Dialogue system

Abstract

This dialogue system performs a dialogue with a user, the dialogue system comprising: an estimation unit that estimates the psychological state of the user; and a dialogue unit that performs a dialogue with the user on the basis of the psychological state of the user.

Description

Dialogue system

[0001] The present disclosure relates to a dialogue system for interacting with a user.

[0002] In recent years, dialogue devices for interacting with users, such as smart speakers, have become widespread. For example, Patent Document 1 discloses a technique for automatically recognizing a user's speech and performing natural language processing to determine the user's intention, and executing an action according to the user's intention. In a dialogue device having the technique described in Patent Document 1, while interacting with the user, it is possible to provide an answer to the user's question to the user or control household appliances such as lighting or air conditioning based on the user's instruction.

[0003] Japanese Patent Application Laid-Open No. 2017-527844

[0004] In a system using a conventional dialogue device, although the convenience when the user performs work can be improved, it is not possible to provide psychological support to the user. If there is a problem with the user's psychological state, the productivity may decrease even in an environment with good convenience.

[0005] The present disclosure solves the above problems and aims to provide a dialogue system that can provide psychological support to people.

[0006] The dialogue system according to the present disclosure is a dialogue system for interacting with a user, and includes an estimation unit that estimates the psychological state of the user, and a dialogue unit that interacts with the user based on the psychological state of the user.

[0007] According to the dialogue system of the present disclosure, it is possible to provide psychological support to the user by interacting with the user based on the psychological state of the user.

[0008] This is a schematic diagram of the dialogue system according to Embodiment 1. This is a schematic diagram of the dialogue device according to Embodiment 1. This is a hardware configuration diagram showing an example of the configuration of the control device according to Embodiment 1. This is a hardware configuration diagram showing an example of the configuration of the control device according to Embodiment 1. This is a flowchart showing the operation flow of the dialogue device according to Embodiment 1. This is a schematic diagram of the dialogue system according to Embodiment 2. This is a schematic diagram of the dialogue device according to Embodiment 2. This is a flowchart showing the operation flow of the dialogue device according to Embodiment 2. This is a schematic diagram of the dialogue device according to Embodiment 3. This is a flowchart showing the operation flow of the dialogue device according to Embodiment 3.

[0009] Hereinafter, embodiments of the dialogue device 1 and the dialogue system 100 equipped with the dialogue device 1 according to this disclosure will be described with reference to the drawings. In each figure, components denoted by the same reference numerals are the same or equivalent components, and this is common throughout the entire specification. Note that the relative dimensions or shapes of the components in each drawing may differ from those of the actual components.

[0010] Embodiment 1. Figure 1 is a schematic diagram of the dialogue system 100 according to Embodiment 1. The dialogue system 100 of this embodiment improves the productivity of user P by engaging in dialogue with user P and supporting user P's psychological state. As shown in Figure 1, the dialogue system 100 of this embodiment consists of a dialogue device 1 and an environmental information measuring device 2. The dialogue device 1 is connected to the environmental information measuring device 2 via wired or wireless communication. For wired or wireless communication, interfaces conforming to communication standards such as Bluetooth®, Wi-Fi®, ZigBee®, LTE®, or LoRaWAN® are used.

[0011] The dialogue device 1 is a personal state support system linked to user P. The dialogue device 1 is, for example, an autonomous robot as shown in Figure 1. The appearance and shape of the dialogue device 1 can be customized by user P. By configuring the dialogue device 1 in this way, user P can become familiar with the dialogue device 1, and a stronger bond is formed between user P and the dialogue device 1.

[0012] The dialogue device 1 is not limited to a robot as shown in Figure 1, but may be a terminal device such as a smartphone or tablet, or a wearable device such as a smartwatch. Alternatively, the dialogue device 1 may be a device realized by embedding or attaching a chip with the functions of the dialogue device 1 into or to a device of the user P's choice. The dialogue device 1 is a device unique to user P and starts operating when user P is authenticated, such as by inputting an identification number, fingerprint authentication, or facial recognition.

[0013] The environmental information measuring device 2 is a device that measures environmental information of the indoor space where user P is present. Environmental information includes, for example, indoor temperature, humidity, CO2 ２ The parameters include at least one of the following: concentration, dust concentration, illuminance, odor intensity, noise level, and number of people present. The environmental information measuring device 2 includes, for example, a temperature sensor to measure the room temperature, a humidity sensor to measure humidity, and CO2. ２ Measuring CO concentration ２ The device includes at least one of the following: a sensor, a dust sensor for measuring dust concentration, an illuminance sensor for measuring illumination, an odor sensor for measuring odor intensity, a sound level sensor for measuring noise levels, and an infrared sensor or camera for detecting the number of people present. The environmental information measured by the environmental information measuring device 2 is transmitted to the dialogue device 1. Note that the environmental information measuring device 2 and the environmental information are not limited to the examples listed above, and other devices and information may also be used.

[0014] In the dialogue system 100 of this embodiment, the dialogue device 1 estimates the psychological state of user P based on environmental information measured by the environmental information measuring device 2 and person information and work information obtained from user P, and supports user P by engaging in dialogue with user P based on the estimated psychological state. The dialogue device 1 of this embodiment estimates the psychological state of user P as a psychological value that is the degree of user P's ease of work or degree of well-being. The degree of ease of work is the degree to which user P can positively engage in work (job satisfaction), motivation (job satisfaction), or the degree of labor economics such as reduced working hours due to improved productivity. For example, this could be work efficiency from the perspective of labor productivity, reduction of working hours, intellectual productivity improvement index, degree of concentration on work, or engagement index from the perspective of motivation. The degree of well-being is the degree to which user P feels comfortable both physically and mentally in that space, and is expressed numerically as, for example, a comfort index, fatigue index, medical expense reduction rate, or sleep index. Furthermore, a higher numerical value indicates a greater ease of work and a more comfortable working environment.

[0015] User P's personal information is information about User P's physical and mental state, and includes at least one of the following: information about User P's state, such as User P's speech and actions, physical and mental condition, and behavioral history; and biometric information such as User P's brain waves, heart rate, pulse wave, body temperature, and posture. User P's speech and actions are words and actions that indicate User P's comfort or psychological state, such as "hot," "cold," or "tired," and physical and mental condition indicates the physical and mental state related to work, such as "unable to concentrate" or "tired." User P's personal information may be input into the dialogue device 1 by User P, or it may be measured by biometric information sensors (not shown), such as a microphone, camera, infrared sensor, electroencephalograph, activity tracker, heart rate monitor, thermometer, or accelerometer, and input into the dialogue device 1. If the dialogue device 1 is a wearable device such as a smartwatch, the dialogue device 1 may be equipped with biometric information sensors for acquiring personal information.

[0016] User P's work information is information related to the work User P is engaged in, and includes, for example, User P's schedule, affiliation, position, work content, and work hours. The schedule is User P's hourly schedule of work or outings. Work content includes, for example, specific tasks such as document creation or meetings, the type of work such as intellectual work (desk work) or manual labor, or the workload such as light work or heavy work. Here, "manual labor" includes work inside buildings such as the work of factory line workers or nurses in hospitals, but does not include work outdoors such as construction sites. User P's work information may be input into the dialogue device 1 by User P, or it may be obtained from an external management device that manages User P's work information. In addition, User P's work content and work hours may be obtained by measurement using biometric information sensors such as a microphone, camera, infrared sensor, activity tracker, or accelerometer, and input into the dialogue device 1. Note that User P's personal information and work information are not limited to the examples listed above, and may include other information.

[0017] Figure 2 is a schematic diagram of the dialogue device 1 according to Embodiment 1. As shown in Figure 2, the dialogue device 1 comprises a control device 11, a storage device 12, an input device 13, and an output device 14.

[0018] The control device 11 includes, as functional units, an information acquisition unit 111, an estimation unit 112, and an interaction unit 113. The control device 11 is composed of a processor such as a CPU (Central Processing Unit), DSP (Digital Signal Processor), or GPU (Graphics Processing Unit) that executes a program stored in dedicated hardware or memory. Figures 3 and 4 are hardware configuration diagrams showing an example of the configuration of the control device 11 according to Embodiment 1.

[0019] As shown in Figure 3, when the control device 11 is dedicated hardware, the control device 11 is composed of a processing circuit 600. The processing circuit 600 may be, for example, a single circuit, a composite circuit, an ASIC (application-specific integrated circuit), an FPGA (field-programmable gate array), or a combination thereof. The information acquisition unit 111, estimation unit 112, and dialogue unit 113 of the control device 11 may be implemented with separate hardware or with a single piece of hardware.

[0020] As shown in Figure 4, when the control device 11 is configured by a processor 601, the information acquisition unit 111, estimation unit 112, and dialogue unit 113 of the control device 11 are realized by software, firmware, or a combination of software and firmware. The software or firmware is written as a program and stored in the memory 602 that constitutes the storage device 12. The processor 601 and the memory 602 are connected to each other so as to be able to communicate via a bus 603. The processor 601 realizes the information acquisition unit 111, estimation unit 112, and dialogue unit 113 by reading and executing the program stored in the memory 602. The control device 11 may have multiple processors 601 and multiple memories 602, which work together to realize each function of the control device 11. Alternatively, some of the functions of the control device 11 may be realized by dedicated hardware, and some may be realized by software or firmware.

[0021] Returning to Figure 2, the information acquisition unit 111 communicates with the environmental information measuring device 2 via wired or wireless communication to acquire environmental information of the indoor space from the environmental information measuring device 2. The information acquisition unit 111 also acquires user P's personal information and work information input via the input device 13. Alternatively, the information acquisition unit 111 may acquire user P's personal information and work information by communicating with a biometric information sensor via wired or wireless communication. Information acquisition by the information acquisition unit 111 may be performed periodically at predetermined time intervals (for example, every 30 minutes), or when instructed by user P. Alternatively, the information acquisition unit 111 may acquire environmental information when user P inputs personal information and work information via the input device 13.

[0022] The estimation unit 112 uses the environmental information, human information, and business information acquired by the information acquisition unit 111, along with the learned estimation model 121 stored in the storage device 12, to estimate a psychological value indicating the psychological state of user P. Specifically, the estimation unit 112 inputs the environmental information, human information, and business information acquired by the information acquisition unit 111 as input data into the estimation model 121, and obtains the psychological value output from the estimation model 121 as the estimation result. The psychological value represents the degree of ease of work or well-being of user P.

[0023] The dialogue unit 113 conducts a dialogue with user P based on the psychological values ​​estimated by the estimation unit 112. Specifically, the dialogue unit 113 conducts a dialogue with user P when the psychological values ​​of user P estimated by the estimation unit 112 satisfy a predetermined dialogue condition. The dialogue condition is that user P's degree of ease of work or degree of well-being, which is a psychological value, is lower than a standard value which is the degree to which work is considered easy or comfortable. In other words, the dialogue unit 113 conducts a dialogue with user P when it is estimated that user P is having difficulty working or is not comfortable.

[0024] The dialogue unit 113 has a speech recognition function that recognizes speech input via the input device 13, a natural language processing function that determines the meaning of the recognized speech, and a speech function that generates dialogue content with user P and outputs it via the output device 14. The dialogue unit 113 generates dialogue content speech data from speech data stored in the speech data DB 122 ("DB" is an abbreviation for database) stored in the storage device 12, and outputs it from the output device 14.

[0025] The dialogue generated by the dialogue unit 113 consists of advice or casual conversation with user P. Advice to user P may include messages encouraging a tired user P to take a break, or messages encouraging a change of environment. Casual conversation may include messages about the current weather or news, or messages about user P's hobbies.

[0026] The dialogue unit 113 may randomly select audio data stored in the audio data DB 122 to be used as dialogue content, or it may generate dialogue content based on the psychological state of user P. For example, if the difference between the estimated psychological value and a standard value, which is the degree to which one is considered easy or comfortable to work with, is small, the dialogue unit 113 may generate casual conversation as dialogue content, and if the difference between the estimated psychological value and the standard value is large, it may generate advice as dialogue content.

[0027] Furthermore, the dialogue unit 113 generates and outputs a response to the user P's utterance input to the input device 13. For example, the dialogue unit 113 selects voice data corresponding to the user P's utterance from the voice data stored in the voice data DB 122 and outputs it from the output device 14.

[0028] The storage device 12 is composed of, for example, volatile or non-volatile semiconductor memory such as RAM (Random Access Memory), ROM (Read Only Memory), flash memory, EPROM (Erasable and Programmable ROM), EEPROM (Electrically Erasable and Programmable ROM), HDD (Hard Disk Drive), SSD (Solid State Drive), or a combination thereof. The storage device 12 stores programs used for estimation and interaction with the control device 11, as well as various data such as calculation formulas and thresholds used for executing said programs.

[0029] Furthermore, the storage device 12 stores the estimation model 121 and the voice data DB 122. The estimation model 121 is a trained model that takes environmental information, person information, and work information as input data and outputs psychological values ​​indicating the psychological state of user P as output data. The estimation model 121 is trained using user P's unique training data. The estimation model 121 is trained, for example, by supervised learning using a neural network.

[0030] An example of the learning method for the estimation model 121 is described below. The estimation model 121 learns the psychological value of user P for each type of environmental information, person information, and work information, using past environmental information, at least one of user P's personal information and work information, and user P's psychological value at that time as learning data. The psychological value of user P in the learning data can be obtained from past feedback from user P. Feedback from user P may be obtained from at least one of the following: information about user P's state, such as their speech, physical and mental state, and behavioral history in the personal information, and biometric information such as user P's brain waves, heart rate, pulse wave, body temperature, and posture, or it may be input from user P via the input device 13.

[0031] For example, suppose the estimation model 121 is trained using training data where the ease of working is 2 when the noise level is 60 dB and user P is doing desk work, and training data where the ease of working is 5 when the noise level is 40 dB and user P is doing desk work. In this case, the trained estimation model 121 will output an ease of working score of 2 as a psychological value when the environmental information input is a noise level of 60 dB and the work information input is that user P is doing desk work. The above is just one example, and in reality, the estimation model 121 is trained to output a psychological value based on multiple factors.

[0032] Even in the same environment, the degree of ease of work and well-being varies from person to person. Furthermore, the degree of ease of work and well-being also varies from person to person depending on the combination of work content and environment. For example, when doing desk work, some people work better in a quiet environment, while others work better in an environment with music playing. Also, even among people who work better in a quiet environment when doing desk work, some work better with music playing when doing physical labor, while others work better in a quiet environment even when doing physical labor. Estimation model 121 can output an estimated result of a psychological value unique to user P, taking into account user P's physical and psychological state, by learning past environmental information, at least one of person information and work information, and user P's degree of ease of work at that time as training data.

[0033] The voice data DB 122 stores various voice data used for dialogue by the dialogue unit 113. The voice data DB 122 includes voice data for giving advice to user P and voice data for casual conversation with user P. The voice data DB 122 communicates with external devices via a network such as the Internet and acquires voice data from external devices.

[0034] The input device 13 is a microphone that acquires the voice of user P. In addition to the microphone, the input device 13 may also include a camera that acquires an image of user P, or an operating device such as a keyboard, mouse, or touch panel operated by user P. The input device 13 receives the voice from user P, as well as at least one of user P's personal information and business information. The input device 13 transmits the input voice to the dialogue unit 113 and transmits the input personal information and business information to the information acquisition unit 111 of the control device 11.

[0035] The output device 14 is a speaker that outputs audio information to the user P. In addition to the speaker, the output device 14 may also include a display device such as a liquid crystal display or an organic EL display that outputs visual information to the user P. The output device 14 outputs the audio generated by the dialogue unit 113.

[0036] Figure 5 is a flowchart showing the operation flow of the dialogue device 1 according to Embodiment 1. First, the information acquisition unit 111 acquires environmental information, person information, and work information from the environmental information measuring device 2 and user P as input information to be input to the estimation model 121 (S1). Next, the estimation unit 112 estimates the psychological value using the learned estimation model 121 stored in the storage device 12 (S2). Here, the estimation unit 112 inputs the environmental information, person information, and work information acquired by the information acquisition unit 111 to the estimation model 121 and obtains the psychological value as output data.

[0037] The dialogue unit 113 determines whether the psychological value estimated by the estimation unit 112 satisfies the dialogue conditions (S3). For example, the dialogue unit 113 determines that the dialogue conditions are met if the psychological value estimated by the estimation unit 112, which is the degree of ease of work or the degree of well-being, is less than a preset threshold. The threshold is a value less than or equal to a standard value that user P considers easy to work with or comfortable, and is stored in the storage device 12. If the psychological value estimated by the estimation unit 112 does not satisfy the dialogue conditions (S3: NO), this process is terminated.

[0038] If the psychological value estimated by the estimation unit 112 satisfies the dialogue conditions (S3: YES), the dialogue unit 113 generates dialogue content including advice or casual conversation for user P (S4), and outputs it as voice from the output device 14 to start a dialogue with user P (S5). Thereafter, the dialogue with user P continues until user P gives an instruction to end.

[0039] As described above, in this embodiment, the dialogue device 1, which is linked to user P, can estimate user P's psychological state and engage in dialogue with user P based on that state, thereby providing psychological support to user P. This can improve user P's psychological state and motivation, leading to increased productivity. Furthermore, by having the dialogue device 1 as the conversation partner, user P can easily talk about work-related problems or their physical and mental state that they previously discussed with others, thereby reducing user P's stress. In addition, by having the dialogue device 1 estimate user P's psychological state and provide advice from an outsider's perspective, user P can become aware of fatigue and emotions that they may not be aware of, and their fatigue and emotions can be improved.

[0040] Embodiment 2. Figure 6 is a schematic diagram of the dialogue system 100A according to Embodiment 2. As shown in Figure 6, the dialogue system 100A of this embodiment differs from Embodiment 1 in that, in addition to the dialogue device 1A and the environmental information measuring device 2, it also includes an environmental control device 3. The environmental control device 3 is connected to the dialogue device 1A via wired or wireless communication.

[0041] The environmental control device 3 is a device that controls the environment of the indoor space where user P is present. The environmental control device 3 is, for example, at least one of an air conditioner, humidifier, dehumidifier, ventilation device, air purifier, lighting device, sound device such as a speaker (including a sound-dampening device), and fragrance diffuser that emits a scent. The air conditioner, humidifier, dehumidifier, and ventilation device control the temperature and humidity of the room. The ventilation device and air purifier control the CO2 in the room. ２It controls the concentration, dust concentration, and odor intensity. The lighting device controls the indoor illuminance. The acoustic device controls the indoor noise level. The aroma diffuser controls the indoor odor intensity. Note that the environmental control device 3 is not limited to the examples listed above, and other devices may also be used.

[0042] FIG. 7 is a schematic configuration diagram of the dialogue device 1A according to the second embodiment. As shown in FIG. 7, the dialogue device 1A includes a control device 11A, a storage device 12, an input device 13, and an output device 14. The configurations and functions of the storage device 12, the input device 13, and the output device 14 are the same as those in the first embodiment.

[0043] The control device 11A includes, as functional units, an information acquisition unit 111, an estimation unit 112, a dialogue unit 113, and a control unit 114. The configurations and functions of the information acquisition unit 111, the estimation unit 112, and the dialogue unit 113 are the same as those in the first embodiment.

[0044] The control unit 114 controls the environmental control device 3 based on the psychological value estimated by the estimation unit 112. Specifically, when the estimated psychological value satisfies the dialogue condition, that is, when the degree of ease of work or well-being of the user P is less than the threshold, the control unit 114 controls the environmental control device 3 so that the degree of ease of work or well-being of the user P is not less than the threshold.

[0045] For example, the control unit 114 controls the acoustic device to play music that allows the user P to relax, or controls the aroma diffuser to diffuse a scent that allows the user P to relax. Alternatively, the control unit 114 controls the temperature, humidity, or CO of the air in the indoor space. ２The air conditioner, air purifier, or ventilation device is controlled so that values such as the concentration become ideal values generally regarded as comfortable. Alternatively, the control unit 114 uses a table or calculation formula that represents the correspondence relationship between the difference between the estimated degree of ease of work or well-being and the threshold value set as the dialogue condition and the adjustment value of the environmental information to obtain the environmental control information to be instructed to the environmental control device 3, and may instruct the environmental control device 3. The adjustment value of the environmental information is the amount of change in the environmental information for making the estimated degree of ease of work or well-being equal to or higher than the threshold value. The control unit 114 obtains environmental control information by adding or subtracting the adjustment value from the current environmental information measured by the environmental information measuring device 2, and instructs the environmental control device 3 to target the environmental control information.

[0046] FIG. 8 is a flowchart showing the operation flow of the dialogue device 1A according to the second embodiment. First, the information acquisition unit 111 acquires environmental information, human information, and business information from the environmental information measuring device 2 and the user P as input information to be input to the estimation model 121 (S11). Subsequently, the estimation unit 112 estimates a psychological value using the learned estimation model 121 stored in the storage device 12 (S12). Here, as in the first embodiment, the estimation unit 112 inputs the environmental information, human information, and business information acquired by the information acquisition unit 111 into the estimation model 121 and obtains a psychological value as output data.

[0047] The dialogue unit 113 determines whether the psychological value estimated by the estimation unit 112 satisfies the dialogue condition (S13). If the psychological value estimated by the estimation unit 112 does not satisfy the dialogue condition (S13: NO), this process ends. On the other hand, if the psychological value estimated by the estimation unit 112 satisfies the dialogue condition (S13: YES), the dialogue unit 113 generates a voice including advice or casual conversation content for the user P (S14), outputs it from the output device 14, and starts a dialogue with the user P (S15).

[0048] Then, the control unit 114 controls the environmental control device 3 so that the degree of ease of work or well-being, which is the psychological value of the user P, becomes equal to or higher than the threshold value (S16).

[0049] As described above, in this embodiment, the dialogue device 1 linked to user P estimates a psychological value indicating user P's psychological state and engages in dialogue with user P based on the estimated psychological value. In addition, by controlling the environment of the indoor space, it is possible to provide user P with support in both physical and mental aspects. This can improve user P's physical and mental state and increase user P's productivity.

[0050] In the above embodiment, the estimation model 121 outputs the user P's psychological value, but the output data of the estimation model 121 is not limited to this. For example, the estimation model 121 in Embodiment 2 may output environmental control information that realizes optimal ease of work in addition to the psychological value. Environmental control information may include, for example, the indoor temperature, humidity, and CO2 level that realizes optimal ease of work for the user P. ２ This includes at least one of the following: concentration, dust concentration, illuminance, odor intensity, and noise level. Note that the environmental control information is not limited to the examples listed above, and may include other information. In this case, the control unit 114 may control the environmental control device 3 based on the environmental control information output from the estimated model 121.

[0051] Embodiment 3. Figure 9 is a schematic diagram of the dialogue device 1B according to Embodiment 3. The dialogue system 100B of Embodiment 3 consists of the dialogue device 1B and the environmental information measuring device 2. The configuration and function of the environmental information measuring device 2 are the same as in Embodiment 1.

[0052] As shown in Figure 9, the dialogue device 1B comprises a control device 11B, a storage device 12, an input device 13, and an output device 14. The configuration and functions of the storage device 12, the input device 13, and the output device 14 are the same as in Embodiment 1.

[0053] The control device 11B has, as functional units, an information acquisition unit 111, an estimation unit 112, a dialogue unit 113, and a business unit 115. The functions of the information acquisition unit 111, the estimation unit 112, and the dialogue unit 113 are the same as in Embodiment 1.

[0054] The operations unit 115 automatically performs tasks based on the psychological values ​​estimated by the estimation unit 112. Specifically, if the estimated psychological values ​​satisfy the dialogue conditions, that is, if the user P's ease of work or well-being level is below a threshold, the operations unit 115 performs simple tasks on behalf of user P. For example, the operations unit 115 has a function to automatically create documents using generation AI and performs tasks such as creating documents on behalf of user P.

[0055] The types of tasks to be automatically performed by the operations unit 115 are, for example, pre-set by user P and stored in the storage device 12. Alternatively, the operations unit 115 may automatically determine which tasks to perform based on user P's schedule or the content of conversations with user P. For example, if user P says, "I need to send an email to Mr. / Ms. XX...", the operations unit 115 will create the email on behalf of user P and request permission from user P to send it.

[0056] Figure 10 is a flowchart showing the operation flow of the dialogue device 1B according to Embodiment 3. First, the information acquisition unit 111 acquires environmental information, person information, and work information from the environmental information measuring device 2 and user P as input information to be input to the estimation model 121 (S21). Next, the estimation unit 112 estimates the psychological value using the learned estimation model 121 stored in the storage device 12 (S22). Here, similar to Embodiment 1, the estimation unit 112 inputs the environmental information, person information, and work information acquired by the information acquisition unit 111 to the estimation model 121 and obtains the psychological value as output data.

[0057] The dialogue unit 113 determines whether the psychological value estimated by the estimation unit 112 satisfies the dialogue conditions (S23). If the psychological value estimated by the estimation unit 112 does not satisfy the dialogue conditions (S23: NO), this process ends. On the other hand, if the psychological value estimated by the estimation unit 112 satisfies the dialogue conditions (S23: YES), the dialogue unit 113 generates audio containing advice or casual conversation content for user P (S24), outputs it from the output device 14, and starts a dialogue with user P (S25). Then, the business unit 115 automatically performs the business on behalf of user P (S26).

[0058] As described above, in this embodiment, the dialogue device 1B associated with user P estimates a psychological value indicating user P's psychological state, engages in dialogue based on the estimated psychological value, and also performs tasks on behalf of user P, thereby reducing user P's workload. As a result, user P's psychological state can be further improved, tasks can be facilitated, and user P's productivity can be increased.

[0059] The above describes the embodiments, but this disclosure is not limited to the above embodiments and can be modified in various ways without departing from the spirit of this disclosure. Furthermore, this disclosure includes all possible combinations of the configurations shown in the above embodiments. For example, the dialogue device 1 may include both the control unit 114 of Embodiment 2 and the business unit 115 of Embodiment 3.

[0060] Furthermore, the assignment of each functional unit in the control device 11 of the dialogue device 1 is not limited to the example of the above embodiment. For example, each functional unit of the control device 11 may be implemented by dividing it among multiple control devices, or a part of the functional unit of the control device 11 may be implemented by an external device that can communicate with the dialogue device 1. In addition, the estimated model 121 and voice data DB 122 stored in the storage device 12 of the dialogue device 1 may be stored in an external storage device such as online storage.

[0061] Furthermore, in the above embodiment, the dialogue unit 113 of the dialogue device 1 engages in voice-based dialogue with user P, but it may also engage in dialogue with user P using visual information such as text or images. In this case, the dialogue unit 113 displays advice or small talk to user P in text or images on the output device 14, which is a display device, based on user P's psychological state. Subsequently, based on the information input by user P via the input device 13, which is an operating device, it generates content to respond to user P and outputs it from the output device 14 to engage in dialogue. Alternatively, the dialogue unit 113 may engage in dialogue with user P using a combination of voice and visual information.

[0062] Furthermore, in the estimation model 121, when estimating psychological values ​​or environmental control information, environmental information, human information, and work information are configured to be input as input data. However, it is sufficient if at least one of these is input, rather than all of them. For example, only environmental information and human information may be input to the estimation model 121, or only human information and work information may be input. In addition, the estimation model 121 may be input spatial information, which includes at least one of the following: the size, height, materials and colors of the walls, floor and ceiling, and the configuration of the furniture in the room where user P is located. Spatial information is input, for example, by user P via the input device 13 and stored in the storage device 12.

[0063] Furthermore, in the above embodiment, the estimation unit 112 is configured to estimate a psychological value indicating the psychological state of user P using the estimation model 121, but it is not limited to this. For example, the estimation unit 112 may estimate the psychological state of user P based on at least one of the person information and work information input by user P. Specifically, the estimation unit 112 may estimate the psychological state of user P from at least one of the biometric information, speech and actions, mental and physical state, and behavioral history included in the person information, and the work content and work time included in the work information. For example, the estimation unit 112 estimates that user P's psychological state is poor, or estimates the psychological value to be lower than the standard value, when user P's biometric information is different from normal, when user P's work efficiency is poor, or when user P's speech and actions are different from usual. The dialogue unit 113 then determines that the dialogue conditions are met when the psychological state is poor or the psychological value is below the threshold, and conducts a dialogue with user P. In this case, the environmental information measuring device 2 and the estimation model 121 can be omitted, and the information acquisition unit 111 only needs to acquire at least one of the human information and business information.

[0064] Furthermore, the psychological value may be the degree of fatigue or the degree of stress. For example, the estimation unit 112 estimates that the stress level is higher than a standard value when user P's brainwaves are disturbed. Alternatively, the estimation unit 112 estimates that the fatigue level is higher than a standard value when user P spends a long time performing the same task. Alternatively, the estimation unit 112 estimates that the stress level is high when user P's emotional fluctuations are intense in their speech and actions. In this case, the dialogue condition is that the psychological value, the degree of fatigue or the degree of stress, is higher than a predetermined threshold, and the threshold is a value greater than the standard value at which user P is considered not tired or stressed. That is, the dialogue unit 113 conducts a dialogue with user P when it is estimated that user P is fatigued or under high stress.

[0065] Furthermore, the dialogue unit 113 in embodiments 1 to 3 may engage in dialogue with user P regardless of user P's psychological state. For example, if user P voices a problem they are having at work, the dialogue unit 113 may offer advice to user P. In addition, the dialogue unit 113 may take user P's personal personality into consideration and engage in dialogue from a "preventive" perspective before it is estimated that user P is having difficulty or discomfort working.

[0066] Furthermore, the control unit 114 in Embodiment 2 may control the environmental control device 3 regardless of the user P's psychological state. For example, if user P says something to himself such as "It's hot," the control unit 114 may automatically start the cooling operation of the air conditioner. Also, the business unit 115 in Embodiment 3 may perform tasks on behalf of user P regardless of user P's psychological state. For example, if user P says something to himself such as "I need to send an email to Mr. / Ms. XX," the business unit 115 may automatically create an email.

[0067] Furthermore, the information acquired by the dialogue device 1 in the above embodiment, the estimation results, and the content of the dialogue with user P may be stored in the storage device 12 or an external storage device as a personalized database for user P. The dialogue device 1 may then retrain the estimation model 121 and generate dialogue content based on the stored information. This configuration makes the dialogue device 1 personally linked to user P.

[0068] 1, 1A, 1B Dialogue device, 2 Environmental information measuring device, 3 Environmental control device, 11, 11A, 11B Control device, 12 Storage device, 13 Input device, 14 Output device, 100, 100A, 100B Dialogue system, 111 Information acquisition unit, 112 Estimation unit, 113 Dialogue unit, 114 Control unit, 115 Business unit, 121 Estimation model, 122 Voice data DB, 600 Processing circuit, 601 Processor, 602 Memory, 603 Bus.

Claims

1. A dialogue system for engaging in conversation with a user, comprising: an estimation unit for estimating the user's psychological state; and a dialogue unit for engaging in conversation with the user based on the user's psychological state.

2. The dialogue system according to claim 1, further comprising an information acquisition unit that acquires at least one of the environmental information of the indoor space in which the user resides, the user's personal information, and the user's work information, wherein the estimation unit takes at least one of the environmental information, the personal information, and the work information as input and estimates the psychological value indicating the user's psychological state using a trained estimation model that outputs a psychological value indicating the user's psychological state, and the estimation model is trained using training data specific to the user.

3. The dialogue system according to claim 1, further comprising an information acquisition unit that acquires at least one of the user's personal information and the user's business information, wherein the estimation unit estimates a psychological value indicating the user's psychological state using at least one of the personal information and the business information.

4. The dialogue system according to claim 2 or 3, wherein the dialogue unit compares the psychological value with a preset threshold and determines whether or not to engage in dialogue with the user.

5. The dialogue system according to any one of claims 2 to 4, wherein the dialogue unit generates dialogue content with the user based on the psychological value.

6. The dialogue system according to any one of claims 2 to 5, wherein the person information includes at least one of the user's state and the user's biometric information, the state information includes at least one of the user's speech, physical and mental state and behavioral history, the biometric information includes at least one of the user's brain waves, heart rate, pulse wave, body temperature and posture, and the work information includes at least one of the user's schedule, work content and work hours.

7. The dialogue system according to any one of claims 2 to 6, wherein the psychological value is the user's degree of ease of work, degree of well-being, degree of fatigue, or degree of stress.

Images