Intelligent interaction method, intelligent interaction system and display equipment

Abstract

The invention discloses an intelligent interaction method, an intelligent interaction system and a display device, the method is applied to the intelligent interaction system, the intelligent interaction system comprises a task manager, an interaction agent and a deep processing agent, and the task manager obtains an interaction task corresponding to an interaction request of a user; the task manager determines the task type of the interaction task; the task manager allocates the interaction task to a deep processing agent under the condition that the task type of the interaction task is a complex task; the deep processing agent executes the interaction task to obtain task execution information of each task execution stage; and in the interaction task execution process, the interaction agent obtains task execution information of the current task execution stage, and feeds back a response result of the interaction request in the current task execution stage to the user based on the task execution information until the interaction task is executed completely. By adopting the method and the device, the requirement of real-time interaction with the user can be met while deep analysis is performed on the complex task, and the user experience is improved.

Description

Technical Field

[0001] The present application relates to the field of intelligent interaction technology, and in particular to an intelligent interaction method, an intelligent interaction system and a display device. Background Art

[0002] With the gradual development of intelligent agent technology, intelligent agents have become the core of natural language processing and complex task processing.

[0003] Currently, intelligent interaction systems primarily utilize a single-agent architecture, meaning they consist of only one agent. Limited by the single agent's limited functionality, these systems are unable to handle diverse tasks (such as weather queries, travel recommendations, and knowledge quizzes). Furthermore, when processing complex tasks, limited by the computing resources and performance of a single agent, these systems struggle to simultaneously meet the demands of both real-time user interaction and in-depth analysis of complex tasks, resulting in a poor user experience.

[0004] Therefore, there is an urgent need for an intelligent interaction method that can simultaneously meet the needs of real-time interaction with users and in-depth analysis of complex tasks. Summary of the Invention

[0005] To address the above issues, the present invention provides an intelligent interaction method, intelligent interaction system, and display device that can perform in-depth analysis of complex tasks while meeting the needs of real-time interaction with users, thereby improving user experience. Specifically, the present invention discloses the following technical solutions:

[0006] In a first aspect, an embodiment of the present application provides an intelligent interaction method, which is applied to an intelligent interaction system, wherein the intelligent interaction system includes a task manager, an interaction agent, and a depth processing agent. The method includes:

[0007] The task manager obtains the interactive task corresponding to the user's interactive request;

[0008] The task manager determines the task type of the interactive task;

[0009] The task manager assigns the interactive task to the depth processing agent when the task type of the interactive task is a complex task;

[0010] The depth processing agent executes the interactive task and obtains task execution information of the interactive task at each task execution stage;

[0011] During the process of the deep processing agent executing the interactive task, the interactive agent obtains the task execution information of the current task execution stage from the deep processing agent, and based on the task execution information of the current task execution stage, feeds back to the user the response result of the interactive request in the current task execution stage until the deep processing agent completes the execution of the interactive task.

[0012] In the above embodiment of the present application, when the task type of the interactive task is a complex task, the task manager assigns the interactive task to the deep processing agent for processing, which can meet the needs of deep analysis of complex tasks. At the same time, during the execution of the interactive task, the interactive agent obtains the task execution information of the interactive task at different task execution stages in real time, and based on the task execution information, feedbacks the response result of the interactive request to the user until the interactive task is completed. This can ensure real-time interaction with the user during the execution of the interactive task, meet the real-time response requirements, and improve the user experience.

[0013] In some embodiments, the method further comprises:

[0014] The task manager assigns the interactive task to the interactive agent when the task type of the interactive task is a simple task;

[0015] The interactive agent executes the interactive task, obtains a task execution result, and based on the task execution result, feeds back a response result of the interactive request to the user.

[0016] In the above embodiment of the present application, for interactive tasks with a simple task type, the task manager assigns the interactive task to an interactive agent. The interactive agent executes the interactive task, obtains the task execution result, and based on the task execution result, provides feedback to the user on the response result of the interactive request. In this way, for interactive tasks with a simple task type, the interactive agent can quickly provide feedback to the user on the response result of the interactive request, thereby improving the user experience.

[0017] In some embodiments, determining the task type of the interactive task includes:

[0018] The task manager performs semantic understanding processing on the interactive task to obtain semantic information of the interactive task;

[0019] The task manager performs task type discrimination processing on the semantic information to obtain the task type of the interactive task.

[0020] In the above embodiment of the present application, the task manager performs semantic understanding processing on the interactive task to obtain the semantic information of the interactive task. Based on the semantic information of the interactive task, the task manager performs task type discrimination processing on the interactive task using a pre-trained task classification model to determine the task type of the interactive task. In this way, the accuracy of determining the task type of the interactive task can be improved.

[0021] In some embodiments, when there are multiple depth processing agents, allocating the interaction task to the depth processing agents includes:

[0022] The task manager obtains the depth processing capability of each depth processing agent and the number of interaction tasks to be processed;

[0023] The task manager determines a target depth processing agent among the depth processing agents based on the depth processing capabilities of the depth processing agents and the number of interactive tasks to be processed;

[0024] The task manager assigns the interactive task to the target depth processing agent.

[0025] In the above embodiment of the present application, when there are multiple depth processing agents, the task manager determines the target depth processing agent among each depth processing agent based on the depth processing capabilities of each depth processing agent and the number of interactive tasks to be processed, and assigns the interactive tasks to the target depth processing agent, thereby improving the task processing speed and task processing accuracy of the interactive tasks.

[0026] In some embodiments, the deep processing agent includes a task management sub-agent, at least one task execution sub-agent, and a tool set. The deep processing agent executes the interactive task and obtains task execution information of the interactive task at each task execution stage, including:

[0027] The task management sub-agent performs task decomposition processing on the interactive task to obtain sub-interaction tasks of the interactive task, and assigns the sub-interaction tasks to the at least one task execution sub-agent;

[0028] The at least one task execution sub-agent calls the tool set to execute the sub-interaction task, and obtains task execution information of the sub-interaction task at each task execution stage;

[0029] The task management sub-agent generates task execution information of the interactive task at each task execution stage based on the task execution information of the sub-interactive task at each task execution stage.

[0030] In the above embodiment of the present application, the task management sub-agent performs task decomposition processing on the interactive task, obtains sub-interaction tasks of the interactive task, and assigns the sub-interaction tasks to at least one task execution sub-agent. The task execution sub-agent calls the tool set to execute the sub-interaction task, and obtains the task execution information of the sub-interaction task at each task execution stage. Based on the task execution information of the sub-interaction task at each task execution stage, the task management sub-agent generates the task execution information of the interactive task at each task execution stage. In this way, by decomposing the interactive task into multiple sub-interaction tasks and executing the corresponding sub-interaction tasks through multiple task execution sub-agents, the processing speed of the interactive task can be improved.

[0031] In some embodiments, the interactive agent obtains task execution information of the current task execution phase from the depth processing agent, including:

[0032] The interactive agent sends a task query request message to the depth processing agent; receives a task query response message fed back by the depth processing agent, wherein the task query response message carries task execution information of the current task execution stage; or

[0033] The interactive agent receives a task information update message sent by the depth processing agent, wherein the task information update message carries task execution information of the current task execution phase; or

[0034] The interactive agent queries the task execution information of the current task execution stage from the shared memory, and the task execution information of the current task execution stage stored in the shared memory is stored by the deep processing agent.

[0035] In the above-mentioned embodiments of the present application, the interactive intelligent agent can obtain the task execution information of the current task execution stage of the interactive task in a variety of ways, thereby ensuring that the interactive intelligent agent can obtain the task execution information of the current task execution stage of the interactive task in a timely manner, and further ensuring that the interactive intelligent agent can timely feedback the response results of the interactive request in the current task execution stage to the user based on the task execution information of the current task execution stage of the interactive task, thereby improving the user experience.

[0036] In some embodiments, the task execution information includes a task description, a task execution stage, and a task execution result. Feedback to the user on a response result of the interaction request in the current task execution stage based on the task execution information of the current task execution stage includes:

[0037] The interactive agent generates interactive content based on the task description, the task execution stage and the task execution result in the task execution information of the current task execution stage;

[0038] The interactive agent feeds back to the user a response result of the interactive request in the current task execution phase based on the interactive content.

[0039] In the above-mentioned embodiments of the present application, the interactive agent generates interactive content based on the task description, task execution stage, and task execution results, and based on the interactive content, provides feedback to the user on the response results of the interactive request. In this way, the interactive agent can generate different interactive content at different task execution stages based on different task execution information, thereby providing the user with different response results of the interactive request, thereby improving the user experience.

[0040] In some embodiments, the method further comprises:

[0041] When the interactive agent determines based on the task execution information that the interactive task has failed to be executed, the interactive agent terminates the execution of the interactive task by the depth processing agent.

[0042] In the above embodiment of the present application, if the interactive agent determines based on the task execution information that the interactive task has failed, it terminates the execution of the interactive task by the depth processing agent. In this way, if the interactive task fails, the interactive agent promptly terminates the execution of the interactive task by the depth processing agent, which can effectively reduce the resource waste of the interactive task on the depth processing agent and speed up the processing of subsequent interactive tasks.

[0043] A second aspect of an embodiment of the present application provides an intelligent interactive system, which includes a task manager, an interactive agent, and a depth processing agent; the task manager, the interactive agent, and the depth processing agent implement the method described in the first aspect.

[0044] In the above embodiment of the present application, when the task type of the interactive task is a complex task, the task manager assigns the interactive task to the deep processing agent for processing, which can meet the needs of deep analysis of complex tasks. At the same time, during the execution of the interactive task, the interactive agent obtains the task execution information of the interactive task at different task execution stages in real time, and based on the task execution information, feedbacks the response result of the interactive request to the user until the interactive task is completed. This can ensure real-time interaction with the user during the execution of the interactive task, meet the real-time response requirements, and improve the user experience.

[0045] A third aspect of the present application provides a display device, including:

[0046] monitor;

[0047] a controller coupled to the display and configured to:

[0048] Get the user's interaction request;

[0049] Processing the interaction request through the intelligent interaction system described in the second aspect to obtain a response result of the interaction request;

[0050] Feedback the response result of the interaction request to the user.

[0051] In the above embodiment of the present application, when the task type of the interactive task is a complex task, the task manager assigns the interactive task to the deep processing agent for processing, which can meet the needs of deep analysis of complex tasks. At the same time, during the execution of the interactive task, the interactive agent obtains the task execution information of the interactive task at different task execution stages in real time, and based on the task execution information, feedbacks the response result of the interactive request to the user until the interactive task is completed. This can ensure real-time interaction with the user during the execution of the interactive task, meet the real-time response requirements, and improve the user experience.

[0052] The embodiments of the present application provide an intelligent interaction method, intelligent interaction system, and display device. The task manager obtains the interactive task corresponding to the user's interactive request; the task manager determines the task type of the interactive task; if the task type of the interactive task is a complex task, the task manager assigns the interactive task to a deep processing agent; the deep processing agent executes the interactive task and obtains task execution information of the interactive task at each task execution stage; during the process of the deep processing agent executing the interactive task, the interactive agent obtains the task execution information of the current task execution stage from the deep processing agent, and based on the task execution information of the current task execution stage, feedbacks the response result of the interactive request at the current task execution stage to the user until the deep processing agent completes the execution of the interactive task. In this way, if the task type of the interactive task is a complex task, the task manager assigns the interactive task to the deep processing agent for processing, which can meet the demand for in-depth analysis of complex tasks. At the same time, during the execution of the interactive task, the interactive agent obtains the task execution information of the interactive task at different task execution stages in real time, and based on the task execution information, feedbacks the response result of the interactive request to the user until the interactive task is completed. This can ensure real-time interaction with the user during the execution of the interactive task, meet the real-time response requirements, and improve the user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0054] Figure 1 is a schematic diagram illustrating an operation scenario between a display device and a control device according to some exemplary embodiments;

[0055] Figure 2 is a block diagram of a hardware configuration of a display device 200 according to some exemplary embodiments;

[0056] Figure 3 is a schematic diagram of software configuration of a display device 200 according to some exemplary embodiments;

[0057] Figure 4 is a schematic structural diagram of an intelligent interactive system according to some exemplary embodiments;

[0058] Figure 5 is a flow chart of an intelligent interaction method according to some exemplary embodiments;

[0059] Figure 6 is a flow chart of an intelligent interaction method according to some other exemplary embodiments;

[0060] Figure 7 is a flow chart of an intelligent interaction method according to some other exemplary embodiments;

[0061] Figure 8 is a flow chart of an intelligent interaction method according to some other exemplary embodiments;

[0062] Figure 9 is a flow chart of an intelligent interaction method according to some other exemplary embodiments;

[0063] Figure 10 is an interaction diagram of an intelligent interaction system according to some exemplary embodiments. DETAILED DESCRIPTION

[0064] The following embodiments are described in detail, with examples illustrated in the accompanying drawings. When the following description refers to the drawings, identical numbers in different figures represent identical or similar elements unless otherwise indicated. The embodiments described in the following embodiments are not intended to represent all possible implementations consistent with the present application. They are merely examples of systems and methods consistent with certain aspects of the present application, as detailed in the claims.

[0065] It should be noted that the brief descriptions of terms in this application are only for the purpose of facilitating the understanding of the embodiments described below, and are not intended to limit the embodiments of this application. Unless otherwise specified, these terms should be understood according to their ordinary and usual meanings.

[0066] In the specification and claims of this application and the accompanying drawings, the terms "first," "second," "third," etc. are used to distinguish similar or similar objects or entities, and are not necessarily intended to limit a particular order or sequence, unless otherwise noted. It should be understood that the terms used in this manner are interchangeable under appropriate circumstances.

[0067] The terms "comprise," "include," and "have," and any variations thereof, are intended to cover but not exclude inclusion; for example, a product or device comprising a list of components is not necessarily limited to all the components expressly listed but may include other components not expressly listed or inherent to such product or device.

[0068] The term "module" refers to any known or later developed hardware, software, firmware, artificial intelligence, fuzzy logic, or combination of hardware and / or software code that is capable of performing the functionality associated with that element.

[0069] In the embodiments of the present application, a display device generally refers to a device capable of displaying images and processing data. For example, a display device includes but is not limited to smart TVs, mobile terminals, computers, monitors, advertising screens, wearable devices, virtual reality devices, augmented reality devices, etc.

[0070] In the embodiments of the present application, the display device generally refers to a device with image display and data processing capabilities. The display device can have various implementation forms, for example, it can be a television, a smart TV, a laser projection device, a monitor, an electronic bulletin board, an electronic table, etc. Figure 1 and Figure 2 This is a specific implementation of the display device of the present application.

[0071] Figure 1 FIG. 1 is a schematic diagram of an operation scenario between a display device and a control device according to some exemplary embodiments. Figure 1 As shown, the user can operate the display device 200 through the smart device 300 or the control apparatus 100 .

[0072] In some embodiments, the control device 100 may be a remote controller. Communication between the remote controller and the display device may include infrared protocol communication, Bluetooth protocol communication, or other short-range communication methods, and the display device 200 may be controlled wirelessly or wired. The user may control the display device 200 by inputting user commands through buttons on the remote controller, voice input, control panel input, and the like.

[0073] In some embodiments, a smart device 300 (such as a mobile terminal, tablet computer, computer, laptop computer, etc.) can also be used to control the display device 200. For example, the display device 200 can be controlled using an application running on the smart device.

[0074] In some embodiments, the display device may not use the aforementioned smart device or control device to receive instructions, but may receive user control through touch or gestures.

[0075] In some embodiments, the display device 200 can also be controlled in a manner other than the control device 100 and the smart device 300. For example, the user's voice command control can be directly received through a module for obtaining voice commands configured inside the display device 200, or the user's voice command control can be received through a voice control device set outside the display device 200.

[0076] In some embodiments, the display device 200 also communicates data with the server 400. The display device 200 may be connected to a local area network (LAN), a wireless local area network (WLAN), or other networks. The server 400 may provide various content and interactions to the display device 200. The server 400 may be a single cluster or multiple clusters, and may include one or more types of servers.

[0077] Figure 2 FIG. 2 is a block diagram of a hardware configuration of a display device 200 according to some exemplary embodiments. Figure 2 As shown, the display apparatus 200 includes at least one of a tuner and demodulator 210 , a communicator 220 , a detector 230 , an external device interface 240 , a controller 250 , a display 260 , an audio output interface 270 , a memory, a power supply, and a user interface.

[0078] In some embodiments, the controller includes a processor, a video processor, an audio processor, a graphics processor, RAM, ROM, and first to nth interfaces for input / output.

[0079] The display 260 includes a display screen component for presenting images, and a driving component for driving image display, a component for receiving image signals output from a controller, and a component for displaying video content, image content, and a menu control interface and a user control UI interface.

[0080] The display 260 may be a liquid crystal display, an OLED display, or a projection display, and may also be a projection device and a projection screen.

[0081] Communicator 220 is a component used to communicate with external devices or servers using various communication protocols. For example, the communicator may include at least one of a Wi-Fi module, a Bluetooth module, a wired Ethernet module, or other network communication protocol chip or a near-field communication protocol chip, as well as an infrared receiver. Display device 200 can use communicator 220 to send and receive control signals and data signals with external control device 100 or server 400.

[0082] The user interface can be used to receive control signals from the control device 100 (such as an infrared remote controller, etc.).

[0083] Detector 230 is used to collect signals from the external environment or external interactions. For example, detector 230 includes a light receiver, a sensor for collecting ambient light intensity; or detector 230 includes an image collector, such as a camera, for collecting external environmental scenes, user attributes, or user interaction gestures; or detector 230 includes a sound collector, such as a microphone, for receiving external sounds.

[0084] The external device interface 240 may include, but is not limited to, any one or more of the following: a high-definition multimedia interface (HDMI), an analog or digital high-definition component input interface (component), a composite video input interface (CVBS), a USB input interface (USB), an RGB port, etc. It may also be a composite input / output interface formed by multiple of the above interfaces.

[0085] The tuner-demodulator 210 receives broadcast television signals via wired or wireless reception, and demodulates audio and video signals, such as EPG data signals, from a plurality of wireless or wired broadcast television signals.

[0086] In some embodiments, the controller 250 and the tuner 210 may be located in different separate devices, that is, the tuner 210 may also be located in an external device of the main device where the controller 250 is located, such as an external set-top box.

[0087] The controller 250 controls the operation of the display device and responds to user operations through various software control programs stored in the memory. The controller 250 controls the overall operation of the display device 200. For example, in response to receiving a user command for selecting a UI object to be displayed on the display 260, the controller 250 can perform operations related to the object selected by the user command. In an embodiment of the present application, the controller 250 obtains the user's interaction request, processes the interaction request through the intelligent interaction system, obtains a response result of the interaction request, and feeds back the response result of the interaction request to the user. Among them, the intelligent interaction system can be set inside the controller 250 or outside the controller 250, and the embodiment of the present application is not limited thereto.

[0088] In some embodiments, the controller includes at least one of a central processing unit (CPU), a video processor, an audio processor, a graphics processing unit (GPU), RAM Random Access Memory (RAM), ROM (Read-Only Memory, ROM), a first interface to an nth interface for input / output, a communication bus (Bus), etc.

[0089] The user may input a user command through a graphical user interface (GUI) displayed on the display 260, and the user input interface receives the user input command through the graphical user interface (GUI). Alternatively, the user may input a user command through a specific voice or gesture, and the user input interface may recognize the voice or gesture through a sensor to receive the user input command.

[0090] A user interface is the medium for interaction and information exchange between an application or operating system and the user. It converts information between its internal form and a user-friendly format. A common user interface is the graphical user interface (GUI), which refers to a graphical user interface related to computer operations. It can be an icon, window, control, or other interface element displayed on an electronic device's display. Controls can include visual interface elements such as icons, buttons, menus, tabs, text boxes, dialog boxes, status bars, navigation bars, and widgets.

[0091] Figure 3 FIG. 2 is a schematic diagram of software configuration of the display device 200 according to some exemplary embodiments. Figure 3 As shown, in some embodiments, the system is divided into four layers, from top to bottom, namely, the application layer (referred to as "application layer"), the application framework layer (referred to as "framework layer"), the system library layer (referred to as "system runtime library layer"), and the kernel layer.

[0092] In some embodiments, at least one application runs in the application layer. These applications can be window programs, system settings programs, clock programs, etc. that come with the operating system, or applications developed by third-party developers. In specific implementations, the application packages in the application layer are not limited to the above examples.

[0093] The application framework layer provides applications with an application programming interface (API) and programming framework. The application framework layer includes predefined functions. The application framework layer acts as a processing center, determining the actions taken by applications in the application layer. Through the API, applications can access system resources and services during execution.

[0094] like Figure 3 As shown, in the embodiment of the present application, the application framework layer includes managers, content providers, etc., wherein the manager includes at least one of the following modules: an activity manager (ActivityManager) is used to interact with all activities running in the system; a location manager (Location Manager) is used to provide system services or applications with access to system location services; a package manager (Package Manager) is used to retrieve various information related to the application packages currently installed on the device; a notification manager (NotificationManager) is used to control the display and clearing of notification messages; a window manager (Window Manager) is used to manage icons, windows, toolbars, wallpapers, and desktop widgets on the user interface.

[0095] In some embodiments, the activity manager is used to manage the lifecycle of each application and common navigation back functions, such as controlling the exit, opening, and backing of an application. The window manager is used to manage all window programs, such as obtaining the display screen size, determining whether there is a status bar, locking the screen, taking screenshots, and controlling display window changes (such as shrinking the display window, shaking the display, distorting the display, etc.).

[0096] In some embodiments, the system runtime layer provides support for the upper layer, namely the framework layer. When the framework layer is used, the Android operating system will run the C / C++ library contained in the system runtime layer to implement the functions to be implemented by the framework layer.

[0097] In some embodiments, the kernel layer is a layer between hardware and software. Figure 3 As shown, the kernel layer includes at least one of the following drivers: audio driver, display driver, Bluetooth driver, camera driver, WIFI driver, USB driver, HDMI driver, sensor driver (such as fingerprint sensor, temperature sensor, pressure sensor, etc.), and power driver, etc.

[0098] It should be noted that the above example is only a simple division of the operating system functions and does not constitute a limitation on the specific operating system form of the display device 200 in the embodiment of the present application. Depending on factors such as the function of the display device and the type of operating system, the number of levels and specific level types contained in the operating system may be expressed in other forms.

[0099] Currently, intelligent interaction systems primarily utilize a single-agent architecture, meaning they consist of only one agent. Limited by the single agent's limited functionality, these systems are unable to handle diverse tasks (such as weather queries, travel recommendations, and knowledge quizzes). Furthermore, when processing complex tasks, limited by the computing resources and performance of a single agent, these systems struggle to simultaneously meet the demands of both real-time user interaction and in-depth analysis of complex tasks, resulting in a poor user experience.

[0100] To facilitate understanding, the intelligent interaction system involved in the embodiments of the present application is first introduced. Figure 4 FIG. 1 is a schematic diagram of the structure of an intelligent interactive system according to some exemplary embodiments. Figure 4 As shown, the intelligent interaction system includes a task manager 410, an interactive agent 420, a depth processing agent 430, and a shared memory 440. Task manager 410 is configured to obtain interactive tasks corresponding to user interaction requests, determine the task type of the interactive tasks, and assign the interactive tasks to depth processing agent 430 if the task type is complex, or to interactive agent 420 if the task type is simple. Depth processing agent 430 is configured to execute interactive tasks of the complex type and obtain task execution information at different execution stages of the interactive tasks. Interactive agent 420 is configured to obtain task execution information at different execution stages of the interactive tasks from depth processing agent 430 during the execution of the interactive tasks, and based on the task execution information, provide feedback to the user on the response to the interactive request. Interactive agent 420 is also configured to execute interactive tasks of the simple type, obtain the task execution results of the interactive tasks, and based on the task execution results, provide feedback to the user on the response to the interactive request. The shared memory 440 is used to store the task execution results obtained by the interactive agent 420 when performing the interactive task and the task execution information of different task execution stages obtained by the depth processing agent 430 when performing the interactive task.

[0101] It should be noted that in the intelligent interactive system, the number of interactive agents 420 and depth processing agents 430 can be one or more, and this embodiment of the present application does not limit this.

[0102] To address this issue, an embodiment of the present application provides an intelligent interaction method, which is applied to an intelligent interaction system. Figure 5 FIG. 1 is a flow chart of an intelligent interaction method according to some exemplary embodiments. Figure 5 The specific steps are as follows:

[0103] Step 501: The task manager obtains the interactive task corresponding to the user's interactive request.

[0104] For example, when a user wishes to interact intelligently with an intelligent interactive system, they may initiate an interaction request to the intelligent interactive system via voice, text, image, video, or a combination of these methods. The intelligent interactive system then receives the user's interaction request. The task manager within the intelligent interactive system then semantically understands the user's interaction request and obtains semantic information about the interaction request. Based on the semantic information, the task manager then generates an interaction task for the interaction request. For example, a user initiates an interaction request to the intelligent interactive system via voice, and the content of the interaction request is "Hello." Accordingly, the task manager generates an interaction task for the interaction request as "Greeting Task." For another example, a user initiates an interaction request to the intelligent interactive system via voice, and the content of the interaction request is "Help me find out where the weather is colder today, Qingdao or Beijing." Accordingly, the task manager generates an interaction task for the interaction request as "Compare the temperatures today in Qingdao and Beijing." After generating the interaction task for the interaction request, the task manager may also generate a task identifier for the interaction task. A task identifier uniquely identifies an interaction task.

[0105] Step 502: The task manager determines the task type of the interactive task.

[0106] Exemplarily, after the task manager generates an interactive task corresponding to an interactive request, the task manager may perform task classification processing on the interactive task and determine the task type of the interactive task. Task types may include simple tasks and complex tasks, and may also include other task types, which are not limited in the embodiments of the present application. The embodiments of the present application are introduced by taking the task types including simple tasks and complex tasks as an example. Other situations are similar and will not be described in detail in the embodiments of the present application. The task type of an interactive task is a simple task, which indicates that the task goal of the interactive task is single, the demand for in-depth analysis is low, the processing time is short, and only the interactive agent is required to process, and no in-depth processing agent is required to process; the task type of an interactive task is a complex task, which indicates that the task goals of the interactive task are diverse, the demand for in-depth analysis is high, the processing time is long, and deep processing agent processing is required. For example, the task type of the interactive task "greeting task" is a simple task, which only requires an interactive agent to process, and no in-depth processing agent to process; the task type of the interactive task "comparing the temperature in Qingdao and Beijing today" is a complex task, which requires a deep processing agent to process.

[0107] In some embodiments, in step 502 above, the process of the task manager determining the task type of the interactive task includes the following steps:

[0108] Step 1: The task manager performs semantic understanding processing on the interactive task to obtain the semantic information of the interactive task.

[0109] Illustratively, after determining the interactive task corresponding to the interactive request, the task manager may perform semantic understanding processing on the task content of the interactive task, extract key task information of the interactive task, and generate semantic information of the interactive task.

[0110] Step 2: The task manager performs task type discrimination processing on the semantic information to obtain the task type of the interactive task.

[0111] Exemplarily, the task manager can use a pre-trained task classification model to perform task type discrimination processing on the interactive task to determine the task type of the interactive task. For example, the task manager can input the semantic information of the interactive task into the task classification model, and the task classification model outputs the task type of the interactive task. The task classification model can be a task classification model based on a vector machine or a task classification model based on deep learning, which is not limited in the embodiments of the present application.

[0112] In the above embodiment of the present application, the task manager performs semantic understanding processing on the interactive task to obtain the semantic information of the interactive task. Based on the semantic information of the interactive task, the task manager performs task type discrimination processing on the interactive task using a pre-trained task classification model to determine the task type of the interactive task. In this way, the accuracy of determining the task type of the interactive task can be improved.

[0113] In step 503 , if the task type of the interactive task is a complex task, the task manager assigns the interactive task to the depth processing agent.

[0114] Exemplarily, after the task manager determines the task type of the interactive task, if the task type of the interactive task is a complex task, it means that the task objectives of the interactive task are diverse, the demand for in-depth analysis is high, the processing time is long, and it needs to be processed by a deep processing agent. Accordingly, the task manager can assign the interactive task to the deep processing agent for processing. Furthermore, after the task manager assigns the interactive task to the deep processing agent, it can send the task identifier of the interactive task to the interactive agent, so that the interactive agent can subsequently obtain the task execution information of the interactive task based on the task identifier of the interactive task. For example, the task type of the interactive task "Compare the temperatures in Qingdao and Beijing today" is a complex task, and the task manager assigns the interactive task to the deep processing agent for processing.

[0115] In some embodiments, the number of deep processing agents in the intelligent interactive system can be one or more, which is not limited in the embodiments of the present application. Figure 6 As shown, in the above step 503, the process of the task manager assigning the interactive task to the depth processing agent includes the following steps:

[0116] Step 601: The task manager obtains the depth processing capability of each depth processing agent and the number of interaction tasks to be processed.

[0117] Exemplarily, when there are multiple deep processing agents, the task manager can obtain the deep processing capabilities and the number of interactive tasks to be processed of each deep processing agent. Among them, the deep processing capabilities of the deep processing agent can represent the comprehensive capabilities of the deep processing agent in multiple dimensions such as computing power, deep analysis capabilities, task processing speed, and task processing accuracy. The number of interactive tasks to be processed by the deep processing agent can represent the number of interactive tasks that the task manager has assigned to the deep processing agent but has not yet processed and / or is processing by the deep processing agent. The stronger the deep processing capability of the deep processing agent, the faster and more accurate the deep processing agent is in processing interactive tasks of complex task types. The more interactive tasks to be processed by the deep processing agent, the longer the waiting time for the interactive task will be after the task manager assigns the interactive task to the deep processing agent.

[0118] In step 602, the task manager determines a target depth processing agent among the depth processing agents based on the depth processing capabilities of the depth processing agents and the number of interaction tasks to be processed.

[0119] Exemplarily, after the task manager obtains the depth processing capabilities and the number of interactive tasks to be processed of each depth processing agent, it can determine a target depth processing agent from each depth processing agent based on the depth processing capabilities and the number of interactive tasks to be processed of each depth processing agent. Specifically, the task manager can determine the target depth processing agent from each depth processing agent based on the depth processing capabilities and the number of interactive tasks to be processed of each depth processing agent according to a pre-set selection priority. For example, if the pre-selected priority is that the priority of depth processing capability is higher than the priority of the number of interactive tasks to be processed, the task manager can prioritize the depth processing agent with the highest depth processing capability among the depth processing agents as the target depth processing agent. If the depth processing capabilities of the depth processing agents are the same, the task manager can prioritize the depth processing agent with the fewest interactive tasks to be processed as the target depth processing agent. Of course, the task manager can determine the score of each depth processing agent based on the depth processing capabilities and the number of interactive tasks to be processed of each depth processing agent, as well as the corresponding weights, and determine the depth processing agent with the highest score as the target depth processing agent.

[0120] It should be noted that those skilled in the art may also determine the target deep processing agent among the deep processing agents based on other performance indicators of the deep processing agent, and this embodiment of the application is not limited thereto. This embodiment of the application takes the example of determining the target deep processing agent among the deep processing agents based on the deep processing capabilities of each deep processing agent and the number of interactive tasks to be processed as an example. Other situations are similar and will not be described in detail in this embodiment of the application.

[0121] Step 603: The task manager assigns the interactive task to the target depth processing agent.

[0122] For example, after the task manager determines the target deep processing agent, it can assign the interactive task to the target deep processing agent for processing. Furthermore, after the task manager assigns the interactive task to the target deep processing agent, it can also send the task identifier of the interactive task and the agent identifier of the target deep processing agent to the interactive agent, so that the interactive agent can subsequently obtain the task execution information of the interactive task from the target deep processing agent based on the task identifier of the interactive task and the agent identifier of the target deep processing agent.

[0123] In the above embodiment of the present application, when there are multiple depth processing agents, the task manager determines the target depth processing agent among each depth processing agent based on the depth processing capabilities of each depth processing agent and the number of interactive tasks to be processed, and assigns the interactive tasks to the target depth processing agent, thereby improving the task processing speed and task processing accuracy of the interactive tasks.

[0124] Step 504: The deep processing agent executes the interactive task and obtains task execution information of the interactive task at each task execution stage.

[0125] Exemplarily, after receiving the interactive task assigned by the task manager, the deep processing agent can conduct an in-depth analysis of the interactive task and obtain the task execution information of the interactive task at each task execution stage. Specifically, the deep processing agent can divide the deep processing process into multiple task execution stages. For example, the task execution stage can include a task thinking stage, a data acquisition stage, a data recording stage, and a data analysis stage. In the task thinking stage, the deep processing agent deeply analyzes and understands the interactive task, and clarifies the task execution purpose and task execution steps of the data acquisition stage; in the data acquisition stage, the deep processing agent executes the interactive task based on the task execution purpose and task execution steps clarified in the task thinking stage, and obtains the task execution result; in the data recording stage, the deep processing agent records the task execution result. In the data analysis stage, the deep processing agent analyzes the task execution result and obtains the final task result of the interactive task. The deep processing agent processes the interactive task at different task execution stages, and the task execution information obtained for the interactive task is also different. The task execution information of an interactive task may include the task execution stage (also called task status, task_status), the task execution purpose (also called task description, task_description), the task execution result (also called task progress, task_progress) and additional information (additional).

[0126] For example, taking the interactive task "Compare today's temperatures in Qingdao and Beijing" as an example, the different task execution information corresponding to different task execution stages is shown in Table 1: in the task thinking stage, the task execution stage in the task execution information of the interactive task is "task thinking stage", the task execution purpose is "Compare today's temperatures in Qingdao and Beijing", and the task execution result is "Need to obtain today's temperature information in Qingdao and Beijing"; in the data acquisition stage, the task execution stage in the task execution information of the interactive task is "data acquisition stage", the task execution purpose is "Get today's temperatures in Qingdao and Beijing", and the task execution result is "Need to call the weather tool set to obtain temperature data"; in the data recording stage, the task execution stage in the task execution information of the interactive task is "data recording stage", the task execution purpose is "Record the obtained temperature data", and the task execution result is "Obtained temperature data: Qingdao 25℃, Beijing 28℃"; in the data analysis stage, the task execution stage in the task execution information of the interactive task is "data analysis stage", the task execution purpose is "Compare today's temperatures in Qingdao and Beijing", and the task execution result is "The temperature in Beijing is higher, 28℃".

[0127] Table 1

[0128]

[0129] In some embodiments, the deep processing agent includes a task management sub-agent, at least one task execution sub-agent and a tool set. Figure 7 As shown, in the above step 504, the depth processing agent executes the interactive task, and the process of obtaining the task execution information of the interactive task at each task execution stage includes the following steps:

[0130] In step 701, the task management sub-agent decomposes the interactive task to obtain sub-interaction tasks of the interactive task, and distributes the sub-interaction tasks to at least one task execution sub-agent.

[0131] Exemplarily, the deep processing agent may include multiple sub-agents and tool sets. The multiple sub-agents may be structured in a tree-like structure. The root node may be a task management sub-agent, which is used to decompose an interactive task into multiple sub-interaction tasks; the leaf node may be a task execution sub-agent, which is used to execute the sub-interaction tasks. Furthermore, the task execution sub-agents may have different functions. For example, there may be a task execution sub-agent for itinerary planning, a task execution sub-agent for transportation planning, a task execution sub-agent for accommodation planning, a task execution sub-agent for scenic spot planning, and a task execution sub-agent for budget planning, etc., which are not limited in the embodiments of the present application. After the deep processing agent receives the interactive task assigned by the task manager, the task management sub-agent in the deep processing agent may perform task decomposition processing on the interactive task to obtain the sub-interaction tasks of the interactive task. The task management sub-agent may then further assign each sub-interaction task to the corresponding task execution sub-agent. The task management sub-agent may decompose the interactive task into sub-tasks based on the hierarchical structure, execution order, functional requirements, and other aspects of the interactive task. The task management sub-agent and the task execution sub-agent can communicate through the model context protocol, which includes a request interface and a return interface. The request interface includes a request text, a request session identifier, and other information. The request text is used to describe the specific content of the request; the request session identifier is used to uniquely identify the request session to track the correspondence between the session request and the session response; and other information includes any other necessary metadata or context information to support the processing of the sub-interaction task. The return interface includes the task running status, task running results, and other information. The task running status indicates the task execution stage of the sub-interaction task; the task running result is used to provide information on the task execution results of the current sub-interaction task; and other information includes any other relevant information or data.

[0132] Step 702: at least one task execution sub-agent calls a tool set to execute a sub-interaction task, and obtains task execution information of the sub-interaction task at each task execution stage.

[0133] Exemplarily, for each task execution sub-agent, after the task execution sub-agent receives the sub-interaction task assigned by the task management sub-agent, it can further call the toolset to execute the sub-interaction task and obtain the task execution information of the sub-interaction task at each task execution stage. After the task execution sub-agent obtains the task execution information of the sub-interaction task at each task execution stage, it can feed back the task execution information of the sub-interaction task at each task execution stage to the task management sub-agent. Among them, the task execution sub-agent and the toolset can communicate through Web communication. It should be noted that when some task execution sub-agents execute sub-interaction tasks, they also need to base their execution on the sub-tasks obtained by other task execution sub-agents.

[0134] Step 703: The task management sub-agent generates task execution information of the interactive task at each task execution stage based on the task execution information of the sub-interactive task at each task execution stage.

[0135] For example, after the task management sub-agent obtains the task execution information of the sub-interaction task at each task execution stage fed back by all or part of the task execution sub-agent, it can generate the task execution information of the interaction task at each task execution stage based on the obtained task execution information of the sub-interaction task at each task execution stage.

[0136] In the above embodiment of the present application, the task management sub-agent performs task decomposition processing on the interactive task, obtains sub-interaction tasks of the interactive task, and assigns the sub-interaction tasks to at least one task execution sub-agent. The task execution sub-agent calls the tool set to execute the sub-interaction task, and obtains the task execution information of the sub-interaction task at each task execution stage. Based on the task execution information of the sub-interaction task at each task execution stage, the task management sub-agent generates the task execution information of the interactive task at each task execution stage. In this way, by decomposing the interactive task into multiple sub-interaction tasks and executing the corresponding sub-interaction tasks through multiple task execution sub-agents, the processing speed of the interactive task can be improved.

[0137] Step 505: During the execution of the interactive task by the deep processing agent, the interactive agent obtains the task execution information of the current task execution stage from the deep processing agent, and based on the task execution information of the current task execution stage, feeds back the response result of the interactive request in the current task execution stage to the user until the deep processing agent completes the execution of the interactive task.

[0138] For example, during the execution of an interactive task by a deep processing agent, the deep processing agent processes the interactive task at different task execution stages, and the resulting task execution information for the interactive task also varies. For each task execution stage, the interactive agent can obtain the task execution information for the interactive task at the current task execution stage. Based on the task execution information for the current task execution stage, the interactive agent can then provide feedback to the user on the response result corresponding to the interactive request at the current task execution stage, until the deep processing agent completes the interactive task.

[0139] In some embodiments, in step 505, the interactive agent may obtain the task execution information of the current task execution phase in various ways. The present embodiment provides three feasible real-time methods, as follows:

[0140] Method 1: The interactive agent sends a task query request message to the deep processing agent and receives a task query response message from the deep processing agent. The task query response message carries the task execution information of the current task execution stage.

[0141] Exemplarily, after the task manager assigns the interactive task to the deep processing agent, it can send the task identifier of the interactive task to the interactive agent. After the interactive agent receives the task identifier of the interactive task sent by the task manager, it can construct a task query request message carrying the task identifier of the interactive task based on the task identifier of the interactive task, and periodically send the task query request message to the deep processing agent. The task query request message is used to instruct the deep processing agent to feedback the task execution information of the interactive task. Accordingly, after the deep processing agent receives the task query request message sent by the interactive agent, it can feedback a task query response message to the interactive agent. The task query response message carries the task execution information of the interactive task in the current task execution stage. In this way, after the interactive agent receives the task query response message fed back by the deep processing agent, it can obtain the task execution information of the interactive task in the current task execution stage.

[0142] Method 2: The interactive agent receives a task information update message sent by the deep processing agent, wherein the task information update message carries the task execution information of the current task execution phase.

[0143] For example, the deep processing agent processes interactive tasks at different task execution stages, and the resulting task execution information for the interactive tasks is also different. Therefore, after the deep processing agent completes a task execution stage and obtains the task execution information corresponding to the current task execution stage, it can construct a task information update message based on the task execution information corresponding to the current task execution stage and send the task information update message to the interactive agent. The task information update message carries the task execution information corresponding to the current task execution stage.

[0144] Method three: the interactive agent queries the task execution information of the current task execution stage from the shared memory, and the task execution information of the current task execution stage stored in the shared memory is stored by the deep processing agent.

[0145] For example, after the task manager assigns the interactive task to the depth processing agent, it can send the task identifier of the interactive task to the interactive agent. The depth processing agent processes the interactive task at different task execution stages, and the task execution information obtained for the interactive task is also different. Therefore, after the depth processing agent completes a certain task execution stage and obtains the task execution information corresponding to the current task execution stage, the depth processing agent can store the task identifier of the interactive task and the task execution information corresponding to the current task execution stage in the shared memory. The interactive agent can periodically query the task execution information corresponding to the interactive task in the current task execution stage from the shared memory based on the task identifier of the interactive task.

[0146] In the above-mentioned embodiments of the present application, the interactive intelligent agent can obtain the task execution information of the interactive task in the current task execution stage in a variety of ways, thereby ensuring that the interactive intelligent agent can obtain the task execution information of the interactive task in the current task execution stage in a timely manner, and further ensuring that the interactive intelligent agent can timely feedback the response result of the interactive request in the current task execution stage to the user based on the task execution information of the interactive task in the current task execution stage, thereby improving the user experience.

[0147] In some embodiments, as Figure 8 As shown, in the above step 505, the task execution information includes the task description, the task execution stage, and the task execution result. The process of the interactive agent feeding back to the user the response result of the interactive request in the current task execution stage based on the task execution information of the current task execution stage includes the following steps:

[0148] Step 801: The interactive agent generates interactive content based on the task description, task execution stage and task execution result in the task execution information of the current task execution stage.

[0149] For example, after the interactive agent obtains the task execution information corresponding to the current task execution stage of the interactive task, the interactive agent can generate the interactive content required for the interactive response with the user based on the task description, task execution stage, and task execution result contained in the task execution information. The interactive agent can generate the interactive content required for the interactive response with the user based on a pre-trained neural network model, or based on other methods, which are not limited in the embodiments of this application.

[0150] For example, taking the interactive task "Compare today's temperatures in Qingdao and Beijing" as an example, the interactive agent generates the following interactive content based on different task execution information corresponding to different task execution stages: in the task thinking stage, the task execution stage in the task execution information of the interactive task is "task thinking stage", the task execution purpose is "Compare today's temperatures in Qingdao and Beijing", and the task execution result is "Need to obtain today's temperature information in Qingdao and Beijing", and the interactive content is "I am analyzing your request to determine the need to compare today's temperatures in Qingdao and Beijing"; in the data acquisition stage, the task execution stage in the task execution information of the interactive task is "data acquisition stage", the task execution purpose is "Get today's temperatures in Qingdao and Beijing", and the task execution result is "Need to call the weather toolset to get Temperature data", the interaction content is "I am obtaining today's temperature data for Qingdao and Beijing, please wait a moment"; in the data recording stage, the task execution stage in the task execution information of the interactive task is "data recording stage", the task execution purpose is "record the obtained temperature data", the task execution result is "obtained temperature data: Qingdao 25℃, Beijing 28℃", and the interaction content is "I have obtained temperature data: Qingdao is 25℃, Beijing is 28℃"; in the data analysis stage, the task execution stage in the task execution information of the interactive task is "data analysis stage", the task execution purpose is "compare the temperatures of Qingdao and Beijing today", the task execution result is "Beijing's temperature is higher", and the interaction content is "after comparison, Beijing's temperature today is higher, 28℃".

[0151] In step 802, the interactive agent feeds back to the user the response result of the interactive request in the current task execution phase based on the interactive content.

[0152] For example, after the interactive agent generates the interactive content required to interact with the user, it can provide feedback to the user on the response results of the interactive request in the current task execution stage through any one of voice, text, image, video, or a combination of multiple methods.

[0153] In the above-mentioned embodiments of the present application, the interactive agent generates interactive content based on the task description, task execution stage, and task execution results, and based on the interactive content, provides feedback to the user on the response results of the interactive request. In this way, the interactive agent can generate different interactive content at different task execution stages based on different task execution information, thereby providing the user with different response results of the interactive request, thereby improving the user experience.

[0154] In some embodiments, Figure 9 FIG. 1 is a flow chart of an intelligent interaction method according to some other exemplary embodiments. Figure 9 As shown, the intelligent interaction method further performs the following steps:

[0155] In step 901 , the task manager assigns the interactive task to the interactive agent when the task type of the interactive task is a simple task.

[0156] For example, after the task manager determines the task type of an interactive task, if the task type is a simple task, it means that the interactive task has a single task objective, requires less in-depth analysis, and has a short processing time. Therefore, it only needs to be processed by an interactive agent, not a deep processing agent. Accordingly, the task manager can assign the interactive task to an interactive agent for processing. For example, if the interactive task "greeting task" is a simple task, the task manager will assign it to an interactive agent for processing.

[0157] In step 902, the interactive agent executes the interactive task, obtains the task execution result, and based on the task execution result, feeds back the response result of the interactive request to the user.

[0158] For example, after receiving an interactive task assigned by a task manager, the interactive agent can execute the task, obtain the task execution result, and based on the task execution result, provide feedback to the user on the response to the interactive request. For example, if the interactive agent receives the interactive task "greeting task," it can retrieve the user information "Name: Cornetto, Age: 6, Location: Qingdao, Gender: Female" from the shared memory. Based on the user information and the preset task processing strategy or rules, it executes the interactive task and obtains the task execution result "Good afternoon, Cornetto. What a nice weather today." Then, based on the task execution result, the interactive agent provides feedback to the user on the response to the interactive request.

[0159] In the embodiment of the present application, for interactive tasks with a simple task type, the task manager assigns the interactive task to an interactive agent. The interactive agent then executes the interactive task, obtains the task execution result, and based on the task execution result, provides feedback to the user on the response result of the interactive request. In this way, for interactive tasks with a simple task type, the interactive agent can quickly provide feedback to the user on the response result of the interactive request, improving the user experience.

[0160] In some embodiments, the intelligent interaction method further performs the following steps: when the interactive agent determines that the interactive task has failed to be executed based on the task execution information, the interactive agent terminates the execution of the interactive task by the depth processing agent.

[0161] For example, after the interactive agent obtains the task execution information corresponding to the current task execution stage of the interactive task, if it determines based on the task execution information that the interactive task execution has failed, such as if the task execution result in the task execution information is empty or the task execution stage is unreachable (i.e., the task cannot be completed or an error occurs), the interactive agent can terminate the execution of the interactive task by the depth processing agent. Furthermore, the interactive agent can also analyze the reasons for the interactive task execution failure based on the task execution information corresponding to each task execution stage and provide feedback on the interactive response to the user.

[0162] In the above embodiment of the present application, if the interactive agent determines based on the task execution information that the interactive task has failed, it terminates the execution of the interactive task by the depth processing agent. In this way, if the interactive task fails, the interactive agent promptly terminates the execution of the interactive task by the depth processing agent, which can effectively reduce the resource waste of the interactive task on the depth processing agent and speed up the processing of subsequent interactive tasks.

[0163] In the method provided by the embodiment of the present application, the task manager obtains the interactive task corresponding to the user's interactive request; the task manager determines the task type of the interactive task; if the task type of the interactive task is a complex task, the task manager assigns the interactive task to the deep processing agent; the deep processing agent executes the interactive task and obtains the task execution information of the interactive task at each task execution stage; during the process of the deep processing agent executing the interactive task, the interactive agent obtains the task execution information of the current task execution stage from the deep processing agent, and based on the task execution information of the current task execution stage, feedbacks the response result of the interactive request at the current task execution stage to the user until the deep processing agent completes the execution of the interactive task. In this way, if the task type of the interactive task is a complex task, the task manager assigns the interactive task to the deep processing agent for processing, which can meet the needs of in-depth analysis of complex tasks. At the same time, during the execution of the interactive task, the interactive agent obtains the task execution information of the interactive task at different task execution stages in real time, and based on the task execution information, feedbacks the response result of the interactive request to the user until the interactive task is completed. This can ensure real-time interaction with the user during the execution of the interactive task, meet the real-time response requirements, and improve the user experience.

[0164] Figure 10 FIG. 1 is an interaction diagram of an intelligent interaction system according to some exemplary embodiments. Figure 10 As shown, the specific processing process includes the following steps:

[0165] Step 1001: The user initiates an interaction request to the intelligent interaction system;

[0166] Step 1002: The task manager obtains the interactive task corresponding to the user's interactive request and determines the task type of the interactive task.

[0167] Step 1003: When the task type of the interactive task is a simple task, the task manager assigns the interactive task to the interactive agent.

[0168] Step 1004: The interactive agent performs the interactive task and obtains the task execution result.

[0169] Step 1005: Feedback an interactive response to the user based on the task execution result.

[0170] Step 1006: When the task type of the interactive task is a complex task, the task manager assigns the interactive task to the depth processing agent.

[0171] Step 1007: The deep processing agent executes the interactive task and obtains task execution information of the interactive task at different task execution stages.

[0172] Step 1008: The interactive agent sends a task query request message to the depth processing agent, wherein the task query request message carries the task identifier of the interactive task.

[0173] Step 1009: The depth processing agent sends a task query response message to the interaction agent, wherein the task query response message carries task execution information of the interaction task.

[0174] Step 1010: The interactive agent feeds back an interactive response to the user based on the task execution information until the interactive task is completed.

[0175] With reference to the foregoing embodiments, an embodiment of the present application further provides an intelligent interaction system, which includes a task manager, an interactive agent, and a depth processing agent; the task manager, the interactive agent, and the depth processing agent implement the above-mentioned intelligent interaction method.

[0176] With reference to the foregoing embodiment, an embodiment of the present application further provides a display device, including:

[0177] monitor;

[0178] A controller is coupled to the display and is configured to:

[0179] Get the user's interaction request;

[0180] The interactive request is processed by the above-mentioned intelligent interactive system to obtain a response result of the interactive request;

[0181] Feedback the response result of the interaction request to the user.

[0182] For ease of explanation, the above description has been made in conjunction with specific embodiments. However, the above discussion of some embodiments is not intended to be exhaustive or to limit the embodiments to the specific forms disclosed above. Based on the above teachings, various modifications and variations can be obtained. The above embodiments are selected and described to better explain the content of this disclosure, thereby enabling those skilled in the art to better use the embodiments.

Claims

1. An intelligent interaction method, characterized in that: The method is applied to an intelligent interactive system, which includes a task manager, an interactive agent, and a depth processing agent. The method includes: The task manager obtains the interactive task corresponding to the user's interactive request; The task manager determines the task type of the interactive task; The task manager assigns the interactive task to the depth processing agent when the task type of the interactive task is a complex task; The depth processing agent executes the interactive task and obtains task execution information of the interactive task at each task execution stage; During the process of the deep processing agent executing the interactive task, the interactive agent obtains the task execution information of the current task execution stage from the deep processing agent, and based on the task execution information of the current task execution stage, feeds back to the user the response result of the interactive request in the current task execution stage until the deep processing agent completes the execution of the interactive task.

2. The method according to claim 1, characterized in that The method further comprises: The task manager assigns the interactive task to the interactive agent when the task type of the interactive task is a simple task; The interactive agent executes the interactive task, obtains a task execution result, and based on the task execution result, feeds back a response result of the interactive request to the user.

3. The method according to claim 1, characterized in that The determining the task type of the interactive task includes: The task manager performs semantic understanding processing on the interactive task to obtain semantic information of the interactive task; The task manager performs task type discrimination processing on the semantic information to obtain the task type of the interactive task.

4. The method according to claim 1, wherein In the case where there are multiple depth processing agents, allocating the interaction task to the depth processing agents includes: The task manager obtains the depth processing capability of each depth processing agent and the number of interaction tasks to be processed; The task manager determines a target depth processing agent among the depth processing agents based on the depth processing capabilities of the depth processing agents and the number of interactive tasks to be processed; The task manager assigns the interactive task to the target depth processing agent.

5. The method according to claim 1, wherein The deep processing agent includes a task management sub-agent, at least one task execution sub-agent and a tool set. The deep processing agent executes the interactive task and obtains task execution information of the interactive task at each task execution stage, including: The task management sub-agent performs task decomposition processing on the interactive task to obtain sub-interaction tasks of the interactive task, and assigns the sub-interaction tasks to the at least one task execution sub-agent; The at least one task execution sub-agent calls the tool set to execute the sub-interaction task, and obtains task execution information of the sub-interaction task at each task execution stage; The task management sub-agent generates task execution information of the interactive task at each task execution stage based on the task execution information of the sub-interactive task at each task execution stage.

6. The method according to claim 1, characterized in that The interactive agent obtains task execution information of the current task execution phase from the depth processing agent, including: The interactive agent sends a task query request message to the depth processing agent; receives a task query response message fed back by the depth processing agent, wherein the task query response message carries task execution information of the current task execution stage; or The interactive agent receives a task information update message sent by the depth processing agent, wherein the task information update message carries task execution information of the current task execution phase; or The interactive agent queries the task execution information of the current task execution stage from the shared memory, and the task execution information of the current task execution stage stored in the shared memory is stored by the deep processing agent.

7. The method according to claim 1, characterized in that The task execution information includes a task description, a task execution stage, and a task execution result. Feedback of a response result of the interaction request in the current task execution stage to the user based on the task execution information of the current task execution stage includes: The interactive agent generates interactive content based on the task description, the task execution stage and the task execution result in the task execution information of the current task execution stage; The interactive agent feeds back to the user a response result of the interactive request in the current task execution phase based on the interactive content.

8. The method according to claim 1, characterized in that The method further comprises: When the interactive agent determines based on the task execution information that the interactive task has failed to be executed, the interactive agent terminates the execution of the interactive task by the depth processing agent.

9. An intelligent interactive system, characterized in that: The intelligent interaction system includes a task manager, an interactive agent and a depth processing agent; the task manager, the interactive agent and the depth processing agent implement the method according to any one of claims 1 to 8.

10. A display device, characterized in that: include: monitor; a controller coupled to the display and configured to: Get the user's interaction request; Processing the interaction request through the intelligent interaction system according to claim 9 to obtain a response result of the interaction request; Feedback the response result of the interaction request to the user.

Images