Information processing device, information processing method, and program

Abstract

This invention addresses the problem of making it possible to, through communication with a user, propose at a suitable timing a task and / or an action according to the current situation of the user.　In a CPU 11 of a management server 1 constituting an information processing system, an information management unit 32 stores a task, manages a hierarchical structure of the task, manages task communication information related to the task in association with the task, and also manages execution status information related to the current execution status of the task and history information related to the past execution history of the task. In addition, on the basis of the execution status information and the history information, an information generation unit 33 generates assistance information for assisting a user who executes the task. Due to this configuration, the above-mentioned problem is solved.

Description

Information Processing Apparatus, Information Processing Method, and Program

[0001] The present invention relates to an information processing apparatus, an information processing method, and a program.

[0002] Conventional ToDo lists and task management systems have been proposed that operate based on information manually input by the user (see, for example, Patent Document 1).

[0003] Japanese Patent Application Laid-Open No. 2011-187054

[0004] However, in the conventional technology, since tasks are presented without considering the user's situation and physical and mental state, it has not always been possible to prompt execution at an optimal timing. Also, in conventional chat systems, communication management is generally performed on a user or group basis, and communication management linked to tasks and ToDos has not been sufficiently realized.

[0005] The present invention has been made in view of such circumstances, and an object thereof is to enable a task proposal according to the user's current situation and various histories to be made at a suitable timing through communication with the user.

[0006] To achieve the above object, one aspect of the present invention is an information processing apparatus including: task storage means for storing tasks; hierarchical structure management means for managing the hierarchical structure of the tasks; and communication management means for associating and managing task communication information related to the tasks with the tasks.

[0007] An information processing method and a program corresponding to the above information processing apparatus according to one aspect of the present invention are also provided as an information processing method and a program corresponding to the information processing apparatus according to one aspect of the present invention.

[0008] According to the present invention, a task and / or action proposal according to the user's current situation or the like can be made at a suitable timing through communication with the user.

[0009] This figure shows an example of the overall configuration of an information processing system including a management server according to one embodiment of the information processing device of the present invention. This is a block diagram showing an example of the hardware configuration of the management server in Figure 1. This is a functional block diagram showing an example of the functional configuration of the management server. This is a flowchart showing an example of the processing flow of the management server up to the point of proposing support information to the user. This figure shows a specific example of the overall configuration of this service. This figure shows a specific example of the user interface displayed on the user terminal in Figure 1

[0010] Hereinafter, this embodiment will be described with reference to the drawings. <Information Processing System S> Figure 1 is a diagram showing an example of the overall configuration of an information processing system S including a management server 1 according to one embodiment of the information processing device of the present invention. The information processing system S is an information processing system configured such that the management server 1, the generation AI server 2, and the user terminal 3 are connected via a network N. The network N is, for example, the Internet, LAN (Local Area Network), VPN (Virtual Private Network), etc.

[0011] The information processing system S can provide a service (hereinafter referred to as "the Service") that supports the user in executing tasks by suggesting tasks and / or actions appropriate to the user's current situation at an appropriate time through communication with the user. In the following description of the Service, the terms "To-Do," "Task," and "Action" may be used. Of these, "To-Do" mainly means "things that need to be done," and "Task" mainly means "the whole set of things to be done." Also, "Action" mainly means "an action to be performed." In this specification, these are not particularly distinguished and are used to mean "things that need to be done."

[0012] [Management Server 1] Management Server 1, which constitutes the information processing system S, is an information processing device that serves as a server for managing the entire information processing system S. Management Server 1 is managed, for example, by a service provider that provides this service. Management Server 1 is capable of executing a predetermined application program that enables the provision of this service. Management Server 1 is capable of transmitting various types of information to the generation AI server 2, user terminal 3, and external sources, and executing various types of processing. Furthermore, Management Server 1 is capable of acquiring various types of information transmitted from the generation AI server 2, user terminal 3, and external sources, and executing various types of processing.

[0013] Management server 1 manages various types of information. For example, management server 1 manages information about the tasks themselves (hereinafter referred to as "task information"). Task information includes, for example, the hierarchical structure of tasks.

[0014] Furthermore, the management server 1 acquires and manages information regarding the current execution status of tasks (hereinafter referred to as "execution status information") for each user. Here, "user" refers to a person who uses this service to execute or attempt to execute a task. The management server 1 enables real-time synchronized management of changes to execution status information among multiple users. Here, "real-time synchronization" means that the exchange of information between users is synchronized with a delay that does not cause practical problems.

[0015] Furthermore, the management server 1 acquires and manages information regarding the past execution history of tasks (hereinafter referred to as "execution history information") for each user. Based on at least a portion of the user's execution status information, execution history information, current status information, and behavior history information, the management server 1 generates information to support the user (hereinafter referred to as "support information").

[0016] Furthermore, the management server 1 acquires and manages information regarding the user's current status (hereinafter referred to as "current status information") and information regarding the user's history of actions (hereinafter referred to as "action history information") for each user. Here, "current status information" refers to information indicating the user's current state (psychological state, environment, direction of thought, etc.), and "action history information" refers to the history of various activities and states that the user has actually performed in the past, regardless of whether or not they are explicitly managed as tasks. A characteristic of this embodiment is that current status information and action history information also include information regarding actions that are not managed as tasks. This information is acquired from applications, devices, or external services linked to this system. Specifically, for example, this includes the execution results of learning applications and organizational management applications linked to this system, schedule information obtained through calendar linkage, user actions not limited to tasks, sensor data acquired from user terminals, and information processed from them.

[0017] A key feature of this embodiment is that the current status information and activity history information include information that is not explicitly managed as a task. For example, activity history information may include the content of learning materials created in the past, specific interactions during the creation of internal presentation materials, and referenced literature. In addition, current status information may include the user's current thought process and psychological state (e.g., concentration level, relaxation level, motivation level). Based on this information, the management server 1 can generate appropriate support information tailored to the user's work context, expertise, and current state. The current status information and activity history information may also include task execution status information and execution history information.

[0018] Furthermore, the management server 1 determines whether the amount of information acquired regarding the user's current status and activity history is sufficient. If the management server 1 determines that the amount of information acquired regarding the user's current status and activity history is sufficient, it derives a task that the user should perform now and proposes the derived task to the user. At this time, along with the derived task, the management server 1 may also present to the user at least one of the reasons why the task should be performed now and the method for performing the task. By having the management server 1 determine whether the amount of information acquired regarding the current status and activity history is sufficient before deriving a task, inappropriate suggestions can be avoided, and high-quality suggestions that are truly suited to the user's situation can be made. By waiting until sufficient information is accumulated, it becomes possible to derive highly accurate tasks based on a comprehensive understanding of the user's current status, past activity patterns, work context, etc.

[0019] Furthermore, the management server 1 provides feedback for deriving the next task based on at least one of the user's execution history information and behavior history information. The management server 1 causes the generation AI model (hereinafter referred to as the "task derivation model") provided by the generation AI server 2 to derive tasks. The task derivation model is a trained model that enables the derivation of the task to be executed now as the output of a response to a predetermined instruction (prompt) input that includes at least a part of the execution status information, execution history information, current situation information, and behavior history information. For example, the task derivation model receives "the user's past task execution history, current situation, and similar past cases" as input and generates "a suggestion for the task to be executed now, the reason for it, and how to execute it" as output. The management server 1 provides feedback for deriving the next task by having the task derivation model learn additional training data that includes the user's history information. This feedback can be implemented, for example, by fine-tuning the parameters of the task derivation model, by adding training data including execution history information and behavior history information to the task derivation model, by incorporating this information into input prompts, or by analyzing this information to improve the suggestions. For example, if a user successfully performs task X, the result is reflected in the model as training data indicating that it was an "appropriate task suggestion." Conversely, if the user does not perform the suggested task Y, the result is reflected in the model as training data indicating that it was an "inappropriate task suggestion." By repeating this feedback, the task derivation model gradually becomes able to make suggestions that are more suited to the user's preferences and circumstances.

[0020] The management server 1 provides a chat interface as a means of interaction, enabling participation between the user and the AI ​​chatbot provided by the AI ​​generation server 2. The management server 1 controls the display of the chat interface on the user terminal 3. The management server 1 interacts with the user via the chat interface and, through that interaction, obtains information regarding the content of the interaction with the user (hereinafter referred to as "interaction information").

[0021] Furthermore, the management server 1 extracts information related to task-related communications (hereinafter referred to as "task communication information") from the dialogue information and manages it in association with the task. The management server 1 also acquires dialogue information entered into the chat interface and generates questions to capture the user's feelings based on the acquired dialogue information.

[0022] Furthermore, the management server 1 extracts the history of conversations related to a task from the conversation information and manages it in association with the task. When a user selects a task that the user intends to execute through user operation, the management server 1 retrieves and displays the history of conversations associated with that task.

[0023] Furthermore, the management server 1 acquires task information from dialogue information based on natural language input in the chat interface. The management server 1 temporarily generates a task from the acquired task information and presents it to the user. The temporarily generated task is managed as an unconfirmed task. When the user who has received the task information performs a confirmation operation, the management server 1 registers the task related to that task information as a task to be executed by the user. The user's confirmation operation is accepted by a predetermined user interface displayed on the user terminal 3, and the task changes from an unconfirmed state to a confirmed state.

[0024] Furthermore, the management server 1 enables the user to break down tasks they have registered to execute into multiple subtasks. When the management server 1 breaks down a task into multiple subtasks, it calculates and manages the progress rate from the execution status information of the task before it was broken down into multiple subtasks, based on the completion status of each of the multiple subtasks. In addition, the management server 1 analyzes the task structure of past projects similar to the current project and proposes a task structure suitable for the current project.

[0025] Furthermore, the management server 1 generates a dedicated chat thread for each task currently running, enabling the management of all task-related communications within the chat thread. In this case, the management server 1 manages the chat thread history as dialogue information and, by referring to the chat thread history of similar tasks in the past, enables it to suggest information related to the currently running task (hereinafter referred to as "task-related information") to the user. The management server 1 also enables it to suggest solutions to the task to the user by referring to the chat thread history of similar tasks in the past. Details of the configuration and processing of the management server 1 will be described later.

[0026] [Generative AI Server 2] The Generative AI Server 2, which constitutes the Information Processing System S, performs some of the processing of the Management Server 1 by utilizing a Large-Scale Language Model (LLM) and deep learning. Specifically, the Generative AI Server 2 provides a Generative AI model that can output a predetermined response when a predetermined instruction (prompt) is input.

[0027] For example, the generating AI server 2 provides a task derivation model that enables the derivation of the task the user should currently perform as an output response to the input of a predetermined instruction (prompt) including execution status information and execution history information. Furthermore, the generating AI server 2 provides an AI chatbot that enables natural dialogue with the user by performing natural language processing (NLP) and machine learning.

[0028] [User Terminal 3] User terminal 3, which constitutes the information processing system S, is an information processing device operated by a user who uses this service. User terminal 3 is composed of, for example, a smartphone, tablet terminal, personal computer, etc. User terminal 3 is capable of executing a predetermined application program that enables the use of this service. User terminal 3 is capable of performing various processes based on various information transmitted from the management server 1 and external sources, as well as various information entered by the user. User terminal 3 is also capable of transmitting various information to the management server 1 and external sources.

[0029] For example, user terminal 3 acquires execution status information and current status information entered by the user and sends it to management server 1. User terminal 3 also displays a chat interface provided by management server 1 and sends the information entered by the user into the chat interface as dialogue information to management server 1. User terminal 3 also acquires responses sent from management server 1 or generation AI server 2 and outputs them as text data or audio data. Details of the configuration and processing of user terminal 3 will be described later.

[0030] The processing performed by the management server 1, the generation AI server 2, and the user terminal 3, which constitute the information processing system S, is merely one example. In other words, as long as the information processing system S as a whole has the functionality to realize the above-mentioned processing, some or all of the functions for realizing the above-mentioned processing may be shared or performed collaboratively within the information processing system S.

[0031] For example, some or all of the functions of the management server 1 may be assigned to other devices within the information processing system S. Alternatively, some or all of the functions of other devices within the information processing system S may be assigned to the management server 1. Furthermore, some or all of the functions of the management server 1 may be transferred to other servers, etc., not shown in the diagram. This facilitates processing within the entire information processing system S and allows for complementary processing.

[0032] <Hardware Configuration> [Hardware Configuration of Management Server 1] Figure 2 is a block diagram showing an example of the hardware configuration of Management Server 1 shown in Figure 1. Management Server 1 includes a CPU (Central Processing Unit) 11, a ROM (Read Only Memory) 12, a RAM (Random Access Memory) 13, a bus 14, an input / output interface 15, an output unit 16, an input unit 17, a storage unit 18 as a storage means, a communication unit 19, and a drive 20.

[0033] The CPU 11 executes various processes according to the program recorded in the ROM 12 or the program loaded from the storage unit 18 into the RAM 13. The RAM 13 also stores data necessary for the CPU 11 to execute various processes. The CPU 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output interface 15 is also connected to this bus 14.

[0034] The input / output interface 15 is connected to an output unit 16, an input unit 17, a storage unit 18, a communication unit 19, and a drive 20. The output unit 16 consists of a display, speaker, etc., and outputs various information as images, sounds, etc. The input unit 17 consists of a keyboard, mouse, touch panel, etc., and accepts input of various information. The storage unit 18 consists of a hard disk, DRAM (Dynamic Random Access Memory), etc., and stores various data. The communication unit 19 communicates with other devices via the aforementioned network N, which is configured as the Internet, etc.

[0035] The drive 20 is appropriately equipped with removable media 21, such as a magnetic disk, optical disk, magneto-optical disk, or semiconductor memory. Programs read from the removable media 21 by the drive 20 are installed in the storage unit 18 as needed. The removable media 21 can also store various types of data stored in the storage unit 18, just like the storage unit 18.

[0036] [Hardware Configuration of User Terminal 3] User terminal 3 has the same hardware configuration as management server 1 shown in Figure 2 above. That is, user terminal 3 has a CPU, ROM, RAM, bus, input / output interface, output unit, input unit, storage unit, communication unit, drive, and removable media, which correspond to the CPU 11, ROM 12, RAM 13, bus 14, input / output interface 15, output unit 16, input unit 17, storage unit 18, communication unit 19, drive 20, and removable media 21 in Figure 2, respectively.

[0037] <Functional Configuration of Management Server 1> Figure 3 is a functional block diagram showing an example of the functional configuration of Management Server 1. When Management Server 1's CPU 11 is operating, the following functions are active: Information Acquisition Unit 31, Information Management Unit 32, Information Generation Unit 33, Determination Unit 34, Task Proposal Unit 35, Task Registration Unit 36, and Task Decomposition Unit 37. Also, when CPU 11 is operating, the following functions are active: Chat Management Unit 38, Dialogue Derivation Unit 39, Feedback Unit 40, and Transmission Control Unit 41.

[0038] Furthermore, the storage unit 18 of the management server 1 is equipped with various databases. For example, there is a task DB 51 that stores task information, and a user DB 52 that stores user information including user interaction information, execution status information, current status information, execution history information, action history information, and support information.

[0039] The information acquisition unit 31 acquires various types of information transmitted from the generation AI server 2, the user terminal 3, and external sources. Information transmitted from the generation AI server 2 to the management server 1 includes, for example, responses. Information transmitted from the user terminal 3 to the management server 1 includes, for example, various types of input information. This input information includes, for example, task execution status information, task execution history information, user status information, and user activity history information.

[0040] The user's current status information and behavioral history information transmitted from user terminal 3 are obtained from, for example, the results of eye-tracking analysis performed on user terminal 3, the user's biometric information detected by wearable devices, the user's location information, information on the user's SNS activity, and the user's chat logs. Here, eye-tracking is a technology that uses cameras and sensors to track and record the movement and direction of the user's eyes. By using eye-tracking technology, it becomes possible to analyze what the user is looking at and what they are interested in. The user's biometric information detected by wearable devices includes, for example, body temperature, pulse rate, and blood pressure. The user's location information includes, for example, GPS (Global Positioning System) location information. Information on the user's SNS activity includes the history of posts on various SNS platforms.

[0041] Furthermore, the information acquisition unit 31 acquires dialogue information entered into the chat interface, which is managed by the chat management unit 38, described later. The information acquisition unit 31 also acquires task communication information and task information contained in the dialogue information.

[0042] The information management unit 32 stores and manages various types of information in the database of the storage unit 18. For example, the information management unit 32 stores and manages task information in the task DB 51. The information management unit 32 also stores and manages user information, including execution status information, current status information, execution history information, action history information, and support information for each user, in the user DB 52. Furthermore, when multiple users are collaborating, the information management unit 32 has a function to synchronize and reflect changes in the task status to each user's interface in real time.

[0043] The information generation unit 33 performs processing to generate various types of information. For example, the information generation unit 33 generates support information based on at least a portion of the execution status information, execution history information, current status information, and behavior history information. The information generation unit 33 also generates questions to capture the user's senses based on the dialogue information. The information generation unit 33 temporarily generates tasks from task information. The information generation unit 33 also generates a dedicated chat thread for each task currently being executed. The information generation unit 33 may also have an AI chatbot generate the above information.

[0044] The determination unit 34 determines the sufficiency of the amount of information on the user's current situation information and behavior history information. Specifically, the determination unit 34 determines whether the amount of information on the user's current situation information and behavior history information is sufficient to meet a predetermined criterion. The determination unit 34 evaluates the sufficiency of the amount of information based on, for example, the number of items of the acquired information, the data volume, the time series range, or the diversity of the information. Specifically, it evaluates conditions such as whether there is behavior history information for the past one month or more, whether there are a plurality of different types of information related to the task (for example, work history, reference materials, communication history, etc.), and whether a plurality of indicators indicating the current state of the user can be acquired as the current situation information. When it is determined that the amount of information is insufficient, the determination unit 34 instructs the information generation unit 33 to collect additional information. The information generation unit 33, for example, generates additional questions for the user and presents them on the chat interface, acquires behavior history information for a wider period, or relaxes the relevance evaluation criteria to collect more diverse information. Each time additional information is collected, the determination unit 34 determines the sufficiency of the amount of information again. This process of determination and additional information collection is repeated until it is determined that the amount of information is sufficient. Only when it is determined that the amount of information is sufficient, the task derivation process by the task proposal unit 35 is started. This avoids inappropriate task proposals based on insufficient information and enables high-quality proposals that truly match the user's situation. Note that an upper limit can be set for the number of repetitions of determination and additional information collection. For example, if the amount of information does not become sufficient even after repeating a predetermined number of times (for example, 5 times), control may be performed to derive a task based on the information available at the current time or temporarily suspend task proposals. The determination unit 34 may cause a generative AI such as an LLM to perform the determination.

[0045] The task proposal unit 35 proposes tasks that the user should currently execute. Specifically, when the determination unit 34 determines that the amounts of the user's current situation information and history information are sufficient, the task proposal unit 35 derives tasks that the user should currently execute and performs control to propose the derived tasks to the user. "Control to propose to the user" refers to control to display the task information of the tasks to be proposed to the user on the user terminal 3 as a proposal target. When proposing a task, the task proposal unit 35 can perform control to present to the user at least one of the reason why the task should be currently executed and the method for executing the task.

[0046] In addition, the task proposal unit 35 analyzes the composition of tasks in past projects similar to the current project and performs control to propose the composition of tasks suitable for the current project to the user. In addition, the task proposal unit 35 can perform control to propose task-related information and solutions to tasks to the user by referring to the history of chat threads of past similar tasks. The task proposal unit 35 may cause the AI chatbot to make various proposals.

[0047] The task registration unit 36 receives the task information temporarily generated by the information generation unit 33 and stores the task in the task DB 51 in an undetermined state. In the undetermined state, the content of the task is dynamically updated according to additional instructions and refinement in the chat. When a confirmation operation by the user is performed, the task registration unit 36 formally registers the task related to the task information as a task to be executed by the user. At this time, the status of the task is changed from undetermined to confirmed.

[0048] The task decomposition unit 37 decomposes the registered task into a plurality of subtasks. The decomposed subtasks are recorded in the task DB 51 with the hierarchical relationship with the parent task and are managed as a hierarchical structure by the information management unit 32.

[0049] The chat management unit 38 manages the chat interface. The chat management unit 38 automatically generates a dedicated chat thread for each task and centrally manages all communication related to the task within the chat thread. In addition, the chat management unit 38 has a function to detect changes in task status and important updates and automatically notify relevant parties within the chat thread.

[0050] The dialogue derivation unit 39, upon user operation, derives the dialogue history associated with a task that the user intends to execute. The dialogue derivation unit 39 makes the derived dialogue history accessible as contextual information. Specifically, the dialogue derivation unit 39 displays the dialogue history on the chat interface for visual reference, and can also use it internally as decision-making material for task execution support, or provide it via API for integration with external systems. Furthermore, the dialogue derivation unit 39 has the function of guiding the chat towards content related to the selected task and automatically associating additional dialogue related to the task with that task. The dialogue derivation unit 39 can derive the dialogue history from the dialogue history managed by the management server 1.

[0051] The feedback unit 40 provides feedback for the task derivation model to derive the next task based on at least one of the user's execution history information and behavior history information. Specifically, the feedback unit 40 provides feedback for the task derivation model to derive the next task by having the task derivation model learn additional training data including the user's history information. This feedback functions as a recursive process that evaluates the appropriateness of the execution reason (why) and execution method (how) in the previous task proposal and improves the content of the next proposal.

[0052] The transmission control unit 41 controls the transmission of various types of information via the communication unit 19 to the generation AI server 2, the user terminal 3, and external sources. For example, the transmission control unit 41 controls the transmission of a predetermined instruction (prompt) containing current status information and history information input to the task derivation model to the generation AI server 2. The instruction (prompt) input to the task derivation model is transmitted in the form of text data or audio data. The transmission control unit 41 also controls the transmission of the task derivation model's response to the user terminal 3.

[0053] <Processing Flow of Management Server 1> Figure 4 is a flowchart showing an example of the processing flow of Management Server 1 up to the point where support information is proposed to the user. Management Server 1 displays a chat interface on the user terminal 3 (Step S1). Management Server 1 interacts with the user via the chat interface (Step S2) and obtains dialogue information through that interaction (Step S3).

[0054] If the management server 1 extracts task communication information from the dialogue information (YES in step S4), it manages the extracted task communication information by associating it with the task (step S5). On the other hand, if task communication information cannot be extracted from the dialogue information (NO in step S4), the management server 1 proceeds to the decision process in step S6.

[0055] When the management server 1 acquires execution status information, execution history information, current status information, and action history information (YES in step S6), it generates support information based on the acquired execution status information, execution history information, current status information, and action history information (step S7), proposes it to the user (step S8), and terminates the process (END). On the other hand, if the management server 1 has not acquired execution status information, execution history information, current status information, and action history information (NO in step S6), it repeats the decision process in step S6.

[0056] <Specific Example> [Specific Example of this Service] Figure 5 is a diagram showing a specific example of the overall configuration of this service. Figure 5 shows the management server 1 and the generation AI server 2 that constitute the information processing system S (see Figure 1) that provides this service, and the user terminal 3 that is operated by user U. User U uses this service to perform tasks. User U, who uses this service, displays the chat interface 300 on the user terminal 3 by operating the user terminal 3.

[0057] The chat interface 300 displayed on the user terminal 3 includes user U and AI chatbot B provided by the generation AI server 2. When user U interacts with AI chatbot B via the chat interface 300, the management server 1 acquires the conversation information. AI chatbot B generates questions to capture user U's senses and incorporates these questions into the conversation with user U.

[0058] Furthermore, the management server 1 acquires current status information and history information of user U from the user terminal 3. Based on the acquired information, the management server 1 generates support information and provides it to user U via the AI ​​chatbot B. Specifically, the support information generated by the management server 1 is first provided to the AI ​​chatbot B as an instruction (prompt). Then, as a response, the AI ​​chatbot B incorporates the provided support information into its dialogue with user U.

[0059] Furthermore, when the management server 1 determines that the amount of current status information and historical information acquired for user U is sufficient, it derives a task that user U should currently perform and proposes the derived task to user U via the AI ​​chatbot B. Specifically, the task proposed by the management server 1 is provided to the AI ​​chatbot B as an instruction (prompt) along with the task information. Then, as a response, the AI ​​chatbot B proposes the task related to the provided suggestion to user U, incorporating it into the conversation with user U.

[0060] [Specific Examples of How This Service Can Be Used] The following are examples of how this service can be used: (Usage in conjunction with a PC workspace) In this usage scenario, the user's work status is observed using a camera built into or attached to the user terminal 3. The user's posture, facial expressions, and eye movements are tracked through the camera, and keyboard and mouse operation patterns are analyzed at the same time. In addition, the content of work on the screen, such as application usage and document creation status, is grasped. Based on this information, when work interruptions or a decrease in concentration are detected, appropriate breaks are suggested, and during long work periods, support information is presented to the user to encourage posture improvement. Furthermore, an efficient work sequence is suggested that takes into account the user's work efficiency.

[0061] (Usage in conjunction with mobile devices) In this usage scenario, a system will be built that works in conjunction with a smartphone or tablet as the user terminal 3. By utilizing the user's location information, the system will provide optimal task suggestions tailored to the situation, such as suggesting checking the day's schedule upon arrival at work. It will also enable efficient information processing while the user is on the move, such as voice input when the user is traveling on a train. Furthermore, biometric information will be continuously monitored through the mobile device, and tasks will be adjusted based on that information.

[0062] (Usage in conjunction with smart home systems) In this usage scenario, a system linked with IoT devices within the home is implemented. Environmental factors such as room temperature, lighting, and noise levels are continuously monitored to maintain an optimal environment. In addition, household tasks are automatically scheduled and their execution is supported. When the user is at home, the system optimizes their daily rhythm while managing their health. This achieves both the maintenance of a comfortable living environment and the management of the user's health.

[0063] (Usage as a personal assistant) In this usage scenario, for example, a small robot constantly accompanies the user. This robot provides advice and interpretation to the user, assists with household chores, and contributes to the user's safety by monitoring their health and detecting suspicious individuals. It also manages lost items and supports the user's daily life. Furthermore, it has the function to understand the user's emotional state and communicate appropriately according to the situation.

[0064] (Usage as a To-Do-Based Chat System) In this usage scenario, unlike traditional person-based chat management, the service manages chat threads on a To-Do basis. A dedicated chat thread is automatically created for each To-Do, and all communication related to the task is centrally managed within that thread.

[0065] When a user creates a To-Do, management server 1 creates a chat thread associated with that To-Do. All users involved in the task can access this thread to exchange information, ask questions, and report progress related to the task. Management server 1 also detects important updates to the To-Do, such as changes in status (from not started to in progress, in progress to completed, etc.), changes in deadlines, changes in assignees, and changes in priority, and automatically notifies these changes within the chat thread. This allows all stakeholders to understand the latest status of the task in real time.

[0066] Furthermore, communication history from similar past tasks is referenced, and relevant information and solutions are suggested. This makes it possible to leverage past knowledge and experience in performing new tasks. In addition, when cross-functional discussions spanning multiple To-Do items are necessary, the linking function between related To-Do items enables information sharing across multiple threads. This allows for the organization and management of task-specific information while maintaining the relationships between tasks.

[0067] Each chat thread features an AI chatbot that provides information, answers questions, and offers suggestions to support the progress of the task. All stakeholders involved in the task can also participate in the chat thread, enabling real-time information sharing and decision-making. This eliminates information bias and allows for comprehensive decision-making.

[0068] This solves the problem of task hierarchies, which is often overlooked in conventional systems. For example, in complex tasks and projects, tasks are often structured hierarchically, and it is necessary to properly manage the relationship between parent tasks (main tasks) and child tasks (subtasks). Conventional systems have been inadequate in managing these hierarchical structures, and information sharing between related subtasks and coordination of progress between parent and child tasks have not been efficient.

[0069] (Usage Utilizing Hierarchical Structure Management Function) To address the above issues, this service is equipped with a hierarchical structure management function. The hierarchical structure management function includes a parent-child relationship definition management function, an inter-hierarchical information transmission function, and a hierarchical access permission management function. The parent-child relationship definition management function clearly defines and visually displays the relationship between parent tasks and child tasks. According to the parent-child relationship definition management function, the completion of a parent task depends on the completion status of all child tasks, and the system automatically calculates the progress rate of the parent task based on the progress of the child tasks. For example, if there are multiple child tasks (e.g., "Requirements Definition," "Design," "Implementation," "Testing," etc.) under the parent task "Project A Completed," the overall progress rate of the parent task is automatically calculated from the progress rate of each child task. Furthermore, if any of the child tasks are delayed, a delay alert is automatically generated for the parent task, thus realizing status coordination between hierarchical levels.

[0070] In addition to the above, each task can be assigned a priority and importance level, and the impact on the parent task can be weighted according to the priority of the child task. This enables realistic project management, where, for example, delays in critical path child tasks significantly impact the progress of the parent task, while delays in tasks with less leeway have less impact.

[0071] The inter-hierarchical information transfer function enables the transfer of information between different levels of the hierarchy. According to this function, important updates and progress of child tasks are automatically notified in the parent task's chat thread. Furthermore, important decisions made in the parent task's chat thread are notified in the chat threads of all related child tasks. This ensures that necessary information is appropriately shared between project managers who have a grasp of the overall picture and practitioners who are focused on individual tasks.

[0072] For example, if a specification change is decided in the "Design" subtask, the details of that change will be automatically notified in the chat thread of the parent task "Project A Completed," allowing for consideration of the impact on the entire project. Also, if a deadline change is decided in the parent task's chat, that information will be notified in the chats of all subtasks, allowing each person in charge to adjust their work plan.

[0073] When automating information transmission, it's not simply a mechanical transfer; AI-based importance assessment is performed. For example, minor progress reports are not notified to the parent task, but risky changes or major problems are immediately notified to higher levels, ensuring appropriate transmission according to the importance of the information.

[0074] The hierarchical access control function manages access permissions according to each hierarchical level. With this function, project managers can view and edit tasks at all levels, while specific task managers can be restricted to access only the subtasks assigned to them. This improves information security and creates an environment where each person can focus on their area of ​​responsibility.

[0075] For example, access to parent tasks containing confidential information for the entire project can be restricted to managers only, while departmental staff can access only the information of their own department's child tasks. Furthermore, by linking organizational hierarchy with permissions, it's possible to set it up so that department heads can view all tasks in their department, while general employees can only view and edit tasks assigned to them.

[0076] This access restriction is linked not only to the ability to view and edit task information, but also to the permission to participate in chat threads. This ensures that only the appropriate stakeholders participate in communication, balancing information security with communication efficiency. These features make it easier to break down and manage complex tasks, improving the efficiency of large-scale project management.

[0077] (Usage as an AI and stakeholder-participatory chat system) A challenge with conventional chat systems is the lack of a mechanism for AI and all stakeholders to collaborate effectively in task-related communication. To address this challenge, this service is equipped with an AI participation function, an automatic stakeholder invitation function, and an optimal team formation suggestion function.

[0078] The AI ​​participation feature allows an AI chatbot to participate in the chat threads for each task. The AI ​​chatbot not only provides information but also offers advice based on the task's progress, automatically collects and presents relevant information, and suggests execution methods. For example, in a chat thread for the "budget formulation" task, the AI ​​chatbot automatically analyzes budget data from similar past projects and presents it as reference information, and instantly performs budget simulations based on figures discussed and shares the results. Furthermore, if a question is raised in the chat, the AI ​​chatbot automatically suggests possible answers, thus supporting human intellectual work.

[0079] Furthermore, AI chatbots can analyze discussions within chats and provide perspectives that humans might miss, such as pointing out important details that have been overlooked or risks that need to be considered. For example, they can promote multifaceted consideration by presenting examples of failures from similar past projects for a proposal that is gaining consensus within the team.

[0080] The automatic stakeholder invitation feature automatically invites all stakeholders involved in each task (e.g., project managers, administrators, and relevant department representatives) to a chat, enabling real-time information sharing and decision-making. For example, when a task called "System Integration Testing" is created, the system automatically identifies the relevant system managers, testers, and quality control department representatives and invites them to the chat. This ensures that all necessary stakeholders are included in the information sharing circle, preventing situations where someone is later overlooked.

[0081] Identifying stakeholders is based not only on static information such as organizational charts and project structure diagrams, but also on multiple dynamic factors. Specifically, for example, management server 1 makes complex decisions such as analyzing the history of participants in similar past tasks to identify personnel with a proven track record in similar tasks, evaluating the actual expertise derived from the skills and expertise information registered in each user's profile and their past work history, and determining the required areas of expertise by analyzing the task content (title, description, related keywords, etc.) using natural language processing. This makes it possible to appropriately invite personnel with actual stakeholders and expertise that cannot be identified by formal job titles or affiliations alone. For example, if a task called "Database Optimization" is created, not only the database personnel listed in the organizational chart will be automatically invited, but also developers who have achieved high results in database-related tasks in the past, and personnel who have registered database tuning skills in their profiles.

[0082] The optimal team composition suggestion function analyzes the statements and behavioral patterns of stakeholders and proposes the most suitable combination of stakeholders for similar tasks. This enables efficient team formation based on past success stories. For example, by analyzing past project data, it can extract insights such as "In this type of development project, the team of A, B, and C is the most efficient and produces the best results," and propose the optimal team composition when launching a new project. It also considers each member's strengths, compatibility, and current workload to achieve a well-balanced team structure.

[0083] Furthermore, by continuously analyzing communication patterns within the chat, it's possible to detect issues such as "insufficient information sharing among these members" or "delays in responding to questions from this person." Based on these analysis results, specific suggestions for improving communication can be made, contributing to improved team performance.

[0084] These features eliminate information bias and enable comprehensive decision-making. Furthermore, the participation of appropriate stakeholders improves the efficiency and quality of task execution. While these usage patterns can be implemented individually, combining multiple patterns allows for more effective support. Each usage pattern provides timely assistance tailored to the user's situation, improving their quality of life. Additionally, data collected through each usage pattern is used to improve machine learning models while respecting privacy. This is expected to improve overall system performance and enable more appropriate support.

[0085] According to the above usage scenarios, it becomes possible to propose tasks and / or actions to users at the optimal timing according to their situation, and to execute those tasks and / or actions. Furthermore, efficient task and / or action management that takes into account the user's physical and mental state can be achieved. In addition, appropriate task and / or action proposals and task and / or action execution based on past performance data can be achieved. Moreover, by managing chat threads on a To-Do basis, information can be aggregated for each task, improving the searchability and reusability of related information. Furthermore, in collaborative task work among multiple users, efficient communication on a To-Do basis becomes possible, improving the productivity of team work. Furthermore, by utilizing the hierarchical structure of tasks, the breakdown and management of complex tasks becomes easier, improving the efficiency of large-scale project management. In addition, a chat system in which AI and all stakeholders participate eliminates information bias, enabling comprehensive decision-making.

[0086] (Usage Mode Integrating Chat Interface and Task Management Function) In addition to the usage modes described above, this service allows for usage modes that integrate the chat interface and task management function. The following describes specific examples of the user interface configuration, processing flow, and technical implementation in this usage mode.

[0087] (User Interface Configuration) The user interface for this usage scenario employs a configuration that links a chat interface and a task management interface. These interfaces share information with each other, allowing users to perform chat communication and task management in an integrated manner. In the chat interface, the conversation history is displayed in chronological order, similar to conventional messaging applications, and it is possible to handle not only text but also multimedia content such as images and charts. In the task management interface, task information extracted from chat content is structured and managed, and attribute information such as the status, priority, and deadline of each task is also managed. Task management allows for organization by category or project, and management in a hierarchical structure, enabling management according to the status of the task.

[0088] (Processing Flow) This usage mode consists of six main processing flows: a task generation process, a task management process, a task decomposition process, a state retention function, a dialogue derivation process, and a task suggestion process. These processes work together to realize an integrated environment for chat and task management. Each processing flow is described below.

[0089] (Task Generation Process) The system analyzes user instructions and questions entered through the chat interface using natural language processing technology. If task-related content is detected, the system automatically generates a temporary action To-Do. For example, in response to the input "I need to compile a market research report by next week," the task "Create Market Research Report" is generated. At this stage, the task is in a provisional state and is displayed as a temporary task in the task management interface.

[0090] Subsequently, the task details in the temporary status are dynamically updated in response to additional instructions and details provided within the chat. For example, if additional information such as "It should also include competitor analysis" is entered, the element "Include competitor analysis" is added to the task details. The user reviews the provisional task details presented in the task management interface, edits them as needed, and then confirms them. Once confirmed, the task is registered as an official task and changes to a state where it is officially managed within the system.

[0091] This process naturally generates tasks within the flow of conversation, and the necessary information is gradually gathered. Particularly important is that task generation occurs automatically without explicit user instruction, thereby reducing cognitive load such as interruptions or switching to other screens.

[0092] (Task Management Process) From the task list displayed in the task management interface, users can select a specific task and change its status. In addition to basic statuses such as "Not Started," "In Progress," "Completed," and "On Hold," custom statuses can also be defined according to the project or workflow. Status changes can be made using the system's provided methods, and a change history is also maintained.

[0093] When a task is selected, the chat history related to that task is displayed in the chat interface to help understand the context. Additional chats related to the selected task can also be initiated, and their content is automatically associated with the task. For example, if a specific task is selected and the user enters "I want to extend the deadline for this task by one week," the system will update the task's deadline attribute.

[0094] Task status changes and property updates are also displayed as notifications in the chat interface. This allows multiple users to share changes in task status in real time, even when multiple users are participating. The notification display format can be adjusted in the settings, and the display can be differentiated according to importance.

[0095] (Task Decomposition Process) For complex tasks and long-term projects, the system provides a function to break them down into multiple smaller tasks (subtasks). Decomposition can be done manually, but system assistance can also be utilized. For example, in response to the instruction "Please divide this project into major stages," the system will suggest appropriate subtask candidates based on patterns from similar past projects and general project management methodologies.

[0096] When a task is broken down, the chat interface displays an explanation of the breakdown and the reasons for it, while the task management interface displays each of the broken-down subtasks. The relationship between parent tasks and subtasks is managed, and dependencies and order between subtasks can also be defined. For example, it is possible to set conditions such as "this subtask can only be started after the previous subtask is completed."

[0097] Each subtask is managed independently and assigned to a different person. The progress of the parent task is automatically calculated and updated based on the completion status of its subtasks. For example, if 50% of all subtasks are completed, the parent task's progress will also be displayed as 50%. This hierarchical task management makes it easier to grasp both the big picture and the details, even in complex projects. A task breakdown template function is also provided, allowing users to reuse past breakdown patterns for routine projects and workflows. This improves work efficiency during project initiation and prevents important steps from being overlooked.

[0098] (State Retention Function) The execution status of each task is continuously maintained and managed persistently regardless of the flow of the conversation. In conventional chat systems, past content gets buried as the conversation progresses, making it difficult to track the status of mentioned tasks. However, in this service, task information is managed independently of the conversation flow. State retention includes basic task attributes (name, description, due date, priority, assignee, etc.), as well as a history of status changes, references to related chat messages, attachments, and comments. This information is persisted in the database and retained across sessions. Even if a user logs back into the system, the previous state is restored, allowing for continued work.

[0099] Furthermore, the state retention feature supports synchronization across multiple devices, accommodating scenarios such as continuing desktop work on a mobile device. Synchronization occurs in real time, with changes made on one device instantly reflected on others. This persistence and synchronization of state enables continuous task management that transcends temporal and spatial constraints. In addition, regular automatic backups minimize the risk of data loss due to system failures or user errors. A restore function from backups is also provided, allowing users to return to a specific point in time.

[0100] (Dialogue Derivation Process) Each task is automatically associated with and recorded in a database the dialogues that have taken place in relation to it from its creation to the present. When a user selects a task, the system automatically searches the past dialogue history related to that task and presents it in the chat interface. The derived dialogues include the background of the task's creation, the details of the requirements, the reasons for status changes, the problems that occurred and their solutions, and the agreements reached by the stakeholders. This allows, for example, if the person in charge changes midway through a long-term project, the new person in charge can easily understand the past history.

[0101] Furthermore, dialogue derivation can be filtered based on the task's execution status and attributes, allowing for the extraction of dialogues relevant to a specific context, such as "Why was the deadline for this task extended?". This feature significantly reduces information retrieval time and improves work continuity. The dialogue derivation function not only displays past dialogues but also guides users appropriately into new dialogues based on them. For example, it can extract unresolved issues and necessary next steps from past dialogues and suggest them as starting points for new dialogues, supporting the efficient progress of tasks.

[0102] (Task Proposal Process) This service accumulates and analyzes data from similar projects executed in the past, including task structure, progress patterns, success stories, and failure stories. Based on this data, it proposes appropriate tasks according to the current status of the project. For example, it proposes a task structure applicable to the current project based on information such as the types of tasks typically required for projects of a particular industry or size, their typical sequential relationships, and the time required. It can also predict the next tasks to be tackled and potentially high-risk processes based on the current project progress, and propose necessary countermeasures in advance.

[0103] Furthermore, it facilitates the sharing of knowledge and experience across multiple projects within the organization. Effective work procedures and solutions from one project are proposed to other projects facing similar challenges, establishing a learning and improvement cycle across the entire organization. Proposed tasks are adopted or modified at the user's discretion, and if adopted, they are registered as official tasks. In addition, the direction of related chats is guided based on task proposals, supporting efficient problem-solving and decision-making.

[0104] (Technical Implementation) The technical implementation of this usage scenario combines the latest natural language processing technology with web application technology. In the backend system, an intent understanding engine utilizing a large-scale language model extracts task-related information from user input and distributes it to the appropriate processing. This intent understanding engine is pre-trained with a large number of task instruction patterns and has the flexibility to handle various expression forms and language styles.

[0105] The data management layer employs a hybrid configuration combining relational and NoSQL databases to efficiently manage both structured task information and unstructured chat text. The data model prioritizes scalability, designed to facilitate future feature additions and integration with other systems. The user interface implementation is designed to be compatible with various platforms and devices, emphasizing seamless integration between the chat interface and the task management interface.

[0106] The WebSocket protocol is used for inter-system communication, enabling real-time status updates and notifications. This ensures consistency and immediacy of the system's status even when multiple users are accessing it simultaneously. Furthermore, an efficient differential synchronization algorithm is implemented for data synchronization during offline work, optimizing network bandwidth usage. In terms of security, end-to-end encryption technology is employed to protect highly confidential business information and project details. Additionally, granular access control features allow for setting view and edit permissions for each task and project.

[0107] This usage model is expected to yield the following significant benefits. These benefits have been confirmed not only through theoretical predictions but also through empirical experiments using a prototype system. Firstly, the task-centric structure of the chat prevents conversations from diverging. In conventional chat systems, conversations often deviated from their original purpose as they progressed, but in this usage model, tasks and their progress are constantly managed, thus maintaining a goal-oriented conversation. In the empirical experiment, a comparison between conventional chat and task-integrated chat for the same task showed that the latter maintained a clearer focus in the conversation, resulting in an average 32% reduction in the time required to achieve the goal. This effect was particularly pronounced in meetings and planning sessions involving multiple participants.

[0108] Secondly, the real-time management of task status improves the efficiency of progress management. In traditional project management, separate meetings and report creation were required to check progress, but with this usage model, task status is always managed in an up-to-date state, allowing managers to immediately grasp the overall situation. Furthermore, delays and issues that need to be addressed as a priority are clearly indicated, making it easier to optimize resource allocation. Evaluations with multiple project managers have reported that the time required for situation assessment was reduced by up to 78% compared to conventional methods.

[0109] Thirdly, the integration of chat and task management allows work to proceed without losing context. Previously, task management and conversation had to be conducted using separate systems, resulting in cognitive load due to the need to switch contexts. However, this usage model provides a seamless environment. As a result, users can concentrate on deeper thinking and creative activities, and 87% of experiment participants reported an improvement in their "flow state" (a psychological state of optimal concentration and immersion) when using the system. In addition, the time it took to recover from work interruptions was reduced by an average of 45%, confirming improved interruption tolerance.

[0110] Fourth, the efficiency of large-scale project management is improved by enabling the decomposition of complex tasks and the management of each step in an integrated environment. Traditional project management tools often lacked coordination between detailed task decomposition and daily communication, leading to inconsistencies and communication errors. In this usage model, task decomposition and communication are organically integrated, allowing for immediate sharing of the execution status and issues of each subtask, enabling appropriate responses. A case study targeting complex product development projects reported a 56% improvement in the early detection rate of quality problems and a 23% reduction in man-hour losses due to rework compared to conventional methods.

[0111] Fifth, the system improves the efficiency of information access through the appropriate derivation of task-related conversations. In conventional systems, it was necessary to search through a vast amount of conversation history to find past decisions and background information related to a task. However, in this usage mode, tasks and related conversations are automatically associated, allowing for immediate access to the necessary information. As a result, information retrieval time has been reduced by an average of 65%, and a significant improvement in efficiency has been observed, especially in complex projects and long-term tasks. It has also been reported that the rate of overlooking relevant information has decreased by 43%, and the quality of decision-making has improved.

[0112] Sixth, the task suggestion function, which utilizes past project information, has the effect of improving the efficiency of knowledge utilization. Previously, knowledge sharing relied on individual experience and was limited to a small scope, making it difficult for learning and improvement to progress throughout the organization. In this usage model, success and failure cases of similar past projects are analyzed, and task configurations suitable for the current situation are proposed, thereby promoting organizational knowledge utilization. In empirical studies, it has been confirmed that this function reduces the time required to start up new projects by an average of 28%, and the rate of avoiding typical failure patterns has improved by 67%.

[0113] Seventh, the persistent retention of task status makes it easier to resume work after interruptions and facilitates collaboration among multiple users. In traditional chat, it was necessary to search for relevant information from past conversations, but in this usage mode, task status is structured and retained, allowing users to immediately resume work from where they left off. In collaboration among multiple users, the status of each person's assigned task is clearly displayed, preventing duplication and omissions of work. An analysis of a distributed development project by a multinational team showed that the information sharing completeness score improved by 34%, and the accuracy of handing over tasks due to time differences improved by 62%.

[0114] Eighth, the automatic recording and referencing of long-term project history is expected to promote organizational knowledge management and learning. Not only the completion status of tasks, but also the reasoning behind decisions, know-how, challenges encountered, and solutions are saved along with the conversation history, making it valuable reference information for similar projects in the future. In particular, the reduction of training costs when new members join and the improvement of the reliability of knowledge transfer have been confirmed through pilot implementations in multiple organizations. As described above, this usage model is not merely an improvement on the chat system, but provides an innovative communication platform with the potential to bring about fundamental efficiency improvements and quality enhancements in business execution.

[0115] (Example 1) As a specific example of this usage, a scenario for preparing a patent application at a patent office will be described in detail with reference to Figures 6 to 14. Figures 6 to 14 show specific examples of the user interface displayed on the user terminal 3 in Figure 1. In this example, the process by which an inventor, a patent attorney, and patent office staff use this service to efficiently prepare patent application documents is shown.

[0116] First, as shown in Figure 6, the system presents a screen with the question, "What are you currently interested in, or what would you like to improve?" As shown in Figure 7, the inventor enters into the chat interface, "I would like to file a patent application for a todo-based chat system." This service analyzes this input and displays the points that need to be organized in the chat interface (left screen in Figure 8), such as "1. Identifying and summarizing the idea, 2. Market research, 3. Detailing the technical aspects."

[0117] This automatically generates temporary tasks for patent application preparation (a ToDo-based chat system). These tasks are assigned an unconfirmed status and are managed as temporary tasks. For example, as shown in the display screen on the right side of Figure 9, tasks such as patent search and technical document review, preparation of patent application documents, consultation with a patent attorney or patent expert, estimation and preparation of patent application fees and related costs, and submission and confirmation of receipt of the patent application are presented.

[0118] Temporarily generated tasks are refined during subsequent conversations. As the conversation progresses in the chat, additional information related to the task may be mentioned. For example, if a user asks, "Shouldn't we also organize the system's features?", the system analyzes this information and adds new task elements, such as "Organize the system's uniqueness and innovativeness," to the existing temporary task. Also, if detailed requirements, deadlines, assignees, and priorities for a task become clear during the conversation, this information is automatically reflected in the task content.

[0119] In this way, the content of the task is dynamically updated according to the flow of the conversation. The task registration unit 36 ​​sequentially updates the content of unconfirmed tasks based on the additional information extracted from the information generation unit 33. As a result, the user does not have to explicitly input task information, and the details of the tasks are automatically enriched through natural dialogue.

[0120] Users can view a list of temporary tasks displayed in the task management interface and edit the details of each task. For example, they can change the task name, set a deadline, and assign a person to complete it. Unconfirmed tasks are visually distinguished from regular tasks by a different display (e.g., a different background color or the addition of an "Unconfirmed" label).

[0121] Once editing is complete, the user can click the "Confirm" button to register the task as an official task. Confirmed tasks are then managed as regular tasks and are subject to progress tracking and subtask breakdown. Users can also delete unconfirmed tasks if they deem them unnecessary. Furthermore, if an unconfirmed task is not checked for a certain period, the system may automatically delete it.

[0122] Next, the user asks, as shown in the left screen of Figure 10, "Shouldn't we start by organizing the system's features?" In response, the system proposes tasks such as "organizing the system's uniqueness and innovativeness, conducting competitive and market research, documenting technical specifications and architecture, developing and testing prototypes, and formulating a strategy for patent applications," as shown in the right screen of Figure 11.

[0123] The dialogue continues, and the user clicks "Organize the system's uniqueness and innovativeness," as shown in the right-hand screen of Figure 12. In response, the system automatically suggests breaking down the task into subtasks. For example, as shown in the right-hand screen of Figure 13, it suggests "Identifying features that can differentiate the system in the market, documenting future innovation points, comparative analysis with competing technologies, collecting user feedback, and evaluating the system's future technological trends."

[0124] Each subtask is assigned an assignee and a deadline, and its status can be selected from options such as not started, in progress, awaiting review, or completed. For example, a comparative analysis of competing technologies is assigned to a patent search specialist and set as a deadline one week from now. When this task is changed to "in progress," a notification will appear in the chat interface, although not illustrated, such as, "Comparative analysis of competing technologies task started (Assigned to: XX, Deadline: May 15th)." This report is automatically associated with the corresponding task, and selecting the task later will display all related conversations in chronological order.

[0125] When a task is completed, its status changes to "Completed." Then, as shown in the right-hand screen of Figure 14, a strikethrough is added to the task. This also automatically updates the progress rate of the parent task, "Organizing the Uniqueness and Innovativeness of the System" (for example, 16.7% complete). In the same way, each subtask progresses sequentially, and when all subtasks are completed, the parent task also becomes "Completed."

[0126] Throughout this entire process, the current state of a task and its associated conversation history are always maintained in a linked manner. For example, if an inventor asks "What were the differences from the prior art?" several weeks later, the relevant task and associated conversation can be immediately referenced. Furthermore, if similar patent applications arise in the future, past task decomposition patterns can be reused as templates.

[0127] Thus, this service enables efficient and high-quality patent application work by managing the complex and multi-stage process of patent application in a structured manner while maintaining a natural flow of conversation.

[0128] (Example 2) As a second specific example of this usage, a project management scenario will be described. Specifically, a management scenario for a new mobile application development project at a software development company will be described. This scenario shows the process of managing a complex project involving members with diverse roles, such as project managers, designers, developers, and QA engineers.

[0129] At the start of the project, the project manager announces via the chat interface, "We are starting a development project for the health management app 'HealthTracker.' Its main functions will be activity tracking, meal tracking, sleep analysis, and goal setting." The system analyzes this input and automatically generates a project task called "HealthTracker App Development," which is displayed in the task management interface.

[0130] Next, the entire team discusses the project details. The product manager shares the target user base and points of differentiation from competing products, and the technical leader proposes an outline of the architecture. This information is automatically accumulated as detailed project task information. After sufficient discussion, the project manager finalizes the tasks and instructs the system to "break down this project into key phases."

[0131] The system proposes the following phase divisions, referencing similar past project patterns: (1) Requirements definition and planning, (2) UI / UX design, (3) Frontend development, (4) Backend development, (5) Testing and quality assurance, and (6) Release preparation. Each phase is further broken down into specific tasks; for example, the "UI / UX design" phase is divided into subtasks such as "User persona creation," "Wireframe design," "Mockup creation," and "Usability testing."

[0132] The project manager reviews the proposed breakdown, makes any necessary adjustments, and then assigns a person and deadline to each task. For example, the "Wireframe Design" task is assigned to UI designer Suzuki, with a deadline of two weeks from now. Once the assignments are complete, the relevant team members receive notifications, and their assigned tasks become visible.

[0133] As development progresses, each member updates the status of their assigned tasks and shares progress and issues via chat. For example, if backend developer XX reports, "We encountered a performance issue when processing large amounts of data while implementing the data synchronization function," it is automatically associated with the relevant task. Based on this report, the technical leader proposes countermeasures, and a new subtask, "Performance Optimization," is added as needed.

[0134] In the daily morning meeting, the overall progress is shared, and tasks requiring special attention (those with approaching deadlines, those with reported problems, etc.) are clearly identified. Each member briefly reports the status of their assigned tasks, and this information is linked to the relevant tasks. Even if there are changes in team members during the project, new members can smoothly join by referring to past conversation history and task status. For example, when a developer who was on sick leave returns, they can immediately check the progress and decisions made during their absence simply by selecting their assigned tasks.

[0135] If a bug is discovered during the testing phase, the QA engineer can create a "bug fix" task and assign it to the relevant developer. Bug details and reproduction steps are shared via chat and automatically associated with the task. After the fix is ​​completed, the developer changes the task status to "fixed," and after retesting by the QA engineer, it becomes "resolved." Upon project completion, all task history and related conversations are saved in a structured format, serving as valuable reference material for similar projects in the future. Furthermore, project retrospective meetings can easily identify areas for process improvement based on analysis of task duration and state transitions.

[0136] In this way, this service plays a role in streamlining team collaboration and information sharing across complex project management environments by integrating structured task management with natural communication, thereby increasing the probability of project success.

[0137] (Example 3) As a specific example of this usage, we will describe a scenario for utilizing task suggestion and conversation derivation. Specifically, we will describe a scenario for utilizing the task suggestion and conversation derivation functions in a software product design review meeting. In this scenario, product managers, designers, developers, testers, etc., participate in the process of reviewing the design of new functions.

[0138] At the start of the meeting, the product manager types "Starting the design review of the remote access feature" into the chat interface. The system automatically generates a task called "Remote Access Feature Design Review" and, after analyzing patterns from similar past review meetings, suggests the following subtasks: (1) Requirements Confirmation, (2) Security Requirements Review, (3) Performance Requirements Review, (4) UI / UX Review, and (5) Implementation Plan Confirmation.

[0139] Of particular note is that the system also identifies issues that frequently occurred in similar past review meetings, and in relation to the "Security Requirements Review" task, it provides a note stating, "In past similar functions, deciding on the authentication method took a long time. We recommend making a decision early." The product manager reviews this suggestion, makes adjustments as needed, and then adopts it as part of the meeting agenda.

[0140] When a meeting begins with the "Requirements Confirmation" subtask, the system automatically searches for past conversations and documents related to this task and retrieves relevant conversation history. For example, it can refer to decisions made at a requirements definition meeting three weeks ago, such as "remote access must be linked to the company VPN," or requests made at a customer interview two months ago, such as "we want support for access from mobile devices."

[0141] Next, when moving to the "Security Requirements Review," the system automatically retrieves and makes available the discussions about authentication methods from similar past feature developments. For example, it displays a past decision such as "Project X adopted a combination of OAuth 2.0 and multi-factor authentication," along with the reasons for that decision. This allows the team to make decisions efficiently by leveraging past knowledge without having to repeat the same discussions.

[0142] If new issues are identified during the review process, for example, if a requirement such as "biometric authentication support is needed on mobile devices" is added, the system will detect this and suggest a new task candidate: "Biometric authentication implementation." Furthermore, based on past development of similar features, the system will also provide information indicating that this task typically requires about two weeks of development time and that the involvement of specific security experts is desirable.

[0143] At the end of the meeting, based on the discussions, the system automatically proposes a task structure for "implementing remote access functionality," presenting standard estimated times and priority levels for each task. This facilitates smooth implementation planning and enables realistic scheduling.

[0144] Later, if the development team has questions about a specific task while implementing it, simply selecting the "Biometric Authentication Implementation" task will retrieve all relevant discussions and decisions from the review meeting, making it easy to review them. Furthermore, when the project is completed three months later, the actual time taken and any issues encountered will be recorded for each task, providing valuable reference data for similar projects in the future.

[0145] This scenario demonstrates that the task suggestion function based on past experience and the efficient derivation of task-related conversations significantly improve the efficiency of review meetings, ensure information continuity, and promote the accumulation and utilization of organizational knowledge.

[0146] <Advantageous Effects of This Embodiment> According to the information processing system S including the management server 1 of the above embodiment, it becomes possible to propose tasks and / or actions according to the user's current situation at an appropriate time through communication with the user. In other words, the information processing system S is a system that organically integrates two essential business functions, communication and task management, in today's rapidly digitizing business environment. The following effects are realized by the information processing system S.

[0147] Firstly, the management server 1 enables task extraction and structuring using natural language processing technology. In conventional task management systems, users had to explicitly input task information, but the information processing system S can automatically extract and structure task-related information from everyday conversations. This reduces the cognitive load on the user while enabling systematic task management.

[0148] Secondly, the management server 1 enables real-time synchronization and state persistence. The information processing system S utilizes distributed system technology to achieve both immediate sharing of task states among multiple users and long-term state retention. In particular, managing the state and synchronization control of a task tree with a complex hierarchical structure is a technically advanced challenge, and its efficient implementation represents an important technical contribution of the present invention.

[0149] Thirdly, the management server 1 enables the establishment of techniques for associating tasks with conversations and deriving them. The technology to appropriately associate conversations related to each task and accurately derive them when needed is valuable as an effective information management method in the age of information overload. In particular, the deriving of related conversation history linked to task selection is advanced in that it makes it possible to instantly refer to highly relevant parts from a large amount of information.

[0150] Fourthly, the management server 1 can analyze past project information and propose tasks. The technology for structuring knowledge and know-how within an organization and utilizing it appropriately according to the current situation contributes to promoting and streamlining organizational learning. From a marketability standpoint, it has a wide range of applicability, as follows: Firstly, in the field of corporate project management, integrating the functions of conventional specialized project management tools and chat tools can result in a more intuitive yet highly functional solution. In particular, in the current work environment where remote work and hybrid work have become commonplace, there is a high demand for a tool that streamlines collaboration among distributed teams.

[0151] Furthermore, by including information that is not explicitly managed as a task in the user's current status information and activity history information, the following effects are achieved by the management server 1 of the above embodiment. Firstly, by utilizing the rich user activity history information, it becomes possible to provide support that deeply understands the user's work context. For example, the user's activity history information includes the specific content of learning materials created in the past, detailed interactions in preparing presentation materials within the company, the content of statements made in meetings, and the history of referenced literature and materials. This information not only shows the execution status and execution history of tasks, but also shows what kind of content the user has worked on and what kind of thought process they have gone through. By analyzing this rich user activity history information, the management server 1 can grasp the user's expertise, areas of interest, and work style, and generate support information that is deeply suited to the user's context, going beyond mere task progress management.

[0152] Secondly, by utilizing user status information, adaptive support tailored to the user's current state becomes possible. For example, if the user's status information indicates that the user is currently thinking about a particular topic, tasks and information related to that topic can be prioritized and suggested. Also, if it is determined that the user is relaxed, tasks that can be started with a light burden can be suggested, while if the user is focused, highly important and complex tasks can be suggested. This allows for suggestions tailored to the user's psychological state and motivation. Through such adaptive support, users can work on tasks at a comfortable pace that suits their own state, resulting in improved work efficiency and continuity.

[0153] Thirdly, even when motivation is low, appropriate support can encourage a return to work. Conventional task management systems present tasks uniformly regardless of the user's psychological state, so when motivation is low, users may ignore the presented tasks, and as a result, stop using the system altogether. In this embodiment, when a decrease in the user's motivation is detected from the user's current status information, tasks that can be tackled with little burden or topics that can boost motivation are first suggested. For example, by presenting related information on areas the user has shown interest in in the past, or suggesting simple tasks that provide a small sense of accomplishment, it is possible to support a gradual transition to more serious work.

[0154] Fourthly, by including behavioral information that is not managed as a task, it becomes possible to comprehensively understand the user's actual activities and provide more realistic support. For example, if a user is continuously learning using a linked learning application, suggesting tasks related to that learning content enables effective support that integrates learning and practice. Also, by understanding meetings and appointments from calendar information, it becomes possible to suggest tasks that take into account the schedule before and after those events. In this way, comprehensive information utilization that goes beyond the framework of task management provides truly useful support that is tailored to the context of the user's actual life and work. Thus, by including diverse information in the user's current status information and behavioral history information, and by creating a system that comprehensively analyzes and utilizes this information, it becomes possible to provide context-adaptive, flexible, and effective support that was not possible with conventional task management systems.

[0155] As described in the above embodiment, the fact that the management server 1 determines the sufficiency of the amount of current status information and behavioral history information before deriving tasks has the following significance. Firstly, proposing tasks based on insufficient information may result in inappropriate suggestions that do not fit the user's actual situation or context. For example, if a complex task is proposed without a sufficient understanding of the user's area of ​​expertise or past work history, it is highly likely that the task will be difficult for the user to perform or will be of no interest to them. If the management server 1 determines that the amount of information is insufficient, it will actively collect additional information. Specifically, it will expand the period it goes back in time to acquire more behavioral history information, or adjust the relevance evaluation criteria to collect a wider range (more) of current status information and behavioral history information. By collecting such additional information, it becomes possible to propose tasks that are feasible and meaningful to the user.

[0156] Secondly, by collecting a wide range of relevant information, it becomes possible to derive tasks that are more contextually appropriate. For example, by broadly collecting the user's past behavior history on similar content or related themes, it becomes possible to more accurately understand the user's work context and direction of thinking. Management server 1 can derive tasks that are truly suited to the user's current situation by considering not only the quantity of information but also the relevance and diversity of the collected information.

[0157] Thirdly, it enables the efficient use of system resources. Since the management server 1 executes task derivation processing only when there is sufficient information, it can avoid executing meaningless processing based on insufficient information. Furthermore, by dynamically adjusting the scope and depth of information collection, it is possible to acquire the necessary information without excess or deficiency, thereby improving the overall efficiency of the system. In this way, a mechanism that determines the sufficiency of the amount of information in the user's current status information and behavioral history information, and proactively collects additional information as needed, can achieve both improved quality of proposals and efficient system operation.

[0158] Fourth, the quality of task suggestions continuously improves through a recursive feedback mechanism based on execution history information and behavior history information. When the management server 1 suggests a task, it presents not only what to do (what), but also why (why) and how to do it (how). The user then evaluates whether these what / why / how were appropriate based on the results of the task execution and feeds this back into the next task derivation model. Through this recursive refinement, the system learns the user's preferences and work style, gradually improving the accuracy of its suggestions. For example, if a user tends to prefer a particular method of execution, the system will prioritize suggesting that method in subsequent attempts. Also, if the reasoning for a suggestion is insufficient, it will be improved to provide more persuasive reasons. Through such continuous improvement, a truly useful and personalized support system for the user will be realized in the long term.

[0159] Furthermore, in the field of education, integrating learner task management and communication can provide an effective learning support environment. Teachers and instructors can provide appropriate guidance while monitoring learners' progress in real time, and learners themselves can structure and manage their own learning process. In the medical and healthcare field, integrating patient care workflow management and communication among medical staff has the potential to improve the quality and safety of the treatment process. Structuring individual patient care plans as tasks and continuously tracking their implementation can also contribute to preventing medical errors. Moreover, in the personal market, it can provide value as a tool that integrates daily task management with communication with family and friends. In particular, it enables efficient information sharing and progress management in areas such as role-sharing within the family and collaborative projects (e.g., moving preparations, travel planning). As described above, this embodiment possesses technological advancement while also having the potential to provide practical value to diverse markets, and can be evaluated as having significant business potential.

[0160] <Other> Although one embodiment of the present invention has been described above, the present invention is not limited to the above-described embodiment, and modifications, improvements, etc., within the scope that can achieve the objective of the present invention are included in the present invention. For example, the service shown in Figure 5 is merely one example of a service to which the present invention can be applied, and is not particularly limited. Similarly, the user interfaces shown in Figures 6 to 14 are merely examples, and are not particularly limited.

[0161] Furthermore, the series of processes described above can be executed by hardware or by software. In other words, the functional configuration described above is merely illustrative and not particularly limiting. That is, it is sufficient for the information processing system to have the functionality to execute the series of processes described above as a whole, and the functional blocks used to realize this functionality are not particularly limited to the examples given above.

[0162] Furthermore, the location of the functional blocks is not particularly limited and can be arbitrary. For example, the functional blocks of management server 1 may be transferred to other devices, or the functional blocks of other devices may be transferred to servers. Also, a single functional block may consist of hardware alone, software alone, or a combination of both.

[0163] When a series of processes are executed by software, the programs that make up that software are installed on a computer or other device from a network or storage medium. The computer may be a computer built into dedicated hardware. Alternatively, the computer may be a computer capable of performing various functions by installing various programs, such as a server, a general-purpose smartphone, or a personal computer.

[0164] Such recording media containing programs may consist not only of removable media (not shown) distributed separately from the main device to provide programs to users, but also of recording media provided to users in a state where they are pre-installed in the main device. Since programs can be distributed via a network, the recording media may be installed on or accessible from a computer connected to or capable of connecting to a network.

[0165] In this specification, the step of describing a program to be recorded on a recording medium includes not only processes that are performed chronologically in that order, but also processes that are not necessarily performed chronologically, but are executed in parallel or individually. Furthermore, in this specification, the term "system" refers to an overall system composed of multiple devices, means, etc.

[0166] In other words, the information processing device to which the present invention is applied can take various forms having the following configurations: (1) That is, a management server 1, which is an example of an information processing device to which the present invention is applied, is an information processing device comprising: task storage means for storing tasks (for example, the information management unit 32 in Figure 3); hierarchical structure management means for managing the hierarchical structure of the tasks (for example, the information management unit 32 in Figure 3); and communication management means for managing task communication information related to the tasks in association with those tasks (for example, the information management unit 32 in Figure 3).

[0167] (2) The system may also be further equipped with execution status management means (for example, the information management unit 32 in Figure 3) for managing execution status information regarding the current execution status of the task.

[0168] (3) The system may also be further equipped with execution history management means (for example, the information management unit 32 in Figure 3) for managing execution history information relating to the past execution history of the task.

[0169] (4) The system may further include support information generation means (for example, the information generation unit 33 in Figure 3) that generates support information to assist the user performing the task based on the execution status information or the execution history information.

[0170] (5) The system may further include a user status acquisition means (for example, the information acquisition unit 31 in Figure 3) that acquires at least one of the current status information relating to the user's current situation and the behavioral history information relating to the history of the user's actions, and a task proposal means that derives the task to be performed now based on at least one of the acquired current status information and behavioral history information and proposes it to the user.

[0171] (6) The task suggestion means may also suggest the task to the user by presenting the derived task along with at least one of the reasons why the task should be performed now and a method for performing the task.

[0172] (7) The user status acquisition means can acquire the current status information from at least one of the following: the analysis results of eye tracking, biometric information detected by the wearable device, location information, SNS activity, and chat logs.

[0173] (8) The system may also further include a feedback means (for example, the feedback unit 40 in Figure 3) that provides feedback for deriving the next task based on at least one of the execution history information and the action history information.

[0174] (9) The system may further include a dialogue acquisition means (for example, an information acquisition unit 31 in Figure 3) that acquires dialogue information relating to the content of the dialogue through dialogue with the user, and a question generation means (for example, an information generation unit 33 in Figure 3) that generates questions to capture the user's feelings based on the dialogue information.

[0175] (10) The system may further include task acquisition means (for example, information acquisition unit 31 in Figure 3) that acquires task information relating to the task from the dialogue information based on natural language input in a chat interface between the user and the AI ​​chatbot as a means of dialogue, and task registration means (for example, task registration unit 36 ​​in Figure 3) that temporarily generates the task from the acquired task information (for example, task generation by the information generation unit 33 in Figure 3) and registers the task confirmed by the user.

[0176] (11) The system may further include a task decomposition means (for example, a task decomposition unit 37 in Figure 3) for decomposing the registered task into a plurality of subtasks, and a progress management means (for example, an information management unit 32 in Figure 3) for calculating and managing a progress rate from the completion status of each of the plurality of subtasks and from the execution status information of the task before it was decomposed into the plurality of subtasks.

[0177] (12) The system may also include a dialogue management means (for example, a dialogue derivation unit 39 in Figure 3) that extracts the history of dialogues related to the task from the dialogue information, manages it in association with the task (for example, managed by the information management unit 32 in Figure 3), and when the task is selected, makes the history of the dialogues associated with the task available for reference as contextual information.

[0178] (13) The task suggestion means can also analyze the configuration of the tasks in past projects stored in the task storage means and suggest to the user a configuration of the task that is suitable for the current project.

[0179] (14) Furthermore, the execution status management means can manage changes to the execution status information in real time and synchronized among multiple users.

[0180] (15) The system may also include a ToDo-based chat management means (for example, the chat management unit 38 in Figure 3) that generates a dedicated chat thread for each task and manages all communications related to that task in the chat thread.

[0181] (16) The task suggestion means may also refer to the chat thread history of similar tasks in the past and suggest at least one of task-related information related to the currently running task and a solution for the task.

[0182] 1: Management Server, 2: Generation AI Server, 3: User Terminal, 11: CPU, 16: Output Unit, 17: Input Unit, 18: Storage Unit, 19: Communication Unit, 31: Information Acquisition Unit, 32: Information Management Unit, 33: Information Generation Unit, 34: Judgment Unit, 35: Task Proposal Unit, 36: Task Registration Unit, 37: Task Decomposition Unit, 38: Chat Management Unit, 39: Dialogue Derivation Unit, 40: Feedback Unit, 41: Transmission Control Unit, 51: Task DB, 52: User DB, N: Network, S: Information Processing System

Claims

1. An information processing device comprising: task storage means for storing tasks; hierarchical structure management means for managing the hierarchical structure of the tasks; and communication management means for managing task communication information related to the tasks, linked to those tasks.

2. An information processing apparatus according to claim 1, further comprising: execution status management means for managing execution status information relating to the current execution status of the task.

3. An information processing apparatus according to claim 2, further comprising: execution history management means for managing execution history information relating to the past execution history of the task.

4. The information processing apparatus according to claim 3, further comprising: support information generation means for generating support information to assist a user performing the task based on the execution status information or the execution history information.

5. The information processing apparatus according to claim 4, further comprising: a user status acquisition means for acquiring at least one of current status information relating to the user's current situation and behavioral history information relating to the user's behavioral history; and a task proposal means for deriving and proposing to the user a task to be performed based on at least one of the acquired current status information and behavioral history information.

6. The information processing apparatus according to claim 5, wherein the task suggestion means proposes to the user, along with the derived task, at least one of the reason why the task should be performed now and a method for performing the task.

7. The information processing apparatus according to claim 5, wherein the user status acquisition means acquires the current status information from at least one of the following: the results of eye tracking analysis, biometric information detected by a wearable device, location information, SNS activity, and chat logs.

8. The information processing apparatus according to claim 5, further comprising a feedback means for providing feedback for deriving the next task based on at least one of the execution history information and the action history information.

9. The information processing apparatus according to claim 5, further comprising: dialogue acquisition means for acquiring dialogue information relating to the content of a dialogue through a dialogue with a user; and question generation means for generating questions for capturing the user's senses based on the dialogue information.

10. The information processing apparatus according to claim 9, further comprising: a task acquisition means for acquiring task information relating to the task from dialogue information based on natural language input in a chat interface between the user and an AI chatbot as a means of dialogue; and a task registration means for temporarily generating the task from the acquired task information and registering the task confirmed by the user.

11. The information processing device according to claim 10, further comprising: task decomposition means for decomposing a registered task into a plurality of subtasks; and progress management means for calculating and managing a progress rate from the execution status information of the task before it was decomposed into the plurality of subtasks based on the completion status of each of the plurality of subtasks.

12. The information processing apparatus according to claim 10, further comprising a dialogue management means for extracting a history of dialogues related to the task from the dialogue information, managing it in association with the task, and making the history of dialogues associated with the task accessible as contextual information when the task is selected.

13. The information processing apparatus according to claim 10, wherein the task suggestion means analyzes the configuration of past tasks stored in the task storage means and suggests to the user a configuration of the task suitable for the current project.

14. The information processing apparatus according to claim 4, wherein the execution status management means manages changes to the execution status information synchronously among multiple users.

15. The information processing device according to claim 5, further comprising a ToDo-based chat management means for generating a dedicated chat thread for each task and managing communication related to the task in the chat thread.

16. The information processing apparatus according to claim 15, wherein the task suggestion means refers to the history of chat threads for similar tasks in the past and suggests at least one of task-related information related to the currently running task and a solution for the task.

17. An information processing method performed by an information processing device, comprising: a step of storing a task; a step of managing the hierarchical structure of the task; and a step of managing task communication information related to the task in association with the task.

18. A program for causing a computer to execute control processing including the steps of: storing a task; managing the hierarchical structure of the task; and managing task communication information related to the task in association with the task.

Images