Task processing method, system and equipment and storage medium

By introducing a dynamic scheduling mechanism of central control module and functional module in the task processing system, the problems of low efficiency and high resource consumption in multi-task processing are solved, the scenario adaptability and user experience are improved, and more efficient Task processing.

CN118963949APending Publication Date: 2024-11-15BEIJING BAIDU NETCOM SCI & TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411035906.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-30
Publication Date
2024-11-15

AI Technical Summary

Technical Problem

When handling multitasking, large models have problems such as low processing efficiency in complex scenarios, poor scenario adaptability, high resource usage and poor user experience. Especially in scenarios where multiple rounds of dialogue are required to clarify the needs, it is difficult for a general model to provide the optimal solution. , resulting in lengthy processing, inefficient and excessive computing resource consumption.

Method used

Using a task processing system including a central control module and multiple functional modules, the central control module executes quality scheduling tasks to appropriate functional modules based on task request information and the tasks of the functional module, and evaluates the task execution results, and dynamically adjusts task allocation. Improve response speed and consistency.

Benefits of technology

By focusing on functional modules to handle tasks in specific scenarios, the efficiency and accuracy of multi-task processing are improved, resource consumption is reduced, user experience is improved, and faster and coherent task response is achieved.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118963949A_ABST
    Figure CN118963949A_ABST
Patent Text Reader

Abstract

The invention provides a task processing method, system and device and a storage medium, the task processing method is applied to the task processing system comprising a central control module and a plurality of function modules, and each function module corresponds to at least one scene; the method comprises the steps that a central control module receives current task request information and determines a scene corresponding to the current task request information; according to the scene and at least one of task execution quality of a function module in the task processing system, scheduling current task request information to the corresponding function module; the function module executes a corresponding task according to the current task request information and outputs a task execution result; and the central control module evaluates the task execution result to determine the task execution quality of the function module. The multi-task processing effect can be improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present disclosure relates to the field of computer technology, and in particular to technical fields such as neural network models, large language models, and intelligent services. Background Art

[0002] With the development of artificial intelligence, the business processing capabilities of neural network models have been greatly improved, and large models that can handle multiple tasks have emerged. At present, large models are generally used in conjunction with plug-ins to handle multiple tasks, and specific functions or services are implemented by adding specific plug-ins. For example, multiple plug-ins are set for a large model, and each plug-in corresponds to a different function or service; when the large model receives a task, the corresponding plug-in is selected to implement the task. Since this type of large model is a general model, it is impossible to give the optimal solution for a specific scenario, resulting in poor multi-task processing effect of the large model. Summary of the invention

[0003] The present disclosure provides a task processing method, apparatus, device and storage medium.

[0004] According to one aspect of the present disclosure, a task processing method is provided, which is applied to a task processing system including a central control module and multiple functional modules, each functional module corresponding to at least one scenario; the method includes:

[0005] The central control module receives the current task request information, determines the scenario corresponding to the current task request information, and dispatches the current task request information to the corresponding functional module according to at least one of the scenario and the task execution quality of the functional module in the task processing system;

[0006] The function module executes the corresponding task according to the current task request information and outputs the task execution result; and,

[0007] The central control module evaluates the task execution results to determine the task execution quality of the functional module. A task processing system includes a central control module and multiple functional modules, each of which corresponds to at least one scenario;

[0008] The central control module is used to receive the current task request information, determine the scenario corresponding to the current task request information; dispatch the current task request information to the corresponding functional module according to at least one of the scenario and the task execution quality of the functional module in the task processing system; and evaluate the task execution result of the functional module to determine the task execution quality of the functional module;

[0009] The function module is used to execute the corresponding task according to the current task request information and output the task execution result.

[0010] According to another aspect of the present disclosure, there is provided an electronic device, comprising:

[0011] at least one processor; and

[0012] a memory communicatively connected to the at least one processor; wherein,

[0013] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute any method in the embodiments of the present disclosure.

[0014] According to another aspect of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to enable the computer to execute any method according to the embodiments of the present disclosure.

[0015] According to another aspect of the present disclosure, a computer program product is provided, including a computer program, which implements any method according to the embodiments of the present disclosure when executed by a processor.

[0016] The present disclosure sets up a task processing system including a central control module and multiple functional modules, and utilizes the system to realize multi-task execution; specifically, the central control module schedules a suitable functional module to execute the task according to the scenario corresponding to the current task request information and the current task execution quality of each functional module in the task processing system; and after the scheduled functional module executes the task, its task execution result is evaluated, and the evaluation result can be used as a basis for the central control module to schedule the subsequent received task request information. Through this cyclic process, it is possible to realize dynamic adjustment of task requests, improve response speed and consistency, and thus improve the multi-task processing effect.

[0017] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present disclosure.

[0019] Figure 1 is a schematic diagram of an application scenario according to an embodiment of the present disclosure;

[0020] Figure 2 is a flowchart of a task processing method according to an embodiment of the present disclosure;

[0021] Figure 3 It is a schematic diagram of an implementation method of task scheduling by a central control module in a task processing method according to an embodiment of the present disclosure;

[0022] Figure 4 is a schematic diagram of an implementation of a task processing method according to an embodiment of the present disclosure;

[0023] Figure 5 is a structural diagram of a task processing system according to an embodiment of the present disclosure;

[0024] Figure 6 is a structural diagram of a task processing system according to an embodiment of the present disclosure;

[0025] Figure 7 A schematic block diagram of an example electronic device 700 that may be used to implement embodiments of the present disclosure is shown. DETAILED DESCRIPTION

[0026] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope of the present disclosure. Similarly, for the sake of clarity and conciseness, the description of well-known functions and structures is omitted in the following description.

[0027] The “and / or” in the embodiments of the present disclosure indicates that three relationships may exist. For example, A and / or B may indicate the following three situations: A exists alone, A and B exist at the same time, and B exists alone. The term “at least one” herein indicates any combination of at least two of any one or more of a plurality of types. For example, at least one of A, B, and C may indicate any one or more elements selected from the set consisting of A, B, and C. The terms “first” and “second” herein refer to and distinguish between multiple similar technical terms, and do not mean to limit the order or to limit the meaning to only two. For example, the first feature and the second feature refer to two types / two features. The first feature may be one or more, and the second feature may also be one or more.

[0028] In the related art, for multi-task scenarios, a large model is generally used in conjunction with plug-ins to handle multiple tasks, and specific functions or services are realized by adding specific plug-ins. For example, multiple plug-ins are set for a large model, and each plug-in corresponds to a different function or service; when the large model receives a task, the corresponding plug-in is selected to realize the task. Since this type of large model is a general model, it is impossible to give the optimal solution for a specific scenario, resulting in poor task processing effect of the large model. In addition, for many specific scenarios, the interaction of plug-ins also requires multiple rounds. For example, in shopping scenarios, it is necessary to ask clearly about many parameters such as model, purpose, color, quantity, brand, price, etc., in order to finally give the product recommendation expected by the user; frequent multiple rounds of interaction consume a lot of computing resources, which limits the ability of the large model to handle multiple tasks at the same time, and multiple rounds of dialogue lead to lengthy and inefficient processing, affecting the user experience. Large models with multiple plug-ins are usually general models, which cannot give the optimal solution for specific scenarios and have limited customization. In order to improve the performance of large models with set categories, it is common practice to write complex prompt words to guide the large model to clarify these scenarios. This will lead to two problems. On the one hand, in order to prevent the large model from having "hallucinations", a large number of constraints and example descriptions are required; on the other hand, the effectiveness of using such general large models will be reduced.

[0029] In summary, the existing large models with multiple plug-ins have at least the following problems when handling multi-scenario tasks:

[0030] Inefficient processing of complex scenarios: For scenarios that require multiple rounds of dialogue to clarify requirements, large models are relied upon to handle all interaction details, resulting in lengthy and inefficient processing.

[0031] Poor scenario adaptability: It is difficult for a general large model to provide the most optimized solution for each specific scenario, and the degree of customization is limited;

[0032] High resource usage: Frequent multi-round interactions consume a lot of computing resources, limiting the model's ability to handle multiple tasks simultaneously;

[0033] Poor user experience: Response delays and frequent information confirmation affect the user experience, especially when multiple tasks are processed concurrently, the user experience is even worse.

[0034] In order to solve the above problems, an embodiment of the present disclosure proposes a task processing method, which can process tasks in different scenarios. Figure 1 is a schematic diagram of an application scenario according to an embodiment of the present disclosure, such as Figure 1As shown, the schematic diagram of the application scenario of the embodiment of the present disclosure may include but is not limited to a terminal device 110 and a task processing system 120, and the terminal device 110 and the task processing system 120 may communicate with each other through any type of wired or wireless network. Specifically, the terminal device 110 may receive information input by a user, or actively obtain information from a user; for example, receiving information input by a user for requesting a task, or using sensors to obtain physical data such as the user's movement speed, body temperature, heart rate, etc. The terminal device 110 may also send the received or actively obtained information to the task processing system 120 through a wired or wireless network with the task processing system 120. The task processing system 120 may be used to receive information from the terminal device 110 and execute the task corresponding to the information.

[0035] The terminal device 110 proposed in the embodiment of the present disclosure includes but is not limited to mobile phones, computers, intelligent voice interaction devices, smart home appliances, vehicle-mounted terminals, game consoles, e-book readers, multimedia playback devices, wearable devices and other electronic devices; the task processing system 120 may include a server or a server cluster. In addition, the embodiment of the present disclosure does not specifically limit the number of terminal devices 110. For example, the application scenario diagram of the embodiment of the present disclosure may include one or more terminal devices 110.

[0036] The disclosed embodiment proposes a task processing method, which can be applied to a task processing system including a central control module and multiple functional modules, each functional module corresponds to at least one scenario, and each functional module can be used to process the task request of at least one scenario corresponding to it. The central control module can also be called a central control module, which can be used as a management and control module for multiple functional modules, and is used to dynamically schedule task allocation according to the most recent activity status and response quality of the functional modules; the central control module receives task request information in real time, and according to the current task request information and the response quality of each functional module, schedules the currently received task request information to the appropriate functional module, thereby realizing dynamic allocation of task request information.

[0037] Figure 2 is a flowchart of a task processing method according to an embodiment of the present disclosure, including:

[0038] S210, the central control module receives the current task request information, determines the scenario corresponding to the current task request information, and dispatches the current task request information to the corresponding functional module according to at least one of the scenario and the task execution quality of the functional module in the task processing system;

[0039] S220, the function module executes the corresponding task according to the current task request information and outputs the task execution result;

[0040] S230: The central control module evaluates the task execution result to determine the task execution quality of the functional module.

[0041] In some embodiments, the central control module continuously receives task request information, and for newly received task request information, the new task request information is scheduled according to the scenario corresponding to the task request information and the task execution quality of each functional module before receiving the task request information, thereby realizing dynamic scheduling of tasks and improving the response speed and consistency of the entire task processing system. For example, in the above process, after step S230 is executed, the execution returns to step S210; the task execution quality of the functional module determined in step S230 will be used as one of the bases for scheduling the task request information in step S210.

[0042] Since multiple functional modules corresponding to different scenarios are set up, each functional module is used to process the task request of at least one scenario corresponding to it. The central control module dispatches the task request information to the corresponding functional module, so that each functional module can focus on a specific scenario, avoiding the problem of not being able to achieve the optimal solution in different scenarios due to the use of a general model, and improving the task processing capability of the entire system.

[0043] Figure 3 FIG. 1 is a schematic diagram of an implementation method of a central control module for performing task scheduling in a task processing method according to an embodiment of the present disclosure. Figure 3 As shown, the central control module controls N functional modules, where N is an integer greater than 1. The central control module continuously receives task request information. When receiving the current task request information, the current task request information is dispatched to the appropriate functional module according to the scenario corresponding to the task request information and the task execution quality of each functional module in the task processing system at the previous moment (i.e., the moment before receiving the current task request information); the functional module executes the task corresponding to the current task request information and outputs the task execution result; the central control module evaluates the task execution result to determine the task execution quality of the functional module, and the task execution quality will be used as one of the bases for scheduling the task request information when the task request information is received at the next moment. In this way, the central control module realizes the dynamic scheduling of each functional module in the system.

[0044] In the disclosed embodiments, the task request information may correspond to a persistent task or a temporary task. For example, online shopping is a persistent task. During the online shopping process, the user will continuously send multiple task request information, such as task request information for realizing a series of requirements such as querying goods, selecting goods, generating orders, and submitting orders; for such task request information, the central control device may continuously assign multiple task request information to the functional module used to realize online shopping, thereby improving the consistency of task realization. For another example, querying the weather forecast is a temporary task. When querying the weather forecast, the user generally only needs to submit one or a small number of task request information, such as entering a text or voice command of "What is the weather like today", or clicking the corresponding function button in the relevant application (Application, APP); for such task request information, the central control device may temporarily assign one or a small number of task request information to the functional module used for weather forecast, thereby improving the flexibility of task realization.

[0045] In an embodiment of the present application, the central control module may include an identifier and a discriminator; wherein the identifier is used to identify the type of current task request information (such as text, image, voice, etc.), and / or key information (such as keywords in text or voice information, text information in an image, face information in an image, etc.); the discriminator is used to discriminate the scene corresponding to the type of the current task request information and / or the key information according to the type of the current task request information and / or the key information, and pre-set discrimination rules, thereby determining the scene corresponding to the current task request information.

[0046] For example, one discrimination rule includes: when it is identified that the current task request information contains words related to weather and keywords used to express an inquiry, it can be determined that the scenario corresponding to the current task request information is querying the weather.

[0047] According to the above-mentioned discrimination rules, when the central control module receives the current task request information that meets the discrimination rules, it can dispatch the current task request information to the functional module corresponding to the scene, such as the functional module responsible for providing weather forecast.

[0048] When scheduling the current task request information, the central control module can directly send the received current task request information to the scheduled functional module, or extract the task requirements from the received current task request information and send the task requirements to the scheduled functional module, thereby reducing the amount of information transmitted between the central control module and the functional module and alleviating the burden of information transmission. For example, the current task request information received by the central control module is a voice sent by the user, and the voice content is "Do I need to bring an umbrella when I go out today?" The central control module can extract the task requirement from the voice content as "Check the weather", and then send the instruction to the scheduled functional module responsible for providing weather forecasts, without sending the complete voice content to the functional module responsible for providing weather forecasts, thereby reducing the amount of information sent by the central control module to the functional module.

[0049] In the embodiment of the present application, the central control module can also be implemented using a neural network model, which uses training samples containing task request information and corresponding scenes for supervised training, and can determine the scene corresponding to the current task request information based on the input current task request information. The specific training method is not repeated here.

[0050] In an embodiment of the present application, the central control module serves as a central controller and may be responsible for receiving and distributing multimodal signals (including but not limited to voice, text, images, videos, etc.), while each functional module serves as an independent model and is specialized in handling tasks of a specific type or field. Each functional module has the ability to continuously respond and provide feedback, and can determine whether more input information is needed.

[0051] In some embodiments, the functional module may include a neural network model. For example, the functional module may include a large language model (LLM), or a multimodal large language model (MLLM). By using a neural network model as a functional module, neural network technology can be used to solve task requests in various scenarios and improve task processing effects. Each functional module is independently responsible for continuously responding and feeding back the results of task execution and whether more information input is required. The task request information involved in the embodiment of the present application may be multimodal, including signals and multimedia information that can be obtained by various sensors. In the embodiment of the present application, the functional module uses an attention mechanism to process the task request information received by each in parallel. The attention mechanism means that each functional module only focuses on the task request information of the scene most relevant to it, and the central control module allocates task request information to each functional module, thereby solving the problem of using a single model to process multiple types of tasks in the prior art and improving the processing efficiency and accuracy of the task processing system.

[0052] In the embodiment of the present application, multiple functional modules can run in parallel to realize the simultaneous processing of different tasks. For example, when processing a shopping scene, a functional module for processing weather queries or a functional module for processing TV program playback can be temporarily called to realize other needs such as querying the weather or the status of TV program playback. The functional module for processing shopping scenes can be run simultaneously with multiple functional modules such as the functional module for processing weather queries and the functional module for processing TV program playback, thereby realizing the parallel implementation of multiple tasks and improving the consistency of tasks and the overall response speed of the system. In the prior art, since a single model is used to process multiple tasks, when the task scenario changes, the single model needs to switch repeatedly between different tasks, resulting in reduced consistency in task execution and a slow response speed of the model.

[0053] In some embodiments, when the central control module performs scene recognition and task scheduling, it dynamically adjusts signal distribution according to the most recent activity status and response quality of the functional module to improve response speed and consistency. The most recently active functional module has a higher priority, that is, under the same conditions, the central control module preferentially allocates task request information to the most recently active functional module, that is, it tends to maintain the continuity of functional module scheduling. For example, when receiving the current task request information, if the task execution quality level of functional module 1 and functional module 2 before the current moment is the same, and functional module 1 is the execution module of the task request information at the previous moment, then the current task request information is allocated to functional module 1.

[0054] In addition, when performing task scheduling, the task execution quality of the functional module is considered, and the task execution quality can be reflected in the consistency, accuracy or effectiveness of the task execution results. Under the same conditions, the central control module preferentially allocates task request information to the functional module with the highest task execution quality. For example, when receiving the current task request information, if neither functional module 1 nor functional module 2 is the execution module of the task request information at the previous moment, and the task execution quality of functional module 2 is higher, then the current task request information is allocated to functional module 2.

[0055] In one example, the central control module evaluates the task execution results to determine the task execution quality of the functional module, including:

[0056] The central control module determines at least one of the consistency, accuracy and effectiveness of the task execution results;

[0057] The central control module determines the task execution quality of the functional module according to at least one of consistency, accuracy and effectiveness.

[0058] The consistency of the task execution results of the functional module may include at least one of the consistency between the task request information received by the functional module and the execution result of the task request information by the functional module, and the correlation between the adjacent task execution results output by the functional module. Taking the large language model as a functional module as an example, the large language model is used to realize the dialogue with the user; if there is no correlation between the question received by the large language model and the answer output by the large language model for the question, and / or, there is no correlation between the multiple answers given by the large language model for multiple questions continuously input by the user, then the consistency of the large language model is poor, indicating that the task execution quality of the large language model for the current task is poor. Among them, the correlation between the aforementioned questions and answers, and the correlation between adjacent answers can be determined by analyzing keywords, etc.; if the keywords in the question and the keywords in the answer do not belong to the same type or field, it can be determined that the correlation between the question and the answer is poor; if the keywords in the adjacent answers do not belong to the same type or field, the correlation between the adjacent answers can be poor.

[0059] The accuracy of the task execution results of a functional module may refer to whether the task execution results conform to common sense or logic.

[0060] The validity of the task execution results of the functional module refers to whether there is substantive feedback on the received task request information. Taking the large language model as a functional module as an example, the large language model is used to realize the dialogue with the user; in response to the user's question, if the large language model gives an answer such as "I don't have this function yet" or "I don't understand what you mean", such an answer is not valid.

[0061] Through the consistency, accuracy and effectiveness of the functional module responses, the task execution quality of each functional module can be evaluated more comprehensively, thereby providing an accurate reference basis for the scheduling of task request information, scheduling the task request information to the most effective functional module, and improving the task processing effect of the entire system.

[0062] In addition to the consistency, accuracy and effectiveness of the responses of multiple functional modules, the central control device can also determine which functional module should respond and process the current task request information according to the importance of the task, which can not only make the task request information of the same scene processed by the same functional module, but also better respond to the needs of multiple scenes. In some embodiments, the importance of the task can be set in advance, and the importance standard can be set according to the specific situation; for example, if the timeliness requirement of a task is higher, the importance of the task is higher. In one example, the user is shopping online, and the task request information corresponding to the online shopping scene is currently being received; when a task request information asking about the weather or a TV program is received, the user's inquiry will not be rejected because the functional module responsible for online shopping is processing the task, but the task will be dispatched to the corresponding functional module responsible for the weather or the functional module responsible for video playback and query; at the same time, the functional module responsible for online shopping is still running. Thus, the entire task processing system can process multiple different tasks simultaneously, making the user experience more natural and smooth.

[0063] In actual applications, the central control module can schedule the functional modules in at least the following ways:

[0064] Method 1:

[0065] After the central control module receives the current task request, if the scenario corresponding to the current task request information is the same as the scenario corresponding to the historical task request information received by the central control module, and the task execution quality of the first functional module meets the predetermined requirements, the task request information is scheduled to the first functional module; wherein the first functional module includes a functional module that executes the historical task request information.

[0066] Method 2:

[0067] After the central control module receives the current task request, if the scenario corresponding to the current task request information is the same as the scenario corresponding to the historical task request information received by the central control module, and the task execution quality of the first functional module does not meet the predetermined requirements; or if the scenario corresponding to the current task request information is different from the scenario corresponding to the historical task request information received by the central control module,

[0068] Determine at least one second functional module corresponding to the scenario corresponding to the current task request information;

[0069] Selecting a function module with the highest task execution quality from the at least one second function module, and dispatching the current task request information to the function module with the highest task execution quality;

[0070] Among them, the first functional module includes a functional module for executing the historical task request information.

[0071] Combining the above two methods, it can be seen that the central control module tends to keep the functional module of the current task unchanged when scheduling. As long as the scenario of the received task request information does not change and the task execution quality of the functional module meets the requirements, the subsequent task request information will continue to be scheduled to the functional module (i.e., method one); when the central control module determines that the scenario of the received task request information has not changed, if the task execution quality of the functional module does not meet the requirements, it is very likely that the scenario of the user request information newly input by the user has changed, resulting in the functional module previously scheduled by the central control module being unable to provide correct feedback in the new scenario (but the central control module has not yet discovered the change), and it is necessary to switch scheduling (i.e., the first case of method two); or, if the central control module has determined that the scenario of the received task request information has changed, the central control module schedules the user request information to other functional modules (i.e., the second case of method two). When the progress is switched, the central control module selects the functional module with the highest quality of historical task execution from the multiple functional modules corresponding to the current scenario.

[0072] It can be seen that the scheduling method proposed in the embodiment of the present disclosure can maintain the continuity of task execution, reduce switching between functional modules, make the system response smoother, and reduce user waiting time; and the central control module can actively determine whether the scene has changed, or use a posteriori method to determine whether the scene has changed (that is, judge according to the task execution quality of the functional module), so that it can respond to task switching in a timely manner, schedule tasks to the currently suitable functional modules, and improve the overall execution effect of the system.

[0073] In some embodiments, the task processing system applied in the embodiment of the present disclosure may also include a multimodal signal integration module, which can receive multimodal signals from multiple information sources, wherein the signal source may include at least one of a multimedia signal receiving device and a sensor, and other devices with the function of receiving and transmitting multimedia signals. The multimodal signal integration module of the task processing system can integrate the multimodal signals to obtain one or more current task request information, and send the one or more task request information to the central control module; thereafter, the central control module can continuously schedule the current task request information in the manner described in the above content. Using this method, the system can realize multimodal information processing and meet the needs of diversified information and request methods.

[0074] Figure 4 FIG. 1 is a schematic diagram of a task processing method according to an embodiment of the present disclosure. Figure 4As shown, the multimodal signal integration module receives multimodal signals from multiple multimedia signal receiving devices, sensors and other information sources, integrates the multimodal signals, continuously generates task request information, and after generating the current task request information, sends the current task request information to the central control module. The central control module schedules the current task request information to the appropriate functional module based on the scenario corresponding to the current task request information and the task execution quality of each functional module. Afterwards, the central control module evaluates and makes decisions based on the task execution results of the functional modules to determine the task execution quality of each functional module, and uses the task execution quality as one of the bases for scheduling after receiving the task request information at the next moment. Figure 4 In the example shown, the central control module can schedule multiple functional modules at the same time. For example, the task request information corresponding to the online shopping scenario is continuously scheduled to the functional module 1 used to implement online shopping, the task request information corresponding to the weather forecast scenario is temporarily scheduled to the functional module 2 used to implement the weather forecast, and the task request information corresponding to the program query is temporarily scheduled to the functional module 3 used to implement the program query; multiple functional modules can execute corresponding functions respectively at the same time.

[0075] In some implementations, the central control module can determine the scene corresponding to the current task request information based on the request type and information type of the current task information. The information type includes but is not limited to images, videos, voices, and texts. In this way, the scene corresponding to the current task request can be determined, so that the current task request can be dispatched to the functional module corresponding to the scene, giving full play to the advantages of each functional module, enhancing the adaptability of the entire task processing system in different scenes, and providing users with more personalized and efficient solutions.

[0076] The disclosed embodiment also provides a task processing system. Figure 5 is a structural diagram of a task processing system according to an embodiment of the present disclosure, including a central control module 510 and multiple functional modules 520; wherein,

[0077] The central control module 510 is used to receive the current task request information, determine the scenario corresponding to the current task request information; dispatch the current task request information to the corresponding functional module according to at least one of the scenario and the task execution quality of the functional module in the task processing system; and evaluate the task execution result of the functional module to determine the task execution quality of the functional module;

[0078] Functional module 520 is used to execute the corresponding task according to the current task request information and output the task execution result.

[0079] In some embodiments, the central control module 510 is used to:

[0080] Determine at least one of the consistency, accuracy and effectiveness of the results of task execution;

[0081] The task execution quality of the functional module is determined based on at least one of consistency, accuracy and effectiveness.

[0082] In some embodiments, the central control module 510 is used to:

[0083] If the scenario corresponding to the current task request information is the same as the scenario corresponding to the historical task request information received by the central control module, and the task execution quality of the first functional module meets the predetermined requirements, dispatching the task request information to the first functional module;

[0084] Among them, the first functional module includes a functional module for executing historical task request information.

[0085] In some embodiments, the central control module 510 is used to:

[0086] When the scenario corresponding to the current task request information is the same as the scenario corresponding to the historical task request information received by the central control module, and the task execution quality of the first functional module does not meet the predetermined requirements; or when the scenario corresponding to the current task request information is different from the scenario corresponding to the historical task request information received by the central control module,

[0087] Determine at least one second functional module corresponding to the scenario corresponding to the current task request information;

[0088] Selecting a function module with the highest task execution quality from at least one second function module, and dispatching the current task request information to the function module with the highest task execution quality;

[0089] Among them, the first functional module includes a functional module for executing the historical task request information.

[0090] The present application also provides a task processing system. Figure 6 is a structural diagram of a task processing system according to an embodiment of the present disclosure, comprising: a central control module 510, a plurality of functional modules 520 and a multi-modal signal integration module 610: wherein,

[0091] The multimodal signal integration module 610 is used to receive multimodal signals from multiple information sources, integrate the multimodal signals according to at least one of the type and content of the multimodal signals to obtain one or more current task request information; and send the one or more current task request information to the central control module.

[0092] In some embodiments, the central control module 510 is used to:

[0093] Identify the request type and information type of the current task request information;

[0094] According to the request type and information type of the current task request information, the scenario corresponding to the current task request information is determined.

[0095] For the description of specific functions and examples of each module and submodule of the device in the embodiment of the present disclosure, reference can be made to the relevant description of the corresponding steps in the above method embodiment, which will not be repeated here.

[0096] In the technical solution disclosed in the present invention, the acquisition, storage and application of personal information of users involved are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0097] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium and a computer program product.

[0098] Figure 7 A schematic block diagram of an example electronic device 700 that can be used to implement an embodiment of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or required herein.

[0099] like Figure 7 As shown, the device 700 includes a computing unit 701, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 702 or a computer program loaded from a storage unit 708 into a random access memory (RAM) 703. In the RAM 703, various programs and data required for the operation of the device 700 can also be stored. The computing unit 701, the ROM 702, and the RAM 703 are connected to each other via a bus 704. An input / output (I / O) interface 705 is also connected to the bus 704.

[0100] A number of components in the device 700 are connected to the I / O interface 705, including: an input unit 706, such as a keyboard, a mouse, etc.; an output unit 707, such as various types of displays, speakers, etc.; a storage unit 708, such as a disk, an optical disk, etc.; and a communication unit 709, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 709 allows the device 700 to exchange data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0101] The computing unit 701 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 701 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 701 performs the various methods and processes described above, such as detection methods. For example, in some embodiments, the detection method may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit 708. In some embodiments, part or all of the computer program may be loaded and / or installed on the device 700 via ROM 702 and / or communication unit 709. When the computer program is loaded into RAM 1003 and executed by the computing unit 701, one or more steps of the detection method described above may be performed. Alternatively, in other embodiments, the computing unit 701 may be configured to perform the detection method in any other appropriate manner (e.g., by means of firmware).

[0102] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0103] The program code for implementing the method of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0104] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0105] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0106] The systems and techniques described herein may be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0107] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship of client and server is generated by computer programs running on respective computers and having a client-server relationship with each other. The server may be a cloud server, a server of a distributed system, or a server combined with a blockchain.

[0108] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and this document does not limit this.

[0109] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the principles of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A task processing method, applied to a task processing system including a central control module and multiple functional modules, each of the functional modules corresponding to at least one scenario; the method comprises: The central control module receives the current task request information and determines the scene corresponding to the current task request information; dispatching the current task request information to a corresponding functional module according to at least one of the scenario and the task execution quality of the functional module in the task processing system; The function module executes the corresponding task according to the current task request information and outputs the task execution result; and, The central control module evaluates the task execution result to determine the task execution quality of the functional module.

2. The method according to claim 1, wherein: The functional module includes a neural network model.

3. The method according to claim 2, wherein: The neural network model includes a multimodal large language model.

4. The method according to any one of claims 1 to 3, wherein: The central control module evaluates the task execution result to determine the task execution quality of the functional module, including: The central control module determines at least one of the consistency, accuracy and effectiveness of the task execution result; The central control module determines the task execution quality of the functional module according to at least one of the coherence, accuracy and effectiveness.

5. The method according to any one of claims 1 to 4, wherein: The dispatching the current task request information to the corresponding functional module according to at least one of the scenario and the task execution quality of the functional module in the task processing system includes: If the scenario corresponding to the current task request information is the same as the scenario corresponding to the historical task request information received by the central control module, and the task execution quality of the first functional module meets the predetermined requirement, dispatching the task request information to the first functional module; Wherein, the first functional module includes a functional module for executing the historical task request information.

6. The method according to any one of claims 1 to 4, wherein: The dispatching the current task request information to the corresponding functional module according to at least one of the scenario and the task execution quality of the functional module in the task processing system includes: In the case where the scenario corresponding to the current task request information is the same as the scenario corresponding to the historical task request information received by the central control module, and the task execution quality of the first functional module does not meet the predetermined requirements; or in the case where the scenario corresponding to the current task request information is different from the scenario corresponding to the historical task request information received by the central control module, Determine at least one second functional module corresponding to the scenario corresponding to the current task request information; Selecting a function module with the highest task execution quality from the at least one second function module, and dispatching the current task request information to the function module with the highest task execution quality; Wherein, the first functional module includes a functional module for executing the historical task request information.

7. The method according to any one of claims 1 to 6, wherein: The task processing system also includes a multimodal signal integration module; Before the central control module receives the current task request information, it also includes: The multimodal signal integration module receives multimodal signals from multiple information sources, and integrates the multimodal signals according to at least one of the type and content of the multimodal signals to obtain one or more current task request information; The multimodal signal integration module sends the one or more current task request information to the central control module.

8. The method according to claim 7, wherein: The plurality of information sources include at least one of a multimedia signal receiving device and a sensor.

9. The method according to any one of claims 1 to 8, wherein: The determining the scenario corresponding to the current task request information includes: The central control module identifies the request type and information type of the current task request information; The central control module determines the scenario corresponding to the current task request information according to the request type and information type of the current task request information.

10. The method according to claim 9, wherein: The information type includes at least one of image, video, voice and text.

11. A task processing system, comprising a central control module and a plurality of functional modules, each of the functional modules corresponding to at least one scenario; wherein: The central control module is used to receive current task request information, determine the scenario corresponding to the current task request information; dispatch the current task request information to the corresponding functional module according to the scenario and at least one of the task execution quality of the functional module in the task processing system; and evaluate the task execution result of the functional module to determine the task execution quality of the functional module; The functional module is used to execute the corresponding task according to the current task request information and output the task execution result.

12. The system according to claim 11, wherein: The central control module is used to: Determining at least one of the consistency, accuracy and effectiveness of the results of the task execution; The task execution quality of the functional module is determined according to at least one of the coherence, accuracy and effectiveness.

13. The system according to claim 11 or 12, wherein: The central control module is used to: If the scenario corresponding to the current task request information is the same as the scenario corresponding to the historical task request information received by the central control module, and the task execution quality of the first functional module meets the predetermined requirement, dispatching the task request information to the first functional module; Wherein, the first functional module includes a functional module for executing the historical task request information.

14. The system according to claim 11 or 12, wherein: The central control module is used to: In the case where the scenario corresponding to the current task request information is the same as the scenario corresponding to the historical task request information received by the central control module, and the task execution quality of the first functional module does not meet the predetermined requirements; or in the case where the scenario corresponding to the current task request information is different from the scenario corresponding to the historical task request information received by the central control module, Determine at least one second functional module corresponding to the scenario corresponding to the current task request information; Selecting a function module with the highest task execution quality from the at least one second function module, and dispatching the current task request information to the function module with the highest task execution quality; Wherein, the first functional module includes a functional module for executing the historical task request information.

15. The system according to any one of claims 11-14, further comprising a multimodal signal integration module; The multimodal signal integration module is used to receive multimodal signals from multiple information sources, integrate the multimodal signals according to at least one of the type and content of the multimodal signals to obtain one or more current task request information; and send the one or more current task request information to the central control module.

16. The system according to any one of claims 11 to 15, wherein: The central control module is used to: Identify the request type and information type of the current task request information; The scenario corresponding to the current task request information is determined according to the request type and information type of the current task request information.

17. An electronic device comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 10.

18. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to cause the computer to execute the method according to any one of claims 1-10.

19. A computer program product comprising a computer program, which, when executed by a processor, implements the method according to any one of claims 1 to 10.