System, presentation method and presentation program

Abstract

An agent system according to the present disclosure includes a control unit that performs a presentation corresponding to a target specified based on information about a state of a sensed user. The control unit performs a plurality of presentations corresponding to the specified target in ascending order of fluctuations in the state of the user.

Description

Technical Field

[0001] This disclosure relates to systems, presentation methods, and presentation procedures. Background Technology

[0002] To date, agent systems exist that make various suggestions to users from terminal devices such as AI speakers. As a related technology in such agent systems, techniques are known for addressing community problems by automatically generating predetermined behavioral models (means for solving problems in the community, predetermined value standards (i.e., values), predetermined lifestyles, etc.) and indirectly prompting behavioral changes based on these behavioral models.

[0003] Reference List

[0004] Patent documents

[0005] PTL 1: JP 2021-182170 A Summary of the Invention

[0006] Technical issues

[0007] However, among the related technologies described above, there are already instances where users are repeatedly given unacceptable suggestions, and there is already a problem that may increase distrust or discomfort with the agent system.

[0008] Therefore, this disclosure provides systems, presentation methods, and presentation procedures that can suppress the increase of user distrust or discomfort with agent systems.

[0009] Note that the problems or objectives described above are merely one of many problems that can be solved or one of many objectives that can be achieved by the various embodiments disclosed in this specification.

[0010] Solution to the problem

[0011] The system according to this disclosure includes a control unit that performs presentation corresponding to a target determined based on information related to the sensed state of the user. The control unit presents multiple presentations corresponding to the determined target in ascending order relative to changes in the user's state. Attached Figure Description

[0012] [ Figure 1 ] Figure 1 This is a block diagram illustrating a configuration example of an agent system according to an implementation method.

[0013] [ Figure 2 ] Figure 2 This is a block diagram illustrating an example of the functional configuration of an agent system according to an implementation method.

[0014] [ Figure 3 ] Figure 3This is an illustrative diagram showing an example of the definition information (solution).

[0015] [ Figure 4 ] Figure 4 This is an explanatory diagram showing an example of the definition information (measures).

[0016] [ Figure 5 ] Figure 5 This is a flowchart illustrating an operational example of an agent system according to an implementation method.

[0017] [ Figure 6 ] Figure 6 This is a flowchart illustrating an example of an operation that makes another proposal.

[0018] [ Figure 7 ] Figure 7 This is a flowchart illustrating an example of the process of creating a new request based on prompts.

[0019] [ Figure 8 ] Figure 8 This is an illustrative diagram showing an example of the definition information (solution) based on Example 1.

[0020] [ Figure 9 ] Figure 9 This is an illustrative diagram showing an example of the definition information (solution) based on Example 2.

[0021] [ Figure 10 ] Figure 10 This is an illustrative diagram showing an example of the proposal based on Example 3.

[0022] [ Figure 11 ] Figure 11 This is an illustrative diagram showing an example of the proposal based on Example 4.

[0023] [ Figure 12 ] Figure 12 This is an explanatory diagram showing an example of the definition information (measures) based on Example 4.

[0024] [ Figure 13 ] Figure 13 This is an explanatory diagram showing an example of the definition information (measures) based on Example 5.

[0025] [ Figure 14 ] Figure 14 This is an explanatory diagram showing an example of the definition information (measures) based on Example 5.

[0026] [ Figure 15 ] Figure 15 This is an explanatory diagram showing an example of the definition information (measures) based on Example 6.

[0027] [ Figure 16 ] Figure 16 This is an explanatory diagram showing an example of the options according to Example 7.

[0028] [ Figure 17 ] Figure 17 This is a hardware configuration diagram showing an example of a computer. Detailed Implementation

[0029] In the following, embodiments of the present disclosure will be described in detail with reference to the accompanying drawings. Note that in this specification and the drawings, components having substantially the same functional configuration are indicated by the same reference numerals, and redundant descriptions thereof will be omitted.

[0030] Furthermore, in this specification and accompanying drawings, similar components of an embodiment can be distinguished from each other by adding at least one different letter character or number after the same reference numeral. However, when multiple similar components do not need to be specifically distinguished from each other, the same reference numerals are assigned to each component. For example, as needed, as in definition information (solution) 132a and definition information (measure) 132b, multiple configurations including substantially the same functional configuration are distinguished (see...). Figure 3 , Figure 4 For example, when there is no need to specifically distinguish between definition information (solution) 132a and definition information (measure) 132b, definition information (solution) 132a and definition information (measure) 132b are simply referred to as definition information 132 (see [link]). Figure 2 ).

[0031] One or more embodiments (including examples, modifications, and application examples) described below can be implemented independently. On the other hand, at least a portion of the various embodiments described below can be suitably combined with at least a portion of another embodiment. These various embodiments may include novel features that differ from each other. Therefore, the various embodiments can contribute to achieving different purposes or solving different problems, and can exhibit different effects.

[0032] System Configuration

[0033] Figure 1 This is a block diagram illustrating a configuration example of an agent system according to an implementation method. For example... Figure 1 As shown, as a configuration example of the agent system 100, there is a configuration in which information processing device 1, such as a server device, and terminal device 2 used by the user are connected to each other via a network 3, such as the Internet or a local area network (LAN).

[0034] Information processing device 1 performs control over terminal device 2 connected via network 3. This control is related to presentation, such as determining the content to be presented to the user by terminal device 2.

[0035] Terminal device 2 can be applied to various devices used by users, such as AI speakers, tablet terminals, smartphones, personal computers (PCs), televisions, air conditioners, and in-vehicle devices (e.g., navigation devices) installed in vehicles for user use.

[0036] In the configuration example of the proxy system 100 described above, the information processing device 1, unlike terminal device 2, performs presentation-related controls, such as determining the content to be presented to the user; however, this disclosure is not limited to this configuration. For example, the proxy system 100 may be configured as a single terminal device 2, wherein terminal device 2 performs presentation-related controls.

[0037] Functional Configuration

[0038] Figure 2 This is a block diagram illustrating an example of the functional configuration of an agent system 100 according to an embodiment. For example... Figure 2 As shown, the agent system 100 includes an input unit 10, a processing unit 11, an output unit 12, and a storage unit 13.

[0039] The input unit 10 is a processing unit that receives user status as input from various sensors of the terminal device 2, such as microphones, cameras, accelerometers, biosensors, and temperature sensors. Specifically, the input unit 10 receives data sensed by the terminal device 2 or by devices installed separately from the terminal device 2, such as webcams, microphones, accelerometers, gyroscopes, heart rate monitors, EEG monitors, or digital thermometers, via the network 3.

[0040] Processing unit 11 is a processing unit that receives information from input unit 10 and provides feedback (presentation) to the user. For example, processing unit 11 may apply a control layer of a general-purpose operating system (OS) that performs rendering control for displaying application windows and event dispatching, such as touch, for multiple contents and each content. Processing unit 11 is an example of a control unit.

[0041] The processing unit 11 includes a user information sensing unit 111, an expectation value grasping unit 112, a problem grasping unit 113, a solution definition unit 114, a measure definition unit 115, a simulation unit 116, and a feedback unit 117.

[0042] The user information sensing unit 111 is a processing unit that determines the user's state based on information received by the processing unit 11 from various sensors, such as microphones, cameras, accelerometers, gyroscopes, biosensors, and temperature sensors. Specifically, the user information sensing unit 111 recognizes the user's movements and expressions based on body or facial information identified from video from the camera. The user information sensing unit 111 recognizes the user's voice based on input speech from the microphone. The user information sensing unit 111 recognizes the user's state based on input signals from various biosensors (brainwave sensors, pulse sensors, body temperature sensors, blood pressure sensors, etc.). The user information sensing unit 111 identifies the user's surrounding environment (e.g., temperature and humidity) based on measurements from digital thermometers, digital humidity sensors, etc. For example, the user's state determined by the user information sensing unit 111 can be represented by numerical values ​​such as a score, indicating the user's mood, temperature, etc. As an example, it is assumed that users with values ​​close to the specified values ​​experience small changes in their state, while users with values ​​far from the specified values ​​experience large changes in their state.

[0043] The expectation value grasping unit 112 is a processing unit that grasps the expected state of the user based on the user's state determined by the user information sensing unit 111. Specifically, when the user is in an angry or sad state based on the user's facial expression determined by the user information sensing unit 111, the expectation value grasping unit 112 grasps the opposite state of happiness or joy as the expected state. When the expectation value grasping unit 112 detects that the user is sleeping or about to sleep (drowsy) based on the user's facial expression determined by the user information sensing unit 111, the expectation value grasping unit 112 grasps the state that the user can sleep well as the expected state.

[0044] When the user's movement, as determined by the user information sensing unit 111, is driving (riding) a car, the expectation value grasping unit 112 grasps the awakened state (a state without motion sickness) as the expected state. When the user's ambient temperature, as determined by the user information sensing unit 111, is outside a predetermined temperature range (e.g., at low or high temperatures), the expectation value grasping unit 112 grasps the state where the temperature is within the predetermined temperature range as the expected state. Note that the above description of grasping the expected state based on the user's state is an example, and this grasping is not limited to the above.

[0045] Problem-solving unit 113 is a processing unit that determines the problems that hinder the desired state of the user as determined by the desired state-solving unit 112, and the proposed options that bring about the desired state. Specifically, problem-solving unit 113 refers to a table in which the proposed problems and options for each desired state are set up, thereby determining the problems and proposed options that hinder the desired state.

[0046] The solution definition unit 114 is a processing unit that defines solutions (methods) to the problem determined by the problem mastering unit 113. Specifically, the solution definition unit 114 defines solutions to the determined problem based on definition information 132 pre-stored in the storage unit 13.

[0047] Figure 3 This is an illustrative diagram showing an example of definition information (solution) 132a. (See diagram below.) Figure 3 As shown, in the definition information (solution) 132a, a solution, category, effect, disadvantage, etc., corresponding to each problem are defined. The solution indicates the content to be presented to the user as a solution to the problem.

[0048] Category indicates the category to which the solution belongs. For example, categories indicate modalities such as "visual, olfactory, and gustatory" and environmental indices such as "temperature and humidity".

[0049] Effect indicators measure the magnitude of the effect (the size of the change in the user's state) when the solution is presented to the user. For example, effect indicators have values ​​of "large," "medium," and "small."

[0050] The disadvantage indicates the magnitude of the disadvantage (the extent of the change in the user's state) when the solution is presented to the user but does not achieve the desired effect. For example, the disadvantage indicates values ​​of "large," "medium," and "small."

[0051] The measure definition unit 115 is a processing unit that defines the proposed measures of the options determined by the problem mastering unit 113. Specifically, the measure definition unit 115 defines the proposed measures for the determined options based on the definition information 132 pre-stored in the storage unit 13.

[0052] Figure 4 This is an explanatory diagram showing an example of definition information (measure) 132b. (See diagram below.) Figure 4 As shown, in the definition information (measures) 132b, for each option, the proposal defines the corresponding measure, category, goal, etc. The measure indicates the content of the measure to be presented to the user for the proposed option. The execution order of the measures is preset in the definition information (measures) 132b (e.g., from top to bottom). Specifically, the execution order is set such that measures that cause small changes to the user's state when executed are ranked higher.

[0053] Category indicates the category to which the measure belongs. For example, when the measure is a change in the temperature environment, the category indicates environment, and when the measure is the presentation of content, the category indicates content.

[0054] An objective indicates the specific goal of an action. For example, when the action is a change in the temperature environment, the objective indicates temperature; and when the action is the presentation of content, the objective indicates commercial advertising.

[0055] return Figure 2 The simulation unit 116 is a processing unit that simulates the user's reaction when a proposed solution to a problem determined by the problem mastering unit 113 or an option determined by the problem mastering unit 113 is presented to the user. Specifically, the simulation unit 116 simulates the user's reaction (change in user state) when a solution or option is presented to the user in the user's current state determined by the user information sensing unit 111, using known simulation techniques. The simulation unit 116 calculates the probability that the solution or option will be accepted as the user's reaction based on the simulation results.

[0056] Feedback unit 117 is a processing unit that, when a solution to a problem determined by problem mastering unit 113 or a measure corresponding to an option determined by problem mastering unit 113 is presented to the user, the processing unit provides feedback on the solution or the measure corresponding to the option based on the user's state determined by user information sensing unit 111.

[0057] Specifically, the feedback unit 117 specifies the content of the feedback based on whether the user's state is close to the expected state grasped by the expected value grasping unit 112 when the solution or measure is presented, and updates the definition information 132 with the specified content.

[0058] For example, when the user's state when presented with the predetermined solution is close to the expected state as determined by the expected value control unit 112, the feedback unit 117 changes the effect value in the definition information (solution) 132a to a larger value. When the user's state when presented with the predetermined solution deviates from the expected state as determined by the expected value control unit 112, the feedback unit 117 changes the value of the disadvantage in the definition information (solution) 132a to a larger value. When the user's state when presented with the predetermined measure is close to the expected state as determined by the expected value control unit 112, the feedback unit 117 moves the execution order of the measure in the definition information (measure) 132b to a higher ranking. When the user's state when presented with the predetermined measure deviates from the expected state as determined by the expected value control unit 112, the feedback unit 117 moves the execution order of the measure in the definition information (measure) 132b to a lower ranking.

[0059] Output unit 12 is a processing unit that presents the solutions or measures defined by processing unit 11 to the user. Specifically, output unit 12 outputs the solutions or measures to terminal device 2 via network 3. Terminal device 2 then presents the solutions or measures output by output unit 12 to the user.

[0060] For example, when the solution or measure is audio output, screen display, etc., terminal device 2 performs audio output or screen display based on the content of the solution. The presentation of the solution or measure is an example, and can be the output of a scent using a fragrance diffuser, etc. For example, terminal device 2 can operate household appliances such as air conditioners based on the presentation of the solution or measure.

[0061] Storage unit 13 is a storage device that stores various types of information from information processing device 1, terminal device 2, or network 3 in a readable / writable manner. Specifically, storage unit 13 stores user information 131 related to the user, and definition information 132 including definition information (solution) 132a and definition information (measure) 132b.

[0062] For example, user information 131 includes information obtained by sensing the user using a microphone, camera device, etc., and received by processing unit 11. User information 131 includes attribute information related to the user's attributes (e.g., at least one of gender, age, nationality, religion, and race). User information 131 includes setting information about the presentation of solutions or measures pre-set by the user (details will be described later). The attribute information and setting information are pre-set by the user via terminal device 2, etc.

[0063] operate

[0064] Figure 5 This is a flowchart illustrating an operational example of the agent system 100 according to an embodiment. For example... Figure 5 As shown, when processing begins, processing unit 11 performs loop processing (S1 to S10) until the system terminates.

[0065] Specifically, the user information sensing unit 111 senses the user's state and obtains (determines) the user's state (S2).

[0066] Next, the expectation value grasping unit 112 grasps the expected state of the user based on the state of the user determined by the user information sensing unit 111 (S3).

[0067] Next, the problem-solving unit 113 identifies the problems that hinder the user's desired state as understood by the desired value-solving unit 112, as well as the proposed options that bring about the desired state.

[0068] Next, the solution definition unit 114 refers to the definition information (solution) 132a and defines one or more solutions (methods) corresponding to the problem determined by the problem mastering unit 113. Here, the solution definition unit 114 can narrow down multiple determined solutions based on the attribute information included in the user information. Specifically, the solution definition unit 114 narrows down the multiple determined solutions to solutions whose content matches the attribute information.

[0069] The measure definition unit 115 refers to the definition information (measure) 132b and defines one or more measures corresponding to the proposed options determined by the problem mastering unit 113. Here, the measure definition unit 115 can narrow down multiple determined measures based on attribute information included in the user information. Specifically, the measure definition unit 115 narrows down the multiple determined measures to measures whose content matches the attribute information.

[0070] Next, processing unit 11 determines whether there is content (a proposed solution or option) to be proposed (presented) (S4). If there is no content to be proposed (S4: No), processing unit 11 skips processing to S10.

[0071] When there is content to be proposed (S4: Yes), the processing unit 11 determines whether the content to be proposed is a proposal for a solution to the problem (S5). When the proposal is a proposal for a solution to the problem (S5: Yes), the processing unit 11 performs the process of making another proposal to confirm the problem (S6) and advances the process to S10.

[0072] Figure 6 This is a flowchart illustrating an example of the process of making another proposal. For example... Figure 6 As shown, when the process begins, the problem mastering unit 113 determines the problem, and the solution definition unit 114 refers to the definition information (solution) 132a to determine the solution corresponding to the problem determined by the problem mastering unit 113 (S20).

[0073] Next, the simulation unit 116 simulates the situation where the determined solution is presented to the user, and obtains the user's state after presentation (S21).

[0074] Next, when the proposal fails to achieve the goal, the processing unit 11 determines whether a new problem has arisen due to the solution based on the simulation results (S22).

[0075] Specifically, when a solution is presented to the user, the expectation value grasping unit 112 grasps the user's expected state after the presentation based on the user's state after the presentation. Next, the problem grasping unit 113 determines whether there are any new problems that hinder the user's expected state after the presentation grasped by the expectation value grasping unit 112.

[0076] When no new problems arise (S22: No), the processing unit 11 advances the processing to S29.

[0077] When a new problem arises (S22: Yes), processing unit 11 refers to definition information (solution) 132a to determine a solution corresponding to the problem (S23), and performs cyclic processing (S24 to S28) that presents solutions sequentially, starting with solutions that have small changes relative to the user's state (i.e., solutions with low effectiveness but low drawbacks). The number of cyclic processing steps is obtained by subtracting 1 from the number of solutions corresponding to the problem obtained through the definition information (solution) 132a.

[0078] Specifically, when the loop processing (S24 to S28) begins, processing unit 11 presents the solution with the least effect among the unimplemented solutions identified in the solutions, because the solution with the least effect is the solution that has a small change relative to the user's state (S25). Specifically, processing unit 11 presents solutions with low effect and low drawbacks. For example, when achieving the same effect, a solution with low drawbacks is a solution that has a small change relative to the user's state. Therefore, processing unit 11 can further reduce changes to the user's state by presenting solutions with low drawbacks.

[0079] Next, the user information sensing unit 111 senses the user's state after the solution is presented and obtains (determines) the user's state (S26).

[0080] Next, processing unit 11 determines whether there are signs of problem resolution based on the user's state obtained in S26 (S27). Specifically, processing unit 11 determines whether the user's state obtained in S26 is close to the user's expected state held by the expected value holding unit 112. For example, when the user's state is close to the user's expected state, processing unit 11 determines that there are signs of problem resolution. If the user's state deviates from the user's expected state, processing unit 11 determines that no signs of problem resolution are seen.

[0081] When no signs of problem resolution are seen (S27: No), processing unit 11 advances the process to S28 and then to the next loop. Therefore, agent system 100 can sequentially present solutions to the user, starting with solutions that have low effectiveness.

[0082] When signs of a problem being resolved are seen (S27: Yes), the processing unit 11 presents the solution with the highest effect among the determined solutions (S29) and ends the processing.

[0083] When the loop processing (S24 to S28) ends, processing unit 11 redefines the request (problem) (S30) and ends the processing.

[0084] return Figure 5 When the content to be proposed is not a proposal for a solution to the problem (S5: No), the processing unit 11 determines whether the content to be proposed is a proposal for an option (S7).

[0085] When the content to be proposed is not a suggestion of an option (S7: No), the processing unit 11 directly presents the content of the proposal to the terminal device 2 (S9) and causes the processing to proceed to S10.

[0086] When the content to be proposed is a proposal of options (S7: Yes), the processing unit 11 performs the process of prompting to create a new request (accepting options) (S8).

[0087] Figure 7 This is a flowchart illustrating an example of the process of creating a new request based on prompts. For example... Figure 7 As shown, when processing begins, processing unit 11 defines options for achieving the desired state (satisfying the user's request) (S40), and performs cyclic processing (S41 to S47) with a number of trials less than a predetermined number of trials (N). The number of trials (N) is preset by the user through terminal device 2, etc.

[0088] When the loop process (S41 to S47) begins, the simulation unit 116 simulates the situation where the defined options are presented to the user (S42).

[0089] Next, the simulation unit 116 calculates the probability of the option being accepted based on the simulation results, and determines whether the calculated probability is equal to or greater than a predetermined value (R[%]) (S43).

[0090] When the probability of an option being accepted is equal to or greater than a predetermined value (S43: Yes), the processing unit 11 presents the option (S46) and ends the processing.

[0091] When the probability of an option being accepted is not equal to or greater than a predetermined value (S43: No), the processing unit 11 refers to the definition information (measure) 132b, determines the measure corresponding to the option according to the execution order, and outputs the measure to the terminal device 2. In the terminal device 2, the output measure is executed (presented) (S44). Next, the user information sensing unit 111 senses the user's state to obtain (determine) the user's state (S45) and causes the processing to proceed to S47.

[0092] In S47, when the number of trials is less than the predetermined number of trials (N), the processing returns to S41. When the number of trials is equal to or greater than the predetermined number of trials (N), the processing unit 11 terminates the processing without any presentation options (by discarding options).

[0093] In the following text, examples 1 to 8 of the agent system 100 according to the embodiments described above will be described.

[0094] Example 1

[0095] In Example 1, “Reduce motion sickness in autonomous driving” is indicated as an example of making another suggestion to confirm the user’s problem.

[0096] Figure 8 This is an illustrative diagram showing an example of the definition information (solution) 132a according to Example 1. (See diagram below.) Figure 8 As indicated in definition information (solution) 132a, examples of solutions for the problem of "motion sickness" include "slowing down", "presenting an odor", and "opening a window".

[0097] Among these solutions, the most effective is "hysteresis reduction through deceleration," but it suffers from a time delay when the user is not experiencing motion sickness. Therefore, agent system 100 executes solutions with lower effectiveness, such as "opening the window" or "presenting an odor (e.g., allowing the inhalation of mint)," based on the defined information (solution) 132a. Next, when motion sickness is seen to lessen, agent system 100 determines that the user is experiencing motion sickness and executes "deceleration." When motion sickness is not seen to lessen, it is determined that the problem is not motion sickness, and the presentation of solutions for motion sickness ceases. Thus, in agent system 100, the increase in user distrust or discomfort with the system can be suppressed.

[0098] Example 2

[0099] In Example 2, "light adjustment for falling asleep" is shown as an example of making another suggestion to confirm the user's problem.

[0100] Figure 9 This is an illustrative diagram showing an example of the definition information (solution) 132a according to Example 2. (See diagram below.) Figure 9As shown in definition information (solution) 132a, examples of solutions to the problem of "sleeping with the lights on" include "turning off the lights in the room", "dimming the lights in the room", "turning off indirect lighting", and "closing the curtains" are included as "light adjustment for falling asleep".

[0101] For example, suppose agent system 100 has detected that a user is reading a book late at night, yawning, and appears sleepy. In this case, agent system 100 presents a solution corresponding to the problem of "sleeping with the lights on." Here, agent system 100 determines whether the user intends to sleep by presenting the solution in stages.

[0102] At this point, the solution category is "light," and therefore "distance from the user" is set as a sub-parameter in the definition information (solution) 132a. The agent system 100 presents solutions with low impact on the user and at a greater distance based on this sub-parameter. Here, when the agent system 100 detects a user increasing the light's brightness midway through the process, the agent system 100 stops the light adjustment because the user does not want to sleep.

[0103] Example 3

[0104] For example, when recommending tourist attractions to users who are planning a trip, there is a problem where users feel uncomfortable if they don't prefer the suggested options. Therefore, this problem is solved by prompting them to create a new request.

[0105] In Example 3, the agent system 100 will, for example, display things that evoke tourist attractions as performances in the vehicle in stages, and specify the proposed content based on whether the user shows interest.

[0106] Figure 10 This is an illustrative diagram showing an example based on the proposal in Example 3. Specifically, Figure 10 Examples of images presented to the user as suggested content for tourist attractions (P1 through P9) are shown.

[0107] like Figure 10 As shown, the agent system 100 presents a suggestion P1 that evokes an animal character from a popular theme park and determines whether the user takes an action that shows interest, such as engaging in conversation related to the suggestion, pointing, or looking at it. If no interest is shown, the agent system 100 presents a suggestion P2 that evokes a character from another theme park or a suggestion P7 that evokes a scent.

[0108] Next, when a user shows interest in proposals P1, P4, and P7, the agent system 100 presents more specific proposals in sequence. For example, when an user shows interest in proposal P1, the agent system 100 presents proposal P2, and when the probability of accepting the destination is equal to or greater than a predetermined value, the agent system 100 presents proposal P3, which is related to the destination (option).

[0109] Similarly, when interest is shown in proposal P4, the agent system 100 presents proposal P5, and when the probability of accepting the destination is equal to or greater than a predetermined value, the agent system 100 presents proposal P6 related to the destination (option). When interest is shown in proposal P7, the agent system 100 presents proposal P8, and when the probability of accepting the destination is equal to or greater than a predetermined value, the agent system 100 presents proposal P9 related to the destination (option).

[0110] Example 4

[0111] In Example 4, "dish suggestions" is indicated as an example prompting the user to make a new request.

[0112] Figure 11 This is an illustrative diagram showing an example of the proposal based on Example 4. Figure 12 This is an explanatory diagram showing an example of definition information (measure) 132b according to Example 4.

[0113] like Figure 12 As defined in measure 132b, suppose that when viewing a travel guide on an AR head-up display (AR-HUD), AR / VR head-mounted display (AR / VR-HMD), etc., the option of eating Chinese food is suggested as an advertisement. In this case, as... Figure 11As shown, the agent system 100 presents a Chinese-inspired advertising image G11, such as a dragon, on an AR-HUD screen G1 in a town, and then recommends Chinese food as an option. In this way, the agent system 100 can increase the probability of choosing Chinese food by reminding the user of it. Here, HUD is a display device that projects information directly into the human visual field. HMD is a display device mounted on the user's head. For example, in HUD and HMD, a semi-transparent display is placed in front of the person's eyes, and therefore the information is displayed in a state superimposed on the person's normal visual field. The advertising image G11 is a virtual object displayed in a state superimposed on real space. In the example described above, a Chinese-inspired virtual advertisement, such as a dragon, is exemplified to remind the user of Chinese food, but the virtual object is not limited to this and can be any object. The user can arbitrarily set which virtual object to display. The virtual object can be a static virtual object in which the virtual object itself does not move, or it can be a dynamic virtual object in which the virtual object itself moves.

[0114] Example 5

[0115] Example 5 describes a scenario where changing the indoor environment, such as temperature and humidity, makes it easy for the user to perform the suggested actions (drinking carbonated water, sleeping). Figure 13 and Figure 14 This is an explanatory diagram showing an example of definition information (measure) 132b according to Example 5.

[0116] like Figure 13 As defined in information (measures) 132b, for example, in the case of suggesting drinking carbonated water for health reasons, agent system 100 increases the air temperature and humidity to make the user want to drink carbonated water. Therefore, agent system 100 can increase the probability that the suggestion (to drink carbonated water) is accepted.

[0117] like Figure 14 As shown in definition information (measures) 132b, similarly, for example, in the case of proposing sleep for rest, agent system 100 increases the air temperature of the environment that makes the user drowsy.

[0118] Example 6

[0119] Example 6 describes a scenario where a commercial is inserted before or after a moving image or during an interval between moving images, allowing the user to easily perform the suggested content (e.g., having a salad for dinner). Figure 15 This is an explanatory diagram showing an example of definition information (measure) 132b according to Example 6.

[0120] like Figure 15As defined in measure 132b, in cases where a user is advised to eat salads due to a lack of vegetables, agent system 100 inserts commercial advertisements for vitamin supplements before or after the user's daily viewing of sports images, or during the intervals between such images. Therefore, agent system 100 does not directly suggest eating vegetables, but rather makes the user aware of the importance of vitamins and makes them want to eat vegetables, thus increasing the probability that the suggestion will be accepted.

[0121] Example 7

[0122] Example 7 describes a situation where the user's options are narrowed down in order to guide the user toward the ultimately desired action to be proposed. Figure 16 This is an explanatory diagram showing an example of the options according to Example 7.

[0123] like Figure 16 As shown, for example, when agent system 100 suggests a walk to a seated user, agent system 100 can present suggestion content (options) P10, P11 with a lower barrier to entry than the walk, such as "recommend going to the kitchen for a drink of water" or "check the contents of the refrigerator." Therefore, agent system 100 can ultimately guide the user to take a walk (go shopping).

[0124] Example 8

[0125] In Example 8, the processing unit 11 of the agent system 100 improves availability by optimizing (updating) predefined definition information 132 according to the user.

[0126] The processing unit 11 of the agent system 100 can use reinforcement learning to sense the user's reaction after a proposal is made, and to evaluate and learn satisfaction as a method to optimize the definition information 132. In this case, there is a risk of making low-confidence proposals to the user during the exploration phase, potentially leading to a loss of user trust. Therefore, the agent system 100 performs exploration (making low-confidence proposals) in the morning and makes high-confidence proposals in the afternoon. Thus, in the latter half of the day, the proposals become high-confidence and high-quality, allowing the user to achieve improved accuracy. In this way, the agent system 100 can perform reinforcement learning while preventing a decrease in the user's overall satisfaction throughout the day.

[0127] Effect

[0128] As described above, the agent system 100 includes a processing unit 11 that performs presentations corresponding to a target determined based on the sensed state of the user. The processing unit 11 presents multiple presentations corresponding to the determined target in ascending order relative to changes in the user's state. Therefore, the agent system 100 performs presentations with small changes relative to the user's state, and thus can suppress situations where unacceptable presentations to the user continue. In other words, in the agent system 100, an increase in the user's distrust or discomfort with the system can be suppressed.

[0129] When the goal is to solve a predetermined problem, if a problem arises that differs from the predetermined problem due to the execution of a presentation, the agent system 100 executes multiple presentations in ascending order of change relative to the user's state. Therefore, the agent system 100 can start executing presentations from those with small changes relative to the user's state (i.e., presentations with a low probability of another problem occurring) and can suppress the increase in the user's distrust or discomfort with the system.

[0130] When the target is a proposed option, the agent system 100 performs multiple presentations corresponding to the proposed option until the probability of the proposed option being accepted becomes equal to or greater than a predetermined value. Therefore, the agent system 100 can present the proposed option after making the proposal more likely to be accepted, and can suppress the increase in user distrust or discomfort with the system due to the proposed option.

[0131] When multiple presentations corresponding to an option proposal are repeated a predetermined number of times, if the probability of accepting the option proposal does not become equal to or greater than a predetermined value, the agent system 100 discards the option proposal without making an option proposal. Therefore, the agent system 100 can suppress the increase of user distrust or discomfort with the system by repeatedly executing the presentations.

[0132] The agent system 100 executes multiple presentations for a determined target based on definition information 132, which defines multiple presentation contents for that target and the impact of the presentation contents on the user's state for each target. Therefore, the agent system 100 can execute multiple presentations for a target based on the predefined definition information 132.

[0133] The agent system 100 updates the definition information 132 based on the user's state sensed when the proposed content is presented. Therefore, when a presentation based on the definition information 132 is performed, the agent system 100 can provide feedback on the user's state.

[0134] The agent system 100 narrows down multiple presentations corresponding to a determined target based on attribute information that indicates the user's attributes. Therefore, the agent system 100 can perform presentations corresponding to the user's attributes (e.g., gender, age, religion, ethnicity, etc.).

[0135] Hardware configuration example

[0136] For example, by having such Figure 17 The computer 1000 with the configuration shown implements information devices such as the information processing device 1 and the terminal device 2 according to each embodiment described above. Figure 10 This is a hardware configuration diagram illustrating an example of a computer 1000. The computer 1000 includes a CPU 1100, RAM 1200, read-only memory (ROM) 1300, a solid-state drive (SSD) 1400, a communication interface 1500, and an input / output interface 1600. The various units of the computer 1000 are connected together via a bus 1050.

[0137] The CPU 1100 operates based on programs stored in ROM 1300 or SSD 1400 and controls each unit. For example, the CPU 1100 deploys programs stored in ROM 1300 or SSD 1400 to RAM 1200 and executes processing corresponding to various programs.

[0138] ROM 1300 stores boot programs such as the Basic Input / Output System (BIOS) executed by CPU 1100 when computer 1000 starts up, and programs that depend on the hardware of computer 1000.

[0139] SSD 1400 is a recording medium that is readable by computer 1000 and non-transitory in storing programs executed by CPU 1100, data used by such programs, etc. Specifically, SSD 1400 is a recording medium that records a communication program according to this disclosure, the communication program being an example of program data 1450. SSD 1400 can also be another non-transitory recording medium, such as a hard disk drive (HDD).

[0140] Communication interface 1500 is an interface for computer 1000 to connect to external network 1550 (e.g., the Internet). For example, CPU 1100 receives data from another device via communication interface 1500 and sends data generated by CPU 1100 to another device.

[0141] Input / output interface 1600 is an interface used to connect input / output devices 1650 and computer 1000 to each other. For example, CPU 1100 receives data from input devices such as touch screens, keyboards, mice, microphones, and cameras via input / output interface 1600. CPU 1100 also sends data to output devices such as monitors, speakers, and printers via input / output interface 1600. Input / output interface 1600 can be used as a media interface for reading programs recorded on a predetermined recording medium. For example, the medium is an optical recording medium such as a Digital Universal Disc (DVD) or Phase Change Rewritable Disc (PD), a magneto-optical recording medium such as a magneto-optical disc (MO), magnetic tape, magnetic recording media, semiconductor memory, etc.

[0142] For example, when computer 1000 is used as information processing device 1 or terminal device 2 according to the embodiment, the CPU 1100 of computer 1000 executes a presentation program loaded onto RAM 1200 to implement the functions of input unit 10, processing unit 11, output unit 12, etc. SSD 1400 stores the presentation program according to this disclosure and data in RAM 1200. Note that CPU 1100 reads program data 1450 from SSD 1400 and executes the program; however, in another example, CPU 1100 may obtain these programs from another device via external network 1550.

[0143] Note that this technology can also have the following configurations.

[0144] (1) A system comprising a control unit configured to perform presentation corresponding to a target determined based on information related to a sensed user state, wherein,

[0145] The control unit is configured to present multiple presentations corresponding to the determined target in ascending order relative to the user's state change.

[0146] (2) According to the system described in (1) above, wherein,

[0147] The information relating to the user's state is acquired by at least one of a microphone, camera, accelerometer, gyroscope, biosensor, temperature sensor, or humidity sensor.

[0148] (3) The system according to (1) or (2) above, wherein,

[0149] The control unit is configured to execute multiple presentations in ascending order relative to the user's state change when a problem different from the predetermined problem arises due to the execution of the presentation, provided that the objective is to resolve a predetermined problem.

[0150] (4) According to the system described in (3) above, wherein,

[0151] The control unit is configured to perform the plurality of presentations a predetermined number of times.

[0152] (5) The system according to (3) or (4) above, wherein,

[0153] The control unit is configured to determine whether there are signs of resolving the predetermined problem based on the information related to the user's state, and to perform presentation based on the determination result.

[0154] (6) The system according to any one of (1) to (5) above, wherein,

[0155] The control unit is configured to, if the target is a proposal of a predetermined option, execute multiple presentations corresponding to the proposal of the option until the probability of the proposal of the option being accepted becomes equal to or greater than a predetermined value.

[0156] (7) The system according to (6) above, wherein,

[0157] The control unit is configured to discard the proposal of the option without making a proposal if the probability of accepting the proposal of the option after the plurality of presentations corresponding to the proposal of the option has been repeated a predetermined number of times does not become equal to or greater than the predetermined value.

[0158] (8) The system according to any one of (1) to (7) above, wherein,

[0159] The control unit is configured to perform multiple presentations for a determined target based on definition information, which defines multiple presentation contents for the target and the impact of the presentation contents on the user's state for each of the targets.

[0160] (9) The system according to (8) above, wherein,

[0161] The definition information defines the magnitude of the change in the user's state when the presented content is presented to the user.

[0162] (10) According to the system described in (9) above, wherein,

[0163] The definition information includes a sub-parameter specifying the magnitude of the impact on the user's state.

[0164] (11) According to the system described in (8) above, wherein,

[0165] The control unit is configured to update the definition information based on the user's state sensed when the presented content is presented.

[0166] (12) The system according to any one of (1) to (11) above, wherein,

[0167] The control unit is configured to narrow down the plurality of presentations corresponding to the determined target based on attribute information indicating the user's attributes.

[0168] (13) The system according to (11) or (12) above, wherein,

[0169] The definition information defines the magnitude of the change in the user's state when the presented content is presented to the user; and

[0170] The control unit is configured to update the magnitude of the user's state change in the definition information based on whether the user's state sensed when the presented content is presented is close to a predetermined state.

[0171] (14) According to the system described in (12) above, the attribute information includes at least one of the user's gender, age, nationality, religion or race.

[0172] (15) According to the system described in (11) above, wherein,

[0173] The control unit is configured to use reinforcement learning to update the definition information, the reinforcement learning evaluating and learning the user's state sensed when the presented content is presented.

[0174] (16) A presentation method, comprising using a computer:

[0175] Execute presentation corresponding to the target determined based on information related to the sensed user's state; and

[0176] Multiple presentations corresponding to the determined target are presented in ascending order relative to the user's state change.

[0177] (17) A presentation program that causes a computer to execute:

[0178] Execute presentation corresponding to the target determined based on information related to the sensed user's state; and

[0179] Multiple presentations corresponding to the determined target are presented in ascending order relative to the user's state change.

[0180] List of reference numerals

[0181] 1. Information processing equipment

[0182] 2 terminal devices

[0183] 3 Networks

[0184] 10 input units

[0185] 11 processing units

[0186] 12 output units

[0187] 13 storage units

[0188] 100 Agent System

[0189] 111 User Information Sensing Unit

[0190] 112 Expected Value Mastery Unit

[0191] 113 Problem Mastery Unit

[0192] 114 Solution Definition Unit

[0193] 115 Measures Definition Unit

[0194] 116 simulation units

[0195] 117 Feedback Unit

[0196] 131 User Information

[0197] 132 Definition Information

[0198] 132a Definition Information (Solution)

[0199] 132b defines information (measures).

[0200] 1000 computers

[0201] 1050 bus

[0202] 1100 CPU

[0203] 1200 RAM

[0204] 1300ROM

[0205] 1400SSD

[0206] 1450 program data

[0207] 1500 communication interface

[0208] 1550 external network

[0209] 1600 Input / Output Interface

[0210] 1650 Input / Output Devices

[0211] G1AR-HUD screen

[0212] G11 advertising image

[0213] Proposal content for P1, P2, P4, P5, P7, and P8

[0214] Suggestions (options) on pages 3, 6, 9, 10, and 11

Claims

1. A system comprising a control unit configured to perform presentation corresponding to a target determined based on information related to a sensed user state, wherein, The control unit is configured to present multiple presentations corresponding to the determined target in ascending order relative to the user's state change.

2. The system according to claim 1, wherein, The information relating to the user's state is acquired by at least one of a microphone, camera, accelerometer, gyroscope, biosensor, temperature sensor, or humidity sensor.

3. The system according to claim 1, wherein, The control unit is configured to execute multiple presentations in ascending order relative to the user's state change when a problem different from the predetermined problem arises due to the execution of the presentation, provided that the objective is to resolve a predetermined problem.

4. The system according to claim 3, wherein, The control unit is configured to perform the plurality of presentations a predetermined number of times.

5. The system according to claim 3, wherein, The control unit is configured to determine whether there are signs of resolving the predetermined problem based on the information related to the user's state, and to perform presentation based on the determination result.

6. The system according to claim 1, wherein, The control unit is configured to, if the target is a proposal of a predetermined option, execute multiple presentations corresponding to the proposal of the option until the probability of the proposal of the option being accepted becomes equal to or greater than a predetermined value.

7. The system according to claim 6, wherein, The control unit is configured to discard the proposal of the option without making a proposal if the probability of accepting the proposal of the option after the plurality of presentations corresponding to the proposal of the option has been repeated a predetermined number of times does not become equal to or greater than the predetermined value.

8. The system according to claim 1, wherein, The control unit is configured to perform multiple presentations for a determined target based on definition information, which defines multiple presentation contents for the target and the impact of the presentation contents on the user's state for each of the targets.

9. The system according to claim 8, wherein, The definition information defines the magnitude of the change in the user's state when the presented content is presented to the user.

10. The system according to claim 9, wherein, The definition information includes a sub-parameter specifying the magnitude of the impact on the user's state.

11. The system according to claim 8, wherein, The control unit is configured to update the definition information based on the user's state sensed when the presented content is presented.

12. The system according to claim 1, wherein, The control unit is configured to narrow down the plurality of presentations corresponding to the determined target based on attribute information indicating the user's attributes.

13. The system according to claim 11, wherein, The definition information defines the magnitude of the change in the user's state when the presented content is presented to the user; as well as The control unit is configured to update the magnitude of the user's state change in the definition information based on whether the user's state sensed when the presented content is presented is close to a predetermined state.

14. The system according to claim 12, wherein, The attribute information includes at least one of the user's gender, age, nationality, religion, or race.

15. The system according to claim 11, wherein, The control unit is configured to use reinforcement learning to update the definition information, the reinforcement learning evaluating and learning the user's state sensed when the presented content is presented.

16. A presentation method, comprising using a computer: Execute presentation corresponding to the target determined based on information related to the sensed user's state; and Multiple presentations corresponding to the determined target are presented in ascending order relative to the user's state change.

17. A presentation program that causes a computer to execute: Execute presentation corresponding to the target determined based on information related to the sensed user's state; and Multiple presentations corresponding to the determined target are presented in ascending order relative to the user's state change.

Images