Mental information input system, program, and method

Abstract

[Problem] To enable deciferring of the true feelings and latent consciousness of a support recipient in cognitive behavioral therapy involving self-observation recording based on a behavior or operation by the support recipient, thereby realizing advanced feedback and behavioral therapy that could not be provided by conventional technology. [Solution] This mental information input method which involves use of a computer in performing self-observation recording based on an behavior or operation by a support recipient, the method comprising: an operation acquisition step in which the computer acquires the behavior or operation by the support recipient as an incomplete operation history and acquires input information determined by the behavior or operation by the support recipient as a determined operation history; an observation record accumulation step in which a self-observation record including the incomplete operation history and the determined operation history is accumulated in an observation record accumulation unit; a correlation feature analysis step in which a correlation feature analysis unit of the computer analyzes correlation between behavioral information and cognitive information of the support recipient on the basis of the self-observation record; and a response generation step in which a response generation unit of the computer generates response information to be provided to the support recipient in accordance with the analysis result in the correlation feature analysis unit.

Description

Mental Information Input System, Program, and Method

[0001] The present invention relates to a mental information input system, program, and method for a client to perform self-observation records in psychotherapy or psychological therapy such as cognitive behavioral therapy or behavioral therapy.

[0002] In recent years, self-monitoring has been advocated as a management technique used in psychotherapy or psychological therapy such as cognitive behavioral therapy or behavioral therapy. According to this self-monitoring, effects such as strengthening the target behavior by recording one's own behavior and activities or boosting one's own motivation can be expected.

[0003] Here, cognitive behavioral therapy (CBT: Cognitive Behavioral Therapy) is a scientific treatment method that systematically improves problem behaviors and symptoms using behavioral therapy techniques based on learning theory and a cognitive approach based on cognitive therapy or logic therapy. In the treatment of cognitive behavioral therapy, the goal is for the client to be able to self-care for problems and symptoms. The therapist and the client share the problems and jointly set therapeutic goals to solve the problems.

[0004] By the way, in recent years, services and systems that support the above-mentioned self-monitoring for cognitive behavioral therapy have been developed through communication networks (see Patent Document 1). In this system, behavioral therapy is implemented based on behavioral records such as text input by the client himself / herself.

[0005] Shinichi Suzuki, Hiroaki Kumano, Yujiro Sano, "Cognitive Behavioral Therapy", 1999, http: / / hikumano.umin.ac.jp / cbt_text.html Japanese Patent Laid-Open No. 2021-015466

[0006] However, the system disclosed in Patent Document 1 had a problem in that it was not possible to provide adequate psychotherapy because, for example, information where the client had initially entered an action record but hesitated or canceled the final decision operation such as sending it—so-called "incomplete operations"—was not properly tracked within the system. In detail, these incomplete operations include content that the client hesitated or canceled as a result of deep thought, and this can be important information that reflects the client's true feelings and deep psychology. However, conventional systems do not have the function to track, record, and analyze such incomplete operations, so it was not possible to interpret important psychological signs such as what the client hesitated about or intentionally modified. As a result, it was difficult to provide in-depth behavioral therapy based on the client's true intentions and emotions.

[0007] Therefore, the present invention aims to provide a mental information input system, program, and method that enables the interpretation of a client's true feelings and subconscious mind by recording and tracking data, including information that the client entered but later canceled, and a history of actions in which the client hesitated, in cognitive behavioral therapy that involves recording self-observation records based on the client's actions or actions. This enables advanced feedback and behavioral therapy that could not be provided by conventional technologies.

[0008] To solve the above problems, the present invention provides a system for inputting mental information including self-observation records based on the actions or operations of a person receiving support, comprising: an incomplete operation acquisition unit that acquires the actions or operations of the person receiving support as an incomplete operation history; a decision operation acquisition unit that acquires input information determined by the actions or operations of the person receiving support as a decision operation history; an observation record storage unit that stores self-observation records including the incomplete operation history and the decision operation history; a correlation feature analysis unit that analyzes the correlation between the behavioral information and cognitive information of the person receiving support based on the self-observation records; and a response generation unit that generates response information to be provided to the person receiving support according to the analysis results in the correlation feature analysis unit.

[0009] Furthermore, the present invention relates to a method for inputting mental information, including a self-observation record based on the actions or operations of a person receiving support, using a computer, comprising: an operation acquisition step in which the computer acquires the actions or operations of the person receiving support as an incomplete operation history and input information determined by the actions or operations of the person receiving support as a determined operation history; an observation record storage step in which the computer stores the self-observation record, including the incomplete operation history and the determined operation history, in an observation record storage unit; a correlation feature analysis step in which the correlation feature analysis unit of the computer analyzes the correlation between the behavioral information and cognitive information of the person receiving support based on the self-observation record; and a response generation step in which the response generation unit of the computer generates response information to be provided to the person receiving support according to the analysis results in the correlation feature analysis unit.

[0010] In the above invention, it is preferable that the pattern calculation unit analyzes the correlation between the behavioral information and cognitive information of the person being supported in accordance with the comparison result of the incomplete operation history and the decision operation history, calculates a feature pattern that characterizes the analyzed correlation, and has a natural language model trained to calculate response information corresponding to the comparison result of the incomplete operation history and the decision operation history and the feature pattern receive the comparison result of the incomplete operation history and the decision operation history and the feature pattern, and the computer's response request unit generates and executes a prompt for the natural language model to generate response information.

[0011] In the above invention, it is preferable that the system further includes an operation pattern analysis unit that analyzes operation patterns including at least one of the following in the input operations of the person being supported: time delay, frequency of re-input, occurrence of deletion operations, and time period of occurrence. The response generation unit estimates the potential psychological state of the person being supported based on the analysis results by the operation pattern analysis unit and determines the content and timing of the response.

[0012] In the above invention, it is preferable that, in both the incomplete operation history and the decided operation history, the computer's operation identification unit uses a machine learning model to estimate the level of awareness of an operation in order to distinguish between the conscious and unconscious operations of the person being supported, and evaluates the cognitive and emotional tendencies of the person being supported based on the identified level of awareness.

[0013] Furthermore, the systems and methods according to the present invention described above can be realized by executing a program of the present invention written in a predetermined language on a computer. That is, by installing the program of the present invention on an IC chip or memory device of a mobile terminal device, smartphone, wearable device, mobile PC or other information processing terminal, or a general-purpose computer such as a personal computer or server computer, and executing it on the CPU, it is possible to construct a system having the above-described functions or to implement a method according to the present invention.

[0014] Furthermore, the program of the present invention can be distributed, for example, via a communication line, and can also be transferred as a package application that runs on a standalone computer by recording it on a computer-readable recording medium. Specifically, this recording medium can be a variety of recording media, such as magnetic recording media like flexible disks and cassette tapes, optical discs like CD-ROMs and DVD-ROMs, and RAM cards. With a computer-readable recording medium on which this program is recorded, the above-described system and method can be easily implemented using a general-purpose computer or a dedicated computer, and the program can be easily stored, transported, and installed.

[0015] As described above, according to the present invention, by acquiring and accumulating a history of incomplete operations and a history of decided operations based on the actions and operations of the person being supported, and by analyzing the correlation between behavioral information and cognitive information, it is possible to grasp the client's behavioral and cognitive tendencies in detail. This makes it possible to generate appropriate responses, provide more individualized support, and reduce the psychological burden on the client.

[0016] Furthermore, the above invention compares the history of incomplete operations with the history of decided operations, calculates feature patterns based on the comparison results, and utilizes a natural language model to generate responses through machine learning, thereby providing appropriate response information tailored to the client's state. This enables rapid response to changes in the client's behavior and cognition, and by using a machine learning model, it is possible to automatically generate responses that are optimal for individual needs, resulting in more accurate support.

[0017] In the above invention, by analyzing the client's operation patterns (such as time delays, frequency of re-entry, and deletion operations) and estimating the client's psychological state based on these patterns, a more effective response can be provided by adjusting the content and timing of the response based on the results of this analysis.

[0018] Furthermore, the above invention identifies conscious and unconscious operations in the history of incomplete operations and the history of decided operations, and evaluates the client's cognitive and emotional tendencies based on this. This enables appropriate evaluation based on the level of awareness regarding operations, thereby improving the accuracy of treatment.

[0019] This is a conceptual diagram showing the overall configuration of the cognitive behavioral therapy feedback system S according to the embodiment. This is a block diagram showing the internal configuration of the management server 3 according to the embodiment. This is a block diagram showing the configuration of the analysis unit within the management server 3 according to the embodiment. This is a block diagram showing the configuration of the feedback processing unit within the management server 3 according to the embodiment. This is a block diagram showing the internal configuration of the client's smartphone 1 according to the embodiment. This is a flowchart showing the overall operation of the cognitive behavioral therapy feedback system according to the embodiment. This is a flowchart showing the operation of the cognitive behavioral therapy feedback system according to the embodiment when feedback is being executed (self-counseling mode). This is an explanatory diagram showing the situation of monitoring user operations according to the embodiment. This is an explanatory diagram showing the analysis situation by monitoring incomplete and decided user operations according to the embodiment. This is an explanatory diagram showing the response selection patterns according to the embodiment.

[0020] The following describes in detail embodiments of the mental information input system, method, and program of the present invention when applied to a cognitive behavioral therapy feedback system S, with reference to the attached drawings. It should be noted that the present invention is not limited to the embodiments described above, and the components can be modified and implemented in practice without departing from the gist of the invention. Furthermore, various inventions can be formed by appropriate combinations of the multiple components disclosed in the above embodiments. For example, some components may be deleted from all the components shown in the embodiments.

[0021] In this embodiment, the mental information input system, method, and program of the present invention are described as being applied to cognitive behavioral therapy, but the present invention is not limited thereto. The present invention can be applied not only to psychotherapies such as cognitive behavioral therapy, but also to fields such as mental care and learning support in companies, educational institutions, and other non-medical environments. For example, it is useful in general mental care, learning support, mental care for students in educational institutions, psychological support for employees in companies, and improving mental health in the workplace environment.

[0022] (Overview of Cognitive Behavioral Therapy) In the cognitive behavioral therapy according to this embodiment, the goal is for the client, who is the person being supported, to be able to self-care for their problems and symptoms. The therapist and the client share the problems, and together they set therapeutic goals and work to solve the problems.

[0023] In the cognitive behavioral therapy according to this embodiment, self-monitoring is performed on the client's smartphone through a dedicated application. Through this self-monitoring, the client records their own actions and events, and also records their thoughts, feelings, and insights regarding those actions and events as mental information, including "automatic thoughts." This employs cognitive behavioral therapy and cognitive restructuring techniques.

[0024] This method of recording mental information such as automatic thoughts provides feedback that encourages self-awareness in the client by showing the causal relationship between automatic thoughts and emotion estimation, such as which words in the client's own "automatic thoughts" (i.e., the client's own "thoughts") were focused on to estimate emotions. This gives the client an opportunity to think for themselves, and a significant psychological effect can be expected.

[0025] Furthermore, in this embodiment, self-monitoring through the application allows for an assessment of cognitive aspects, such as observing what kinds of thoughts frequently arise in what situations, how confident the client is in the actions required in that situation, how they perceive the results of their actions, and how they think about future prospects. In this embodiment, self-observation records are assigned as homework, instructing clients to record what kinds of thoughts arise in actual situations. Based on these records, analysis is performed to identify common thoughts observed in many situations and what kinds of cognitions strongly influence behavior and mood.

[0026] Based on the analysis results of the self-monitoring, parents, doctors, educators, therapists, etc., send messages such as advice and counseling. A dedicated application with a counseling interface is installed on the psychologist's (Dr1) smartphone (1d), allowing for viewing necessary information, entering documents, and sending messages (such as feedback to the client). Furthermore, in sending counseling, advice, and progress reports as feedback to the client via message, this system accumulates and analyzes the client's self-observation records and flexibly selects a response from the generative AIGA1 or the psychologist (Dr1) according to the client's situation, thereby reducing the human burden on the psychologist (Dr1) while enabling appropriate support tailored to the psychological state of the client (CLa-c).

[0027] (Overall Configuration of the Cognitive Behavioral Therapy Feedback System) Figure 1 shows the overall configuration of the cognitive behavioral therapy feedback system S according to this embodiment. The system according to this embodiment is a system for providing cognitive behavioral therapy to multiple clients CL (CLa to c), and as shown in Figure 1, it is generally composed of smartphones 1 (1a to c) used by each client CL and a management server 3 installed on the internet 2. In this embodiment, the smartphone 1 is described as an example of an information processing terminal device, but any information processing terminal device such as a personal computer or tablet PC can be used.

[0028] Furthermore, the mental information input system and method according to the present invention can be realized by executing the program of the present invention, which is written in a predetermined language, on a computer. In this embodiment, the mental information input system of the present invention can be constructed by executing the mental information input program of the present invention on the CPUs of the smartphone 1 (1a to c) and the management server 3.

[0029] In this embodiment, the management server 3 is a server that processes the progress of cognitive behavioral therapy. It can be implemented as a single server device or a group of multiple server devices. By executing software, it virtually constructs multiple functional modules on the CPU, and each functional module works together to perform processing. Furthermore, the management server 3 can send and receive data via the Internet 2 using its communication function, and can also display web pages via browser software using its web server function.

[0030] Smartphone 1 (1a, 1b) is a portable information processing terminal device that utilizes wireless communication. It communicates wirelessly with relay points such as wireless base stations 22, allowing users to receive communication services such as voice calls and data communication while on the move. Examples of communication methods for this mobile phone include 3G (3rd Generation), LTE (Long Term Evolution), and 5G. In addition to communication functions, Smartphone 1 is equipped with various other functions, including a digital camera, application software execution, and location information acquisition functions such as GPS (Global Positioning System). Such information processing terminal devices include mobile computers such as tablet PCs, as well as IoT devices such as smartwatches and smart speakers.

[0031] Furthermore, this smartphone 1 is equipped with a liquid crystal display as a display unit for displaying information, and is equipped with operation devices such as operation buttons for the user to perform input operations. These operation devices include a touch panel, which is superimposed on the liquid crystal display and is an input unit that acquires operation signals such as touch operations that specify the coordinate position on the liquid crystal display. Specifically, this touch panel is an input device that receives operation signals by pressure, electrostatic detection, etc., from touch operations using the user's fingertip or a pen, and is configured by superimposing a liquid crystal display that displays graphics and a touch sensor that receives operation signals corresponding to the coordinate position of the graphics displayed on the liquid crystal display.

[0032] In this embodiment, the client CL (CLa to c) uses a smartphone 1 (1a to c). The client's smartphone 1 has a client application installed and functions as a dedicated client terminal. When a guardian or other person monitors the client's mental health treatment, the guardian runs a guardian application on their smartphone. The guardian application includes functions such as receiving and viewing alerts from the client application. For example, in addition to a dedicated application for receiving and viewing alerts, it also includes communication applications such as an email application, a browser application, and a messaging application.

[0033] (Internal Structure of Each Device) Next, the internal structure of each device constituting the cognitive behavioral therapy feedback system S described above will be explained. Figures 2 to 4 show the internal configuration of the management server 3 according to this embodiment, and Figure 5 shows the internal configuration of the smartphone 1 according to this embodiment. In this explanation, the term "module" refers to a functional unit composed of hardware such as devices or equipment, software that has the same function, or a combination thereof, for achieving a predetermined operation.

[0034] (1) Management Server First, the internal configuration of the management server 3 will be explained. The management server 3 is a server device located on the Internet 2, and is capable of sending and receiving data with each smartphone 1 via the Internet 2. Specifically, the management server 3 includes a communication interface 31 for data communication via the Internet 2, an authentication unit 32 for authenticating the authority of users and user terminals, a cognitive behavioral therapy execution unit 33 and a response generation unit 34 for executing the overall progress processing of cognitive behavioral therapy, an information distribution unit 36 ​​for distributing various content data to each client, a billing processing unit 37, and various database groups 35 (35a to 35d).

[0035] The database group 35 includes an AI learning database 35a for storing the learning results of artificial intelligence, a user database 35b for storing information about users such as clients and their guardians, a cognitive behavioral therapy database 35c for storing worksheets executed during work mode related to cognitive behavioral therapy, serial content delivered in lesson mode, scenario data related to lesson progress processing, etc., and an observation record database 35d as an observation record storage unit for storing data related to monitoring for each client. Each of these databases may be a single database, or it may be divided into multiple databases and relationships may be set up to link the data together, creating a relational database.

[0036] The Cognitive Behavioral Therapy Database 35c is a storage device that stores a series of interconnected content, such as teaching materials and videos related to psychotherapy or psychological therapy, used in lesson mode. This Cognitive Behavioral Therapy Database 35c also stores data such as scenario data and worksheets.

[0037] The information stored in the user database 35b includes identification information that identifies clients and guardians, or identifiers (user ID, device ID) that identify the mobile device used by the client, and authentication information linked to passwords, etc. It also includes personal information of the user linked to the user ID, as well as the model of the device. In addition, the user database 35b stores authentication history (access history) for each user or user device, observation results regarding the progress of lessons and work for each client, status, points, usage history, etc., as well as payment information, in relation to the cognitive behavioral therapy database 35c.

[0038] The observation record database 35d is primarily a database that stores observation data by associating occurrence record information and automated thought information for each client. The information stored in this observation record database 35d consists of self-observation records collected from a large number of clients through worksheets. In this embodiment, each client's actions or operations are included in the self-observation record as an incomplete operation history, and the input information determined by each client's actions or operations is included as a decision operation history. In addition, occurrence record information, behavioral information, cognitive information, and characteristic keywords are also stored in the self-observation record for each client, all of which are associated with each other.

[0039] The authentication unit 32 is a module that establishes a communication session with each smartphone 1 via the communication interface 31 and performs authentication processing for each established communication session. This authentication process involves obtaining authentication information from the smartphones 1 (1a to c) of the client or their guardian who is accessing the system, referring to the user database 35b to identify the accessor, and authenticating their authority. The authentication results from the authentication unit 32 (user ID, authentication time, session ID, etc.) are sent to the cognitive behavioral therapy execution unit 33 and are also stored as authentication history in the user database 35b via the authentication unit 32. In addition, the authentication unit 32 also performs authentication for access from supporters participating in the operation of cognitive behavioral therapy, such as staff and SEs on the operating side, as well as smartphones 1 and counselors, and grants the access rights assigned to each accessor.

[0040] Furthermore, the authentication unit 32 is equipped with a location information management unit 32a, which acquires the current time and current location of the authenticated client from the smartphone 1 and records it in the user database 35b, as well as recording it in the observation record database 35d via the observation record unit 331 as an observation record of the client.

[0041] The cognitive behavioral therapy execution unit 33 is a module that performs the necessary processing for cognitive behavioral therapy and conducts treatment. It collects and analyzes various observation records through an application installed on the smartphone used by each client, and controls the generation system AIGa1 to provide information necessary for treatment. The cognitive behavioral therapy execution unit 33 also has the function of conducting cognitive behavioral therapy by generating various event processing to develop work mode and lesson mode according to cognitive behavioral therapy scenario data. It executes a script that includes certain rules, logic, and algorithms, and generates event processing such as lessons, work, mini-games, and movie playback.

[0042] The cognitive behavioral therapy scenarios implemented in this cognitive behavioral therapy implementation unit 33 include: self-monitoring to enable objective self-viewing; cognitive restructuring to correct cognitive distortions; behavioral activation as an approach to boost mood by activating behavior to recover from the decline in activity levels that occurs when depressed or feeling down; sleep care as care for insomnia that often occurs when depressed or feeling down (specifically, avoiding naps to increase sleep pressure and maintaining a regular sleep schedule, and waking up at a set time in the morning even if you can't sleep); assertion (communication skills) learning, which involves learning assertion skills to effectively communicate opinions to others, as the majority of depression and stress stem from interpersonal relationships; and problem-solving skills, which involves learning how to deal with troubles and worries that are sources of stress.

[0043] In this embodiment, the cognitive behavioral therapy execution unit 33 on the management server 3 cooperates with the cognitive behavioral therapy execution unit 141 on the smartphone 1 through the synchronization processing unit 361 and the synchronization processing unit 145 on the smartphone 1. The management server 3 performs some of the cognitive behavioral therapy progress processing and work analysis processing, while the cognitive behavioral therapy execution unit 141 on the smartphone 1 performs some of the graphic processing and event processing.

[0044] For example, on the management server 3 side, based on the progress of the scenario, the lesson progress level of the client, as well as the classification, current location, current time, etc. of the user, it predicts the event processing that may occur, causes the occurrence conditions to occur on the management server 3 side, sends the conditions to the smartphone 1 side, and based on the occurrence conditions received from the management server 3, the actual occurrence of event processing and the graphic processing therefor can be executed on the smartphone 1 side.

[0045] In addition, the cognitive behavioral therapy execution unit 33 according to the present embodiment has an alert monitoring function, constantly monitors the alert signals transmitted from the smartphone 1 side, and when an alert signal is detected, it can send an alert message to the registered contact (phone number, email, etc.). In particular, the cognitive behavioral therapy execution unit 33 includes an observation recording unit 331, an analysis unit 332, and an AI control unit 333 as modules related to cognitive behavioral therapy.

[0046] (Observation recording unit 331) The observation recording unit 331 is a module that collects self-observation records based on the actions or operations of each client through the smartphones 1a to 1c used by each client. The observation records recorded by this observation recording unit 331 include, in addition to the text data input by the client, the video of the client's expression, or the blood flow, voice, location information history, etc. of the client, as well as the occurrence record information of external factors and the client's actions and cognitions regarding the external factors, which are also included in the automatic thinking information.

[0047] In the present embodiment, the observation recording unit 331 has a function of collecting self-observation records based on the actions or operations of each client through the smartphones 1a to 1c used by each client. In the present embodiment, as a module related to the collection of this self-observation record, it has an input reception unit 331a, modules 331b to 331d that execute acquisition processing of various information, and an information control unit 331e.

[0048] The input reception unit 331a is a module that receives inputs from clients and is a module that selects appropriate processing according to the type of the input. The inputs received by this input reception unit 331a are classified and then stored as a history. As a result, the reaction classification unit classifies the reaction patterns of the clients and further classifies the attributes of the clients according to the combination of reaction patterns for each external factor. According to the classification by this input reception unit 331a, the reaction classification unit 332b executes appropriate analysis processing.

[0049] And this input reception unit 331a includes an unfinished operation acquisition unit 331b, a determination operation acquisition unit 331c, and a sensor information acquisition unit 331d as modules that execute each acquisition process.

[0050] The unfinished operation acquisition unit 331b is a module that acquires actions or operations by each client as an unfinished operation history, and the determination operation acquisition unit 331c is a module that acquires input information determined by actions or operations by each client as a determination operation history.

[0051] When the user performs an operation such as text input on the client's smartphone, the unfinished operation acquisition unit 331b has a mechanism to trace unfinished operations, which are operation actions such as input errors, rewriting, and deletion, as shown in FIGS. 8 and 9, before pressing the send or determination button. Examples of the functions and processes for acquiring this unfinished operation are as follows.

[0052] - Real-time event listening: Touch operation events (text input, gesture operations, deletion, focus movement, etc.) on the client's app are monitored in real time, and this information is sent to the incomplete operation acquisition unit 331b as incomplete operation history at regular intervals or after specific actions (such as stopping after input) and acquired. - Asynchronous communication: To synchronize input operations from the client with the server in real time, each input is sent to the incomplete operation acquisition unit 331b via asynchronous communication, and the server traces sequentially what the user is inputting. - Input log retention and analysis: The incomplete operation acquisition unit 331b records the content entered by the client, deleted characters, and rewritten parts as input logs in the incomplete operation history, records each operation with timestamped data, and tracks the changes from the content initially entered by the user in chronological order.

[0053] Furthermore, the decision operation acquisition unit 331c acquires and records the data included in the aforementioned incomplete operation history as a decision operation history when the client finally performs a completion operation such as clicking the send button, making a decision operation, or saving operation, by adding a completion flag to the data and distinguishing it from the incomplete operation history.

[0054] The sensor information acquisition unit 331d is a module that traces operations other than text input, and acquires sensor signals detected on the client terminal as one of the incomplete operations. For example, operations such as attempting to select an option from a pull-down menu or checkbox on the client terminal but ultimately interrupting the selection without completing it, operations such as starting to drag a file or object but stopping the operation before dropping it, operations such as starting to scroll a page or content but stopping the operation before reaching a specific area, and operations such as attempting to click a link or button but not actually completing the transition, are recorded in the incomplete operation history by the sensor information acquisition unit 331d as behavioral data when the user hesitated or stopped their actions midway.

[0055] The information control unit 331e is a module that acquires data and signals as information, calculates characteristic patterns, and records them in various databases according to the calculated patterns. It integrally acquires data and signals received from multiple sensors and input devices as information, and uses algorithms to analyze and extract fluctuations, trends, or other characteristic elements of the signals and data. The resulting patterns are recorded in various predefined databases, forming a foundation for utilizing the data for purposes such as feedback control or report generation.

[0056] (Analysis Unit 332) The analysis unit 332 is a module that analyzes the client's behavior based on self-observation records acquired and recorded in accordance with the progress of cognitive behavioral therapy by the cognitive behavioral therapy execution unit 33. For example, it functions as a correlation feature analysis unit that analyzes the correlation between each client's behavioral information and cognitive information. The analysis unit 332 also has the function of collecting work entered by the client in accordance with the progress of the cognitive behavioral therapy lessons and work and performing an evaluation of the user according to the cognitive behavioral therapy written in that work.

[0057] The analysis of this client can be performed on the management server 3, on the smartphone 1, or jointly by both the management server 3 and the smartphone 1. The analysis unit 332 manages the analysis results performed or stored on the management server 3 and the analysis results performed and stored on the smartphone 1, compares the results from both, and, if necessary, delivers all or part of the analysis results to the smartphone 1 to synchronize them. To elaborate on the analysis unit 332, as shown in Figure 4, the analysis unit 332 is equipped with an automatic thought acquisition unit 332a, an emotion generation unit 332c, a cognitive distortion evaluation unit 332f, and an evaluation execution unit 332g.

[0058] The automatic thought acquisition unit 332a is a module that acquires various types of information input by the client via the smartphone 1. The automatic thoughts acquired here include not only behavioral records entered by the client themselves, but also biological information such as blood flow, body temperature, and heart rate input from the biological information analysis unit 332l. The means of acquiring information in the automatic thought acquisition unit 332a include reading data recorded by the observation recording unit 331 and receiving data directly from the smartphone 1 via the communication interface 31.

[0059] The automatic thought acquisition unit 332a is equipped with a response classification unit 332b, which is a module that classifies the data acquired by the automatic thought acquisition unit 332a and sends it to the emotion generation unit 332c and the cognitive distortion evaluation unit 332f.

[0060] The response classification unit 332b also has the function of classifying the client's response patterns based on occurrence record information, behavioral information, cognitive information, and characteristic keywords stored in the observation record database 35d. In particular, in this embodiment, when the client inputs self-monitoring text, the response classification unit 332b acquires actions or operations by the client that have not yet resulted in a return, transmission, or decision operation as "incomplete operations," and acquires completed operations and their input information as "decision operations."

[0061] In this embodiment, the response classification unit 332b comprises a comparison unit 334a, a pattern calculation unit 334b, and an operation pattern analysis unit 334c. The comparison unit 334a is a module that compares the incomplete operation history with the decision operation history, and the pattern calculation unit 334b ​​is a module that analyzes the correlation between the client's behavioral information and cognitive information according to the comparison result by the comparison unit 334a, and calculates a characteristic pattern that characterizes the analyzed correlation. The operation pattern analysis unit 334c is a module that analyzes operation patterns that include at least one of the following in the client's input operations: time delay, frequency of re-input, occurrence status of deletion operations, or time period of occurrence.

[0062] The emotion generation unit 332c is a module that generates emotion data by classifying observation data stored in the statistical data storage unit according to a pre-learned model which is a correlation pattern between the occurrence record information and the automatic thought information. The emotion generation unit 332c according to this embodiment includes a pre-learned model management unit 332d. The pre-learned model management unit 332d is a module that generates and manages a learning model in advance by the generation system AIGa1. The emotion data generated by this emotion generation unit 332c is input to the judgment unit 332e.

[0063] The cognitive distortion evaluation unit 332f is a module that compares the automated thought information of a specific client with the emotional data generated by the generation system AIGa1 and evaluates the client's cognitive distortion based on the degree of agreement between them. In this embodiment, it includes a determination unit 332e. The determination unit 332e is a module that compares the automated thought information of a specific client with the emotional data, classifies the client's automated thought information based on the degree of agreement between them, and calculates the deviation rate from the emotional data related to that classification. The deviation rate calculated by the cognitive distortion evaluation unit 332f is input to the evaluation execution unit 332g.

[0064] The evaluation execution unit 332g is a module that performs comprehensive evaluation processing based on the results of evaluation and analysis of acquired automatic thoughts, emotions, and cognitive distortions. Specifically, it comprises a causal relationship analysis unit 332h, a change extraction unit 332i, a qualitative evaluation calculation unit 332j, and a self-counseling management unit 332k.

[0065] The causal relationship analysis unit 332h is a module that extracts keywords included in the occurrence record information and emotion data used to calculate the deviation rate as feature keywords, weights the extracted feature keywords according to the deviation rate calculated by the comparison judgment unit, and records the feature keywords as causal relationship data based on the frequency of that weighting.

[0066] Furthermore, the causal relationship analysis unit 332h has the function of analyzing the correlation between external factors and the client's behavior and cognition regarding those external factors, and extracting characteristic keywords that characterize the analyzed correlation. The characteristic keywords extracted here are associated with occurrence record information, behavioral information, and cognitive information and stored in the observation data storage units, databases 35a to d.

[0067] The change extraction unit 332i is a module that extracts statistical numerical changes in a time series for predetermined items from the information stored in the observation record database 35d. When selecting a response, it automatically generates these statistical numerical changes as response information according to the client's psychological state. Examples of numerical changes extracted by this change extraction unit 332i include: - Changes in emotional intensity: The emotional intensity recorded by the client through self-monitoring (for example, evaluating emotions such as anxiety and anger on a scale from 0 to 10) is analyzed in a time series, and changes in the numerical values ​​indicating how the intensity has decreased or increased are obtained. - Frequency of negative automatic thoughts: The frequency of negative automatic thoughts is recorded, and it is evaluated whether the frequency has decreased with the progress of treatment. For example, if the number of negative thoughts per week has decreased, the numerical value evaluated as a treatment outcome changes. - Frequency of problematic behaviors: The frequency of maladaptive behaviors (for example, overeating and avoidance behaviors) is tracked, and the decrease is confirmed numerically. The number of times a specific behavior occurs in a day is recorded, and the decrease in that number over time is analyzed to obtain numerical changes and behavioral activation indicators. The frequency of positive behaviors (e.g., socializing with friends or relaxation activities) by the client is recorded, and the increase or decrease in these behaviors is quantified. These numerical changes function as important indicators for objectively evaluating the effectiveness of treatment, and by utilizing the generative system AIGA1 for more efficient analysis, they become basic data for providing feedback to the client and determining the next treatment step.

[0068] The qualitative evaluation calculation unit 332j is a module that objectively evaluates the degree to which a client's cognitive distortions have improved. Specifically, it measures the improvement in the client's cognitive distortions based on the change in the deviation rate calculated by the judgment unit 332e. This change in the deviation rate is considered as "improvement" if the deviation rate has decreased, and the degree of change is input as qualitative evaluation data to the user database 35b and the billing processing unit 37. The billing processing unit 37 calculates the degree of improvement for each client based on the qualitative evaluation data and performs billing processing according to that degree.

[0069] The self-counseling management unit 332k is a module that classifies the client's response patterns based on the occurrence record information, automated thought information, and deviation rate stored in the observation record database 35d. The feedback processing unit has the function of selecting and presenting comments according to the response patterns classified by the self-counseling management unit 332k.

[0070] (AI Control Unit 333) The AI ​​Control Unit 333 is a module that analyzes input from the client and generates an appropriate response by performing natural language processing (NLP) in cooperation with an external AI service. The AI ​​Control Unit 333 works in cooperation with an external NLP service to establish a dialogue with the user and generates a response that corresponds to the characteristics (age, gender, personality, experience, knowledge, etc.) of AI characters such as doctors, counselors, and teachers realized by NLP. To achieve this, it refers to the AI ​​learning database 35a and the cognitive behavioral therapy database 35c, and performs processes such as generating information requests (prompts, etc.), verifying responses, outputting information, and reflecting the dialogue results in the database. In particular, the AI ​​Control Unit 333 receives the comparison results and feature patterns between the incomplete operation history and the decided operation history, and functions as a response request unit that generates and executes prompts to the natural language model to generate response information.

[0071] Furthermore, the AI ​​control unit 333 also has a function to predict external factors that have occurred in accordance with information on the current time and current location, in cooperation with an external AI service (in this case, the generation system AIGa1) during the response generation process in the response generation unit 34, and the analysis unit 332 has a function to generate feedback information necessary for processing in the feedback processing unit 344, using behavioral information, cognitive information, and related feature keywords associated with the external factors predicted by the AI ​​control unit 333 as intermediate seeds.

[0072] The AI ​​control unit 333 makes predictions by referring to table data that lists the expected timing, time, location, speed of movement, and acceleration for each external factor. Based on the current time, current location, speed of movement, acceleration, and direction, it predicts external factors that are likely to occur at that time, given the client's current location and situation. For example, if the client is moving at a certain speed or higher during commuting hours, there is a high probability that the client is on a train. The AI ​​control unit 333 then refers to table data regarding potential troubles, changes in "thoughts," "emotions," and "physical condition" that may occur on trains during commuting hours, and extracts candidate external factors that may occur.

[0073] (Response Generation Unit 34) The response generation unit 34 is a module that generates response information to be provided to the client in accordance with the analysis results in the cognitive behavioral therapy execution unit 33, including correlation feature analysis of the analysis unit 332. For example, it executes a process to sequentially provide a series of content according to a scenario that describes event processing to advance a series of consecutive content. In this embodiment, the scenario is advanced based on each client's lesson progress and the number of times worksheets have been created. This scenario data is script data that describes the development of lesson mode and work mode.

[0074] Furthermore, the response generation unit 34 has the function of synchronizing the cognitive behavioral therapy progress processing on the smartphone 1 and the cognitive behavioral therapy progress processing on the management server 3 according to this scenario data. Specifically, the management server 3 predicts event processing that may occur based on the progress of the lesson in the scenario data, generates the conditions for its occurrence on the management server 3, and transmits those conditions to the smartphones 1a to 1c.

[0075] Furthermore, the response generation unit 34 has functions such as character control, chat execution, and worksheet generation to generate responses for interacting with the client in order to execute a cognitive behavioral therapy scenario.

[0076] The character control function generates feedback based on behavioral information, cognitive information, and characteristic keywords related to external factors predicted by the generation system AIGa1. The feedback generated here, along with other advice, is sent to the output control unit 147 on the smartphone 1 and displayed as dialogue for the psychologist character, etc. This character control function generates display data for displaying animations and CG of people, creatures, etc., on the display 13a on the smartphone 1. The display data is generated by combining graphic data, image data, text data, video data, audio, and other data.

[0077] The chat execution function generates display data for displaying the conversational exchange on the display 13a of the smartphone 1. This chat execution function sends feedback from the generation system AIGa 1 and human intervention to the client as messages from the conversation partner, and also receives messages from the client as automated thoughts. Messages received by this chat execution function will continue the conversation according to the prepared scenario if they fall within that scenario. If an irregular response that deviates from the scenario is received, the re-evaluation execution function of the observation recording unit 331 can request re-evaluation and re-analysis to modify or replace the scenario, allowing the conversation to continue.

[0078] The re-evaluation execution function of the observation record unit 331 monitors the interactions in chats and worksheets entered by the client through each module within the observation record unit 331. In this monitoring, if keywords or indirect behaviors listed in the monitoring list stored in the cognitive behavioral therapy database 35c are detected in the interactions, a request for re-evaluation is sent to the analysis unit 332. The analysis unit 332 then analyzes the frequency of occurrence of the relevant keywords and indirect behaviors according to the client's response type, determines whether an alert is necessary, the danger level, and the urgency, and if the danger level exceeds a certain level, it has the function to send an alert to psychologists, advisors, etc., via the human operation unit 343. The alert from this re-evaluation execution function may also be sent directly to the smartphones of the client's guardians, clinical psychologists, assigned advisors, etc.

[0079] The worksheet generation function of the observation recording unit 331 records self-observation records (cognitive behavioral therapy such as behavioral records and cognitive records) entered based on the client's operations, and executes a worksheet as a work mode for analyzing the recorded self-observation records. In this worksheet, information on the occurrence of external factors, behavioral information regarding the client's actions in response to the external factors that occurred, and cognitive information regarding the client's perception of the external factors are recorded based on the client's operations.

[0080] Specifically, the response generation unit 34 includes an automatic response execution unit 341, a response selection unit 342, a human operation unit 343, and a feedback processing unit 344. The automatic response execution unit 341 is a module that automatically executes a response based on the analysis results by the analysis unit 332 using the generation system AIGa1. In this embodiment, the automatic response execution unit 341 also has the function of generating numerical changes in a time series for a predetermined item from the information stored in the observation record database 35d, which has been extracted by the change extraction unit 332i of the analysis unit 332, as response information (feedback) for the automatic response, according to the psychological state of the client.

[0081] On the other hand, the human operation unit 343 is a module that accepts predetermined human operations when a human, such as a psychologist, doctor, or counselor, operates it to generate feedback. It generates feedback that includes content such as text and audio entered by the responder through an information terminal (PC or smartphone) used by the responder, such as a psychologist. In particular, the human operation unit 343 is a module for human responders to provide feedback when it is difficult for the generation system AIGa1 to respond. Through this human operation unit 343, it connects to an information terminal used by the responder, such as a psychologist, and enables the psychologist to send advice and other information to the client.

[0082] In this embodiment, the system also includes a function to send data necessary for intervention from templates and databases to the PC used by the psychologist, for the purpose of standardizing and applying feedback patterns performed by human psychologists. Furthermore, the content of the automated response generated by the generation system AIGa1 when the automated response execution unit 341 responds can be referenced as data necessary for human intervention, such as templates.

[0083] Furthermore, even when the response selection unit 342 selects an automatic response by the automatic response execution unit 341, the human operation unit 343 also functions as an approval execution unit, sending a request for human approval regarding the content of the automatic response according to the client's psychological state. The response generation unit 34 generates response information (feedback) reflecting the approval result from this approval execution unit function and notifies the client.

[0084] The response selection unit 342 is a module that selects an automatic response by the generation system AIGa1 in the automatic response execution unit 341 or a human response in the human operation unit 343 based on the analysis results by the analysis unit 332. The response generation unit 34 generates response information to be provided to the client based on the correlation characteristics analyzed by the analysis unit 332 and the selection result by the response selection unit 342. The response selection unit 342 determines the optimal response method according to the client's psychological state, performs an automatic response by the automatic response execution unit 341 if the generation system AIGa1 is able to respond, and sends a request for a human response through the human operation unit 343 if a human response is necessary under specific circumstances. The response generation unit 34 generates response information based on the human response operated in response to this request.

[0085] The feedback processing unit 344 is a module that generates comments in the form of questions for the user, based on weighted characteristic keywords and deviation rates included in the causal relationship data, and presents them to the client. The feedback processing unit 344 generates feedback using keywords related to the external factors predicted by the AI ​​control unit 333, and prompts the user to input information about the predicted external factors in the form of questions from the character.

[0086] (Information distribution unit 36) The information distribution unit 36 ​​is a module that, based on the client's lesson progress and under the control of the cognitive behavioral therapy execution unit 33, distributes the analysis results performed by the analysis unit 332 to each smartphone 1 via the communication interface 31 in order to synchronize them.

[0087] The synchronization processing unit 361 of the information distribution unit 36 ​​is equipped with a trigger generation function that instructs the timing for presenting feedback to the client. This trigger generation function is a module that instructs the timing of event processing, such as automatically launching the application on the smartphone 1, outputting a notification from the application, or having a character automatically start speaking within the application, as a way of presenting feedback. For example, it outputs feedback prompting the client to take the next lesson after a certain amount of time has passed since the end of the previous lesson.

[0088] (2) Smartphone 1 (1a-c) Next, the internal configuration of the client's smartphone 1 will be described. In this embodiment, clients CLa-c each use smartphone 1a-c. The hardware configuration of these smartphones 1a-c is basically the same, and the client's smartphone 1 functions as a dedicated client terminal by having a client application installed. The parent's smartphone only needs to have the function to receive alerts from the client application and view their contents. For example, in addition to a dedicated application for receiving and viewing alerts, it may have communication applications such as an email application, a browser application, or a messaging application.

[0089] This section describes the client's smartphone 1. As shown in Figure 5, the client's smartphone 1 is equipped with a communication interface 11, an input interface 12, an output interface 13, an application execution unit 14, and memory 15. The hardware configuration is the same for the psychologist's and guardian's smartphones.

[0090] The communication interface 11 is a communication interface for data communication and has the function of contactless communication via wireless or other means, as well as contact (wired) communication via cables, adapters, etc. The input interface 12 is a device for inputting user operations, such as a mouse, keyboard, operation buttons, or touch panel 12a. The output interface 13 is a device for outputting video and sound, such as a display or speaker. In particular, this output interface 13 includes a display 13a such as an LCD display, and this display unit is superimposed on the touch panel 12a, which is the input interface.

[0091] Memory 15 is a storage device that stores the OS (Operating System), firmware, programs for various applications, and other data. This memory 15 stores user IDs or terminal IDs that identify users or terminals, as well as lesson data and worksheets downloaded from the management server 3, and input data processed by the application execution unit 14. In particular, in this embodiment, memory 15 stores scenario data obtained from the management server 3, movie data used in lesson mode, and monitoring lists referenced when alerts are issued.

[0092] The application execution unit 14 is a module that executes applications such as general operating systems, game applications, and browser software, and is usually implemented by a processing unit such as a CPU. In particular, in the client's smartphone 1, the cognitive behavioral therapy feedback program according to the present invention is executed in this application execution unit 14, and a display data generation unit 146, an output control unit 147, a location information acquisition unit 148, and a cognitive behavioral therapy execution unit 141 are virtually constructed.

[0093] The cognitive behavioral therapy execution unit 141 is a module that executes lesson mode and work mode as event processing according to the control of the analysis unit 332. In this embodiment, the cognitive behavioral therapy execution unit 141 has the function of sequentially unfolding story mode and game mode according to the rules, logic, and algorithms related to cognitive behavioral therapy described in the scenario, and generating various event processing to advance cognitive behavioral therapy. It synchronizes with the cognitive behavioral therapy execution unit 33 on the management server 3 and generates event processing such as playing lesson movies and inputting into worksheets according to the scenario progress, the user's lesson progress, time, and current location.

[0094] Furthermore, in this embodiment, the cognitive behavioral therapy execution unit 141 is configured to cooperate with the cognitive behavioral therapy execution unit 33 on the management server 3 side. The management server 3 performs part of the cognitive behavioral therapy progress processing, while the cognitive behavioral therapy execution unit 141 on the smartphone 1 side processes part of the graphic processing and event processing. For example, the management server 3 generates the conditions for event occurrence, sends these conditions to the smartphone 1, and the smartphone 1 processes the actual event generation (trigger generation) and the corresponding graphic processing. The cognitive behavioral therapy execution unit 141 includes an automatic thought acquisition unit 142, a biometric information acquisition unit 143, an operation identification unit 144, and a synchronization processing unit 145.

[0095] The automatic thought acquisition unit 142 is a module that acquires text, images, and other biometric information from the client through the input interface 12, camera, and other sensors provided in the smartphone 1. The automatic thoughts acquired here include behavioral records entered by the client themselves, images (of the client's facial expressions) captured by the camera 12b, location information acquired by the location information acquisition unit 148, signals obtained from various sensors such as acceleration sensors, and biometric information such as blood flow, body temperature, and heart rate measured by other sensors and acquired by the biometric information acquisition unit 143.

[0096] The operation identification unit 144 is a module that uses a machine learning model to estimate the level of awareness of an operation in order to distinguish between the client's conscious and unconscious operations in both the incomplete operation history and the decided operation history, and evaluates the client's cognitive and emotional tendencies based on the identified level of awareness.

[0097] The synchronization processing unit 145 is a module that synchronizes and links processing on the smartphone 1 with processing on the management server 3. The display data generation unit 146 is a module that generates display data to be displayed on the display 13a. The display data is generated by combining graphic data, image data, text data, video data, audio, and other data.

[0098] The location information acquisition unit 148 is a module that acquires information regarding the current time and current location as current information. In this embodiment, current location information is input from GPS or the like, and current time information is input from the timing unit. The location information acquisition unit 148 has the function of acquiring and recording location information indicating the position of the device. This location information acquisition function includes, for example, a method of detecting the device's position by signals from satellites, such as GPS, or a method of detecting the position by radio wave strength from a mobile phone base station 22 or a Wi-Fi communication access point, as shown in Figure 1.

[0099] Furthermore, the output control unit 147 is equipped with a trigger function that controls the timing of presenting feedback according to instructions from the management server 3. This trigger function is a module that instructs the timing of event processing, such as automatically starting the application, outputting notifications from the application, or having characters automatically start speaking within the application, as a way of presenting feedback. For example, it outputs feedback prompting the user to take the next lesson when a certain amount of time has elapsed since the end of the previous lesson.

[0100] In this embodiment, feedback is triggered by input of current location, movement speed, current time, and elapsed time. However, the present invention is not limited to this. For example, external factors may be predicted or trigger data generated based on detection data from detection devices such as a camera, thermometer, heart rate monitor, and microphone connected to the input interface 12. These detection devices can be provided in a wearable terminal connected via wireless communication.

[0101] (Operation of the Cognitive Behavioral Therapy Feedback System S) The mental information input method of the present invention can be implemented by operating the cognitive behavioral therapy feedback system S described above. Figures 6 to 8 show the operation of the cognitive behavioral therapy feedback system S. Note that the processing procedure described below is merely an example, and each process may be modified as much as possible. Furthermore, depending on the embodiment, steps in the processing procedure described below can be omitted, replaced, and added as appropriate.

[0102] (1) As shown in the overall operation diagram 6, first, the user terminal, smartphone 1, launches a cognitive behavioral therapy feedback application to input automated thoughts, which are then acquired by the management server 3 (S201, S101). In addition, during this automated thought input, incomplete operations are monitored and recorded. In this monitoring of incomplete operations, any operational actions such as input errors, rewriting, or deletion before pressing the send or confirm button during the automated thought input process are traced as incomplete operations.

[0103] For example, as shown in Figure 8, in the initial stages of the input operation, a record of behavior based on negative self-perception, such as "I was tired from the morning and couldn't concentrate on work," was initially entered. However, this entry was partially deleted and rewritten to reflect positive self-perception, such as "I felt anxious from the morning, but I was able to make progress on my work." This entry was further corrected later, and ultimately, it mentions that the user recognized the negative mental burden of "feeling anxious," acknowledged the causal relationship of "losing concentration," and spontaneously adopted the solution of "taking a walk to clear their head," but ultimately arrived at the mental result of "not feeling refreshed." Upon receiving such automated thought input and monitoring and recording of the incomplete operations during that process, the management server 3 references the pre-trained model (S102) and generates emotion data (S103).

[0104] Then, the automatic thoughts entered by the client themselves are compared with the emotional data generated by the generative system AIGa1, and the discrepancy rate is calculated (S104), and an evaluation of cognitive distortion is performed (S105). Next, based on this evaluation of cognitive distortion, various analyses of evaluations are performed (S106). In this embodiment, analyses of correlations and causal relationships between external events and automatic thoughts, qualitative evaluations regarding the recovery or progression of cognitive distortion, and diagnoses for self-counseling are performed.

[0105] For example, in the example shown in Figure 8, the initial negative self-perception input is deleted, then corrected to a positive description, and finally the decision is reached based on the result that the individual attempted to resolve the issue themselves but was unsuccessful. As for the evaluation of this behavior, as shown in Figure 9, although negative automatic thoughts are initially expressed in the initial input, they are later deleted, suggesting that such thoughts are likely to be the root cause of daily stress and anxiety, indicating a high degree of mental dissociation. Subsequently, further corrections are made, and while anxiety and fatigue were initially emphasized, the final description emphasizes positive cognitions such as "I tried to concentrate" and "I tried to change my mood by taking a walk." In this case, human intervention is often necessary, so a positive response should be given depending on the situation and the degree of dissociation, such as "You were able to change your mood by taking a walk. The connection between your actions and emotions is becoming clearer within you, which is a good sign."

[0106] Furthermore, in the example in Figure 8, the description of the behavior "I couldn't concentrate on my work" is ultimately corrected to a self-assessment of "I was able to make progress on my work." Since this also indicates a positive trend, the AI-generated automated response may be limited to objective evaluations and numerical values ​​only, such as "It is highly likely that 'cognitive restructuring,' a common process in cognitive behavior, has progressed by about 7.2%." In this case, as shown in Figure 9, the numerical changes related to cognitive restructuring over time may be extracted and returned as response information in the form of a flag diagram showing these numerical changes.

[0107] By referring to this self-monitoring input and the history of incomplete operations, the underlying psychological state is estimated. In the example shown in Figure 8, the initial entries made by the client often express their emotions and thoughts honestly, and if they are deleted, it may reflect the core emotions or problems they are actually experiencing. In response to this, a human intervention might be used to reply, "The act of deletion itself may be part of your self-defense mechanism. When there is resistance to expressing negative emotions or a desire not to show weakness to others, it is possible that there are further emotions or unresolved issues behind it."

[0108] On the other hand, depending on the client's condition, as the description progresses, the AI ​​may diagnose that the client is showing signs of recovery, such as finding solutions or coping behaviors (trying to take a walk) rather than simply being overwhelmed by negative emotions, and then, as shown in Figure 9, it may respond with an automated AI response such as, "As a result of the treatment, there are signs of improvement of about 84% compared to last month," along with objective numerical changes such as graphs.

[0109] Based on the results of these various analyses, an appropriate feedback mode is selected (S107), and feedback processing is performed through the selected mode to generate output information (S108). Figure 10 shows the feedback mode patterns of the generation system AIGa1 and smartphone1.

[0110] Basically, feedback to the client is provided through human intervention by a psychologist or other Dr. 1, including psychological advice, counseling, encouragement, and reports on the healing process. However, depending on the client's psychological state, for example, if they are showing signs of recovery and it is acceptable for the generative system AIGa 1 to respond mechanically and automatically, or if the client is uncomfortable with interpersonal interactions, the response selection unit 342 determines and selects whether to correspond to the generative system AIGa 1 or to the psychologist or other Dr. 1, as shown in Figure 10(a). Here, based on the analysis results by the analysis unit 332, a response selection step is executed in which the response selection unit selects either an automatic response by the generative system AIGa 1 or a human response. As a result, based on the correlation characteristics analyzed by the analysis unit 332 and the selection result by the response selection unit 342, the response generation unit 34 generates response information to be provided to the client and notifies the client.

[0111] Furthermore, the response selection unit 342 automatically responds when it is possible to respond using only the generation system AIGa1, depending on the client's psychological state. However, as shown in Figure (b), if a human response is necessary under certain circumstances, the generation system AIGa1 sends a request for a human response, and the response generation unit 34 transmits response information generated by the human operation performed in response to this request to the client. In this way, by intervening with human response as needed, both the efficiency of AI and the emotional support of humans are utilized.

[0112] On the other hand, in cases where the client shows a strong tendency towards recovery, a mechanical and automated response by AI may actually allow the treatment to proceed more smoothly. In such cases, as shown in Figure (c), a graph showing numerical changes that succinctly indicate the healing tendency (recovery tendency) is sent to the client as a response by the generation system AIGA1. Specifically, the change extraction unit 332i extracts numerical changes in predetermined items according to the time series from observation record information, including numerical values ​​indicating the client's state, stored in the observation record database 35d, and generates these numerical changes as response information, which is sent to the client as a response by the generation system AIGA1 without human intervention. This allows the client to quantitatively grasp the changes in their psychological state and the tendency toward healing, reducing the interpersonal burden on the client while enabling more accurate psychological support.

[0113] Furthermore, as described above, even when automatic responses are possible solely by the generation system AIGA1, some clients may not want to rely entirely on the AI's responses. Therefore, as shown in Figure (d), even when an automatic response by the generation system AIGA1 is selected, the response selection unit 342, depending on the client's psychological state, sends a request for human approval to the psychologist or other Dr1 regarding the content of the automatic response through the approval execution function of the automatic response execution unit 341 or the human operation unit 343. The response generation unit 34 generates response information reflecting the approval result of the approval execution function, and the information distribution unit 36 ​​executes the notification in accordance with the human approval. This allows the content of the generation system AIGA1's responses to be confirmed by the human psychologist or other Dr1, and allows for human adjustments if necessary. This increases the client's confidence in the AI's automatic responses and provides the client with consistency and a sense of security.

[0114] Furthermore, if the client requires a human response, or if the client strongly desires a human response, a human response by a psychologist (Dr. 1) is selected. In this case, as shown in Figure (d), the human operation unit 343 accepts the human operation from the psychologist (Dr. 1) and, in response to that human operation, presents a hypothetical automated response as a response proposal, showing what kind of response the generation system AIGa 1 would have given, which can then be referenced. As a result, the burden on the psychologist (Dr. 1) performing the human response is reduced, human error is avoided, and up-to-date and comprehensive responses can be achieved. When the response selection unit 342 makes a selection, in this embodiment, in analyzing the client's psychological state, the correlation between behavioral information and cognitive information is analyzed by considering the client's history of incomplete operations and decision operations in their behavior and operations.

[0115] The response information generated in this way is delivered to the client's smartphone 1, and after waiting for feedback response processing from the smartphone 1, synchronization processing is performed with the application on the smartphone 1 (S109 and S202), and the communication session proceeds while sending and receiving the data necessary for feedback, and feedback is performed in various modes (S204). Thereafter, each of the above processes S201 to S203 and S101 to S109 is repeated until the application terminates ("Y" in S204) ("N" in S204).

[0116] (2) Feedback Execution The process related to the feedback execution in step S203 described above will be explained in detail. Figure 7 shows the sequence during feedback execution. As shown in the figure, first, the management server 3 performs an evaluation of cognitive distortion (S301), and then it is determined whether the evaluation value (here, the "deviation rate" mentioned above) is greater than the threshold (S302). In this step S302, if it is determined that the evaluation value exceeds the threshold ("Y" in S302), it is determined that cognitive distortion has occurred, and this determination result is reflected in the subsequent feedback.

[0117] On the other hand, in step S302, if it is determined that the evaluation value is below the threshold ("N" in S302), it is determined that no cognitive distortion has occurred, and this determination is reflected in subsequent feedback. After the evaluation value determination is completed, the pattern rules are referenced according to the evaluation value, and scenarios necessary for each mode, such as for characters or chat, are generated according to the feedback mode (S304).

[0118] After generating this scenario, the system initiates a response to the client for cognitive reconstruction (S305), and the smartphone 1 executes event processing for this cognitive reconstruction (S401). In this event processing, the generation system AIGa 1 executes a response, which can be, for example, a dialogue with a character generated by the generation system AIGa 1, or a chat dialogue with the generation system AIGa 1. In accordance with this event processing, the management server 3 provides the client with support from the generation system AIGa 1 or a human through dialogue with the character or chat (S306). Support here can include functions within the application such as "asking questions," "prompting awareness," and "encouraging messages," as well as outputting messages through other messaging applications, email, smart speakers, etc.

[0119] Then, based on the client's responses during this cognitive reconstruction, the causal relationship between the automatic thought sentences and emotions is estimated, and the relevant sections are highlighted (S403). At this time, if it is necessary to observe the client's facial expressions, biometric information is also acquired through functions such as video chat (S404). In addition, the responses from the client during the interaction in this cognitive reconstruction event processing are acquired by the management server 3 as new automatic thoughts and become subject to re-evaluation.

[0120] Subsequently, self-counseling is performed (S405). Here, automatic thoughts and worksheets are entered, and biometric information such as voice, video, blood flow, and heart rate is acquired as needed (S406). These acquired automatic thoughts and biometric information are transmitted to the management server 3 and recorded (S407, S307). Upon receiving this automatic thoughts and biometric information, a decision is made as to whether re-evaluation is necessary (S308). If necessary ("necessary" in S308), the processing from the evaluation of cognitive distortion (S301) onward is executed again. On the other hand, if re-evaluation is not necessary ("unnecessary" in S308), or when the mode is to be terminated ("Y" in S408), step S203 is terminated.

[0121] (Effects and Effects) As described above, according to this embodiment, by accumulating and analyzing the client's self-observation records and flexibly selecting responses from the generative system AIGa1 or the psychologist Dr1, it becomes possible to reduce the human burden on the psychologist Dr1 while providing appropriate support according to the psychological state of the client CLa to c, and to concentrate human resources on clients who truly need human counseling.

[0122] Furthermore, in this embodiment, the system automatically selects a generative AIGA1 or a human response from a psychologist or other professional (Dr1) according to the client's psychological state. By intervening or acknowledging as needed, both the efficiency of AI and the emotional support of humans can be utilized. This facilitates human intervention in situations where AI-based responses are deemed inappropriate, thereby enabling appropriate psychotherapy.

[0123] Furthermore, in automated responses using the AIGA1 generation system, the time-series changes in the client's behavior and psychological state are analyzed and mechanically provided to the client, allowing the client to quantitatively understand their healing progress. This reduces the interpersonal burden on the client while enabling more accurate psychological support. On the other hand, even when a human response is selected, the AI's automated response can be referenced, reducing the burden on psychologists and other professionals (Dr1), avoiding human error, and enabling the execution of the latest and most comprehensive responses. This allows for rapid and effective human responses based on AI support.

[0124] In this embodiment, by acquiring and accumulating a history of incomplete and decided operations based on the client's behavior and actions, and analyzing the correlation between behavioral and cognitive information, it becomes possible to understand the client's behavioral and cognitive tendencies in detail. This makes it possible to generate appropriate responses, provide more individualized support, and reduce the client's psychological burden.

[0125] Furthermore, in this embodiment, by comparing the history of incomplete operations and the history of decided operations, and calculating feature patterns based on the comparison results, a natural language model is utilized to generate responses through machine learning, thereby providing appropriate response information that matches the client's state. This enables rapid response to changes in the client's behavior and cognition, and by using a machine learning model, it is possible to automatically generate responses that are optimal for individual needs, thereby achieving more accurate support.

[0126] In this embodiment, by analyzing the client's operation patterns (such as time delays, frequency of re-entry, and deletion operations) and estimating the client's psychological state based on these patterns, a more effective response can be provided by adjusting the content and timing of the response based on the analysis results. Furthermore, in this embodiment, by identifying conscious and unconscious operations in the history of incomplete operations and the history of decided operations, and evaluating the client's cognitive and emotional tendencies based on these, it becomes possible to make appropriate evaluations based on the level of awareness regarding operations, thereby improving the accuracy of treatment.

[0127] It should be noted that the present invention is not limited to the embodiments described above, and the components can be modified and implemented in practice without departing from the spirit of the invention. Furthermore, various inventions can be formed by appropriately combining the multiple components disclosed in the above embodiments. For example, some components may be deleted from all the components shown in the embodiments.

[0128] CLa-c...Client Ga1...Generative AI Dr1...Psychologist, etc. S...Cognitive Behavioral Therapy Feedback System 1...Smartphone 1a-c...Smartphone 2...Internet 3...Management Server 11...Communication Interface 12...Input Interface 12a...Touch Panel 12b...Camera 13...Output Interface 13a...Display 14...Application Execution Unit 15...Memory 22...Wireless Base Station 31...Communication Interface 32...Authentication Unit 32a...Location Information Management Unit 33...Cognitive Behavioral Therapy Execution Unit 34...Response Generation Unit 35...Database Group 35a...AI Learning Database 35b...User Database 35c...Cognitive Behavioral Therapy Database 35d...Observation Record Database 36...Information Distribution Unit 37...Billing Processing Unit 141...Cognitive Behavioral Therapy Execution Unit 142...Automatic Thought Acquisition Unit 143...Biometric Information Acquisition Unit 144...Operation Identification Unit 145...Synchronization Processing Unit 146...Display data generation unit 147...Output control unit 148...Location information acquisition unit 331...Observation recording unit 331a...Input reception unit 331b...Incomplete operation acquisition unit 331c...Decision operation acquisition unit 331d...Sensor information acquisition unit 331e...Information control unit 331f...Comparison unit 331g...Pattern calculation unit 331h...Operation pattern analysis unit 332...Analysis unit 332a...Automatic thought acquisition unit 332b...Reaction classification unit 332c...Emotion generation unit 332d...Pre-trained model management unit 332e...Judgment unit 332f...Evaluation unit 332g...Evaluation execution unit 332h...Causal relationship analysis unit 332i...Change extraction unit 332j...Qualitative evaluation calculation unit 332k...Self-counseling management unit 332l...Biological information analysis unit 333...AI control unit 334a...Comparison unit 334b...Pattern calculation unit 334c...Operation pattern analysis unit 341...Automatic response execution unit 342...Response selection unit 343...Human operation unit 344...Feedback processing unit 361...Synchronization processing unit

Claims

1. A system for inputting mental information, including self-observation records based on the actions or operations of a person receiving support, comprising: an incomplete operation acquisition unit that acquires the actions or operations of the person receiving support as an incomplete operation history; a decision operation acquisition unit that acquires input information determined by the actions or operations of the person receiving support as a decision operation history; an observation record storage unit that stores self-observation records including the incomplete operation history and the decision operation history; a correlation feature analysis unit that analyzes the correlation between the behavioral information and cognitive information of the person receiving support based on the self-observation records; and a response generation unit that generates response information to be provided to the person receiving support according to the analysis results in the correlation feature analysis unit.

2. The mental information input system according to claim 1, further comprising: a comparison unit that compares the incomplete operation history with the decision operation history; a pattern calculation unit that analyzes the correlation between the behavioral information and cognitive information of the person to be supported according to the comparison result by the comparison unit and calculates a feature pattern that characterizes the analyzed correlation; a natural language model that has been trained to calculate response information according to the comparison result between the incomplete operation history and the decision operation history and the feature pattern; and a response request unit that receives the comparison result between the incomplete operation history and the decision operation history and the feature pattern, generates and executes a prompt for the natural language model to generate the response information.

3. The mental information input system according to claim 1 or 2, further comprising an operation pattern analysis unit that analyzes operation patterns including at least one of the following in the input operations of the person being supported: time delay, frequency of re-input, occurrence of deletion operations, and time period of occurrence, wherein the response generation unit estimates the potential psychological state of the person being supported and determines the content and timing of the response based on the analysis results by the operation pattern analysis unit.

4. The mental information input system according to any one of claims 1 to 3, further comprising an operation identification unit that uses a machine learning model to estimate the level of awareness of an operation in order to distinguish between conscious and unconscious operations of the person being supported in both the history of incomplete operations and the history of decided operations, and characterized in that the cognitive and emotional tendencies of the person being supported are evaluated based on the level of awareness identified by the operation identification unit.

5. A program for inputting mental information, including self-observation records based on the actions or operations of a person receiving support, characterized in that the computer functions as: an incomplete operation acquisition unit that acquires the actions or operations of the person receiving support as an incomplete operation history; a decision operation acquisition unit that acquires input information determined by the actions or operations of the person receiving support as a decision operation history; an observation record storage unit that stores self-observation records including the incomplete operation history and the decision operation history; a correlation feature analysis unit that analyzes the correlation between the behavioral information and cognitive information of the person receiving support based on the self-observation records; and a response generation unit that generates response information to be provided to the person receiving support according to the analysis results in the correlation feature analysis unit.

6. The mental information input program according to claim 5, characterized in that the computer is further configured to function as a response request unit that receives the comparison results of the incomplete operation history and the decision operation history, generates and executes prompts for the natural language model, and generates the response information according to the comparison results of the incomplete operation history and the decision operation history and the feature patterns, and the computer is configured to function as a response request unit that receives the comparison results of the incomplete operation history and the decision operation history and the feature patterns, and generates and executes prompts for the natural language model to generate the response information.

7. The mental information input program according to claim 5, wherein the computer further functions as an operation pattern analysis unit that analyzes operation patterns including at least one of the following in the input operations of the person being supported: time delay, frequency of re-input, occurrence of deletion operations, and time period of occurrence, and the response generation unit estimates the potential psychological state of the person being supported and determines the content and timing of the response based on the analysis results by the operation pattern analysis unit.

8. The mental information input program according to any one of item 5, characterized in that the computer further functions as an operation identification unit that estimates the level of awareness of an operation using a machine learning model in order to identify differences between conscious and unconscious operations of the person being supported in both the history of incomplete operations and the history of decided operations, and evaluates the cognitive and emotional tendencies of the person being supported based on the level of awareness identified by the operation identification unit.

9. A method for inputting mental information, including a self-observation record based on the actions or operations of a person receiving support, comprising: an operation acquisition step in which the computer acquires the actions or operations of the person receiving support as an incomplete operation history and acquires input information determined by the actions or operations of the person receiving support as a determined operation history; an observation record storage step in which the computer stores the self-observation record, including the incomplete operation history and the determined operation history, in an observation record storage unit; a correlation feature analysis step in which the computer's correlation feature analysis unit analyzes the correlation between the behavioral information and cognitive information of the person receiving support based on the self-observation record; and a response generation step in which the computer's response generation unit generates response information to be provided to the person receiving support according to the analysis results in the correlation feature analysis unit.

10. The mental information input method according to claim 9, further comprising: a pattern calculation step in which the computer's pattern calculation unit analyzes the correlation between the behavioral information and cognitive information of the person to be supported in accordance with the comparison result of the incomplete operation history and the decision operation history, and calculates a feature pattern that characterizes the analyzed correlation; and a response request step in which a natural language model trained on machine learning to calculate response information corresponding to the comparison result of the incomplete operation history and the decision operation history and the feature pattern receives the comparison result of the incomplete operation history and the decision operation history and the feature pattern, and the computer's response request unit generates and executes a prompt for the natural language model to generate response information.

11. The mental information input method according to claim 9, further comprising an operation pattern analysis step in which an operation pattern analysis unit analyzes an operation pattern that includes at least one of the following in the input operation of the person being supported: time delay, frequency of re-input, occurrence of deletion operations, or time period of occurrence, wherein the response generation step estimates the potential psychological state of the person being supported and determines the content and timing of the response based on the analysis results by the operation pattern analysis unit.

12. The mental information input method according to claim 9, further comprising an operation identification step in which the computer's operation identification unit estimates the level of awareness of an operation using a machine learning model in order to distinguish between conscious and unconscious operations of the person being supported, in both the incomplete operation history and the decided operation history, and the cognitive and emotional tendencies of the person being supported are evaluated based on the level of awareness identified in the operation identification step.

Images