Reaction investigation apparatus, reaction investigation program, and reaction investigation method

Abstract

This invention provides a reaction investigation device, a reaction investigation program, and a reaction investigation method that make it easier to investigate how simulated personalities are affected and reacted to investigative influence information. [Solution] The reaction investigation device 2 includes a first simulated personality data acquisition unit 10 that acquires simulated personality data as first simulated personality data, a learning unit 11 that causes the simulated personality corresponding to the first simulated personality data to learn investigation influence information, a second simulated personality data acquisition unit 12 that acquires the second simulated personality data corresponding to the simulated personality after it has learned the investigation influence information, and a difference output unit 13 that outputs the difference obtained by comparing first comparison information based on the first simulated personality data and second comparison information based on the second simulated personality data.

Description

Technical Field

[0001] The present invention relates to a reaction investigation apparatus, a reaction investigation program, and a reaction investigation method.

Background Art

[0002] Conventionally, when it is difficult to directly conduct an experiment on a person in order to predict or investigate what kind of influence or reaction a subject will show when the subject is exposed to certain influence information.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] Even if one wants to show information that strongly affects the mind of an experimental subject and investigate the influence and reaction thereof, considering the possibility of affecting the experimental subject, it is difficult to directly conduct an experiment on a human to find out what kind of influence and reaction will be shown. For example, it is difficult and restricted both ethically and socially to conduct an experiment on a human experimental subject to determine whether the subject has resistance to fear.

[0005] On the other hand, in recent years, as shown in Patent Document 1, a technique for generating a simulated personality such as a virtual character using generative AI has been proposed. Therefore, the inventor of the present invention has intensively studied whether it is possible to investigate and estimate the influence and reaction of certain information on a person by using a simulated personality.

[0006] The present invention has been made to solve such problems, and an object thereof is to provide a reaction investigation apparatus, a reaction investigation program, and a reaction investigation method that can easily investigate how a simulated personality is affected and reacts by investigation influence information. [Means for solving the problem]

[0007] To achieve the above objective, according to one embodiment of the present invention, a reaction investigation device for investigating reactions to a simulated personality comprises: a first simulated personality data acquisition unit that acquires simulated personality data as first simulated personality data; a learning unit that causes the simulated personality corresponding to the first simulated personality data to learn investigational influence information; a second simulated personality data acquisition unit that acquires the second simulated personality data corresponding to the simulated personality after it has learned the investigational influence information; and a difference output unit that outputs the difference obtained by comparing first comparison information based on the first simulated personality data and second comparison information based on the second simulated personality data. According to the embodiment of the present invention formed in this manner, a first simulated personality data acquisition unit acquires simulated personality data as first simulated personality data, a second simulated personality data acquisition unit acquires the second simulated personality data corresponding to the simulated personality after learning the investigative influence information, and a difference output unit outputs the difference obtained by comparing the first comparison information based on the first simulated personality data with the second comparison information based on the second simulated personality data. This makes it possible to investigate the difference between the first comparison information and the second comparison information after learning the investigative influence information. Therefore, it becomes easier to investigate what kind of influence or reaction the simulated personality receives, does not receive, or is less likely to receive due to the investigative influence information.

[0008] According to one embodiment of the present invention, preferably, a reaction investigation program for investigating reactions to a simulated personality, wherein the computer functions as: a first simulated personality data acquisition unit that acquires simulated personality data as first simulated personality data; a learning unit that causes the simulated personality corresponding to the first simulated personality data to learn investigational influence information; a second simulated personality data acquisition unit that acquires second simulated personality data corresponding to the simulated personality after it has learned the investigational influence information; and a difference output unit that outputs the difference obtained by comparing first comparison information based on the first simulated personality data and second comparison information based on the second simulated personality data. According to the embodiment of the present invention formed in this manner, a computer acquires first simulated personality data as first simulated personality data using a first simulated personality data acquisition unit, acquires second simulated personality data corresponding to the simulated personality after learning the investigative influence information using a second simulated personality data acquisition unit, and outputs the difference obtained by comparing first comparison information based on the first simulated personality data with second comparison information based on the second simulated personality data using a second simulated personality data. This makes it possible to investigate the difference between the first comparison information and the second comparison information after learning the investigative influence information. Therefore, it becomes easier to investigate what kind of influence or reaction the simulated personality receives, does not receive, or is less likely to receive from the investigative influence information.

[0009] According to one embodiment of the present invention, preferably, is a reaction investigation method for investigating reactions to a simulated personality, comprising: a first simulated personality data acquisition step of acquiring simulated personality data as first simulated personality data; a learning step of causing the simulated personality corresponding to the first simulated personality data to learn investigation influence information; a second simulated personality data acquisition step of acquiring second simulated personality data corresponding to the simulated personality after it has learned the investigation influence information; and a difference output step of outputting the difference obtained by comparing first comparison information based on the first simulated personality data and second comparison information based on the second simulated personality data. According to the embodiment of the present invention formed in this manner, the first simulated personality data acquisition step of the reaction investigation method acquires simulated personality data as first simulated personality data, the second simulated personality data corresponding to the simulated personality after learning the influence information for investigation is acquired in the second simulated personality data acquisition step, and the difference output step outputs the difference obtained by comparing the first comparison information based on the first simulated personality data with the second comparison information based on the second simulated personality data. This makes it possible to investigate the difference between the first comparison information and the second comparison information after learning the influence information for investigation. Therefore, the reaction investigation method makes it easier to investigate what kind of influence or reaction the simulated personality receives, does not receive, or is less likely to receive from the influence information for investigation. [Effects of the Invention]

[0010] According to the reaction investigation apparatus, reaction investigation program, and reaction investigation method of the present invention, it is possible to easily investigate what kind of effects and reactions a simulated personality receives from the influence information used for investigation. [Brief explanation of the drawing]

[0011] [Figure 1] This is a schematic diagram showing an outline of a reaction investigation system according to one embodiment of the present invention. [Figure 2] This figure shows how a reaction investigation device according to one embodiment of the present invention connects to a server or the like via the internet. [Figure 3] This is a block diagram showing the configuration of a reaction investigation apparatus according to one embodiment of the present invention. [Figure 4] This is a block diagram showing the configuration of the control unit of a reaction investigation apparatus according to one embodiment of the present invention. [Figure 5] This figure shows the schematic configuration of a reaction investigation apparatus according to one embodiment of the present invention and the relationship between the data and the apparatus. [Figure 6] This figure shows the schematic configuration of a reaction investigation apparatus according to one embodiment of the present invention and the relationship between the data and the apparatus. [Figure 7] This figure shows the schematic configuration of a reaction investigation apparatus according to one embodiment of the present invention and the relationship between the data and the apparatus. [Figure 8] This figure shows the schematic configuration of a reaction investigation apparatus according to one embodiment of the present invention and the relationship between the data and the apparatus. [Figure 9] This figure shows the schematic configuration of a reaction investigation apparatus according to one embodiment of the present invention and the relationship between the data and the apparatus. [Figure 10] This figure shows a flowchart of a reaction investigation method related to a reaction investigation apparatus according to one embodiment of the present invention. [Figure 11] This figure shows an example of the screen of the application software for a reaction investigation apparatus according to one embodiment of the present invention. [Figure 12] This figure shows an example of the screen of the application software for a reaction investigation apparatus according to one embodiment of the present invention. [Figure 13]It is a diagram showing an example of the screen of the application software of the reaction investigation device according to an embodiment of the present invention.

Embodiments for Carrying out the Invention

[0012] Hereinafter, with reference to the accompanying drawings, the reaction investigation device 2 according to an embodiment of the present invention will be described. The embodiments of the present disclosure are described as examples, and it is clear to those skilled in the art that many modifications, changes, and substitutions can be made within the spirit and scope of the present invention. Therefore, the present invention is not limited to the disclosed embodiments, and various modifications, changes, etc. are possible in its form and details without departing from the scope of the claims. Also, the components disclosed in the specification can be freely combined.

[0013] As shown in FIG. 1, the reaction investigation device 2 according to an embodiment of the present invention is provided in the reaction investigation system 1 and has a function of investigating the reaction of the investigation influence information F with respect to the simulated personality D (see FIGS. 1 and 5). For example, the reaction investigation device 2 has a function of causing a simulated personality D simulating a human to learn investigation influence information F, such as text information, image information, video information, or composite information thereof, and investigating the reaction of what kind of influence is given to the simulated personality D. The reaction investigation system 1 includes the reaction investigation device 2 and the simulated personality generation unit 3.

[0014] The simulated personality generation unit 3 has the function of generating simulated personality data D0 using a large-scale language model (LLM) 20, for example, a character-LLM. The large-scale language model 20 is composed of application software with AI functionality. The large-scale language model 20, for example, a character-LLM, has the function of generating simulated personalities D, such as characters or simulated personalities, using predetermined application software, such as ChatGPT, Meta AI Studio, or Character AI. The simulated personality D may be a virtual character, whether tangible or intangible. The simulated personality D is a simulated personality represented on a computer, and can also be called a pseudo-personality. In recent years, technologies have emerged that can form simulated personalities while assigning personality traits to characters. For example, it is said that the simulated personality D generated using a large-scale language model can generate a simulated personality with thoughts and personality traits very similar to those of the subject (test subject). Simulated personality D is, for example, a simulated personality or virtual character that is a copy of the subject (test subject). While a higher level of completeness in Simulated Personality D is certainly more effective in investigating its influence, even a less complete version can be effective in showing certain trends. The large-scale language model 20 is inputted with attribute information of the person (test subject) that serves as the basis for Simulated Personality D, such as age, gender, height, weight, place of origin, experience, memories, and knowledge. The large-scale language model 20 may also be inputted with basic personality information, such as personality, persona, purpose, behavior, tone of voice, conversation style, records of upbringing, records of interactions with friends, and records of information dissemination on social media, etc. The large-scale language model 20 may also be inputted with information such as mental illness tendency information, disease tendency information, and medical history in order to reflect somewhat specific thought relationships. The large-scale language model 20 can generate a virtual simulated personality D by inputting the above information into a predetermined application software. For example, when a subject, such as a person who wants to generate their own simulated personality, inputs their own information into a predetermined application software, a virtual simulated personality D that simulates the subject is generated. The simulated personality D may generate the simulated personalities of others. The data constituting the main part for indicating the virtual simulated personality D is described as simulated personality data D0. The simulated personality data D0 representing the simulated personality D generated by the simulated personality generation unit 3 is acquired as the first simulated personality data D1 by the first simulated personality data acquisition unit 10 of the reaction survey device 2. For example, the simulated personality data D0 is in the state generated by the simulated personality generation unit 3, and the simulated personality data in the state acquired by the first simulated personality data acquisition unit 10 becomes the first simulated personality data D1. Investigations such as the influence on the simulated personality D can also be carried out as part of investigation reaction experiments, psychological experiments, mental experiments, etc. Therefore, investigation reaction experiments, psychological experiments, mental experiments, etc. with the simulated personality D as the subject can be carried out.

[0015] The simulated personality generation unit 3 is formed by, for example, a computer such as a server. The simulated personality generation unit 3 may be composed of a desktop computer, a laptop computer, a smartphone, a tablet terminal, or other information processing devices. As shown in FIGS. 1 and 5, the simulated personality generation unit 3 is electrically connected to the reaction survey device 2. The simulated personality generation unit 3 may be connected to the reaction survey device 2 via the Internet 5 or the like described later. The simulated personality generation unit 3 may be configured integrally with the reaction survey device 2 as described later as a modification example.

[0016] The reaction investigation device 2 is formed by, for example, a computer such as a server. The reaction investigation device 2 may be composed of a desktop computer, a laptop computer, a smartphone, a tablet terminal, or other information processing device. As shown in Figure 2, the reaction investigation device 2 is connected to the server device 4 via the internet 5. As will be described later, the reaction investigation device 2 may be formed by separating a part of it into the server device 4. The server device 4 may be other internet connection devices besides the server device. Also, the server device 4 may be omitted. The server device 4 may also be composed of a computer, a smartphone, a tablet terminal, or other information processing device. As a modification, the reaction investigation device 2 may be configured as an integrated reaction investigation device with the server device 4, as will be described later. For example, all or part of the control unit 26 of the reaction investigation device 2, which will be described later, may be provided on the server device 4 via the internet. In this case, the control processing of the control unit 26 may be executed on the server device 4, and the execution result may be displayed on the monitor 21 of the user's terminal via the internet 5. That is, the reaction investigation device 2 may be composed of a computer device on the user's side and a server device 4 via the internet 5.

[0017] As shown in Figure 3, the reaction investigation device 2 includes a monitor 21 as an output unit having a screen that outputs and displays images and videos, an input unit including an input device 22 such as a keyboard or mouse that allows the user to input instructions such as text, a communication device 23 as a communication unit that sends and receives data with other devices, and a control unit 26 that controls the monitor 21, the input device 22, the communication device 23, and the processor 24 and storage device 25 described later. The devices are connected by a bus 27. The reaction investigation device 2 includes a processor 24 such as a CPU and a storage device 25 such as memory to execute the control unit 26. For example, a user operates an application displayed on the monitor 21 of such a reaction investigation device 2, and the functions of the reaction investigation device 2 are executed.

[0018] As shown in Figure 1, the monitor 21 displays the application launched by the execution of the reaction investigation device 2 (execution of a predetermined program) on its screen. For example, the monitor 21 displays a screen that accepts information input via the input device 22, which will be described later. Also, for example, the monitor 21 displays a field where information related to the question can be entered. The communication device 23 includes a communication interface for sending and receiving signals with external devices such as the server device 4 via the Internet 5, either wirelessly or via wired connection.

[0019] As shown in Figure 4, the control unit 26 incorporates a processor 24 such as a CPU and a storage device 25 such as memory, and controls connected devices based on a predetermined control program recorded in the memory. The electrical connection between the control unit 26 and other devices may be entirely or partially connected by wireless communication such as infrared communication or other methods. The control unit 26 may be electrically connected to an external server device 4 via the internet. That is, a part of the control unit 26 may be located on a server device 4 at a physically distant location via the internet 5, or it may be provided on the server device 4 in the form of a program. For example, the control unit 26 and the server device 4 may function as a single control unit even if they are located separately. The control unit 26 stores predetermined programs that execute each function, but it is not necessarily required to store all programs; some or all of them may be stored via the internet on other server devices 4, for example, on the cloud. The control unit 26 may control not only the reaction investigation device 2, but also the simulated personality generation unit 3 and the reaction investigation system 1.

[0020] The storage device 25 is composed of an HDD (Hard Disk Drive). The storage device 25 may also be composed of, for example, RAM (Random Access Memory), ROM (Read Only Memory), SSD (Solid State Drive), USB memory, memory card, etc. The storage device 25 stores a predetermined program and a predetermined dataset. The predetermined program is a reaction investigation program that investigates reactions to a simulated personality, and it causes the computer to function as: a first simulated personality data acquisition unit 10 that acquires simulated personality data as first simulated personality data; a learning unit 11 that teaches the simulated personality D corresponding to the first simulated personality data D1 investigation influence information F, for example, data; a second simulated personality data acquisition unit 12 that acquires second simulated personality data D2 corresponding to the simulated personality D after it has learned the investigation influence information; and a difference output unit 13 that outputs the difference obtained by comparing first comparison information E1 based on the first simulated personality data D1 and second comparison information E2 based on the second simulated personality data D2. The predetermined program may be configured to cause the computer to function as one of the functional units of the control unit 26. Alternatively, the predetermined program may be configured to cause the computer to function as one of the reaction investigation devices described later in (1) to (12). The predetermined program may be configured such that all or part of it is stored in the server device 4 and the output results are displayed on the reaction investigation device 2.

[0021] The specified program may be recorded on a computer-readable recording medium. This computer-readable recording medium may be, for example, a storage device such as an HDD, RAM, ROM, or SSD, or other data-storage-capable recording medium such as a USB memory stick, SD card, or DVD. Each program, while recorded on a computer-readable recording medium, can be executed by a computer and perform its specified function.

[0022] Next, as shown in Figure 4, the functions of the control unit 26 of the reaction investigation device 2 will be described. The control unit 26 of the reaction investigation device 2 includes a first simulated personality data acquisition unit 10, a learning unit 11, a second simulated personality data acquisition unit 12, a difference output unit 13, and an evaluation unit 14.

[0023] The first simulated personality data acquisition unit 10 acquires simulated personality data D0 as first simulated personality data D1. The first simulated personality data acquisition unit 10 acquires simulated personality data D0 generated by the large-scale language model 20 as first simulated personality data D1. As described above, simulated personality data D0 is generated using the large-scale language model 20, for example, the Character Large-Scale Language Model (Character-LLM). The first simulated personality data acquisition unit 10 receives simulated personality data D0 from the simulated personality generation unit 3. By generating simulated personality data D0 using the large-scale language model 20, it is possible to increase the completeness of the simulated personality as a character and its similarity to a human personality. In addition, by generating simulated personality data D0 using the large-scale language model 20, simulated personalities can be generated quickly and easily. Furthermore, by using the large-scale language model 20, a large amount of records such as records of upbringing can also be reflected in the formation of the simulated personality. The simulated personality data acquired by the first simulated personality data acquisition unit 10 becomes the first simulated personality data D1. The simulated personality data D0 generated by the simulated personality generation unit 3 may vary in format and form depending on the application software. The first simulated personality data acquisition unit 10 tolerates a certain degree of variation in the simulated personality data D0 and recognizes the first comparison information E1 as a parameter for comparing the differences of the first simulated personality data D1 described later. Since the first simulated personality data acquisition unit 10 only needs to recognize the first comparison information E1 necessary for comparison, it can tolerate a relatively wide range of differences in the substance and format of the simulated personality data D0. The format and form of the first simulated personality data D1 can be freely changed according to the predetermined application software. The first simulated personality data acquisition unit 10 recognizes the first comparison information E1, such as parameter information, for comparing the differences of the first simulated personality data D1 described later.

[0024] The first simulated personality data acquisition unit 10 can, for example, recognize first comparison information E1 based on the first simulated personality data D1. The first comparison information E1 is obtained based on the first simulated personality data D1 through predetermined questions or through external input, etc. The first simulated personality data acquisition unit 10 includes a first comparison information setting unit 15 that sets first comparison information E1 based on the first simulated personality data D1 based on a predetermined question (a question entered into the target information Q) that is either set in advance or input externally to the first simulated personality data D1. The first comparison information E1 set by the first comparison information setting unit 15 can be set to any content, such as impressions or emotions, as will be described later. The first comparison information E1 is content derived in correspondence with the first simulated personality data D1, and as will be described later, the second comparison information E2 is content derived in correspondence with the second simulated personality data D2. A comparison is made between the first simulated personality data D1 and the second comparison information E2. Furthermore, the first comparison information setting unit 15 can obtain first comparison information E1 based on the first simulated personality data D1 by asking the first simulated personality data D1 a predetermined question, which is the target information Q. This allows the target information Q to be freely defined, and the first comparison information E1, which serves as the comparison target, can be set relatively freely and relatively easily.

[0025] The first simulated personality data acquisition unit 10 can acquire first impression information as first comparison information E1, for example, based on the first simulated personality data D1, and indicating an impression of a certain target information Q (for example, information to be investigated). That is, the first simulated personality data acquisition unit 10 can acquire first impression information based on the target information Q that asks for an impression. The first impression information is subjective evaluation information or image information held about a certain target information Q, for example, the information being asked about. The first impression information may include linguistic information, numerical information, image information, etc. The first simulated personality data acquisition unit 10 acquires first impression information using the first comparison information setting unit 15. For example, if the first comparison information setting unit 15 asks a question (target information Q) about an impression of a melon, it is conceivable that first impression information (first comparison information E1) indicating that the melon is sweet and delicious will be acquired based on the first simulated personality data D1. In this way, first impression information (first comparison information E1) based on the first simulated personality data D1 can be obtained for various types of target information Q. A question about a particular type of target information Q can also be called a question for setting comparison information.

[0026] The first simulated personality data acquisition unit 10 can acquire first emotional information as first comparison information E1, for example, based on the first simulated personality data D1, and indicating an emotion towards a certain object information Q. That is, the first simulated personality data acquisition unit 10 can acquire first emotional information from object information Q that asks about emotions. The first emotional information is information about the overall mental response to a certain object information, for example, the information being asked about. In other words, the first emotional information can be said to be a subjective experience that arises in response to external events or internal thoughts and memories. The first emotional information may include linguistic information, numerical information, image information, etc.

[0027] The first simulated personality data acquisition unit 10 can acquire, for example, first fear emotion information as first comparison information E1, which is based on the first simulated personality data and indicates a fear emotion towards a certain object information Q. That is, the first simulated personality data acquisition unit 10 can acquire first fear emotion information from object information Q that asks about fear emotions. The first fear emotion information is information about the fear emotion one has towards a certain object information, for example, the information being asked about. In other words, the first fear emotion information may be expressed in different terms, such as a feeling of wanting to avoid something psychologically or a feeling of stress. The first fear emotion information is information that is answered by a predetermined target information Q based on the first simulated personality data, and can be provisionally set to one value. The first fear emotion information may include linguistic information, numerical information, image information, etc.

[0028] The first simulated personality data acquisition unit 10 can acquire, for example, first anger emotion information that indicates anger towards a certain target information Q, based on the first simulated personality data, as first comparison information E1. That is, the first simulated personality data acquisition unit 10 can acquire first anger emotion information from target information Q that asks about anger emotions. The first anger emotion information is information about the anger emotion felt towards a certain target information, for example, the information that was asked about. In other words, the first anger emotion information may be expressed in different terms, such as feelings of dissatisfaction or feelings of stress. The first anger emotion information is information that is answered by asking a question about a predetermined target information Q based on the first simulated personality data, and can be provisionally set to one value. The first anger emotion information may include linguistic information, numerical information, image information, etc. Furthermore, the first simulated personality data acquisition unit 10 is not limited to impression information, emotional information, fear, and anger as described above, but can acquire first comparative information E1 from various perspectives and compare it with second comparative information E2.

[0029] The learning unit 11 trains the simulated personality D, which corresponds to the first simulated personality data D1, with the research influence information F. The learning unit 11 is equipped with a research influence information input unit 17 as a form into which the research influence information F can be input. The research influence information input unit 17 is executed on application software. The research influence information input unit 17 may accept information not only through direct input, but also by specifying information on a server, etc. The learning unit 11 uses the research influence information F received by the research influence information input unit 17 as data to influence and elicit reactions from the simulated personality D. The research influence information F can consist of text, images, etc. For example, the research influence information F may be information from SNS posts that induce specific emotions, or information from pages on specific websites. The research influence information F is not limited to correct information, but may also be misinformation, misinformation about a specific ethnic group, etc. For example, by training the simulated personality D with misinformation as research influence information F and examining its effects, it is possible to investigate the tolerance to and effects of misinformation.

[0030] The second simulated personality data acquisition unit 12 acquires the second simulated personality data D2, which corresponds to the simulated personality D after it has learned the research influence information F. The second simulated personality data D2 reflects the influence and reactions that the research influence information F has on the simulated personality D corresponding to the first simulated personality data D1. It is assumed that the second simulated personality data D2 may or may not be influenced by the research influence information F. The simulated personality D corresponding to the second simulated personality data D2 will reflect the result of whether or not the research influence information has influenced or not, or whether or not it has a reaction to the simulated personality D corresponding to the first simulated personality data D1. The second simulated personality data acquisition unit 12 receives the second simulated personality data D2, which is based on the first simulated personality data D1, from the first simulated personality data acquisition unit 10 and the learning unit 11. The simulated personality data acquired by the second simulated personality data acquisition unit 12 becomes the second simulated personality data D2. The second simulated personality data acquisition unit 12 is basically configured in the same manner and format as the first simulated personality data acquisition unit 10, and recognizes the second comparison information E2 of the second simulated personality data D2, which corresponds to the first comparison information E1 of the first simulated personality data D1. The second simulated personality data acquisition unit 12 only needs to recognize the second comparison information E2 necessary for comparison, so it can tolerate differences in the substance and format of the simulated personality data D0 over a relatively wide range. The format and form of the second simulated personality data D2 can be freely changed according to the predetermined application software. The second simulated personality data acquisition unit 12 recognizes the second comparison information E2, such as parameter information, which compares the differences in the second simulated personality data D2 described later.

[0031] The second simulated personality data acquisition unit 12 can, for example, recognize second comparison information E2 based on the second simulated personality data D2. The second comparison information E2 is obtained by asking the second simulated personality data D2 the same predetermined question (target information Q) that is used to obtain the first comparison information E1 from the first simulated personality data D1. The second simulated personality data acquisition unit 12 includes a second comparison information setting unit 16 that sets second comparison information E2 based on the second simulated personality data D2 using the same predetermined question (see Figure 11) as the first comparison information setting unit 15 (the question entered into the target information Q). The second comparison information E2 set by the second comparison information setting unit 16 can be set to any content, such as impressions or emotions, as will be described later. The second comparison information E2 is content derived in correspondence with the second simulated personality data D2. It is assumed that the first comparison information E1 and the second comparison information E2 will be the same answer information unless new additional information is learned. Furthermore, the predetermined question asked by the first comparison information setting unit 15 and the predetermined question asked by the second comparison information setting unit 16 are basically the same question, and if necessary they are substantially the same, and differences are permitted in non-essential parts. This is to compare the first simulated personality data D1 and the second simulated personality data D2 by comparing the first comparison information E1 and the second comparison information E2 as answers to the same question. Furthermore, the second comparison information setting unit 16 can obtain the second comparison information E2 based on the second simulated personality data D2 based on the target information Q, which is the same predetermined question as the first comparison information setting unit 15. This allows the first comparison information E1 and the second comparison information E2 to be set based on the same target information Q, making it easier to compare the two. It also makes it easier to clarify the commonalities between the two and to clarify the differences.

[0032] The second simulated personality data acquisition unit 12 can acquire, for example, second impression information that indicates an impression of a certain target information Q, based on the second simulated personality data D2, as second comparison information E2. That is, the second simulated personality data acquisition unit 12 can acquire second impression information from the target information Q that asks for an impression. The second impression information is subjective evaluation information or image information held about a certain target information Q, for example, the information being asked about. The second impression information may include linguistic information, numerical information, image information, etc. The second simulated personality data acquisition unit 12 acquires the second impression information using the second comparison information setting unit 16. The certain target information Q that leads to the second impression information is set in common with the certain target information Q that leads to the first impression information. The certain target information Q that leads to the second impression information may be read when the first impression information is derived, or it may be input again when the second impression information is acquired. For example, if the second comparison information setting unit 16 asks a question (target information Q) similar to the first comparison information setting unit 15 regarding impressions of melons, it is conceivable that a second impression information (second comparison information E2) indicating that melons are sweet and delicious will be obtained based on the second simulated personality data D2. It is also possible that, based on input from the learning unit 11, the second comparison information setting unit 16 may obtain a second impression information, as described later, that "eating melons makes you live longer," based on the second simulated personality data D2. In this way, second impression information (second comparison information E2) based on the second simulated personality data D2 can be obtained for various types of target information Q. A question regarding a certain type of target information can also be called a comparison information setting question. Various types of target information Q to be compared can be set.

[0033] The second simulated personality data acquisition unit 12 can acquire second emotional information as second comparison information E2, based on the second simulated personality data D2, and indicating emotions towards a certain object information Q. That is, the second simulated personality data acquisition unit 12 can acquire second emotional information using the object information Q that asks about emotions. The second emotional information is information about the overall mental response to a certain object information, for example, the information being asked about. In other words, the second emotional information can be described as a subjective experience that arises in response to external events or internal thoughts and memories. The second emotional information may include linguistic information, numerical information, image information, etc. The object information Q that leads to the second emotional information is set in common with the object information that leads to the first emotional information. The object information Q that leads to the second emotional information may be read at the time the second emotional information is derived, or it may be input again at the time the second emotional information is acquired.

[0034] The second simulated personality data acquisition unit 12 can acquire, for example, second fear emotion information as second comparison information E2, which is based on the second simulated personality data and indicates a fear emotion towards a certain object information Q. That is, the second simulated personality data acquisition unit 12 can acquire second fear emotion information using object information Q that asks about fear emotions. The second emotion information may include language information, numerical information, image information, etc. The object information Q that leads to the second fear emotion information is set in common with the object information Q that leads to the first fear emotion information. The object information Q that leads to the second fear emotion information may be read at the time the second fear emotion information is derived, or it may be input again at the time the second fear emotion information is acquired.

[0035] The second simulated personality data acquisition unit 12 can acquire, for example, second anger emotion information as second comparison information E2, which is based on the second simulated personality data and indicates anger towards a certain target information Q. That is, the second simulated personality data acquisition unit 12 can acquire second anger emotion information using target information Q that asks about anger emotions. The second anger emotion information may include language information, numerical information, image information, etc. The certain target information Q that leads to the second anger emotion information is set in common with the certain target information Q that leads to the first anger emotion information. The certain target information Q that leads to the second fear emotion information may be read when the second fear emotion information is derived, or it may be input again when the second anger emotion information is acquired.

[0036] The difference output unit 13 outputs the difference between the first comparison information E1 based on the first simulated personality data D1 and the second comparison information E2 based on the second simulated personality data D2. As shown in Figure 6, the difference output unit 13 includes an impression difference output function unit 31 that outputs the difference between, for example, the first impression information E1a and the second impression information E2a. The impression difference output function unit 31 can output and represent the difference regarding impression information. The impression difference output function unit 31 can represent the difference regarding impression information on the monitor 21 using language representation, image representation, etc., as will be described later. This makes it possible for the user of the reaction survey device to recognize the difference between the first impression information E1a and the second impression information E2a relatively early and easily. The impression difference output function unit 31 makes it easier to recognize the difference between the first impression information E1a and the second impression information E2a after learning the survey influence information F. Therefore, it becomes easier to investigate how the simulated personality D is affected, not affected, or less affected by the investigative influence information F regarding its impression of the target information Q. Furthermore, the impression difference output function 31 allows for investigation into how the simulated personality D is affected in terms of its impression of the target information Q.

[0037] The difference output unit 13 includes an emotion difference output function unit 32 that outputs the difference between the first emotion information E1b and the second emotion information E2b. The emotion difference output function unit 32 can output and express the difference regarding emotion information. The emotion difference output function unit 32 can express the difference regarding emotion information on the monitor 21 using language expressions, image expressions, etc., as described later. This makes it possible for the user of the reaction investigation device to recognize the difference between the first emotion information E1b and the second emotion information E2b relatively early and easily. The emotion difference output function unit 32 makes it easier to recognize the difference between the first emotion information E1b and the second emotion information E2b after learning the investigation influence information F. Therefore, it is possible to investigate how the simulated personality D is affected, not affected, or less affected by the investigation influence information F regarding emotions towards the target information Q. Furthermore, the emotion difference output function unit 32 makes it possible to investigate how the simulated personality is affected in terms of emotions regarding the target information Q.

[0038] As shown in Figure 8, the difference output unit 13 includes a fear emotion difference output function unit 33 that outputs the difference between the first fear emotion information E1c of the first simulated personality data D1 and the second fear emotion information E2c of the second simulated personality data D2. The fear emotion difference output function unit 33 makes it easier to recognize the difference between the first fear emotion information E1c and the second fear emotion information E2c after learning the investigative influence information. Therefore, it becomes easier to investigate how the simulated personality D is affected, not affected, or less affected by the investigative influence information F regarding its impression of the target information Q. Furthermore, the fear emotion difference output function unit 33 makes it possible to investigate how the simulated personality D is affected in terms of fear emotions towards the target information Q.

[0039] The difference output unit 13 includes an anger emotion difference output function unit 34 that outputs the difference between the first anger emotion information E1d of the first simulated personality data D1 and the second anger emotion information E2d of the second simulated personality data D2. The anger emotion difference output function unit 34 allows for the investigation of the difference between the first anger emotion information E1d and the second anger emotion information E2d after learning the research influence information. Therefore, it becomes easier to investigate how the simulated personality is affected, not affected, or less affected by the research influence information F regarding its impression of the target information. Furthermore, the anger emotion difference output function unit 34 allows for the investigation of how the simulated personality is affected in terms of anger emotions towards the target information.

[0040] As described above, the difference output unit 13 outputs the difference between the first simulated personality data D1 and the second simulated personality data D2, for example, the difference between the first impression information and the second impression information. As shown in Figure 4, the difference output unit 13 includes a language output unit 35, a numerical output unit 36, and an image output unit 37.

[0041] The language output unit 35 outputs the differences in linguistic form. This makes it easier to grasp the differences between the first simulated personality data D1 and the second simulated personality data D2 through linguistic expression. In addition, the language output unit 35 can express differences in qualitative information such as emotional information and impression information linguistically.

[0042] The numerical output unit 36 ​​outputs the difference in numerical form. This makes it easier to grasp the difference between the first simulated personality data D1 and the second simulated personality data D2 in numerical form. Furthermore, the numerical output unit 36 ​​can express the differences in information such as emotional information and impression information relatively objectively using numerical values.

[0043] The image output unit 37 outputs the differences using image representations. This makes it easier to grasp the differences between the first simulated personality data D1 and the second simulated personality data D2 through visual representations. Furthermore, the image output unit 37 can visually represent differences in information such as emotional information and impression information. Therefore, the image output unit 37 makes it easier to imagine the differences between the first simulated personality data D1 and the second simulated personality data D2 through visual representations. In addition, the image output unit 37 makes it easier to share images with people who use different languages.

[0044] The reaction investigation device 2 may also include an evaluation unit 14 that outputs evaluation information based on the differences output by the difference output unit 13. The evaluation unit 14 can, for example, analyze the differences output by the difference output unit 13 and output the evaluated information on the monitor 21. For example, as shown in Figure 12, the evaluation unit 14 can output trends and evaluations of the differences, such as "the second impression information is worse than the first impression information." This provides evaluation information based on the differences output by the difference output unit 13, making it easier to grasp the meaning of the differences between the first comparison information E1 based on the first simulated personality data and the second comparison information E2 based on the second simulated personality data. With such an evaluation unit 14, it is possible to estimate and display not only the differences between the two but also the meanings based on those differences.

[0045] Next, as shown in Figure 10, a series of operations of the reaction investigation method for investigating reactions to a simulated personality will be described. As shown in Figure 10, in step S1 of the reaction investigation method, the control unit 26 executes a first simulated personality data acquisition step S1, which acquires simulated personality data D0 as first simulated personality data D1. The first simulated personality data acquisition unit 10 of the control unit 26 receives simulated personality data D0 from the simulated personality generation unit 3. Simulated personality data D0 is data generated by the simulated personality generation unit 3, and its format is set relatively freely. The first simulated personality data acquisition unit 10 recognizes the simulated personality data D0 and acquires it as first simulated personality data D1. After the first simulated personality data acquisition unit 10 acquires the first simulated personality data D1, the first simulated personality data acquisition unit 10 acquires first comparison information E1 for a certain target information (e.g., investigation target information) Q using the first comparison information setting unit 15. The first simulated personality data acquisition unit 10 displays the first input form 41 for target information Q on the monitor 21 and accepts input for target information Q. When information about the subject to be investigated is entered into target information Q, impression information etc. that the simulated personality of the first simulated personality data D1 possesses about the subject to be investigated is output as first comparison information E1 on the first output form 42. For example, if the input for target information Q is a question about the impression of melons, and this question is sent to the simulated personality D of the first simulated personality data D1, the simulated personality of the first simulated personality data D1 will generate the impression information "Melons are sweet and delicious," and the first simulated personality data acquisition unit 10 outputs this as first comparison information E1. When step S1 is completed, the control unit 26 proceeds to S2. Furthermore, prior to step S1, preparation steps such as preparing the reaction investigation system 1 and reaction investigation device 2 to be operational, and a step of generating simulated personality D using the simulated personality generation unit 3 may be optionally incorporated into the reaction investigation method.

[0046] In step S2, the control unit 26 executes a learning step S2 to cause the simulated personality D corresponding to the first simulated personality data D1 to learn the investigative influence information F. In the learning step S2, the control unit 26 causes the learning unit 11 to cause the simulated personality D corresponding to the first simulated personality data D1 to learn the investigative influence information F. The learning unit 11 displays a learning unit input form 43 on the monitor 21 that accepts input of the investigative influence information F. The learning unit 11 learns the investigative influence information F entered into the learning unit input form 43. For example, if the information "Eating melons makes you live longer" is entered as investigative influence information F, this investigative influence information F "Eating melons makes you live longer" can be taught to the simulated personality D corresponding to the first simulated personality data D1. Compared to the state before the first simulated personality data D1 was exposed to the investigative influence information F, the state after exposure to the investigative influence information F may or may not be affected. The state of the first simulated personality data D1 after it has come into contact with the investigative influence information F becomes the second simulated personality data D2.

[0047] The research influence information F that the learning unit 11 is trained on can be in various forms, such as text and images. Furthermore, the learning unit 11 can freely set the number of times and duration of exposure to the research influence information F. For example, it is possible to train the learning unit 11 to the research influence information F three times within a week (exposure to the information). The research influence information F can be correct information, misinformation, or information with unclear meaning. In this way, it is possible to investigate how the simulated personality D receives and is affected by various types of information. It is also possible to investigate the simulated personality D's tolerance for misinformation and information of uncertain veracity. In relation to susceptibility to influence, for example, tolerance to brainwashing and mind control can also be investigated. When step S2 is completed, the control unit 26 proceeds to S3.

[0048] In step S3, the control unit 26 executes the second simulated personality data acquisition step S3 to acquire the second simulated personality data D2 corresponding to the simulated personality D after it has learned the investigation influence information F. The second simulated personality data acquisition unit 12 of the control unit 26 acquires the first simulated personality data D1, which has been trained with the investigation influence information F by the learning unit 11. As a result, the second simulated personality data acquisition unit 12 acquires the second simulated personality data D2 corresponding to the simulated personality D after it has learned the investigation influence information F. The second simulated personality data D2 is the data corresponding to the first simulated personality data D1. After the second simulated personality data acquisition unit 12 acquires the second simulated personality data D2, the second comparison information setting unit 16 of the second simulated personality data acquisition unit 12 sets the second comparison information E2 based on the second simulated personality data D2 using the target information Q, which is a predetermined question. The second comparison information E2 is set by the second comparison information setting unit 16 using the same target information Q that has already been queryed by the first comparison information setting unit 15. The second simulated personality data acquisition unit 12 displays the second comparison information E2 for the target information Q on the second output form 45 (see Figure 12) on the monitor 21. In the second simulated personality data acquisition unit 12, the target information Q is basically the same as that for the first simulated personality data acquisition unit 10, so the second input form 44 for the target information Q is assumed to have the same target information Q provisionally entered without any input (shown by a dashed line in Figure 9). The second output form 45 corresponds to the second simulated personality data D2. The impression information etc. that the simulated personality of the second simulated personality data D2 possesses regarding the target information Q is output as the second comparison information E2 on the second output form 45. For example, if, as described above, a question about impressions of melons is accepted as input to target information Q, and this question is posed to simulated personality D of the second simulated personality data D2, the simulated personality of the second simulated personality data D2 will generate impression information such as "Melons are sweet and delicious, and eating melons makes you live longer," and the second simulated personality data acquisition unit 12 will output this as second comparison information E2. In this case, it is considered that the second simulated personality data D2 was influenced by learning the research influence information F "Eating melons makes you live longer" by the learning unit 11. As an example involving a different personality, if, as described above, a question about impressions of melons is accepted as input to the target information Q, and this question is posed to the simulated personality of the second simulated personality data D2, then simulated personality D of the second simulated personality data D2 might generate the impression information "Melons are sweet and delicious," and the second simulated personality data acquisition unit 12 might output this as the second comparison information E2. In this case, it is considered that simulated personality D was not influenced by the learning of the research influence information F "Eating melons makes you live longer" by the learning unit 11. Another example involving a different personality is when, as mentioned above, the system accepts a question about impressions of melons as input to the target information Q, and asks the simulated personality of the second simulated personality data D2. Simulated personality D of the second simulated personality data D2 might generate impression information such as, "Melons are sweet and delicious, and they might even be good for longevity," and the second simulated personality data acquisition unit 12 might output this as second comparison information E2. In this case, it is possible that simulated personality D has been slightly influenced by the learning of the research influence information F, "Eating melons makes you live longer," by the learning unit 11. Thus, it is expected that responses will vary depending on the personality of simulated personality D. In this survey, there is no particularly desired result or correct answer, and it is possible to openly investigate responses to the research influence information F, etc. When step S3 is completed, the control unit 26 proceeds to S4.

[0049] In step S4, the control unit 26 executes a difference output step S4, which outputs the difference between the first comparison information E1 based on the first simulated personality data D1 and the second comparison information E2 based on the second simulated personality data D2. The difference output unit 13 of the control unit 26 compares the first comparison information E1 and the second comparison information E2 to extract or analyze the difference. The difference output unit 13 of the control unit 26 outputs the difference between the first comparison information E1 and the second comparison information E2 to the monitor 21, etc. The difference output unit 13 outputs the difference between the first comparison information E1 and the second comparison information E2 to the difference result output form 46 of the monitor 21. For example, according to the above example, the difference between E1 "Melons are sweet and delicious" and E2 "Melons are sweet and delicious, and eating melons makes you live longer" is output by the difference output unit 13 as "Eating melons makes you live longer". At this time, the difference output unit 13 can output, for example, the difference between the first impression information E1a and the second impression information E2a with respect to impression information, using the impression difference output function unit 31. At this time, the difference output unit 13 can output, for example, the difference between the first emotion information E1b and the second emotion information E2b with respect to emotion information, using the emotion difference output function unit 32. At this time, the difference output unit 13 can output, for example, the difference between the first fear emotion information E1c and the second fear emotion information E2c with respect to fear emotion information, using the fear emotion difference output function unit 33. At this time, the difference output unit 13 can output, for example, the difference between the first fear emotion information E1d and the second fear emotion information E2d with respect to anger emotion information, using the anger emotion difference output function unit 34. The control unit 26 may also output evaluation information based on the differences described above using the evaluation unit 14. Furthermore, the control unit 26 may also output evaluation information obtained by analyzing and evaluating the differences output by the difference output unit 13 via the evaluation unit 14. When step S4 is completed, the control unit 26 proceeds to the end.

[0050] An example of one embodiment of the present invention may be provided in the following embodiments.

[0051] (1) A reaction investigation device for investigating reactions to a simulated personality, comprising: a first simulated personality data acquisition unit that acquires simulated personality data as first simulated personality data; a learning unit that causes the simulated personality corresponding to the first simulated personality data to learn investigational influence information; a second simulated personality data acquisition unit that acquires the second simulated personality data corresponding to the simulated personality after it has learned the investigational influence information; and a difference output unit that outputs the difference obtained by comparing first comparison information based on the first simulated personality data and second comparison information based on the second simulated personality data.

[0052] (2) The reaction investigation apparatus according to (1), further comprising an evaluation unit that outputs evaluation information based on the difference output by the difference output unit.

[0053] (3) The reaction investigation apparatus according to (1), wherein the first simulated personality data acquisition unit acquires simulated personality data generated by a large-scale language model as the first simulated personality data.

[0054] (4) The reaction investigation apparatus according to (1), wherein the first simulated personality data acquisition unit acquires first impression information as first comparison information, based on the first simulated personality data and indicating an impression of a certain target information; the second simulated personality data acquisition unit acquires second impression information as second comparison information, based on the second simulated personality data and indicating an impression of a certain target information; and the difference output unit comprises an impression difference output function unit that outputs the difference obtained by comparing the first impression information and the second impression information.

[0055] (5) The reaction investigation apparatus according to (1), wherein the first simulated personality data acquisition unit acquires first emotion information indicating an emotion towards a certain target information based on the first simulated personality data, the second simulated personality data acquisition unit acquires second emotion information indicating an emotion towards a certain target information based on the second simulated personality data, and the difference output unit comprises an emotion difference output function unit that outputs the difference obtained by comparing the first emotion information and the second emotion information.

[0056] (6) The reaction investigation apparatus according to (1), wherein the first simulated personality data acquisition unit acquires first fear emotion information indicating fear emotion towards a certain target information based on the first simulated personality data, the second simulated personality data acquisition unit acquires second fear emotion information indicating fear emotion towards a certain target information based on the second simulated personality data, and the difference output unit comprises a fear emotion difference output function unit that outputs the difference obtained by comparing the first fear emotion information of the first simulated personality data with the second fear emotion information of the second simulated personality data.

[0057] (7) The reaction investigation apparatus according to (1), wherein the first simulated personality data acquisition unit acquires first anger emotion information indicating anger towards certain target information based on the first simulated personality data, the second simulated personality data acquisition unit acquires second anger emotion information indicating anger towards certain target information based on the second simulated personality data, and the difference output unit has an anger emotion difference output function that outputs the difference obtained by comparing the first anger emotion information of the first simulated personality data with the second anger emotion information of the second simulated personality data.

[0058] (8) The reaction investigation apparatus according to (1), wherein the first simulated personality data acquisition unit includes a first comparison information setting unit that sets the first comparison information based on the first simulated personality data by asking a predetermined question about the first simulated personality data.

[0059] (9) The reaction investigation apparatus according to (1), wherein the second simulated personality data acquisition unit includes a second comparison information setting unit that sets the second comparison information based on the second simulated personality data using the same predetermined questions as those asked by the first comparison information setting unit.

[0060] (10) The reaction investigation apparatus according to (1), wherein the difference output unit is equipped with a language explanation unit that outputs the difference in linguistic expression.

[0061] (11) The reaction investigation apparatus according to (1), wherein the difference output unit is equipped with a numerical explanation unit that outputs the difference in numerical form.

[0062] (12) The reaction investigation apparatus according to (1), wherein the difference output unit is an image output unit that outputs the difference in the form of an image.

[0063] (13) A reaction investigation program for investigating reactions to a simulated personality, wherein the computer functions as: a first simulated personality data acquisition unit that acquires simulated personality data as first simulated personality data; a learning unit that causes the simulated personality corresponding to the first simulated personality data to learn investigational influence information; a second simulated personality data acquisition unit that acquires second simulated personality data corresponding to the simulated personality after it has learned the investigational influence information; and a difference output unit that outputs the difference obtained by comparing first comparison information based on the first simulated personality data and second comparison information based on the second simulated personality data.

[0064] (14) A reaction investigation method for investigating reactions to a simulated personality, comprising: a first simulated personality data acquisition step of acquiring simulated personality data as first simulated personality data; a learning step of causing the simulated personality corresponding to the first simulated personality data to learn information for investigation; a second simulated personality data acquisition step of acquiring second simulated personality data corresponding to the simulated personality after it has learned the information for investigation; and a difference output step of outputting the difference obtained by comparing first comparison information based on the first simulated personality data and second comparison information based on the second simulated personality data.

[0065] The embodiments for carrying out the present invention are not limited to those described above, and further variations can be applied. Various alternative embodiments and examples will be apparent to those skilled in the art based on the disclosed technology. As a modified example, the reaction investigation device 2 may include a difference change output unit that outputs change information for a first difference output by the difference output unit when the learning unit 11 causes the simulated personality D to learn the first investigation influence information, and a second difference output by the difference output unit when the learning unit 11 causes the simulated personality D to learn a second investigation influence information that is modified from the first specific information data with a predetermined concept, with respect to a second investigation influence information different from the above. [Explanation of Symbols]

[0066] 1: Reaction Investigation System 2: Reaction investigation apparatus 10: First Simulated Personality Data Acquisition Unit 11: Learning Department 12: Second Simulated Personality Data Acquisition Unit 13: Difference Output Section 14: Evaluation Department 15:First comparison information setting section 16:Second comparison information setting section 20: Large-scale language models 31: Impression difference output function section 32: Emotional Difference Output Function Unit 33: Fear emotion difference output function unit 34: Emotional Difference Output Function Unit

Claims

1. A reaction investigation device for investigating reactions to a simulated personality, A first simulated personality data acquisition unit acquires simulated personality data as first simulated personality data, A learning unit that causes the simulated personality corresponding to the first simulated personality data to learn the influence information for investigation, A second simulated personality data acquisition unit acquires second simulated personality data corresponding to the simulated personality after learning the aforementioned impact information for investigation, A reaction investigation device comprising a difference output unit that outputs the difference obtained by comparing first comparison information based on the first simulated personality data and second comparison information based on the second simulated personality data.

2. Furthermore, the reaction investigation apparatus according to claim 1 is further comprising an evaluation unit that outputs evaluation information based on the difference output by the difference output unit.

3. The reaction investigation apparatus according to claim 1, wherein the first simulated personality data acquisition unit acquires simulated personality data generated by a large-scale language model as the first simulated personality data.

4. The first simulated personality data acquisition unit acquires, as first comparison information, first impression information that is based on the first simulated personality data and indicates an impression of a certain target information, The second simulated personality data acquisition unit acquires, as second comparison information, second impression information that is based on the second simulated personality data and indicates an impression of a certain target information. The reaction investigation apparatus according to claim 1, wherein the difference output unit includes an impression difference output function unit that outputs the difference obtained by comparing the first impression information and the second impression information.

5. The first simulated personality data acquisition unit acquires first emotional information that indicates emotions towards a certain target information, based on the first simulated personality data. The second simulated personality data acquisition unit acquires second emotional information that indicates emotions towards a certain target information, based on the second simulated personality data. The reaction investigation apparatus according to claim 1, wherein the difference output unit comprises an emotion difference output function unit that outputs the difference obtained by comparing the first emotion information and the second emotion information.

6. The first simulated personality data acquisition unit acquires first fear emotion information that indicates fear emotions towards a certain target information, based on the first simulated personality data. The second simulated personality data acquisition unit acquires second fear emotion information that indicates fear emotions towards a certain target information, based on the second simulated personality data. The reaction investigation apparatus according to claim 1, wherein the difference output unit includes a fear emotion difference output function unit that outputs the difference obtained by comparing the first fear emotion information of the first simulated personality data with the second fear emotion information of the second simulated personality data.

7. The first simulated personality data acquisition unit acquires first anger emotion information that indicates anger towards a certain target information, based on the first simulated personality data. The second simulated personality data acquisition unit acquires second anger emotion information that indicates anger towards a certain target information, based on the second simulated personality data. The reaction investigation apparatus according to claim 1, wherein the difference output unit comprises an anger emotion difference output function unit that outputs the difference obtained by comparing the first anger emotion information of the first simulated personality data with the second anger emotion information of the second simulated personality data.

8. The reaction investigation apparatus according to claim 1, wherein the first simulated personality data acquisition unit includes a first comparison information setting unit that sets first comparison information based on the first simulated personality data by asking a predetermined question about the first simulated personality data.

9. The reaction investigation apparatus according to claim 1, wherein the second simulated personality data acquisition unit includes a second comparison information setting unit that sets the second comparison information based on the second simulated personality data using the same predetermined questions as those asked by the first comparison information setting unit.

10. The reaction investigation apparatus according to claim 1, wherein the difference output unit comprises a language explanation unit that outputs the difference in linguistic expression.

11. The reaction investigation apparatus according to claim 1, wherein the difference output unit comprises a numerical explanation unit that outputs the difference in numerical form.

12. The reaction investigation apparatus according to claim 1, wherein the difference output unit comprises an image output unit that outputs the difference in the form of an image representation.

13. A reaction survey program that investigates reactions to a simulated personality, using a computer, A first simulated personality data acquisition unit acquires simulated personality data as first simulated personality data, A learning unit that causes the simulated personality corresponding to the first simulated personality data to learn the influence information for investigation, A second simulated personality data acquisition unit acquires second simulated personality data corresponding to the simulated personality after learning the aforementioned impact information for investigation, A reaction survey program that functions as a difference output unit that outputs the difference obtained by comparing first comparison information based on the first simulated personality data and second comparison information based on the second simulated personality data.

14. A reaction survey method for investigating reactions to simulated personalities, The first simulated personality data acquisition step involves the computer acquiring simulated personality data as the first simulated personality data, A learning step in which the computer causes the simulated personality corresponding to the first simulated personality data to learn information for investigation, A second simulated personality data acquisition step, in which the computer acquires second simulated personality data corresponding to the simulated personality after it has learned the investigation impact information, A reaction investigation method comprising a difference output step in which the computer outputs the difference obtained by comparing first comparison information based on the first simulated personality data with second comparison information based on the second simulated personality data.

Images