating deviations from the benchmark, etc.

The organizational driver change column 21h is a column that outputs key drivers that have changed among the three key drivers, dividing them into important increases and important decreases. The organizational driver evaluation column 21i is a column that outputs a written evaluation of key drivers that have changed among the three key drivers.

By viewing the work driver change column 21e, the work driver evaluation column 21f, the organizational driver change column 21h, and the organizational driver evaluation column 21i, respondent A can easily grasp the changes obtained from the analysis results of the latest questionnaire regarding respondent A's psychological state toward the work and/or organization of the department (sales) to which he/she belongs, his/her willingness to contribute to the work and/or organization, and the factors he/she considers important within his/her willingness to contribute to the work and/or organization. The viewing authority column 21j is a column that outputs the groups for which respondent A has viewing authority.

Figure 22 is a diagram showing a group analysis result output screen. The engagement score may be output by outputting the analysis result of one respondent, but it may also be calculated and output not only for each respondent but also for each specified group. Figure 22(a) is a comparative diagram showing the analysis result output screen of the department (sales) to which respondents A, B, and C belong. The first column 22a is a column for outputting the score (score) for each specified element indicating the engagement of respondent A, which is calculated according to the importance obtained from the analysis result of the questionnaire, for the first and second times. The second column 22b is a column for outputting the score (score) for each specified element indicating the engagement of respondent B, which is calculated according to the importance obtained from the analysis result of the questionnaire, for the first and second times.

The third column 22c is a column that outputs the score for each predetermined element indicating the engagement of respondent C, calculated according to the importance obtained from the analysis results of the questionnaire, for the first and second times. The importance calculation column 22d is a column that outputs the importance of each predetermined element for respondents A, B, and C. The group column 22e is a column that outputs the score for each predetermined element indicating the engagement of a predetermined group (sales), calculated according to the importance obtained from the analysis results of the questionnaire, and the importance of each predetermined element for the predetermined group (sales).

Figure 22 (b) is a scatter diagram showing the analysis result output screen for the department (sales) to which respondents A, B, and C belong. The CPU 101 displays a scatter diagram in which the score of a specific group (sales) and the importance of a specific group (sales) are plotted on a two-dimensional graph with the axes representing the score and the importance of the specific group (sales) according to the importance of respondents A, B, and C registered in the importance database 310. By viewing such a scatter diagram, for example, it is possible to grasp the level of importance of employees in the organization and to determine whether improvement is necessary. In this way, the main system 100 is provided with a notification means for notifying the respondent of their psychological state regarding their work and/or the organization, their willingness to contribute, and the factors they consider important.

The score for engagement can be calculated by adding up the weighted values for the importance of the respondent's organization, work, superior, and workplace, and the importance of the items related to each diagnostic area. The score for engagement may be calculated not only for each respondent, but also for each question, organization, department of an organization, group, or attribute of the respondent. The score or ranking of engagement may be displayed, for example, in comparison with other departments, groups, or attributes in the same organization, or in comparison with departments, groups, or attributes of organizations of the same regulations. By displaying in this way, if the work engagement of a department or employee of an organization is low compared to the same organization, it can be determined that some kind of activity is necessary, for example. By displaying in this way, if the work engagement of a department or employee of an organization is high compared to the same organization, it can be determined that it is time to expand the business.

By utilizing the information analysis system W of the first embodiment of the present invention, a small number of questionnaires (16 questions) can be administered over a short period of time (e.g., every four weeks) with different questions for each respondent. Such a questionnaire reduces the burden on the respondent and is likely to obtain cooperation from the respondent even if it is administered relatively frequently (every four weeks). Furthermore, by administering the questionnaire relatively frequently, it becomes easier to gauge the psychological state of each respondent regarding their work and/or organization, the state of each respondent's willingness to contribute to their work and/or organization, and what factors each respondent considers important.

Second Embodiment
In the first embodiment described above, different questions were sent to each respondent as selected question information, but in the second embodiment, in addition to the selected question information, an action plan devised by the respondent is sent to the respondent in a specified department (sales) as action plan information. In addition to the selected questions, the respondent can evaluate the action plan on the respondent terminal. In the second embodiment, candid opinions can be collected from the respondent who is the person in the field who implemented the action plan, and the respondent can review the effectiveness of the action plan.

The main system 100 of the second embodiment performs the following process.
(1) Action plans are accepted from terminals of respondents of a predetermined group (for example, respondent terminal 3a) and stored in an activity plan information database.
(2) Questions for evaluation of the action plan are presented to the terminals of respondents in a predetermined group (for example, the respondent terminals 3a, 3b, and 3c) to request an evaluation of the action plan.
(3) Evaluations of the action plan are accepted from the terminals of respondents of a predetermined group (for example, respondent terminals 3a, 3b, 3c) and stored in an activity evaluation information database.

[Second analysis result output process]
Fig. 23 is a flow chart for explaining the procedure for the second analysis result output process performed by the main system 100. The second analysis result output process executes a process for requesting input of an activity plan and a process for outputting an evaluation of an activity plan that has already been input, in addition to a process for outputting analysis results for each respondent in response to a request from each respondent. Note that the second analysis result output process is the above-mentioned analysis result output process (see Fig. 18) to which certain processes have been added. For this reason, only the processes unique to the second analysis result output process, which are surrounded by a dashed line in Fig. 23, will be explained.

When the main system 100 finishes the feature output process (S202), the CPU 101 determines whether or not there is an activity plan (S203). That is, the CPU 101 determines whether or not there is activity evaluation information in the activity evaluation information database 114. If there is no activity evaluation information in the activity evaluation information database 114 (S203: N), the CPU 101 generates activity plan information input screen data in the RAM 102 (S204). In the analysis result transmission process (see FIG. 20), the CPU 101 transmits the activity plan information input screen data together with the analysis result output screen data to the respondent terminals 3a, 3b, and 3c.

If the activity evaluation information database 114 contains activity evaluation information (S203: Y), the CPU 101 reads the activity evaluation information from the activity evaluation information database 114 (S205). When the CPU 101 reads the activity evaluation information (e.g., see FIG. 10b), it generates activity evaluation information output screen data in the RAM 102 (S206). In the analysis result transmission process (see FIG. 20), the CPU 101 transmits the activity evaluation information output screen data together with the analysis result output screen data to the respondent terminals 3a, 3b, and 3c.

[Activity plan information input screen]
FIG. 24 is a diagram showing the activity plan information input screen. After confirming the analysis results, the respondents A, B, and C of the group (sales) can devise a common activity plan for the group, and one person representing the group (e.g., respondent A) can input it as activity plan information. When the CPUs of the respondent terminals 3a, 3b, and 3c receive the activity plan information input screen data together with the analysis result output screen data from the main system 100, they store it in RAM. Then, the analysis result output screen is displayed on the output section using the stored employee registration screen data. In this case, in addition to the line graph and score distribution, the transition column 24a displays a "Start" button 24b for displaying the activity plan information input screen. When the "Start" button 24b is selected by respondent A, the CPU of the respondent terminal 3a displays the activity plan information input screen 24c on the output section using the stored activity plan information input screen data.

When the respondent A inputs the activity plan information in the text input section 24d of the activity plan information input screen 24c, the CPU of the respondent terminal 3a stores the activity plan information in the RAM. When the respondent A selects, for example, the "Share survey results" button 24e, the CPU of the respondent terminal 3a transmits the activity plan information stored in the RAM to the main system 100. When the CPU 101 receives the activity plan information from the respondent terminal 3a, it registers it in the activity plan information database 113. In this example, the respondent A has set an activity plan that was devised by the respondents A, B, and C in the group (sales) to "check each other's health conditions every Monday morning" (see Figures 9(a) and 24). In this way, the respondent terminal is provided with an activity plan information receiving means that receives the activity plan proposed by the respondent as activity plan information. In this example, only one activity plan is input by the respondent A on behalf of the group, but one or more people may input one or more activity plans.

[Analysis result output screen]
25 is a diagram showing the analysis result output screen. After answering the questionnaire and the evaluation of the activity plan, the respondents A, B, and C of the group (sales) commented on the activity plan that they had already devised within the group. When the respondent terminals 3a, 3b, and 3c receive the activity evaluation information output screen data together with the analysis result output screen data from the main system 100, the CPUs of the respondent terminals 3a, 3b, and 3c store the data in the RAM. Then, the stored employee registration screen data is used to display an analysis result output screen on the output unit. In this example, in addition to the line graph and score distribution, the activity evaluation information output screen is also displayed in the transition column 25a. A "2 wedo" button 25b for displaying the screen is displayed. When the "2 wedo" button 25b is selected by respondent A, the CPU of the respondent terminal 3a The output screen data is used to display the activity evaluation information output screen (FIG. 26) on the output section.

[Activity evaluation information output screen]
26 is a diagram showing an activity evaluation information output screen. In this example, in addition to a line graph and a score distribution, the transition field 26a displays already-entered activity evaluation information as a review display field 26b. The review display field 26b displays an activity plan and its evaluation according to the activity evaluation information already registered in the activity evaluation information database 114. For example, as shown in FIG. 10B, the activity evaluation information registered in the activity evaluation information database 114, for "discuss with everyone," has been highly rated by three people, so that as shown in the review display field 26b of FIG. 26, a smiley face mark 26c is lit and a comment field 26d is displayed indicating that there are no improvement plans.

On the other hand, as shown in FIG. 10(b), the activity evaluation information registered in the activity evaluation information database 114 has received low ratings from three people for "reflection", so as shown in the review display field 26b of FIG. 26, the trouble mark 26e is lit, and the fact that the improvement plan is 3 is displayed in the comment field 26f. Also, when the comment field 26f is selected by respondent A, comments for each respondent are displayed (not shown). In this example, the comments displayed are, for example, "Shouldn't we propose a more specific reflection plan?", "Shouldn't we be more specific?", and "Shouldn't we all check the reflection?", which are registered in the improvement plan item in FIG. 10(b). In this way, in the second embodiment, the evaluation of the executed activity plan is displayed in an easy-to-understand manner. Note that the evaluation of the activity plan is expressed by lighting the smiley face mark 26c and the troubled mark 26e, but the slider 26g between the smiley face mark 26c and the troubled mark 26e may be used to express whether there are more high or low rated responses.

[Second question selection process]
Fig. 27 is a flow chart for explaining the procedure for the main system 100 to perform the second question selection process. The second question selection process executes a process for selecting question information to be asked to each respondent, as well as a process for inputting an evaluation of an already inputted activity plan. Note that the second question selection process is the above-mentioned question selection process (see Fig. 14) to which a predetermined process has been added. For this reason, only the process specific to the second question selection process, which is surrounded by a dashed line in Fig. 27, will be explained.

When the main system 100 finishes executing the multiple choice question machine learning process (S103), the CPU 101 judges whether or not there is an activity plan (S107). That is, the CPU 101 judges whether or not there is activity plan information in the activity plan information database 113. If there is activity plan information in the activity plan information database 113 (S107: Y), the CPU 101 reads the activity plan information from the activity plan information database 113 (S108). When the CPU 101 reads the activity plan information (for example, see FIG. 9(a)), it generates activity evaluation information input screen data in the RAM 102 (S109). After generating the activity evaluation information input screen (S109), if there is no activity evaluation information in the activity plan information database 113 (S107: N), the CPU 101 again monitors whether the date and time of the "distribution timing" has been reached (S101).

[Second questionnaire processing]
Fig. 28 is a flow chart for explaining the procedure for the main system 100 to carry out the second questionnaire process. The second questionnaire process is a process of sending different questions to each respondent and questions for evaluating the activity plan, and receiving answers from each respondent and answers evaluating the activity plan. Note that the second questionnaire process is the above-mentioned questionnaire process (see Fig. 16) to which a certain process has been added. Therefore, only the process unique to the second questionnaire process will be explained.

First, the CPU 101 generates question input screen data for each respondent in the RAM 102 from the selected question information generated in the RAM 102 in the selected question machine learning process (S103), and transmits it to each respondent terminal 3a, 3b, 3c together with the activity evaluation information input screen data generated in the RAM 102 in the second question selection process. When the CPUs of the respondent terminals 3a, 3b, 3c receive the question input screen data and the activity evaluation information input screen data for each respondent A, B, C from the main system 100, they store them in the RAM, and first display the question input screen for each respondent A, B, C on the output section using the question input screen data (see FIG. 17). When the respondent A, B, C answers all the questions and, for example, a confirmation button (not shown) is selected to check whether there are any input errors, the activity evaluation information input screen data is then used to display the activity evaluation information input screen on the output section (see FIG. 29).

[Activity evaluation information input screen]
29 is a diagram showing an example of an activity evaluation information input screen. The activity evaluation information input screen is a page where respondents A, B, and C who executed the action plan input their evaluation in response to a question about the action plan that the main system 100 transmits to the respondent terminals 3a, 3b, and 3c. Depending on the content entered here, the answer information stored in the answer information database 112 is updated (see FIG. 10(a)).

Figure 29(a) is a review selection screen on which the respondent selects whether or not to review the activity plan. When the "Answer" button 29a on the review selection screen is selected by respondents A, B, and C, the CPUs of respondents A, B, and C output an evaluation input screen. Figure 29(b) is an evaluation input screen on which the respondent inputs an evaluation of the activity plan. The activity plan of the activity plan information that is the subject of the review is displayed in the target activity plan information field 29b (see Figure 9(a)). In this embodiment, the answer input field allows two ranks, for example, "Room for improvement!" and "Good to do!" to be input by selecting buttons 29c and 29d. However, buttons 29c and 29d in the answer input field may have other names, and more than two ranks may be set.

Respondents A, B, and C select button 29c or 29d from the input section of respondent terminal 3a, 3b, and 3c. When respondents A, B, and C input their answers from the evaluation input screen, the CPU of respondent terminal 3a, 3b, and 3c stores the answers in RAM as activity evaluation information. The input answers are converted into numerical values according to the conversion data stored in the storage device of respondent terminal 3a, 3b, and 3c, depending on which button 29c or 29d was selected (e.g., button 29c = 1, button 29d = 0, etc.).

When the "Good job!" button 29c or the "Room for improvement!" button 29d on the evaluation input screen is selected by the respondents A, B, C, the CPUs of the respondents A, B, C reflect this in the activity evaluation information and store it in the RAM. When the "Good job!" button 29c or the "Room for improvement!" button 29d on the evaluation input screen is selected by the respondents A, B, C, the CPUs of the respondents A, B, C output an improvement proposal input screen. Figure 29(c) shows the improvement proposal input screen on which the respondent inputs improvement proposals for the activity plan. The improvement activity plan information column 29e displays the activity plan for the activity plan information that is deemed to have room for improvement (see Figure 9(a)).

When respondents A, B, C input activity evaluation information into the text input section 24d of the activity plan information input screen 24c, the CPU of the respondent terminal 3a, 3b, 3c reflects this in the activity evaluation information and stores it in RAM. When respondents A, B, C select the "skip" button 29g without inputting activity evaluation information into the text input section 24d of the activity plan information input screen 24c, the CPU of the respondent terminal 3a, 3b, 3c reflects this in the activity evaluation information and stores it in RAM.

[Response information reception SC102]
When the respondents A, B, and C answer all the questions, evaluate the questions that review the activity plan, and select a confirmation button (not shown) to check for input errors, for example, the CPU of the respondent terminal 3a, 3b, and 3c transmits the answer information and activity evaluation information stored in the RAM to the main system 100. When the CPU 101 receives the answer information for each of the respondents A, B, and C from the respondent terminal 3a, 3b, and 3c, it registers it in the answer information database 112 as shown in Figure 8(a) or 8(b). When the CPU 101 receives the activity evaluation information from the respondent terminal 3a, 3b, and 3c, it registers it in the activity evaluation information database 114 as shown in Figure 10(a).

For example, when respondent A selects the "I'm glad I did it!" button 29c, the CPU of the respondent terminal 3a stores a "1" in the high rating of the activity evaluation information rating (1) and stores this in the RAM as activity evaluation information. The CPU of the respondent terminal 3a then transmits the activity evaluation information stored in the RAM to the main system 100. When the CPU 101 receives the activity evaluation information from the respondent terminal 3a, it registers it in the activity evaluation information database 114 (see FIG. 10(a)).

For example, if respondent B selects the "Room for improvement!" button 29d and does not input a proposed improvement plan, the CPU of the respondent terminal 3b stores "0" in the low evaluation of evaluation (2) of the activity evaluation information, and stores this in the RAM as the activity evaluation information. The CPU of the respondent terminal 3b then transmits the activity evaluation information stored in the RAM to the main system 100. When the CPU 101 receives the activity evaluation information from the respondent terminal 3b, it registers it in the activity evaluation information database 114 (see FIG. 10(a)).

For example, when respondent C selects the "Room for improvement!" button 29d and inputs "Isn't it more essential to check engagement and not just health conditions?" as an improvement plan proposal, the CPU of the respondent terminal 3c stores "0" in the low rating of rating (3) of the activity evaluation information, stores "Isn't it more essential to check engagement and not just health conditions?" as an improvement proposal for rating (3) of the activity evaluation information, and stores this in RAM as activity evaluation information. The CPU of the respondent terminal 3c then transmits the activity evaluation information stored in RAM to the main system 100. When the CPU 101 receives the activity evaluation information from the respondent terminal 3c, it registers it in the activity evaluation information database 114 (see FIG. 10(a)).

Using the information analysis system W of the second embodiment of the present invention, the respondent can register an action plan after checking the analysis results of the department (sales) to which he or she belongs. When an action plan has been registered, the respondent can evaluate the action plan on the respondent terminal in addition to answering multiple choice questions. The department (sales) to which the respondent belongs can collect candid opinions from the respondent, who is the person in the field who implemented the action plan, and can look back on the respondent's confirmation of the effectiveness of the action plan, and can use this as reference for the next action plan to be registered.

[Counting process]
30 is a flow chart for explaining the procedure of the main system 100 performing the tallying process. The tallying process is a process of tallying the answer information and outputting the tallying result in response to a request from a person having a certain authority (e.g., the personnel manager terminal 2, the respondent A). The CPU 101 monitors whether or not there is a tallying request from a person having a certain authority (e.g., the personnel manager terminal 2, the respondent A) (S301). For example, when there is a tallying request from the personnel manager terminal 2 (S301: Y), the CPU 101 transmits the answer information from the answer information database 112 to the tallying system 200 and requests the tallying system 200 to execute the tallying of the answer information (S302).

When the tallying system 200 has finished tallying the response information (S302), the CPU 101 generates tally result output screen data in RAM 102 for displaying a tally result output screen, and sends this to the human resources officer terminal 2. When the CPU of the human resources officer terminal 2 receives the tally result output screen data, it stores it in RAM and uses it to display the tally result output screen on the output section. The tally result output screen displays the tally results in tables, graphs, etc. for each respondent, question, organization, department of an organization, group, or respondent attribute. Note that in the tally results displayed on the tally result output screen, the names of the respondent are anonymous. When the tallying system 200 has finished processing, the CPU 101 monitors whether there is a tally request from a person with certain authority (e.g., the human resources officer terminal 2, respondent A).

Although the embodiment of the present invention has been described above, the present invention is not limited to this embodiment, and can be modified within the scope of the purpose and intent of the present invention. The present invention has been described as storing each piece of data in a table format, but this data does not necessarily have to be expressed in a data structure based on a table. For example, the data may be expressed in a data structure such as a list, DB, queue, or other method.

In this example, the selected question information is predicted from the importance of one respondent and the amount of remaining information of one respondent through a selected question information model. However, the selected question information may be predicted from either the importance of one respondent or the amount of remaining information of one respondent through a selected question information model. In this example, the selected question information is predicted from the importance of one respondent and the amount of remaining information of one respondent through a selected question information model. However, the selected question information may be selected according to the frequency and/or item by updating the frequency of question information from the importance of one respondent and/or the amount of remaining information of one respondent. In this way, the main system 100 may be provided with a frequency update means for updating the frequency of question information.

In this example, the number of questions asked to the respondent is described as 16 or 32, but the number of questions need not be limited to this number. For example, the number of questions may be changed as appropriate by the person managing the information analysis system W or the human resources officer H who uses the information analysis system W. Furthermore, the human resources officer H who uses the information analysis system W may generate, update, and delete the respondent information as appropriate. Furthermore, the person managing the information analysis system W may generate, update, and delete the question information as appropriate. In this way, the main system 100 may be provided with a storage means for updating the respondent information and question information.

In this example, the "questions" prepared in the question information are intended to measure the respondent's psychological state regarding their work and/or the organization, the respondent's willingness to contribute to their work and/or the organization, and the factors that each respondent considers important. In particular, the purpose is to measure through the questionnaire whether the respondent, an employee, has a positive and fulfilling psychological state regarding work and work engagement characterized by any one or a combination of energy, enthusiasm, and/or absorption.

However, if the respondents are not employees but people who belong to a specific organization, the "questions" prepared in the question information may be aimed at measuring not only work-related matters but also the respondent's psychological state toward the organization, the respondent's motivation to contribute to the organization, the factors that each respondent considers important, etc.

In addition, the "questions" are not limited to these, but may also be prepared to measure employee satisfaction, which is overall satisfaction with the work environment, salary, employee benefits, etc.; motivation, which is the motivation to take action; and the expectation of waiting for something to happen in the future. In this way, the selected question information may be predicted based on any one or a combination of importance, amount of remaining information, employee satisfaction, motivation, and expectation.

1 Information analysis device

Claims (8)

a storage means for storing respondent information for identifying a respondent and question information for asking the respondent questions regarding duties, self-growth, health, human relations, support, recognition, philosophical strategy, organizational climate, and environment as factors that affect the willingness to contribute to the organization;
an answer information receiving means for receiving, as answer information, a value of each element that affects the willingness to contribute, which is answered by a respondent in response to a question from a respondent terminal;
an analysis means for analyzing the answer information received by the answer information receiving means, and generating an analysis result using factor analysis or Bayesian statistics to determine which elements the respondent considers important and the importance of deriving a causal relationship from the answer information according to a time series transition of the value of each element in the past from the answer information for each respondent;
a question selection means for selecting question information to be asked to each respondent according to the importance, which is an analysis result obtained by the analysis means;
and a question presenting means for presenting the question information for each answerer selected by the question selecting means to an answerer terminal for each answerer. 2. The information analysis device according to claim 1,
The question information comprises a frequency, which is the degree to which the question is distributed to the respondent;
The information analyzing device according to claim 1, wherein the question selecting means selects question information to be asked to each respondent according to the analysis result obtained by the analyzing means and a frequency of the question information. 3. The information analysis device according to claim 2,
and a frequency update means for updating a frequency of question information according to the importance of the question information. The information analysis device according to any one of claims 1 to 3,
The analysis means includes:
an importance model generating means for generating an estimation model by machine learning, using the answer information of all respondents as teacher data, and estimating parameters expressing the relationship between the answer information of all respondents and the importance using factor analysis or Bayesian statistics ;
and an importance output means for outputting the importance of one respondent by analyzing answer information of one respondent using the estimation model generated by the importance model generation means. 5. The information analysis device according to claim 4,
The analysis means includes:
a residual information amount model generation means for generating an estimation model by machine learning, which uses answer information of all respondents as teacher data, determines how long ago a question was selected from the answer information of all respondents and the answer information of each respondent, and estimates parameters expressing the relationship between the residual information amount, which is information for determining whether a question to be selected for a respondent is new or old , using factor analysis, Bayesian statistics, or clustering ;
and a remaining information amount output means for outputting a remaining information amount of one respondent by analyzing answer information of one respondent using the estimation model generated by the remaining information amount model generation means. 6. The information analysis device according to claim 5,
The analysis means includes:
a selective question model generating means for generating an estimation model by machine learning, the estimation model estimating a parameter expressing a relationship between the importance of all respondents and/or the amount of remaining information of all respondents and the selective question information of all respondents, using the importance of all respondents and/or the amount of remaining information of all respondents as teacher data;
and a selective question output means for outputting selective question information to one respondent by analyzing the importance of one respondent and/or the amount of remaining information of one respondent using the estimation model generated by the selective question model generation means. a storing step of storing respondent information for identifying the respondent and question information for asking the respondent questions regarding duties, self-growth, health, human relations, support, recognition, philosophy strategy, organizational culture, and environment as factors that affect the willingness to contribute to the organization;
an answer information receiving step of receiving, as answer information, a value for each factor that affects the willingness to contribute, which is answered by a respondent in response to a question from the respondent terminal;
an analysis step of analyzing the answer information received by the answer information receiving step, and generating an analysis result using factor analysis or Bayesian statistics to determine which elements the respondent considers important and the importance of deriving a causal relationship from the answer information according to a time series transition of the value of each element in the past from the answer information for each respondent;
a question selection step of selecting question information to be asked to each respondent according to the importance that is the analysis result obtained by the analysis step;
and a question presenting step of presenting the question information for each respondent selected in the question selection step to a respondent terminal for each respondent. A storage function for storing respondent information for identifying the respondent and question information for asking the respondent questions about duties, self-growth, health, human relations, support, recognition, philosophy strategy, organizational culture, and environment as factors that affect the willingness to contribute to the organization;
an answer information receiving function that receives, as answer information, values of each factor that affects the willingness to contribute, which is answered by the respondent in response to the question from the respondent's terminal;
an analysis function that analyzes the answer information received by the answer information receiving function, and generates an analysis result using factor analysis or Bayesian statistics to determine which elements the respondent considers important and the importance of deriving causal relationships from the answer information according to the time series transition of the values of each element in the past from the answer information for each respondent;
a question selection function that selects question information to be asked to each respondent according to the importance that is an analysis result obtained by the analysis function;
A question presentation function that presents the question information for each respondent selected by the question selection function to the respondent terminal for each respondent.

Images