Action suggestion device, action suggestion method, and program

Abstract

An action suggestion device is provided that can generate optimal actions that match the state of the environment in which a user acts. [Solution] A management computer 11 having a processor 17 that executes processes, the processor 17 being configured to execute a first process of obtaining prompts for user P2 when he acts in the environment, a second process of determining the state of the environment in which user P2 acts, a third process of extracting actions that user P2 can perform in the environment based on the state of the environment, a fourth process of estimating features including comfort, efficiency, safety, and economy for user P2, assuming that user P2 will perform the actions extracted in the third process, and a fifth process of generating optimal actions that user P2 can perform in the state of the environment based on the prompts obtained in the first process and the features estimated in the fourth process.

Description

[Technical Field]

[0001] The present disclosure relates to an action suggestion device, an action suggestion method, and a program that suggest an action to a user according to the state of the user's environment. [Background technology]

[0002] An example of an action suggestion device that suggests an action to a user according to the state of the user's environment is disclosed in Patent Document 1. The action suggestion device disclosed in Patent Document 1 aims to provide a technology that increases the availability of a device equipped with an artificial intelligence model by making it possible to respond to changes in indicators requested by the user.

[0003] To achieve the above-mentioned objective, the action suggestion device disclosed in Patent Document 1 includes an acquisition unit, a generation unit, and an output unit. The acquisition unit acquires information about the situation of an instruction target as situation information. The instruction target is, for example, an unmanned aerial vehicle, a car, a person, or a robot. The generation unit generates suggested instruction content for the instruction target's behavior according to the situation information using an artificial intelligence model generated based on the relationship between the situation around the instruction target and the instruction target's behavior based on preset indicators. The generation unit also generates suggested instruction content according to the situation information using multiple artificial intelligence models each based on multiple mutually different indicators. Furthermore, the output unit outputs the generated suggested instruction content. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Publication No. 2022-80325 Summary of the Invention [Problem to be solved by the invention]

[0005] The inventors of the present application recognized that the behavior suggestion device described in Patent Document 1 uses an artificial intelligence model generated based on the relationship between the behavior of the target based on pre-set indicators, and therefore there is still room for improvement in terms of generating optimal behavior that matches the state of the environment in which the user is acting.

[0006] An object of the present disclosure is to provide an action suggestion device, an action suggestion method, and a program that are capable of generating an optimal action that matches the state of the environment in which the user is acting. [Means for solving the problem]

[0007] This embodiment discloses an action suggestion device having a processor that executes processes, wherein the processor executes a first process of obtaining prompts for a user to take an action in an environment; a second process of determining the state of the environment in which the user will take an action; a third process of extracting actions that the user can take in the environment based on the state of the environment; a fourth process of estimating features including comfort, efficiency, safety, and economy of the user, assuming that the user will take the action extracted in the third process; and a fifth process of generating optimal actions that the user can take in the state of the environment based on the prompts obtained in the first process and the features estimated in the fourth process. [Effects of the Invention]

[0008] According to this embodiment, the processor generates an optimal action that the user can perform in the environmental state based on the prompt and features including user comfort, efficiency, safety, and economy. This allows the processor to generate an optimal action that matches the environmental state and the prompt generated by the user. This improves the accuracy of the processor's generation of the optimal action. [Brief explanation of the drawings]

[0009] [Figure 1] FIG. 1 is a conceptual diagram illustrating an example of the configuration of an action suggestion system including an action suggestion device. [Figure 2] FIG. 2 is a conceptual diagram illustrating an example of the configuration of a management server that is an example of an action suggestion device. [Figure 3] 1 is a flowchart illustrating an example of an action suggestion method executed in the action suggestion system. DETAILED DESCRIPTION OF THE INVENTION

[0010] (Example of the configuration of an action suggestion system) In this embodiment, several specific examples of an action suggestion system, an action suggestion device, an action suggestion method, and a program are described with reference to the drawings. Fig. 1 shows an example of an action suggestion system 10. The action suggestion system 10 is composed of a management computer 11, a terminal device 12, an external computer 13, etc. The management computer 11 is connected to the terminal device 12 and the external computer 13 via a network 14. The external computer 13 has a processor, a storage device, a communication device, etc., and provides a website to the network 14.

[0011] The management computer 11 is a computer managed and operated by an administrator P1. The terminal device 12 is operated and used by a user P2. Multiple terminal devices 12 operated by different users P2 may be provided. The user P2 may be an individual, a company representative, a government representative, etc. The external computer 13 is managed by an administrator P3 separate from the administrator P1 of the management computer 11. Multiple external computers 13 managed by different administrators P3 may be realized.

[0012] The network 14 is a communication line that transmits various types of information, and is composed of base stations, communication satellites, communication antennas, communication equipment (access points), repeaters, communication circuits, communication cables (copper wires, optical fibers), etc. Repeaters include modems (devices that convert analog signals into digital signals), hubs (concentrators), routers (devices that connect terminals to the Internet), etc. The communication line includes one or more communication means of wireless communication or wired communication. Wireless communication includes short-range wireless communication. Short-range wireless communication includes, for example, wireless LAN (Wi-Fi (registered trademark)) and Bluetooth (registered trademark). Wireless communication includes electrical signals, optical signals, radio waves, infrared rays, satellite communication, etc. Signals used in communication include digital signals and analog signals.

[0013] (Example of management server configuration) The management computer 11 is a computer realized by various hardware and software. The software includes an operating system and applications. The management computer 11 may have hardware such as a main body (casing), a ROM (Read Only Memory) 16, a processor 17, a main memory device 18, an auxiliary memory device 19, and a communication device 20. The ROM 16 is a memory for reading data only, and data is written to the ROM 16 at the time of manufacturing and remains unchanged. The ROM 16 stores programs, such as an IPL (Initial Program Loader), that are executed first when the management computer 11 is started. The ROM is a non-volatile memory device.

[0014] The processor 17 is provided inside the main body of the management computer 11, and may be configured as a central processing unit (CPU (Central Processing Unit)) that integrates an arithmetic unit (arithmetic circuit) and a control unit (control circuit). The processor 17 is communicably connected to the ROM 16, main memory device 18, auxiliary memory device 19, communication device 20, etc. via a communication bus 21. The processor 17 comprehensively controls other devices and circuits provided inside the main body and devices and circuits provided outside the main body.

[0015] Furthermore, instead of or in addition to a central processing unit, processor 17 may include an arithmetic processing circuit such as a digital signal processor, an application specific integrated circuit (ASIC), a graphics processing unit (GPU), or a micro processing unit (MPU). The GPU processes image data and audio data.

[0016] The processor 17 executes various processes by running a program. The processes executed by the processor 17 include calculation, judgment, learning, comparison, identification, classification, control, generation, etc. The processes executed by the processor 17 include reading a program from the auxiliary storage device 19, running a program read from the auxiliary storage device 19 in the main storage device 18, retrieving information and data from the auxiliary storage device 19 and processing them in accordance with the program, processing information and data acquired from the external computer 13, processing information and data acquired from the terminal device 12, storing the results of processing executed in the main storage device 18 in the auxiliary storage device 19, processing information and data to be transmitted to the terminal device 12, etc.

[0017] The main memory device 18 is a volatile memory device, and is realized by, for example, RAM (Random Access Memory). The main memory device 18 functions as a work area and a buffer area for storing programs, data, instructions, etc. when the processor 17 processes and executes the programs, data, instructions, etc. retrieved from the auxiliary memory device 19. The main memory device 18 is a non-transitory storage medium.

[0018] The auxiliary storage device 19 is a non-transitory storage medium. The auxiliary storage device 19 may also be understood as storage. The auxiliary storage device 19 may be provided inside the main body, or may be provided outside the main body and connected to the communication bus 21 by wireless or wired communication. The auxiliary storage device 19 has a larger capacity than the main storage device 18. The auxiliary storage device 19 operates in response to input and output commands from the processor 17.

[0019] The auxiliary storage device 19 is configured using one or more of the following methods: magnetic storage, optical storage, magneto-optical storage, semiconductor storage, etc. The magnetic storage method is realized by a hard disk, floppy disk, magnetic tape, etc. The optical storage method is configured to irradiate the surface of the storage medium 19A with laser light to record data using light or heat, and read the reflected laser light with an optical head.

[0020] The optical storage method is realized by an optical drive and storage medium 19A. Optical drives include a Combo drive, a Super Combo drive, a Multi drive, a Super Multi drive, a Hyper drive, etc. Storage medium 19A includes a Compact Disc, a Digital Video Disc, a Blu-ray Disc, etc. The magneto-optical method is realized by, for example, a magneto-optical disc.

[0021] The magnetic storage method and the optical storage method have a head and a storage medium 19A. The storage medium 19A may be configured to be attachable to and detachable from the main body. The semiconductor storage method may be defined as a flash memory. Examples of flash memory include an SD memory card, a USB flash drive, and a solid state drive. The flash memory can also be understood as a storage medium 19A configured to be attachable to and detachable from the main body. The auxiliary storage device 19 stores non-transitory programs, various information, and data.

[0022] 2 by running non-transitory programs and cooperating with various pieces of hardware that make up the management computer 11. The processor 17 realizes, for example, a website management unit 22, a user information processing unit 23, an acquired information processing unit 24, an artificial intelligence unit 25, and the like.

[0023] When a web application is run by processor 17, website management unit 22 provides a website at a predetermined uniform resource locator (URL) on network 14. When terminal device 12 accesses management computer 11 via a website, website management unit 22 realizes the functions of transmitting a web page written in HTML (Hyper Text Markup Language) format to terminal device 12, acquiring information and data transmitted from terminal device 12, and transmitting information and data to terminal device 12.

[0024] The user information processing unit 23 processes the user information acquired from the terminal device 12 to register the user, and stores the user information and the results of the user registration in the auxiliary storage device 19. If the user P2 is an individual, the user information includes the name, address, workplace location, age, gender, hobbies, product purchase history, occupation, email address, an account that identifies the individual, etc. The account includes an identification number and a password.

[0025] If user P2 is a company official, the user information includes the company name, company location, company industry, email address, account identifying the company, etc. The account identifying the company includes an identification number and password, etc. If user P2 is a government official, the user information includes the name of the government agency or ministry, the location of the government agency or ministry, the public business or duties handled, email address, and an account identifying the government agency or ministry. The account includes an identification number and password.

[0026] The acquired information processing unit 24 processes information acquired from the terminal device 12 to acquire auxiliary information other than user information, and processes various information and data acquired from the external computer 13. The acquired information processing unit 24 also stores the auxiliary information and various information and data acquired from the external computer 13 in the auxiliary storage device 19. The auxiliary information acquired from the terminal device 12 may include location information of user P2, biometric information of user P2, emotional information of user P2, prompts, information on the user P2's behavior, etc. Of the auxiliary information, the location information of user P2, biometric information of user P2, prompts, and information on the user P2's behavior will be described later.

[0027] The acquired information processing unit 24 processes information and data acquired from the terminal device 12 to indirectly acquire emotional information of the user P2. The acquired information processing unit 24, in cooperation with the artificial intelligence unit 25 and the auxiliary storage device 19, has the function of estimating emotional information of the user P2 from image information and image data acquired from the terminal device 12 and converting the information into text data. The emotions of the user P2 include joy, satisfaction, anger, sadness, calmness, composure, enjoyment, surprise, comfort, discomfort, relief, relaxation, confusion, etc. These emotions can be estimated by processing the eyelid opening, gaze, eyebrow movement, presence or absence of nose wrinkles, mouth movement, mouth opening, pupil opening, etc., contained in an image of the upper body including the face of the user P2. These emotions can also be estimated from the upper body movements of the user P2, such as whether or not the head is tilted, whether or not the user nods, and hand movements. The estimated emotions are each quantified, and a table or graph can be generated.

[0028] In addition, the technology for analyzing the emotions of user P2 based on image information including the face of user P2 is a publicly known technology, as shown in, for example, Patent Gazette No. 6868422, Patent Gazette No. 6703893, Patent Gazette No. 6042015, Patent Gazette No. 3953024, Patent Gazette No. 4458888, JP 2015-229040, JP 2018-32164, JP 2022-139436, JP 2020-184216, etc., and therefore a detailed explanation thereof will be omitted.

[0029] Details of the various information and data that the management computer 11 acquires from the external computer 13 will be described later. The prompt is information that the user P2 inputs into the terminal device 12 when the user P2 attempts to acquire the optimal action from the management computer 11. The management computer 11 acquires the prompt from the terminal device 12. Details of the prompt will be described later.

[0030] When the program is executed, the artificial intelligence unit 25 executes various processes in cooperation with the auxiliary storage device 19. The processes executed by the artificial intelligence unit 25 include a learning process (learning stage) and an inference process (inference stage). The information and data used by the artificial intelligence unit 25 to execute the processes include image data, sound data, text data, various models, etc.

[0031] In a learning process, for example, a machine learning process, the artificial intelligence unit 25 processes input learning target data using a learning model, thereby outputting a processing result (output data). The artificial intelligence unit 25 also calculates the error between the processing result and correct answer data using a loss function, and trains the learning model based on the error. Specifically, it adjusts weights, biases, etc. to reduce the error. In an inference process, the artificial intelligence unit 25 processes inference target data using the learning model, thereby outputting an inference result.

[0032] The processing of input data performed by machine learning includes extracting features from the input data. The features of the input data include classification and regression. Classification involves classifying and identifying features. Regression involves grasping the trends of classified features. The learning model is stored in the auxiliary storage device 19, and as machine learning is repeated, the learning model is trained as needed.

[0033] The machine learning performed by the artificial intelligence unit 25 may include three types: supervised learning, unsupervised learning, and reinforcement learning. In supervised learning, a learning model is trained by comparing output data with correct answer data. Unsupervised learning is a learning method using learning data that does not include correct answers, and is a method of classifying the learning data into groups of data with similar features (such as regularity).

[0034] Reinforcement learning involves repeatedly processing input data with a learning model to obtain output data, thereby generating "actions to achieve this goal." Specifically, it may generate optimal actions to achieve the goal. The reinforcement learning performed by the artificial intelligence unit 25 includes the following four-stage process.

[0035] In the first stage, the artificial intelligence unit 25 processes input data using a learning model and determines the state of the environment of the user (acting subject / agent) P2. The real world in which the user P2 acts corresponds to the environment, and the weather in the real world, the biometric information of the user P2, the current location of the user P2, economic information in the real world, traffic information in the real world, product purchase information by the user P2, etc. correspond to the state of the environment in the real world.

[0036] In the second stage, the artificial intelligence unit 25 uses the learning model to determine actions that the user P2 can perform in the current environment. In the example of stock trading, the actions would be which stocks the user P2 will buy and sell, at what stock prices, and in how many shares. The learning model may determine actions that the user P2 can perform in the current environment under predetermined conditions in response to a prompt received from the terminal device 12.

[0037] In the third stage, the artificial intelligence unit 25 uses the learning model to predict the level of reward that can be obtained by adopting the "feasible action determined in the second stage." In the stock trading example, the profit resulting from trading stocks corresponds to the level of reward. Note that the technical meaning of reward varies depending on the current environment, the feasible action, and the content of the prompt obtained from the terminal device 12.

[0038] Then, in the fourth stage, the artificial intelligence unit 25 repeats the process of training the learning model so that it can select actions that will result in relatively larger (maximized) rewards in the future (practical stage) based on the results of the environmental judgment, the actions that can be performed, and the prediction results of the degree of reward that can be obtained by adopting the actions that can be performed.

[0039] For example, the learning model may be trained so that the value (Q-value / state-action value) of user P2's execution in the environment becomes relatively large, preferably so that the value is maximized. The Q-value is a function including an intermediate reward (immediate reward) in the short term and a long-term reward (expected value) in the long term. In practice, it is difficult to calculate the expected value. For this reason, the management computer 11 proposes an optimal action to the terminal device 12, and user P2 executes the optimal action. Next, the terminal device 12 sends the results of the execution to the management computer 11. The artificial intelligence unit 25 updates the Q-value by evaluating the results of the execution obtained from the terminal device 12.

[0040] Furthermore, when the processor 17 runs the program, the artificial intelligence unit 25 in this embodiment has the function of, in the inference stage, having the learning model execute processing using various information, data and prompts obtained from the terminal device 12, various information and data obtained from the external computer 13, and various information and data stored in the auxiliary storage device 19, and automatically generating the optimal behavior for the user P2.

[0041] The artificial intelligence unit 25 can use a neural network as a learning model to realize machine learning. A neural network has an input layer, an intermediate layer (hidden layer), and an output layer. The intermediate layer is configured to convert input information into a form that is useful for the output layer. The artificial intelligence unit 25 can also perform deep learning to realize the intermediate layer of the neural network. In deep learning, the intermediate layer is multi-layered with multiple types of stages (nodes), and weights and biases are assigned to functions (features) contained in the information. Deep learning automatically indicates the functions (features) that the intermediate layer focuses on, making it possible to reduce noise in the information and data input to the learning model.

[0042] Furthermore, the artificial intelligence unit 25 may use the functions of a large-scale language model to process prompts, generate sentences that suggest optimal actions, etc. Large-scale language models (LLMs) are a type of generative artificial intelligence specialized in natural language processing (NLP). Large-scale language models are language models that realize advanced natural language generation (LNG) and are constructed through deep learning of massive amounts of text data. Natural language processing processes natural language by combining processes such as morphological analysis, syntactic analysis, semantic analysis, context analysis, and intent analysis, enabling machine translation, text summarization, speech recognition, etc.

[0043] A language model is a system that predicts the next word after a given word by understanding the patterns of natural language used by humans and analyzing the probability of word occurrence. Examples of large-scale language models include ChatGPT (Chat Generative Pre-trained Transformer) (registered trademark), PaLM2 (registered trademark), and LaMDA2 (registered trademark). Words and sentences used as instructions to be input into these large-scale language models are called prompts.

[0044] When functioning as a generative artificial intelligence unit, the artificial intelligence unit 25 can realize text generation, image generation, video generation, and voice generation. Text generation is a function that, when a prompt (in text format), such as an instruction or question, is input, analyzes the content and intent of the prompt and uses information and data stored in the auxiliary storage device 19 to automatically generate a response (in text format) to the prompt. Because a large-scale language model is used, natural and highly accurate responses can be obtained.

[0045] Image generation is a function that automatically generates an original image that is close to the image of the input prompt (text format) by using information and data stored in the auxiliary storage device 19. Known examples of artificial intelligence that generates images include Stable Diffusion, Mid Journey, and Dali-2.

[0046] Video generation is a function that automatically generates an original video that closely matches the image of an input prompt (text format) by utilizing information and data stored in the auxiliary storage device 19. Gentoo, for example, is a known example of an artificial intelligence that generates videos.

[0047] Voice generation is a function that automatically generates original voice data (optimal actions, responses, music, etc.) for an input prompt (voice or text format) using information and data stored in the auxiliary storage device 19.

[0048] The auxiliary storage device 19 stores information and data resulting from processing executed by the processor 17, and the information stored in the auxiliary storage device 19 also includes hypertext data that constitutes web pages on a website.

[0049] 2, the auxiliary storage device 19 is realized by including a program storage unit 26, a user information storage unit 27, a model storage unit 28, a various information storage unit 29, a processing result storage unit 30, etc. The program storage unit 26 stores applications run by the processor 17 and various programs run by the processor 17. The applications stored in the program storage unit 26 include non-transitory programs, manuals, setting files, files for storing data, data, various libraries, etc.

[0050] The user information storage unit 27 stores user information and registration information acquired from the terminal device 12. The model storage unit 28 stores learning models, large-scale language models, etc. The various information storage unit 29 stores information and data acquired from the external computer 13, prompts and auxiliary information acquired from the terminal device 12, map data, etc. The processing result storage unit 30 stores various processing results performed by the processor 17, various judgment results, generated optimal actions, etc.

[0051] The communication device 20 includes devices, equipment, and standards that connect the management computer 11 to the network 14 via at least one of a wireless communication system and a wired communication system. The hardware that constitutes the communication device 20 may be configured, for example, by a LAN card, a network adapter, a network interface card, a communication cable, a communication antenna, a communication port, a modem, a hub (line concentrator), a router, etc.

[0052] (Terminal device configuration) The terminal device 12 is a computer used and operated by the user P2. The terminal device 12 includes, for example, a portable computer and a fixed computer. The portable computer includes a smartphone and a tablet computer. The fixed computer includes a desktop computer, a notebook computer, etc.

[0053] The terminal device 12 is connected to the management computer 11 and the external computer 13 via a network 14. As shown in FIG. 1 , the terminal device 12 includes a main body (casing), a processor 40, a main storage device 41, an auxiliary storage device 42, an input device 43, an output device 44, a current position detection device 45, a biometric information detection device 46, a communication device 47, and the like. The processor 40 may be provided inside the main body and configured as a central processing unit (CPU) that integrates an arithmetic device (arithmetic circuit) and a control device (control circuit). The processor 40 is communicatively connected to the main storage device 41, the auxiliary storage device 42, the input device 43, the output device 44, the current position detection device 45, the biometric information detection device 46, and the communication device 47 via a communication bus 48. The processor 40 is configured to comprehensively control other devices and circuits provided inside the main body and devices and circuits provided outside the main body.

[0054] The processor 40 processes information and data by running non-transitory programs. Specifically, the processor 40 processes text data, sound data, image data, etc. Text data includes characters, numbers, sentences, symbols, figures, etc. Image data includes videos, moving images, still images, photographs, etc. The processes executed by the processor 40 include calculations, judgments, comparisons, controls, etc.

[0055] The processing performed by the processor 40 includes processing for reading information and data from the auxiliary storage device 42, processing of information and data obtained from the management computer 11, updating of information and data stored in the auxiliary storage device 42, and processing for sending information and data processed by the processor 40 to the management computer 11.

[0056] The main memory device 41 is a volatile memory device, and functions as a work area and a buffer area for storing programs, data, instructions, etc. retrieved from the auxiliary memory device 42 when the processor 40 executes processing of programs, data, instructions, etc. The main memory device 41 is a non-transitory storage medium.

[0057] The auxiliary storage device 42 is a non-transitory storage medium. Non-transitory applications are stored in the auxiliary storage device 42. The auxiliary storage device 42 also stores information and data to be processed by the processor 40. The information and data stored in the auxiliary storage device 42 include the processing results of the processor 40, various data, the operation contents of the input device 43, control signals to the output device 44, etc.

[0058] The processor 40 is configured to execute various processes by reading and executing non-transitory programs stored in the auxiliary storage device 42. The processor 40 is configured to process, for example, information and data input by the input device 43, and process the current location of the user P2 detected by the current location detection device 45. The processor 40 is configured to process biometric information detected by the biometric information detection device 46, and process information and data output from the communication device 47. The processor 40 is configured to process information and data input via the communication device 47. Furthermore, the processor 40 is configured to store the processing results in the auxiliary storage device 42.

[0059] The auxiliary storage device 42 has a larger capacity than the main storage device 41, and operates in response to input and output commands from the processor 40. The auxiliary storage device 42 is realized, for example, by a head and a non-transitory storage medium 42A. The head writes information and data to the storage medium 42A and reads information and data from the storage medium 42A. The storage medium 42A is realized by a magnetic disk, an optical disk, a flash memory, etc. An example of a magnetic disk is a hard disk drive. An example of an optical disk is a compact disk, a digital video disk, a Blu-ray disk, etc. A flash memory is a type of semiconductor memory, and examples of flash memory include an SD memory card, a USB flash drive, a solid-state drive, etc. The storage medium 42A may be configured to be attachable to and detachable from the main body.

[0060] The input device 43 is operated when inputting information and data into the terminal device 12, when executing various processes in the terminal device 12, when sending various information and data from the terminal device 12 to the network 14, when the terminal device 12 obtains various information and data via the network 14, etc.

[0061] The input device 43 is a device operated by the user P2, and may be composed of devices such as a keyboard 43A, a mouse 43B, a microphone 43C, a camera 43D, and a display 43E. The display 43E may be realized by, for example, a liquid crystal display or an organic electroluminescence display. Images, videos, figures, characters, operation buttons, operation tabs, and the like are displayed on the screen of the display 43E. The display 43E may also be understood as a monitor. The user P2 can input a prompt in the form of text information by operating one or more of the keyboard 43A, the mouse 43B, and the operation buttons and operation tabs displayed on the screen of the display 43E. The user P2 can input a prompt in the form of audio information via the microphone 43C.

[0062] The microphone 43C is an electronic component that converts acquired sound data into an electrical signal and outputs it. The microphone 43C is connected to the communication bus 48 by either a wireless or wired communication system. The camera 43D is connected to the communication bus 48 by either a wireless or wired communication system. The camera 43D captures an object and generates an image. The image includes a video, a still image, a photograph, etc. The image generated by the camera 43D is processed by the processor 40 and stored in the auxiliary storage device 42.

[0063] By using the input device 43, the user P2 can input a destination in the map data, select a travel route in the map data, access the network 14, access the management computer 11, transmit information and data to the management computer 11, select the optimal action obtained from the management computer 11, etc. Note that some of the information and data processed by the terminal device 12 may be transmitted in real time from the terminal device 12 to the management server 123 when the input device 43 is not being operated.

[0064] The output device 44 includes devices such as a display 43E, a printer 60, and an audio output device 61. The display 43E is the same as the display 43E of the input device 43. The audio output device 61 is a device that converts electrical signals contained in information and data into sound information and outputs it. The audio output device 61 is realized by a speaker, headphones, earphones, etc. The audio output device 61 is connected to the communication bus 48 by wireless communication or wired communication. The printer 60 prints information and data on the screen of the display 43E on paper using toner or ink and outputs it.

[0065] The current position detection device 45 is composed of, for example, a detection circuit and a determination circuit. The current position detection device 45 processes signals received from artificial satellites via the communication device 47 and signals from a gyro sensor 45A and a magnetic sensor 45B provided in the terminal device 12. Based on the results of this processing and the map data stored in the auxiliary storage device 42, the current position detection device 45 can determine the current position and movement route of the terminal device 12 in the real world by replacing the map data with the map data. User P2 can display the map on the display 43E and visually check his or her current position and movement route.

[0066] Furthermore, the biological information detection device 46 may detect the heart rate (pulse rate) of the user P2 per predetermined time and the number of breaths of the user P2 per predetermined time. Methods for detecting the heart rate of the user P2 include a method of obtaining the heart rate from a wristwatch-type sensor 46A or the like worn by the user P2, and a method of indirectly estimating the heart rate from image information from the camera 43D. The technology for estimating the heart rate from image information is publicly known, as disclosed in Japanese Patent No. 6717424, Japanese Patent Laid-Open No. 2016-190022, Japanese Patent Laid-Open No. 2022-141984, etc., and therefore a detailed description thereof will be omitted.

[0067] The number of breaths per predetermined time of user P2 can be indirectly estimated from the image information of camera 43 D. The technology for estimating the number of breaths of a subject from image information is publicly known, as disclosed in Japanese Patent No. 4284538 and Japanese Unexamined Patent Publication No. 2016-190022, and therefore a detailed description thereof will be omitted.

[0068] The communication device 47 includes devices, equipment, and standards that connect the terminal device 12 to the network 14 via at least one of a wireless communication system and a wired communication system. The devices and equipment that connect the terminal device 12 to the network 14 may include cables, antennas, routers, communication circuits, communication ports, communication connectors, communication hubs, etc.

[0069] (An example of how to suggest actions) An example of the action suggestion method performed by the action suggestion system 10 is shown in Fig. 3. Fig. 3 shows an integrated view of the learning process and inference process performed by the management computer 11. The learning process performed by the management computer 11 includes steps S10 to S17 in Fig. 3. The learning process performed by the management computer 11 does not include steps S18 to S23 in Fig. 3.

[0070] In step S30, the terminal device 12 may send various types of information to the management computer 11. The various types of information sent from the terminal device 12 to the management computer 11 include user information and auxiliary information. In step S40, the external computer 13 may send various types of information to the management computer 11. The various types of information sent from the external computer 13 to the management computer 11 will be described later.

[0071] The inference process performed by the management computer 11 includes steps S10 to S13 and also steps S18 to S23 in Fig. 3. The inference process performed by the management computer 11 does not include steps S14 to S17 in Fig. 3.

[0072] In step S10, the management computer 11 acquires various information from the terminal device 12 and also acquires various information from the external computer 13. The various information includes instructions, information, and data. The information acquired by the management computer 11 from the terminal device 12 includes user information, auxiliary information, etc. The auxiliary information is associated with the account of user P2.

[0073] The prompt is text data that the user P2 creates when the user P2 attempts to obtain the optimal action from the management computer 11.

[0074] The various types of information that the management computer 11 acquires from the external computer 13 in step S10 may include, for example, trend information contained in a social networking service (SNS), weather information for the area where user P2 is active, economic data, and traffic information for the area where user P2 is personally active. Trend information contained in an SNS may include information on events that are currently trending in society, information on popular products, information on popular fashion, and the like.

[0075] The economic data may include price information in the area where user P2 is active, market trends in the area where user P2 is active, stock price information, real estate price information, production volume of goods by companies, etc. Market trends may include, for example, the consumer price index. Stock price information may include the stock price of each stock traded on the stock market, fluctuations in the stock price of each stock, the number of purchases of each stock, etc. Traffic information in the area where user P2 is active may include road congestion conditions, trouble and delay information for public transportation, and delivery times in logistics. The purchasing history of user P2 may include the user P2's past product purchase history. The product purchase history may be from either an EC (electronic commerce) site or a physical store. In addition, the products include tangible goods, intangible goods, services, etc.

[0076] In addition, the various data that the management computer 11 acquires from the external computer 13 in step S10 may include breaking news (fashion event in Shibuya), information analyzing user P2's emotions (best sale on SNS), energy consumption (power consumption) at user P2's home or company, schedule information for user P2 or the company, and health risk information for user P2.

[0077] Furthermore, the various data that the management computer 11 acquires from the external computer 13 in step S10 may include company inventory data information (clothing store inventory rate 50%), public transportation operation information (trains running normally), restaurant congestion information (cafe seats available), weather warnings (no disasters), exchange rate (weak yen: 150 yen to the dollar), ticket inventory information (50 tickets remaining for the event), information on user P2's medical appointments (no hospital appointments), and information on work tasks for user P2 or the company (number of unprocessed tasks).

[0078] Furthermore, the various data acquired by the management computer 11 from the external computer 13 in step S10 may include crime information in the area where user P2 is active (Shibuya is safe), information detected by an environmental sensor (air cleanliness), and information about the delivery status of a company's products (package delivery scheduled for 12 o'clock). The various data acquired by the management computer 11 from the external computer 13 in step S10 may also include information posted by an influencer (a celebrity posts that they are going to a sale) and progress information about user P2's learning (English study is 50% complete). The various data acquired by the management computer 11 from the external computer 13 in step S10 may also include information about congestion at tourist spots (for example, Harajuku is slightly crowded), electricity market prices (for example, 1 kWh = 30 yen), and information about the condition of user P2's pet (the dog is healthy and can be walked).

[0079] Furthermore, the various data that the management computer 11 acquires from the external computer 13 in step S10 may include climate change indexes, space weather, local event density, road construction information, school event schedules, public facility utilization rates, internet connection speeds, game play status, music trends, movie box office revenues, book sales rankings, cooking recipe popularity, sports game results, fashion trends, detailed local weather forecasts, medical institution congestion levels, traffic accident rates, public Wifi (registered trademark) utilization rates, local population density, electricity demand forecasts, water usage, gas consumption, garbage collection schedules, local SNS activity, disaster recovery progress, election-related information, public transportation occupancy rates, etc.

[0080] Furthermore, the various data that the management computer 11 acquires from the external computer 13 in step S10 may include bicycle sharing usage rate, local temperature change rate, pollen dispersion forecast, local event participation rate, tourist inflow, local economic growth rate, job information update rate, local infrastructure utilization rate, traffic signal optimization data, local energy self-sufficiency rate, public service satisfaction, local cultural event frequency, local health statistics, educational facility utilization rate, local crime prevention rate, local volunteer activities, local traffic safety index, local disaster preparedness, local environmental protection index, local education outcome index, local employment rate, local tourism satisfaction, local medical access, local traffic accident prevention rate, local energy efficiency, etc.

[0081] In step S11, the management computer 11 determines whether an "error" that prevents normal processing has occurred. For example, the management computer 11 may determine an error if the processing time of the acquired data exceeds a predetermined value (default 1 second), and may determine that there is no error (normal) if the processing time of the acquired data is equal to or less than the predetermined value. For example, the management computer 11 may determine an error if the cache of the main storage device 18 exceeds 2 GB in 5 minutes, and may determine that there is no error (normal) if the cache of the main storage device 18 is equal to or less than 2 GB in 5 minutes. The management computer 11 may determine an error in step S11 if, for example, the data log record, i.e., one or more of the data input values ​​or score calculation results processed by the artificial intelligence unit 25, are missing or abnormal values. On the other hand, if there are no missing values ​​in either the data input values ​​or the score calculation results processed by the artificial intelligence unit 25 and they are normal values, the management computer 11 may determine that there is no error (normal) in step S11.

[0082] If the management computer 11 determines Yes in step S11, it performs a process in step S12 in which a substitute value is used as the score for each piece of data, and then proceeds to step S10. For example, for user P's location information, the last location information from the previous data acquisition may be used as a substitute value. For weather information, for example, the average value for the same month and day over the past 10 years may be used as a substitute value. For user P2's biometric information, a standard value for a user of the same age, sex, approximately the same weight, approximately the same height, and the same occupation may be used as a substitute value. For economic information, the average value for user P2's area may be used as a substitute value. For traffic information, a predicted value may be used as a substitute value. For purchase history, the average value of the purchase history of another user who has the same age, sex, hobbies, etc. as user P2 may be used as a substitute value. In step S12, the substitute values ​​other than user P's location information are predefined fixed values ​​or estimated values, and are pre-stored in the auxiliary storage device 19.

[0083] If the management computer 11 determines No in step S11, it performs normalization and complementation in step S13. The normalization is a preprocessing step that eliminates data duplication and enables consistent handling of data. For example, various data may be converted into a numerical value (score) ranging from 0.1 to 1.0. The more popular or talked about SNS information is, the higher the numerical value is converted into. The more weather information is converted into a numerical value when conditions are such that user P2's behavior is relatively likely to be promoted, safety is relatively high, visibility is relatively good, etc. For example, sunny is converted into a numerical value of 1.0, cloudy is converted into a numerical value of 0.8, rain is converted into a numerical value of 0.6, fog is converted into a numerical value of 0.3, snow is converted into a numerical value of 0.2, etc.

[0084] The biometric data of user P2 is converted into a relatively larger numerical value, for example, the smaller the difference between user P2's heart rate and the average heart rate by age. The biometric data of user P2 is converted into a relatively larger numerical value, for example, the smaller the difference between user P2's blood pressure and the average blood pressure by age. The location information of user P2 is converted into a relatively larger numerical value, for example, the shorter the distance between the current location and the destination location. The economic data is converted into a relatively larger numerical value, for example, the lower the price index.

[0085] For example, the traffic information is converted into a relatively larger numerical value the lower the road congestion rate. For example, the purchase history is converted into a relatively larger numerical value the greater the number of times user P2 has purchased goods, the number of products purchased, etc. The imputation process performed in step S13 is to impart missing values ​​of various data with substitute values. The substitute values ​​used in step S13 have a different technical meaning from the substitute values ​​used in step S12. The substitute values ​​used in step S13 are values ​​used to impart missing values ​​before the normalization process. In step S13, any of a fixed value, an estimated value (predicted value), a previous value, an average value, etc. is selected as the substitute value depending on the situation.

[0086] In step S14, the management computer 11 may perform one or more of the following determinations: determining whether data is missing; and determining whether the data is abnormal. Even if data is available, the management computer 11 may determine in step S14 that data is missing if the data does not meet the criteria required for suggesting optimal actions, i.e., the quality criteria required for score calculation. As a premise for the determination in step S14, the number and type of data required for score calculation are predetermined. Necessary data may include, for example, trend information included in SNS, weather information for the area where user P2 is active, biometric data of user P2, location information of user P2, economic data, traffic information for the area where user P2 is active, and purchase history of user P2.

[0087] Then, in step S13, the weather data is supplemented with the immediately preceding value, but if a sudden change in the weather occurs after that and the immediately preceding value and the current state do not match, it is determined in step S14 that the data is missing.

[0088] Furthermore, if some data has been complemented but other data is missing and all element data required for score calculation is not available, management computer 11 may determine that data is missing in step S14. Furthermore, if the time elapsed from the issuance of the complement request in step S13 to the completion of the complement process exceeds a predetermined time, for example, one second, management computer 11 may consider the substitute value used in the complement process to have expired and determine that data is missing in step S14.

[0089] Furthermore, the management computer 11 determines whether the acquired data is Data mean ±3σ If the data deviates from this range, it is judged to be abnormal. Here, "3σ" means the area range of 99.7% relative to the total area of ​​the histogram, which is 100%, and "σ" means the standard deviation.

[0090] If management computer 11 determines "Yes" in step S14, that is, if one or more of the following determinations are true: there is a data loss or there is a data abnormality, management computer 11 performs processing in step S15 and proceeds to step S13. In step S15, management computer 11 performs processing using a substitute value for each piece of data.

[0091] The substitute value used in step S15 has a different technical meaning from the substitute value used in step S12 and the substitute value used in step S13. The substitute value used in step S15 is the real-time data value detected immediately before the determination of Yes in step S14. In other words, the substitute value used in step S15 is a dynamic data value, not a fixed value.

[0092] For the location information of user P2, the management computer 11 may use the last location information at the time of data acquisition on the previous day as a substitute value. For economic information, the management computer 11 may use the average value in user P2's area as a substitute value. For traffic information, the management computer 11 may use a predicted value of traffic information as a substitute value. For purchase history, the management computer 11 may use the average value of the purchase history of another user who is the same age, sex, hobby, etc. as user P2 as a substitute value.

[0093] If the determination in step S14 is No, the management computer 11 analyzes each piece of data in time series using a long short-term memory (LSTM) model and predicts the future of each piece of data in step S16. In the long short-term memory model, data with relatively high importance is retained for a long period of time, and data with relatively low importance is deleted.

[0094] Furthermore, in step S17 following step S16, the management computer 11 trains the learning model by performing reinforcement learning. Specifically, Q-learning, which is one of the policy-off time-difference methods, may be performed. The management computer 11 may train the learning model using a past dataset as learning target data (training data). First, the management computer 11 generates a past dataset based on information about user P2's past execution actions for the "optimal action" generated based on the future prediction performed in step S16. The information about user P2's execution actions includes the value of a reward previously awarded by the artificial intelligence unit 25 for the result of user P2's execution of the proposed optimal action.

[0095] For example, the management computer 11 inputs weather information, traffic information, and the results of actual behavior as initial data, and based on future predictions, suggests behavior to the user P2, such as indoor movement by the user P2, as an optimal behavior.The management computer 11 then evaluates the "time saved" as a result of the user P2, who has acquired the optimal behavior, actually moving indoors.Based on the evaluation results, the Q value (reward) is updated and the weight of the score calculation is adjusted.

[0096] More specifically, the management computer 11 may suggest indoor travel based on a future prediction that the weather information indicates rain and the traffic information indicates congestion. Then, when the management computer 11 determines that the travel time of user P2 can be reduced to 80% of the estimated travel time of user P2 (100%) before the optimal action was suggested, the management computer 11 automatically adjusts the safety score to increase its weight. The travel time is the time required for user P2 to reach the destination from a predetermined location. The data described in the processes and judgments from step S10 to step S17 above corresponds to learning target data in the learning process and to inference target data in the inference process.

[0097] In the inference process performed by the management computer 11, the process skips from step S13 to step S18. In step S18, the management computer 11 determines whether the scores of the various data items that are prerequisites for generating an optimal action are all present and whether the elapsed time since the request to generate an optimal action was made is within a predetermined time. The request to generate an optimal action is made, for example, when the management computer 11 receives a prompt from the terminal device 12. The predetermined time is, for example, one second.

[0098] If the determination in step S18 is No, the management computer 11 proceeds to step S19 to execute default standby processing, and then proceeds to step S10. In the default standby, the intermediate reward is Q value = 0.5 The key is to set

[0099] If the management computer 11 determines "Yes" in step S18, it inputs the prompt and various real-time data acquired from the terminal device 12 into the learning model in step S20, and starts the process of generating an optimal behavior "score S" by natural language processing. The algorithm by which the management computer 11 generates the optimal behavior is, for example, Score S = Σ (element value × weight) It is expressed as:

[0100] Here, the element value is a weighted average of the numerical values ​​indicating each element for evaluating the priority (measure) of an action, such as comfort, efficiency, safety, economy, and action timing. The management computer 11 may determine the numerical value indicating comfort from the numerical values ​​of real-time data such as weather information, road congestion rate, train delay information, air cleanliness, and pollen dispersion status. The management computer 11 may determine the numerical value indicating efficiency from the numerical values ​​of real-time data such as location information, economic information, and traffic information.

[0101] The management computer 11 may determine a numerical value indicating safety from the numerical value of real-time data such as weather information, disaster information, crime information, traffic information, etc. The management computer 11 may determine a numerical value indicating economic efficiency from the numerical value of real-time data such as economic information, energy consumption, exchange rates, etc. The management computer 11 may determine a numerical value indicating action timing from the numerical value of real-time data such as exchange rates, road construction information, regional disaster recovery rates, learning progress, etc.

[0102] The larger the numerical value indicating each element, the higher the comfort, the higher the efficiency, the higher the safety, the higher the economy, and the higher the appropriateness of the timing of the action, respectively. For example, the shorter the distance between the current location and the destination location, the larger the numerical values ​​indicating efficiency and timing, contributing to optimal action by user P2. In other words, "optimal action" means "action that maximizes the score S after comprehensively considering multiple evaluation elements." The weight is a numerical value set according to a prompt obtained from the terminal device, and is set within the range of 0.0 to 1.0.

[0103] The management computer 11 may determine the numerical value indicating comfort from real-time data such as weather information, road congestion rate, train delay information, air cleanliness, pollen dispersion status, etc. The management computer 11 may determine the numerical value indicating efficiency from real-time data such as location information, economic information, traffic information, etc.

[0104] The management computer 11 may determine a numerical value indicating safety from real-time data such as weather information, disaster information, crime information, traffic information, etc. The management computer 11 may determine a numerical value indicating economic efficiency from real-time data such as economic information, energy consumption, exchange rates, etc. The management computer 11 may determine a numerical value indicating action timing from real-time data such as exchange rates, road construction information, regional disaster recovery rates, learning progress, etc.

[0105] An example of an optimal action generated by the management computer 11 is as follows: The management computer 11 may generate an optimal action for a specific individual when the user P2 travels from their current location to their destination. Based on a prompt acquired from the terminal device 12, the management computer 11 may process inference target data such as location information, weather information, and traffic information using a learning model to generate, as the optimal action, a travel route and means of travel that can reduce travel time by 10 minutes compared to the current travel method and route, cost 500 yen or less, and have a safety rating of 80% or more.

[0106] The management computer 11 may generate optimal actions for delivering goods to a company. The management computer 11 may process inference target data such as weather information and traffic information using a learning model based on prompts acquired from the terminal device 12, and generate, as optimal actions, a delivery route and delivery routine that can reduce inventory costs by 5% from current levels, shorten delivery times by 20% compared to conventional delivery times, and maintain customer satisfaction.

[0107] The management computer 11 may generate optimal actions for the government when setting up evacuation shelters in the event of a disaster. The management computer 11 may generate optimal actions, such as the location of the evacuation shelter, the method of procuring materials, and the construction method for setting up the evacuation shelter, so that traffic congestion time can be reduced by 30%, the cost of setting up the evacuation shelter can be reduced by 10% or less, and the resident safety rate can be increased to 90% or more. Here, a 30% reduction in traffic congestion time means that the total time that traffic congestion occurs within a specified period of time, for example, within 24 hours, can be reduced by 30% compared to the conventional method.

[0108] "Keeping shelter construction costs within 10%" means keeping the cost within 10% of the standard cost of constructing a conventional shelter. "A resident safety rate of 90% or more" means that a safe evacuation environment is secured, with a resident safety rate of 100%. For example, an environment in which all residents can evacuate immediately and there are sufficient supplies is a resident safety rate of 100%, while an environment in which 90% of residents can evacuate and supplies are nearly sufficient is a resident safety rate of 90%.

[0109] The management computer 11 may generate, for individuals, the optimal behavior when attending an event, the optimal behavior when deciding whether to buy a product now, the optimal behavior for reducing home power consumption, etc. The management computer 11 may generate, for companies, the optimal behavior for increasing profits, the optimal behavior for improving customer satisfaction, the optimal behavior for increasing ticket sales, etc. The management computer 11 may generate, for governments, the optimal behavior for reducing medical expenses, the optimal behavior for improving public safety, the optimal behavior for reducing air pollution, the optimal behavior for stabilizing prices, etc.

[0110] In step S20, the artificial intelligence unit 25 performs a trial-and-error process to calculate the score S for two or more different combinations of data and information in order to generate an optimal behavior. Then, the artificial intelligence unit 25 outputs the behavior with the highest score among the multiple types of behavior obtained as a result of the trial and error as the optimal behavior. Note that in step S20, Score S = Σ (element value × weight) In this case, the plurality of optimal actions may differ in at least part of the content of the action.

[0111] In step S21, following step S20, the management computer 11 determines whether an optimal action was generated within a predetermined time, e.g., within one second. In other words, the optimal action generated by the artificial intelligence unit 25 is the action with the highest score among the actions generated by trial and error within the predetermined time. In other words, there is no threshold or upper limit for the score corresponding to the optimal action. Each time an optimal action is generated in step S21, the score of the optimal action may be the same or may vary. Furthermore, a maximum number of actions to be generated by trial and error may be set in advance, and the action with the highest score among the actions generated within the maximum number may be treated as the optimal action. If the management computer 11 determines "Yes" in step S21, it stores the generated optimal action in the auxiliary storage device 19 in step S22 and transmits the generated optimal action to the terminal device 12.

[0112] If the management computer 11 determines No in step S21, it outputs a standard phrase (template) for when an optimal action cannot be generated, for example, "Confirming the situation," in step S23, and proceeds to step S10. Standard phrases are stored in advance in the auxiliary storage device 19.

[0113] In step S31, the terminal device 12 may acquire the optimal action from the management computer 11 and output the optimal action from the output device 44. In step S32, the terminal device 12 performs an operation to select the optimal action acquired from the management computer 11. If there are multiple optimal actions, the user P2 can select one by operating the terminal device 12. In step S32, the terminal device 12 may input an actual action by the user P2 corresponding to the selected optimal action. The action performed by the user P2 is an actual action performed by the user P2 in the real world. In step S30, the terminal device 12 transmits information and data including the selected optimal action and the action performed by the user P2 to the management computer 11. In addition, the terminal device 12 stores information and data including the selected optimal action and the action performed by the user P2 in the auxiliary storage device 42.

[0114] (Effects of the embodiment) The management computer 11 generates an optimal action that the user P2 can take in the environmental state based on the prompts acquired from the terminal device 12 and feature quantities including the comfort, efficiency, safety, and economy of the user P2. This makes it possible to generate an optimal action that matches the state of the real world and the prompts generated by the user P. This improves the accuracy of generating the "optimal action" performed by the artificial intelligence unit 25 in the inference process.

[0115] Furthermore, the management computer 11 trains the learning model used in processing by the artificial intelligence unit 25 by evaluating the behavioral results of previously generated optimal behaviors performed by the user P2 in the past. Therefore, the accuracy of generating the "optimal behavior" performed by the artificial intelligence unit 25 in the inference processing is further improved.

[0116] Furthermore, the management computer 11 trains the learning model based on future predictions of the environmental state. This further improves the accuracy of the "optimal behavior" generated by the artificial intelligence unit 25 in the inference process. This also improves the efficiency of the behavior and life of the user P2. Furthermore, compared to conventional technology, this system is more multifunctional and has improved processing speed.

[0117] (Another example of the behavior suggestion system configuration) The terminal device 12 may have the configuration of the management computer 11 shown in Fig. 2. Specifically, the processor 40 may realize the configuration and functions of the processor 17 shown in Fig. 2. Furthermore, the auxiliary storage device 42 may realize the configuration and functions of the auxiliary storage device 19 shown in Fig. 2. With this configuration, the terminal device 12 can execute steps S10 to S21 and step S23 shown in Fig. 3. In other words, the behavior suggestion system 10 can generate optimal behaviors in the terminal device 12 even without the management computer 11. The effect of implementing the behavior management method shown in Fig. 3 in the terminal device 12 is similar to the effect of implementing the behavior management method shown in Fig. 3 in the management computer 11.

[0118] (supplementary explanation) An example of the technical meaning disclosed in this embodiment is as follows. For example, the management computer 11 and the terminal device 12 are each an example of an action suggestion device. The processors 17 and 40 are each an example of a processor. Step S10 shown in FIG. 3 is a first process, a second process, , 4th processing Step S20 is an example of the third process. , 5th processing and This is an example of the seventh process. The flowchart shown in Fig. 3 is an example of a method for suggesting an action.

[0119] The processors 17 and 40 are configured to realize, by running respective programs, a first processing unit that obtains prompts for the user when he or she acts in the environment, a second processing unit that determines the state of the environment in which the user acts, a third processing unit that extracts actions that the user can perform in the environment based on the state of the environment, a fourth processing unit that estimates features including user comfort, efficiency, safety, and economy, assuming that the user will perform the actions extracted in the third processing, and a fifth processing unit that generates optimal actions that the user can perform in the state of the environment based on the prompts obtained in the first processing and the features estimated by the fourth processing unit.

[0120] The fifth processing unit realized by the processors 17 and 40 is configured to generate an optimal action that the user can perform using a learning model. The processors 17 and 40 are configured to respectively realize a sixth processing unit that acquires the action results when the user previously performed the optimal action generated in the past, and a seventh processing unit that trains the learning model by evaluating the action results.

[0121] The seventh processing unit realized by the processors 17 and 40 is configured to include a process of training a learning model based on a future prediction of the state of the environment.

[0122] The computer that constitutes the management computer 11 may be either a single computer or a distributed computer made up of multiple devices. The computer that constitutes the terminal device 12 may be either a single computer or a distributed computer made up of multiple devices.

[0123] The management computer 11 is configured by one or more computers selected from the group consisting of a server, a supercomputer, a mainframe, a server, a workstation, etc. In this embodiment, the program that operates the management computer 11 may be stored in the auxiliary storage device 19. In another example configuration of the action suggestion system, the program may be stored in the auxiliary storage device 42 of the terminal device 12.

[0124] This embodiment discloses the following characteristic configuration: An action suggestion system including a terminal device operated by a user and a management computer communicatively connected to the terminal device via a network, wherein the management computer executes a first process of acquiring, from the terminal device, a prompt for the user to act in an environment, a second process of determining the state of the environment in which the user will act, a third process of extracting an action that the user can perform in the environment based on the state of the environment, a fourth process of estimating feature quantities including comfort, efficiency, safety, and economy of the user, assuming that the user will perform the action extracted in the third process, and a fifth process of generating an optimal action that the user can perform in the state of the environment based on the prompt acquired in the first process and the feature quantities estimated in the fourth process.

[0125] This embodiment discloses a non-transitory storage medium having stored thereon a non-transitory program for causing a computer to execute processes, the program causing the computer to execute the following: a first process for acquiring a prompt for a user to act in an environment; a second process for determining the state of the environment in which the user will act; a third process for extracting actions that the user can perform in the environment based on the state of the environment; a fourth process for estimating features including comfort, efficiency, safety, and economy of the user, assuming that the user will perform the actions extracted in the third process; and a fifth process for generating optimal actions that the user can perform in the state of the environment based on the prompt acquired in the first process and the features estimated in the fourth process. Note that the program can also be understood as a program product. [Industrial Applicability]

[0126] The present embodiment can be used as an action suggestion device, an action suggestion method, and a program. [Explanation of symbols]

[0127] 10...action suggestion system, 11...management computer, 12...terminal device, 17,40...processor

Claims

1. An action suggestion device including a processor for executing a process, The processor: a first process of acquiring a user's current location and destination, a route from the user's current location to the destination, the user's destination, weather information, traffic information, and a prompt for the user to take an optimal action; a second process of determining the user's behavior in the real world from the user's current location, the user's destination, and the user's travel route, and determining the state of the real world in which the user is behaving from weather information and traffic information; a third process for generating, as an optimal action for the user, a travel route and a travel method that can reduce travel time compared to the user's current travel method and travel route and that costs less than a predetermined cost, based on the prompt obtained in the first process and the state of the real world determined in the second process, when the user moves from a current location to a destination; The action suggestion device is configured to execute the above.

2. The action suggestion device according to claim 1, A fourth process of acquiring information about the delivery status of the product by the user; a fifth process of generating a delivery route and a delivery routine that can reduce inventory costs of the product compared to the current situation and shorten delivery time compared to conventional delivery times as an optimal behavior of the user; The action suggestion device is configured to execute the above.

3. The action suggestion device according to claim 1, A sixth process of acquiring information on the user setting up a shelter; a seventh process for generating an optimal action of a shelter installation location, a material procurement method, and a shelter installation method so that the cost of the user installing the shelter is within a predetermined percentage of the cost of installing a conventional shelter; The action suggestion device is configured to execute the above.

4. 1. A computer-implemented method for suggesting an action, comprising: The computer a first process of acquiring a user's current location and destination, a route from the user's current location to the destination, the user's destination, weather information, traffic information, and a prompt for the user to take an optimal action; a second process of determining the user's behavior in the real world from the user's current location, the user's destination, and the user's travel route, and determining the state of the real world in which the user is behaving from weather information and traffic information; a third process for generating, as an optimal action for the user, a travel route and a travel method that can reduce travel time compared to the user's current travel method and travel route and that costs less than a predetermined cost, based on the prompt obtained in the first process and the state of the real world determined in the second process, when the user moves from a current location to a destination; A method of proposing action to carry out the above.

5. A non-transitory program that causes a computer to execute a process, The computer, a first process of acquiring a user's current location and destination, a route from the user's current location to the destination, the user's destination, weather information, traffic information, and a prompt for the user to take an optimal action; a second process of determining the user's behavior in the real world from the user's current location, the user's destination, and the user's travel route, and determining the state of the real world in which the user is behaving from weather information and traffic information; a third process for generating, as an optimal action for the user, a travel route and a travel method that can reduce travel time compared to the user's current travel method and travel route and that costs less than a predetermined cost, based on the prompt obtained in the first process and the state of the real world determined in the second process, when the user moves from a current location to a destination; A program is a configuration that executes a program.

Images