Intelligent question answering system, question answering processing method and device, equipment, storage medium and computer program product

Abstract

The invention discloses an intelligent question answering system, a question answering processing method and device, equipment, a storage medium and a computer program product. The system comprises a scheduling module, an intelligent center module and a task execution module, and execution states and execution details of the scheduling module, the intelligent center module and the task execution module are visualized for a terminal user in the whole process. Wherein the scheduling module is used for receiving a first request sent by the terminal user; the first request comprises the target problem to be processed; the scheduling module is also used for analyzing the to-be-processed target problem to obtain intention analysis information; the intelligent center module is used for determining a planning task category based on the intention analysis information and determining an executable capability matched with the planning task category; and the task execution module is used for executing the executable capability matched with the planning task category to obtain reply information for the to-be-processed target question.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of artificial intelligence (AI), and in particular to an intelligent question answering system, a question answering processing method and device, equipment, a storage medium and a computer program product. BACKGROUND

[0002] With the continuous development and maturity of large model technology, intelligent question answering systems based on large models have sprung up like mushrooms, aiming to provide users with more intelligent and comprehensive services.

[0003] However, all kinds of intelligent question answering systems on the market at home and abroad generally have a black box problem. This black box phenomenon is that the processing process of the intelligent question answering system is not transparent, that is, the user cannot understand how the intelligent question answering system processes the problem and gets the answer, which makes it difficult for the user to evaluate the reliability of the intelligent question answering system, and also limits the user's understanding and trust of the intelligent question answering system, thereby restricting the application and development of the intelligent question answering system in related technologies and hindering the improvement of user experience. SUMMARY

[0004] To solve the technical problems in the related art, the embodiments of the present application provide an intelligent question answering system, a question answering processing method and device, equipment, a storage medium and a computer program product.

[0005] To achieve the above-mentioned purposes, the technical solutions of the embodiments of the present application are as follows:

[0006] In a first aspect, the embodiments of the present application provide an intelligent question answering system, which comprises a scheduling module, an intelligent hub module and a task execution module, the execution state and execution details of the scheduling module, the intelligent hub module and the task execution module are visualized to the terminal user throughout; wherein,

[0007] The scheduling module is configured to receive a first request sent by the terminal user; the first request is used to request a reply to a target problem to be processed, and the first request comprises the target problem to be processed;

[0008] The scheduling module is further configured to analyze the target problem to be processed to obtain intention analysis information;

[0009] The intelligent hub module is configured to determine a planning task category based on the intention analysis information, and determine an executable capability matched with the planning task category;

[0010] The task execution module is configured to execute the executable capability matched with the planning task category to obtain reply information for the target problem to be processed.

[0011] In a second aspect, an embodiment of the present application provides a question and answer processing method, the method comprising:

[0012] receiving a first request sent by a terminal user, the first request being used to request a reply to a target question to be processed, and the first request comprising the target question to be processed;

[0013] analyzing the target question to be processed to obtain intention analysis information;

[0014] determining a planning task category based on the intention analysis information, and determining an executable capability matched with the planning task category;

[0015] executing the executable capability matched with the planning task category to obtain reply information for the target question to be processed.

[0016] In a third aspect, an embodiment of the present application further provides a question and answer processing apparatus, the apparatus comprising:

[0017] a receiving unit configured to receive a first request sent by a terminal user, the first request being used to request a reply to a target question to be processed, and the first request comprising the target question to be processed;

[0018] an analyzing unit configured to analyze the target question to be processed to obtain intention analysis information;

[0019] a first determining unit configured to determine a planning task category based on the intention analysis information;

[0020] a second determining unit configured to determine an executable capability matched with the planning task category;

[0021] an executing unit configured to execute the executable capability matched with the planning task category to obtain reply information for the target question to be processed.

[0022] In a fourth aspect, an embodiment of the present application further provides a question and answer processing device, comprising a processor and a memory for storing a computer program capable of running on the processor;

[0023] When the processor runs the computer program, the processor is configured to perform the steps of the question and answer processing method according to the embodiments of the present application.

[0024] In a fifth aspect, an embodiment of the present application further provides a storage medium having a computer program stored thereon, and the computer program is configured to perform the steps of the question and answer processing method according to the embodiments of the present application when executed by a processor.

[0025] In a sixth aspect, an embodiment of the present application further provides a computer program product comprising a computer program which, when executed by a processor, implements the steps of the question and answer processing method described in the embodiments of the present application.

[0026] The intelligent question and answer system, question and answer processing method, device, storage medium and computer program product provided by the embodiments of the present application, the intelligent question and answer system comprises a scheduling module, an intelligent hub module and a task execution module, the execution state and execution details of the scheduling module, the intelligent hub module and the task execution module are visualized for a terminal user throughout; wherein the scheduling module is configured to receive a first request sent by the terminal user; the first request is configured to request a reply to a target problem to be processed, and the first request comprises the target problem to be processed; the scheduling module is further configured to analyze the target problem to be processed to obtain intention analysis information; the intelligent hub module is configured to determine a planning task category based on the intention analysis information, and determine an executable capability matched with the planning task category; and the task execution module is configured to execute the executable capability matched with the planning task category to obtain reply information for the target problem to be processed. By adopting the scheme of the embodiments of the present application, the execution state and execution details of each module in the processing process of the intelligent question and answer system are visualized, so that the processing process of the innovative intelligent question and answer system is no longer a “black box”, but a transparent process, which enables the terminal user to clearly understand how the intelligent question and answer system processes a problem and obtains an answer, thereby enhancing the understanding and trust of the terminal user for the intelligent question and answer system; in addition, by analyzing the target problem to be processed to obtain intention analysis information, and determining a planning task category and an executable capability matched with the planning task category based on the intention analysis information, and then executing the executable capability matched with the planning task category to obtain reply information, in this way, the reply information output by the intelligent question and answer system will be more accurate and reliable, which helps to improve the reliability of the intelligent question and answer system and the satisfaction of the terminal user. BRIEF DESCRIPTION OF DRAWINGS

[0027] Figure 1 FIG. 1 is a schematic diagram of a ChatGPT intelligent question and answer system in the related art;

[0028] Figure 2 FIG. 2 is a schematic diagram of a domestic Wenxin Yiyang intelligent question and answer system in the related art;

[0029] Figure 3 FIG. 3 is a schematic diagram of the composition structure of the intelligent question and answer system according to an embodiment of the present application;

[0030] Figure 4 FIG. 4 is a schematic diagram of the flow of the question and answer processing method according to an embodiment of the present application;

[0031] Figure 5An architecture schematic diagram of the intelligent question-answering system of the embodiment of the present application;

[0032] Figure 6 An architecture schematic diagram of the intelligent question-answering system of the embodiment of the present application;

[0033] Figure 7 An architecture schematic diagram of the intelligent question-answering system of the embodiment of the present application;

[0034] Figure 8 An architecture schematic diagram of the intelligent question-answering system of the embodiment of the present application;

[0035] Figure 9 An architecture schematic diagram of the intelligent question-answering system of the embodiment of the present application;

[0036] Figure 10 An architecture schematic diagram of the intelligent question-answering system of the embodiment of the present application;

[0037] Figure 11 An architecture schematic diagram of the intelligent question-answering system of the embodiment of the present application;

[0038] Figure 12 An architecture schematic diagram of the intelligent question-answering system of the embodiment of the present application;

[0039] Figure 13 An architecture schematic diagram of the intelligent question-answering system of the embodiment of the present application;

[0040] Figure 14 An architecture schematic diagram of the intelligent question-answering system of the embodiment of the present application;

[0041] Figure 15 An architecture schematic diagram of the intelligent question-answering system of the embodiment of the present application; DETAILED DESCRIPTION

[0042] The present application will be further described below in conjunction with the accompanying drawings and embodiments.

[0043] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used in the description herein is for describing particular embodiments only and is not intended to be limiting of the application.

[0044] At present, with the continuous development and maturity of large model technology, intelligent question and answer systems based on large models have sprung up like mushrooms, aiming to provide users with more intelligent and comprehensive services. However, large models also face some challenges in practical applications, such as being limited to natural language answers. In order to make up for these shortcomings, intelligent question and answer systems usually combine small models or other plugins / tools to provide more accurate and reliable answers.

[0045] However, whether at home or abroad, various intelligent question and answer systems on the market, such as Figure 1 the related art ChatGPT intelligent question and answer system diagram (this ChatGPT intelligent question and answer system is developed by OpenAI abroad), or Figure 2 the related art domestic Wen Xin Yi Yan intelligent question and answer system diagram, generally have the black box problem, that is, the intermediate processing process from problem input to answer output in the intelligent question and answer system is not transparent, that is, when interacting with the intelligent question and answer system, the user can only see the input question and the obtained answer, but the process of how the system converts the question into the answer is like a black box and cannot be directly observed by the user, making it difficult for the user to evaluate the reliability of the system and limiting the user's understanding and trust of the system, thereby restricting the application and development of the related art intelligent question and answer system and hindering the improvement of user experience.

[0046] Based on this, the embodiments of the present application propose an innovative intelligent question and answer system. In various embodiments of the present application, the intelligent question and answer system visualizes the execution state and execution details of each module in the processing process, so that the processing process of the innovative intelligent question and answer system is no longer a "black box", but a transparent process, enabling the end user to clearly understand how the intelligent question and answer system processes the question and obtains the answer, enhancing the end user's understanding and trust of the intelligent question and answer system. In addition, by analyzing the target question to be processed to obtain intention analysis information, and based on the intention analysis information, determining the planning task category and determining the executable ability matching the planning task category, and then obtaining the reply information by executing the executable ability matching the planning task category, in this way, the reply information output by the intelligent question and answer system will be more accurate and reliable, which helps to improve the reliability of the intelligent question and answer system and the satisfaction of the end user.

[0047] The embodiments of the present application provide an intelligent question and answer system, Figure 3 The composition structure diagram of the intelligent question and answer system of the embodiments of the present application is shown as Figure 3 The system includes a scheduling module 31, an intelligent hub module 32, and a task execution module 33, and the execution state and execution details of the scheduling module 31, the intelligent hub module 32, and the task execution module 33 are visualized to the end user throughout; wherein

[0048] The scheduling module 31 is configured to receive a first request sent by the terminal user, wherein the first request is used to request a reply to a target problem to be processed, and the first request comprises the target problem to be processed.

[0049] The scheduling module 31 is further configured to analyze the target problem to be processed to obtain intention analysis information.

[0050] The intelligent core module 32 is configured to determine a planning task category based on the intention analysis information, and determine an executable capability matched with the planning task category.

[0051] The task execution module 33 is configured to execute the executable capability matched with the planning task category to obtain reply information for the target problem to be processed.

[0052] Here, the execution state and execution details of the scheduling module 31, the intelligent core module 32 and the task execution module 33 are visualized to the terminal user, which means that the execution state and execution details of the intermediate processing steps in the processing process of the intelligent question answering system and the results of the intermediate processing steps are visualized to the terminal user, so that the terminal user can clearly understand how the intelligent question answering system processes the target problem to be processed and obtains the reply information, and the understanding and trust of the terminal user for the intelligent question answering system are enhanced.

[0053] Here, the intelligent question answering system provides a visual user interface, and the terminal user can input the target problem to be processed in the user interface to request the intelligent question answering system to reply to the target problem to be processed to obtain corresponding reply information.

[0054] Here, the target problem to be processed input by the terminal user usually contains the intention of the intelligent question answering, and the intention of the intelligent question answering, i.e., the intention analysis information, can be obtained by analyzing the target problem to be processed by the scheduling module 31.

[0055] In actual application, after the target problem to be processed is analyzed preliminarily to obtain the intention analysis information, the information carrier type of the intention analysis information can be judged and the object can be stored.

[0056] Based on this, in an embodiment, the scheduling module 31 is further configured to determine the information carrier type of the intention analysis information.

[0057] The system further comprises an object storage module, wherein the object storage module is configured to store the information carrier type of the intention analysis information.

[0058] Here, the information carrier type of the intention analysis information can include, but is not limited to, text, pictures, audio files, video files, etc.

[0059] In an embodiment, the intelligent hub module 32 includes a task planning module 321 and a capability processing module 322; wherein,

[0060] The task planning module 321 is configured to convert the intention analysis information into a planning task category related to the intention analysis information.

[0061] The capability processing module 322 is configured to determine the executable capability matching the planning task category based on the planning task category related to the intention analysis information.

[0062] Here, after receiving the task planning request uploaded by the scheduling module 31, the intelligent hub module 32 forwards the task planning request to a general control module (also referred to as a general control unit) in the intelligent hub module 32, and controls the execution of the two sub-service modules of the task planning module 321 and the capability processing module 322 by the general control module.

[0063] Here, the task planning module 321 is the core module of the intelligent hub module 32, which is used for one-time global planning of the intention analysis information, i.e., converting the intention analysis information into a related task to obtain a corresponding planning task category, for example, converting the intention analysis information into a planning task category of image processing.

[0064] It should be noted that the planning task category related to the intention analysis information includes one or more categories, and when including multiple categories, after obtaining multiple planning task categories related to the intention analysis information, the association relationship between the multiple planning tasks corresponding to the multiple planning task categories can also be determined.

[0065] In an embodiment, the capability processing module 322 includes a capability screening module and an intelligent transaction module; wherein,

[0066] The capability screening module is configured to screen the planning task category related to the intention analysis information to obtain a corresponding sub-task category of the candidate capability.

[0067] The intelligent transaction module is configured to determine the executable capability matching the planning task category based on the sub-task category of the candidate capability.

[0068] Here, the capability screening process is to further refine the determined planning task category to a sub-task category (i.e., a sub-task category of the candidate capability), for example, assuming that the planning task category obtained by converting the intention analysis information to a task category is image processing, the task-image processing is further refined to an optical character recognition (OCR) recognition capability, and then the sub-task category of the candidate capability is further refined to a category-matched executable capability, for example, the OCR recognition capability is refined to a capability-ninth-day general OCR recognition capability.

[0069] It should be noted that the task planning module 321 mainly plans through the AI large model, and the capability processing module 322 mainly determines the executable capability matched with the planning task category according to the capability of each category and the output result (i.e., the planning task category related to the intention analysis information) of the task planning module 321.

[0070] In actual application, the intelligent transaction module can refine the sub-task category of the candidate capability to the executable capability matched with the planning task category based on an intelligent transaction strategy.

[0071] Based on this, in an embodiment, the intelligent transaction module is specifically used for:

[0072] determining an intelligent transaction strategy; and based on the intelligent transaction strategy, performing intelligent transaction processing on the sub-task category of the candidate capability to obtain the executable capability matched with the planning task category.

[0073] Here, intelligent transaction refers to the process of refining the sub-task category of the candidate capability to an executable capability based on an intelligent transaction strategy, that is, the intelligent transaction strategy here is actually a capability selection strategy, for example, after the planning task category related to the intention analysis information is screened for capability, the obtained sub-task category of the candidate capability is an OCR recognition capability, and the OCR recognition capability includes a ninth-day general OCR recognition capability and an Ali OCR recognition capability, wherein the ninth-day general OCR recognition capability has a low price and the Ali OCR recognition capability has high performance, and then through intelligent transaction, it can be determined whether to use the ninth-day general OCR recognition capability (determined through a price priority strategy) or the Ali OCR recognition capability (determined through a performance priority strategy) as the executable capability matched with the planning task category. Wherein, which priority strategy is determined through intelligent transaction.

[0074] In actual application, the intelligent hub module 32 further includes a result summarization module. After receiving a summarization request uploaded by the scheduling module 31, the result summarization module executes the summarization request, including summarization of execution processes and execution results of the task planning module 321 and the capability processing module 322, and summarization of output results.

[0075] Based on this, in an embodiment, the intelligent hub module 32 further includes a result summarization module. The result summarization module is configured to summarize the planning task category related to the intention analysis information, the executable capability matched with the planning task category, and the reply information for the target problem to be processed respectively to obtain corresponding summarization results.

[0076] Here, after obtaining the summarization results by executing the summarization request, the result summarization module can output the summarization results to the terminal user, for example, display the summarization results on a user interface of the intelligent question answering system to provide the terminal user.

[0077] The intelligent question answering system in the related art generally has a black box problem. This black box also manifests as poor interactivity, i.e., the user cannot effectively feedback on the answering process of the intelligent question answering system, which limits the ability of the intelligent question answering system to further optimize and learn. To solve this technical problem, the intelligent question answering system provided in the embodiments of the present application provides an interactive mode in which the user can participate, i.e., the user is allowed to modify the answering process of the intelligent question answering system and then execute the modified process.

[0078] Based on this, in an embodiment, the system further includes a dialogue management module. The dialogue management module is configured to save the reply information for the target problem to be processed and implement interaction with the terminal user.

[0079] The dialogue management module is configured to save the reply information for the target problem to be processed and implement interaction with the terminal user.

[0080] Here, the interaction with the terminal user implemented by the dialogue management module can include interaction with the terminal user on the front end to complete functions such as deletion, editing, and searching of the answering process of the intelligent question answering system.

[0081] For example, the dialogue management module implements interaction with the terminal user to edit the answering process of the intelligent question answering system.

[0082] In an embodiment, the dialogue management module is specifically configured to:

[0083] In a case where the executable capability matching the planning task category meets a set condition, a target capability selected by the end user is acquired; the target capability and the executable capability matching the planning task category are strongly associated with the planning task category;

[0084] The executable capability matching the planning task category is changed to the target capability.

[0085] The task execution module is further configured to execute the target capability to obtain reply information for the target problem to be processed.

[0086] Here, for the target capability and the executable capability matching the planning task category being strongly associated with the planning task category, the association degree of the target capability and the executable capability with the planning task category can be greater than or equal to an association degree threshold value, and in this case, the two can be considered to be strongly associated, where the association degree threshold value can be set according to actual requirements, which is not limited in the embodiments of the present application. For the executable capability matching the planning task category meeting the set condition, it can be understood that the process of converting the planning task category related to the intent analysis information into the executable capability fails, and in this case, the terminal user is prompted with an error reason. The terminal user changes the selected target capability, for example, the terminal user changes the capability A (corresponding to the executable capability matching the planning task category) recommended by the intelligent question and answer system to the capability B (corresponding to the target capability), where the capability A and the capability B are strongly associated with the planning task category, and then the task execution module re-executes the changed target capability to obtain the reply information for the target problem to be processed, which embodies the interactivity of the intelligent question and answer system.

[0087] In an embodiment, the system further comprises a capability center module; wherein

[0088] The capability center module is configured to register various categories of capabilities, and aggregate the capability information of the various categories of capabilities to obtain aggregated capability information.

[0089] It should be noted that the capability center module can also be responsible for the function of capability calling, specifically, the capability processing module 322 can call and acquire the corresponding category of capability from the capability center module, that is, acquire the executable capability matching the planning task category.

[0090] Here, the intelligent question answering system proposed in the embodiments of the present application provides multiple service modes for terminal users, including a view mode and a developer mode, etc. Through the view mode, the terminal user can be provided with the processing process of the intelligent question answering system after dynamic rendering, so that the terminal user can not only understand the execution progress of each step, but also clearly understand how the intelligent question answering system processes the question and obtains the answer, thereby enhancing the understanding and trust of the terminal user for the intelligent question answering system.

[0091] Based on the foregoing intelligent question answering system, the embodiments of the present application further provide a question answering processing method, which is applied to a question answering processing device, Figure 4 FIG. 1 is a flowchart of a question answering processing method according to an embodiment of the present application; and Figure 4 As shown in the figure, the question answering processing method comprises the following steps.

[0092] Step 401: receiving a first request sent by a terminal user; the first request is used to request a reply to a target question to be processed, and the first request comprises the target question to be processed.

[0093] Here, the execution state and execution details of each step executed by the question answering processing device are visualized to the terminal user throughout the process. The intelligent question answering system is embedded in the question answering processing device, that is, the execution state and execution details of the intermediate processing steps in the processing process of the intelligent question answering system, as well as the results of the intermediate processing steps, are visualized to the terminal user. This enables the terminal user to clearly understand how the intelligent question answering system processes the target question to be processed and obtains the reply information, thereby enhancing the understanding and trust of the terminal user for the intelligent question answering system.

[0094] Here, the intelligent question answering system provides a visual user interface, through which the terminal user can input the target question to be processed and request the intelligent question answering system to reply to the target question to be processed to obtain corresponding reply information.

[0095] Step 402: analyzing the target question to be processed to obtain intention analysis information.

[0096] Here, since the target question to be processed input by the terminal user usually contains the intention of intelligent question answering, the question answering processing device can obtain the intention of intelligent question answering, i.e. the intention analysis information, by analyzing the target question to be processed.

[0097] In actual application, after the question answering processing device preliminarily analyzes the target question to be processed to obtain the intention analysis information, the type of information carrier of the intention analysis information can be further judged and object storage can be performed.

[0098] Based on this, in an embodiment, the method further comprises: determining the information carrier type of the intention analysis information; and storing the information carrier type of the intention analysis information.

[0099] Here, the information carrier type of the intention analysis information can include, but is not limited to, text, pictures, audio files, video files, etc.

[0100] Step 403: Based on the intention analysis information, determine the planning task category, and determine the executable ability matching the planning task category.

[0101] In an embodiment, the determination of the planning task category based on the intention analysis information comprises: task category conversion of the intention analysis information to obtain the planning task category related to the intention analysis information.

[0102] The determination of the executable ability matching the planning task category comprises: based on the planning task category related to the intention analysis information, determining the executable ability matching the planning task category.

[0103] Here, the intention analysis information is converted into a relevant task through one-time global planning task, i.e., the intention analysis information is converted into a relevant task to obtain a corresponding planning task category, for example, the planning task category obtained by converting the intention analysis information into a task category is image processing.

[0104] It should be noted that the planning task category related to the intention analysis information includes one or more categories, and when multiple categories are included, after obtaining multiple planning task categories related to the intention analysis information, the association relationship between multiple planning tasks corresponding to multiple planning task categories can also be determined.

[0105] In an embodiment, the determination of the executable ability matching the planning task category based on the planning task category related to the intention analysis information comprises: ability screening of the planning task category related to the intention analysis information to obtain a corresponding sub-task category of the candidate ability; and based on the sub-task category of the candidate ability, determining the executable ability matching the planning task category.

[0106] Here, the ability screening process is to further refine the determined planning task category to a sub-task category (i.e., a sub-task category of a candidate ability), for example, assuming that the planning task category obtained by converting the intention analysis information into a task category is image processing, then the task-image processing is further refined to an OCR recognition ability, and then the sub-task category of the candidate ability is further refined to a category-matching executable ability, for example, the OCR recognition ability is refined to the ability-Nine-day general OCR recognition ability.

[0107] In actual application, the question and answer processing apparatus can refine the sub-task category of the candidate ability into the executable ability matching the planning task category based on an intelligent transaction strategy.

[0108] Based on this, in an embodiment, the determining of the executable ability matching the planning task category based on the sub-task category of the candidate ability comprises: determining an intelligent transaction strategy; and performing intelligent transaction processing on the sub-task category of the candidate ability based on the intelligent transaction strategy to obtain the executable ability matching the planning task category.

[0109] Here, the intelligent transaction refers to the process of refining the sub-task category of the candidate ability into the executable ability based on an intelligent transaction strategy, that is, the intelligent transaction strategy here is actually an ability selection strategy. For example, in the ability screening for the planning task category related to the intent analysis information, the obtained sub-task category of the candidate ability is the OCR recognition ability, and the OCR recognition ability includes the general OCR recognition ability of Jiutiantong and the OCR recognition ability of Ali, wherein the price of the general OCR recognition ability of Jiutiantong is low, and the performance of the OCR recognition ability of Ali is high. Therefore, through intelligent transaction, it can be determined whether to use the general OCR recognition ability of Jiutiantong (determined by the price priority strategy) or the OCR recognition ability of Ali (determined by the performance priority strategy) as the executable ability matching the planning task category. Wherein, which priority strategy is determined through intelligent transaction.

[0110] In an embodiment, the method further comprises: summarizing the planning task category related to the intent analysis information, the executable ability matching the planning task category, and the reply information for the target problem to be processed respectively to obtain corresponding summary results.

[0111] Here, after obtaining the summary results, the question and answer processing apparatus can also output the summary results to the terminal user, for example, display the summary results on the user interface of the intelligent question and answer system to provide the terminal user.

[0112] In an embodiment, the method further comprises: registering various types of abilities, and aggregating the ability information of the various types of abilities to obtain aggregated ability information.

[0113] Here, the intelligent question and answer system proposed in the embodiments of the present application provides multiple service modes for the terminal user, including a view mode and a developer mode. Through the view mode, the terminal user can be provided with the dynamic rendering of the processing process of the intelligent question and answer system, so that the terminal user not only can understand the execution progress of each step, but also can clearly understand how the intelligent question and answer system processes the question and obtains the answer, thereby enhancing the understanding and trust of the terminal user for the intelligent question and answer system.

[0114] Step 404: performing the executable capability matched with the planning task category to obtain reply information for the target problem to be processed.

[0115] The intelligent question and answer system in the related art generally has a black box problem, and the black box problem also manifests as poor interactivity, i.e., the user cannot effectively feedback the answering process of the intelligent question and answer system, which limits the ability of the intelligent question and answer system to further optimize and learn. In order to solve the technical problem, the intelligent question and answer system embedded in the question and answer processing device to which the question and answer processing method proposed by the embodiments of the present application is applied provides an interactive mode in which the user can participate, i.e., the user is allowed to modify the answering process of the intelligent question and answer system and then execute the modified process.

[0116] Based on this, in an embodiment, after the executable capability matched with the planning task category is performed to obtain the reply information for the target problem to be processed, the method further includes saving the reply information for the target problem to be processed and realizing interaction with the terminal user.

[0117] Here, the interaction with the terminal user can include interaction with the terminal user in front end to complete functions such as deletion, editing, and searching of the answering process of the intelligent question and answer system.

[0118] Exemplarily, taking editing of the answering process of the intelligent question and answer system as an example, the process of realizing the interaction with the terminal user is described.

[0119] In an embodiment, the interaction with the terminal user includes:

[0120] In the case where the executable capability matched with the planning task category meets a set condition, a target capability selected by the terminal user is obtained; the target capability and the executable capability matched with the planning task category are strongly associated with the planning task category; the executable capability matched with the planning task category is changed to the target capability; and the target capability is executed to obtain the reply information for the target problem to be processed.

[0121] Here, for the target ability and the executable ability matching the planning task category, the association degree between the target ability and the executable ability and the planning task category is greater than or equal to an association degree threshold, and the association between the two can be referred to as strong association, where the association degree threshold can be set according to actual needs, and the embodiments of the present application are not limited herein. For the executable ability matching the planning task category that meets the set condition, it can be understood that the process of converting the planning task category related to the intent analysis information into an executable ability fails, and the terminal user is prompted for the error reason at this time. The terminal user changes the selected target ability, for example, the terminal user changes the ability A (corresponding to the executable ability matching the planning task category) recommended by the intelligent question and answer system to the ability B (corresponding to the target ability), where the ability A and the ability B are strongly associated with the planning task category, and then the target ability is re-executed by the task execution module, and the reply information for the target problem to be processed is obtained, which embodies the interactivity of the intelligent question and answer system.

[0122] The execution state and execution details of each step of the question and answer processing device are visualized to the terminal user, so that the processing process of the question and answer processing device embedded with the intelligent question and answer system is no longer a "black box", but a transparent process, so that the terminal user can clearly understand how the intelligent question and answer system processes the problem and obtains the answer, and the understanding and trust of the terminal user for the intelligent question and answer system are enhanced. In addition, by analyzing the target problem to be processed to obtain the intent analysis information, determining the planning task category based on the intent analysis information, and determining the executable ability matching the planning task category, and then executing the executable ability matching the planning task category to obtain the reply information, the final reply information will be more accurate and reliable, which helps to improve the reliability of the intelligent question and answer system and the satisfaction of the terminal user.

[0123] The present application will be described below in conjunction with application examples.

[0124] The current market intelligent question and answer system generally has a black box problem, and this black box performance is as follows:

[0125] The processing process is not transparent: the user cannot understand how the intelligent question and answer system processes the problem and obtains the answer, which leads to the user's distrust of the intelligent question and answer system;

[0126] Poor interactivity: the user cannot effectively feedback the answer process of the intelligent question and answer system, which limits the further optimization and learning ability of the intelligent question and answer system;

[0127] Insufficient explainability: the intelligent question and answer system cannot explain the basis and logic of its answer to the user, which affects the credibility of the user to the answer.

[0128] The above problems restrict the application and development of intelligent question answering systems and hinder the improvement of user experience. In order to solve the above problems, the present application proposes an innovative intelligent question answering system that is interpretable and interactive. The system has the following characteristics:

[0129] Transparent processing process: the intelligent question answering system of the present application visualizes the key steps and intermediate results in the processing process, so that the user can clearly understand how the intelligent question answering system arrives at the answer.

[0130] Interactive interaction mode: the intelligent question answering system of the present application allows the user to modify the system answer process and execute again after modification.

[0131] Strong explanation ability: the intelligent question answering system of the present application can explain its answer construction in detail, helping the user to understand the logic and basis behind the answer.

[0132] Figure 5 The intelligent question answering system of the present application is based on a Holistic Artificial Intelligence Operation System (HAIOS). The HAIOS adopts an AI large model intelligent question answering system to provide services for users. As shown in Figure 5 The architecture of the intelligent question answering system mainly includes three layers: an application layer, a system layer, and a capability layer. The upper layer (application layer) provides a unified entrance for intelligent services, and the lower layer (capability layer) collaborates with various AI capabilities, computing power capabilities, network capabilities, and other capabilities. The middle layer (system layer) is mainly responsible for completing the processing process of the intelligent question answering system.

[0133] In the implementation of the system layer, the problem input by the application layer (corresponding to the aforementioned target problem to be processed) can be processed through the modules of intent analysis, task planning, capability selection, end-to-end optimization, algorithm network scheduling, simulation verification and feedback, task execution, and summary output. Finally, the result (corresponding to the aforementioned reply information for the target problem to be processed) is provided to the user through the application layer. The execution status and details of the intermediate processing steps in the above modules are visible to the user throughout the process. The following describes several important modules among them:

[0134] Intention analysis module: the user's input question is analyzed to obtain the intention of intelligent question answering (corresponding to the aforementioned intention analysis information), and the information carrier type of the intention of intelligent question answering is judged and object storage is performed. The information carrier type can include but is not limited to text, pictures, audio files, video files, etc.

[0135] Task planning module: the core module of the HAIOS intelligent question answering system, used to convert the intention of intelligent question answering into relevant tasks, and obtain the corresponding planning task category, for example, the planning task category is image processing, and the association relationship of multiple planning tasks can be determined.

[0136] Capability selection module: the capability selection module includes capability screening and AI Exchange (corresponding to the aforementioned intelligent transaction), wherein the capability screening process is to further refine the task planning result (i.e., the planning task category related to the intention analysis information) to a sub-task category, for example, refining the task-image processing to an OCR recognition capability; the AI Exchange is to refine the sub-task category to an executable capability through a capability selection strategy (corresponding to the aforementioned intelligent transaction strategy), for example, refining the OCR recognition capability to the capability-nine-day general OCR recognition capability.

[0137] Task execution module: according to the task selection result (i.e., the selected executable capability), sequentially calling the capability reasoning interface, executing the selected executable capability, and obtaining the output result (corresponding to the aforementioned reply information for the target problem to be processed).

[0138] Summary output module: summarizing and polishing the input, execution process, and output result, and finally outputting to the user through the application layer of the HAI OS intelligent question answering system.

[0139] It should be noted that the end-to-end optimization module, the algorithm network scheduling module, the simulation verification and feedback module in the system layer are not scheduled for implementation, and these three modules do not affect the implementation of the scheme of the present application, therefore, the present application will not be specifically expanded.

[0140] Figure 6 The micro-service architecture diagram for the intelligent question answering system of the embodiment of the present application, the core part of the whole micro-service architecture is the service layer, the functions of several service modules of the service layer are described below.

[0141] Dialogue management: responsible for saving question and answer information (corresponding to the aforementioned target problem to be processed and reply information), and interacting with the front end to complete functions such as dialogue deletion, editing, and searching.

[0142] OS scheduler: responsible for receiving the question and answer request (corresponding to the aforementioned first request) sent by the front end, and completing the control and intermediate information recording of the whole process through scheduling other services in the system.

[0143] AI Exchange service: responsible for managing and determining the AI Exchange strategy (corresponding to the aforementioned intelligent transaction strategy), and refining the task planned by the intelligent hub into an executable capability.

[0144] Capability Center: responsible for completing the registration of various types of capabilities, capability information aggregation, capability invocation, and other functions.

[0145] Intelligent Hub: responsible for completing task planning, capability selection, result summary, and other functions.

[0146] File Storage: responsible for completing the object storage of the information carrier type of the intent of intelligent question and answer, and other functions.

[0147] Among them, the intelligent hub is the core module of the service layer in the micro-service architecture of the intelligent question and answer system, mainly based on the LLM-Based Agent (Large Language Model-Based Agent) construction idea, taking the large model as the intelligent engine, and realizing the central control system that can make autonomous decisions and iterate.

[0148] Figure 7 The intelligent hub service process of the embodiment of the present application is shown in FIG. 1, wherein the intelligent hub receives the task planning request sent by the OS scheduling module (corresponding to the aforementioned scheduling module), and forwards the task planning request to the total control unit in the intelligent hub. The total control unit controls the following various sub-service modules to realize the entire service process: Figure 7

[0149] Task Planning Module: according to the environmental perception and intent analysis result, one-time global planning of the task is performed, and the task category (major category) is output. The task planning module is mainly planned by the large model;

[0150] Capability Selection Module: according to the planned task category, the capability executed by the execution module is determined. The capability selection module mainly determines the executable capability according to the capability situation of the capability center and the output result of the task planning module (corresponding to the aforementioned planning task category related to the intent analysis information);

[0151] Result Summary Module: including the summary of the execution process and execution result of the task planning module and the capability selection module, and the summary of the final result (i.e. the reply information). The result summary is polished and optimized by the large model and displayed on the front end.

[0152] It should be noted that the functions of other modules involved in Figure 7 can be understood by referring to the above, which will not be described here.

[0153] Figure 8 ​This diagram illustrates the interaction flow between various service modules of the intelligent question-answering system according to an embodiment of this application. This interaction flow describes the process by which the various service modules of the intelligent question-answering system handle user-submitted questions. Through this service interaction flow, the processing procedures of each service module of the intelligent question-answering system can be visually displayed to the user, demonstrating the explainability of the intelligent question-answering system. Figure 8 As shown, the interaction process includes the following steps:

[0154] Steps 1-4 belong to the "Intent Resolution" step in the intelligent question answering system architecture. When the processing is completed or an error occurs, the user can view the processing status and related information through the front end.

[0155] Steps 5-10 belong to the [Task Planning] step in the intelligent question answering system architecture. When the process is successful, the task planning view and the explanation of each task are output; when the process fails, the task planning error node and the reason for the error are output.

[0156] Steps 11-14: These belong to the "Capability Selection" step in the intelligent question-answering system architecture. They transform task categories into executable capabilities. When processing is successful, they output a capability selection view and descriptions of the tasks and capabilities. When processing fails, they output the capability that was selected incorrectly and the reason for the error.

[0157] Steps 15-24: These belong to the "Task Execution" step in the intelligent question-answering system architecture. When execution is successful, the capability execution view and capability call log are output; when processing fails, the error reason for calling the capability is output.

[0158] Steps 25-29 belong to the "Summary Output" step in the intelligent question-answering system architecture. When the execution is successful, the answer information is displayed to the user through the message body; when the processing fails, the reason for the failure is displayed.

[0159] It should be noted that when users customize the solution, steps 15-29 will be changed. By changing the selection capabilities, users can achieve personal intervention in the final output, thereby demonstrating the interactivity of the intelligent question-and-answer system.

[0160] In practical applications, the HAIOS intelligent question-answering system provides users with multiple service modes, including view mode and developer mode, for example, Figure 9 The diagram shown illustrates the task execution process in view mode, as follows: Figure 10 The diagram shown illustrates the task planning process in developer mode. Figure 11The task execution process in the developer mode is shown in the schematic diagram. In the view mode, the user can be provided with the processing process of task execution and the processing result of the intelligent hub in a dynamic rendering manner. In this way, the user can not only clearly understand the execution progress, but also understand how the HAIOS intelligent question and answer system implements the problem step by step. When an execution exception occurs, the corresponding error execution node and execution details can be directly located.

[0161] In addition, the intelligent question and answer system execution scheme can be customized, that is, the user can customize the scheme on the basis of the recommended scheme of the system, such as Figure 12 The intelligent question and answer system execution scheme customization schematic diagram is shown. For example, the user can change the system recommended capability A to capability B, wherein the capability A and the capability B are respectively strongly associated with the task categories planned in the task planning process. After determining the customized scheme and submitting and executing, the intelligent question and answer system will re-perform the capability execution on the capability B, sequentially call the capabilities according to the user's selection, and summarize the output after the scheme customization. Figure 13 The detail page schematic diagram of the customized scheme is shown in the middle. Through the detail page, the user can clearly understand the capability execution process and the capability execution state and other detail information.

[0162] It should be noted that in actual application, the user can repeatedly adjust the executed capability and submit the adjusted capability for execution. In this way, the user can participate in the processing process of the intelligent question and answer system, thereby reflecting the interactivity of the intelligent question and answer system.

[0163] The present application proposes an interpretable and interactive intelligent question and answer system, which aims to provide transparent, efficient and customized question and answer services for users. The core innovation of the system lies in that:

[0164] Intelligent hub architecture: the intelligent hub is the "brain" of the HAIOS intelligent question and answer system, which is a task planning and capability selection scheme driven by a large model.

[0165] Layered service interaction system: the system abstracts the front-end and back-end services in layers to form a clear service calling relationship. Through this phased design, the system provides a basis for interpretability and interactivity.

[0166] Visual user interface: the system provides view mode, developer mode and scheme customization functions, so that the user can clearly understand the running process of the system and customize the scheme according to the user's own needs, thereby providing an intuitive experience for the interpretability and interactivity of the system.

[0167] Compared with the related art, the scheme of the present application has the following advantages:

[0168] 1) Break the way of providing services of the conventional large model intelligent question and answer system on the market, and innovatively propose an intelligent question and answer system with process visualization, that is, the processing process of the intelligent question and answer system is no longer a "black box", and the output result will be more reliable, which helps to improve the reliability of the intelligent question and answer system and the satisfaction of the user;

[0169] 2) Innovatively introduce an interactive function to support the user to participate in the process of outputting answers by the intelligent question and answer system, that is, the intelligent question and answer system is no longer a problem solved by the system itself, but also a result of the joint participation of the user, and the user's stickiness is also improved.

[0170] In order to realize the question and answer processing method of the embodiment of the application, the embodiment of the application further provides a question and answer processing device, Figure 14 The schematic diagram of the composition structure of the question and answer processing device of the embodiment of the application is shown in Figure 14 As shown in the figure, the question and answer processing device comprises:

[0171] A receiving unit 1401 is configured to receive a first request sent by a terminal user; the first request is used to request to reply to a target question to be processed, and the first request comprises the target question to be processed;

[0172] An analysis unit 1402 is configured to analyze the target question to be processed to obtain intention analysis information;

[0173] A first determination unit 1403 is configured to determine a planning task category based on the intention analysis information;

[0174] A second determination unit 1404 is configured to determine an executable ability matched with the planning task category;

[0175] An execution unit 1405 is configured to execute the executable ability matched with the planning task category to obtain reply information for the target question to be processed.

[0176] In an embodiment, the first determination unit 1403 is specifically configured to:

[0177] perform task category conversion on the intention analysis information to obtain a planning task category related to the intention analysis information;

[0178] The second determination unit 1404 is specifically configured to:

[0179] determine the executable ability matched with the planning task category based on the planning task category related to the intention analysis information.

[0180] In an embodiment, the second determination unit 1404 comprises a capability screening subunit and a determination subunit; wherein,

[0181] The capability screening sub-unit is configured to perform capability screening on the planning task category related to the intention analysis information, to obtain a sub-task category of a candidate capability.

[0182] The determination sub-unit is configured to determine the executable capability matching the planning task category based on the sub-task category of the candidate capability.

[0183] In an embodiment, the determination sub-unit is specifically configured to:

[0184] determine an intelligent transaction strategy, and perform intelligent transaction processing on the sub-task category of the candidate capability based on the intelligent transaction strategy, to obtain the executable capability matching the planning task category.

[0185] In an embodiment, the apparatus further includes a summary unit, and wherein

[0186] The summary unit is configured to respectively summarize the planning task category related to the intention analysis information, the executable capability matching the planning task category, and the reply information for the target problem to be processed, to obtain a corresponding summary result.

[0187] In an embodiment, the apparatus further includes a saving unit and an interaction unit, and wherein

[0188] The saving unit is configured to save the reply information for the target problem to be processed after the execution unit 1405 executes the executable capability matching the planning task category to obtain the reply information for the target problem to be processed.

[0189] The interaction unit is configured to realize interaction with the end user.

[0190] In an embodiment, the interaction unit is specifically configured to:

[0191] In a case where the executable capability matching the planning task category meets a set condition, obtain a target capability selected by the end user; the target capability and the executable capability matching the planning task category are strongly associated with the planning task category; change the executable capability matching the planning task category to the target capability; and execute the target capability to obtain the reply information for the target problem to be processed.

[0192] In an embodiment, the apparatus further includes a registration summary unit, and wherein

[0193] The registration and summarization unit is configured to register various types of capabilities and summarize capability information of the various types of capabilities to obtain summarized capability information.

[0194] In an embodiment, the apparatus further comprises a third determination unit and a storage unit, wherein

[0195] The third determination unit is configured to determine an information carrier type of the intention analysis information.

[0196] The storage unit is configured to store the information carrier type of the intention analysis information.

[0197] In actual application, the receiving unit 1401 can be implemented by a communication interface in the question and answer processing apparatus; the analysis unit 1402, the first determination unit 1403, the second determination unit 1404 and the execution unit 1405 can be implemented by a processor in the question and answer processing apparatus.

[0198] It should be noted that the question and answer processing apparatus provided in the above embodiments is only taken as an example for the division of the above program modules in the question and answer processing, and in actual application, the above processing can be completed by different program modules according to needs, that is, the internal structure of the apparatus is divided into different program modules to complete all or part of the above processing. In addition, the question and answer processing apparatus and the question and answer processing method provided in the above embodiments belong to the same concept, and the specific implementation process is detailed in the question and answer processing method, which will not be repeated here.

[0199] Based on the hardware implementation of the above program modules, and in order to implement the question and answer processing method of the embodiments of the present application, the embodiments of the present application further provide a question and answer processing device, Figure 15 The hardware component structure of the question and answer processing device of the embodiments of the present application is shown in FIG. 15. Figure 15 As shown in FIG. 15, the question and answer processing device 1500 comprises:

[0200] The communication interface 1501 can interact with other devices for information exchange;

[0201] The processor 1502 is connected with the communication interface 1501 to realize information exchange with other devices, and is configured to run a computer program to execute the question and answer processing method provided above, and the computer program is stored on the storage 1503.

[0202] Specifically, the communication interface 1501 is configured to receive a first request sent by a terminal user; the first request is used to request to reply to a target question to be processed, and the first request comprises the target question to be processed.

[0203] The processor 1502 is configured to analyze the target question to be processed to obtain intention analysis information, determine a planning task category based on the intention analysis information, determine an executable capability matched with the planning task category, execute the executable capability matched with the planning task category, and obtain reply information for the target question to be processed.

[0204] It should be noted that the specific processing procedures of the communication interface 1501 and the processor 1502 can be understood with reference to the above-mentioned question and answer processing method.

[0205] Of course, in actual applications, various components in the question and answer processing device 1500 are coupled together through the bus system 1504. It can be understood that the bus system 1504 is used to realize the connection and communication between the components. The bus system 1504 includes not only a data bus, but also a power bus, a control bus, and a status signal bus. However, for the purpose of clear illustration, all kinds of buses are marked as the bus system 1504 in the figure. Figure 15

[0206] The memory 1503 in the embodiment of the present application is used to store various types of data to support the operation of the question and answer processing device 1500. Examples of the data include any computer programs used for operation on the question and answer processing device 1500.

[0207] The question and answer processing method disclosed in the above embodiment of the present application can be applied to or implemented by the processor 1502. The processor 1502 can be an integrated circuit chip with signal processing capability. In the implementation process, each step of the above question and answer processing method can be completed by integrated logic circuits or instruction forms of software in the processor 1502. The processor 1502 mentioned above can be a general-purpose processor, a digital signal processor (DSP), or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. The processor 1502 can implement or execute the question and answer processing methods, steps, and logic block diagrams disclosed in the embodiments of the present application. The general-purpose processor can be a microprocessor or any conventional processor. In combination with the steps of the question and answer processing method disclosed in the embodiments of the present application, the execution can be directly completed by a hardware decoding processor, or completed by a combination of hardware and software modules in the decoding processor. The software module can be located in a storage medium, which is located in the memory 1503, and the processor 1502 reads the information in the memory 1503 and combines the hardware to complete the steps of the above question and answer processing method.

[0208] ​In an exemplary embodiment, the question and answer processing device 1500 can be implemented by one or more Application Specific Integrated Circuits (ASICs), DSPs, Programmable Logic Devices (PLDs), Complex Programmable Logic Devices (CPLDs), Field-Programmable Gate Arrays (FPGAs), general-purpose processors, controllers, microcontrollers (MCUs), microprocessors, or other electronic elements for performing the aforementioned question and answer processing method.

[0209] It can be understood that the memory 1503 of the embodiments of the present application can be a volatile memory or a non-volatile memory, and can also include both volatile and non-volatile memories. Among them, the non-volatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), a ferromagnetic random access memory (FRAM), a flash memory, a magnetic surface memory, an optical disc, or a compact disc read-only memory (CD-ROM); the magnetic surface memory can be a disk memory or a tape memory. The volatile memory can be a random access memory (RAM) used as an external cache. By way of example but not limitation, many forms of RAM can be used, such as static random access memory (SRAM), synchronous static random access memory (SSRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM), and direct memory bus random access memory (DRRAM).The memory 1503 described in the embodiments of the present application is intended to include, but not limited to, these and any other suitable types of memories.

[0210] In the example embodiments, the embodiments of the present application also provide a storage medium, specifically a computer readable storage medium, such as the memory 1503 storing a computer program executable by the processor 1502 in the question and answer processing device 1500 to complete the steps of the question and answer processing method described in the foregoing embodiments of the present application. The computer readable storage medium can be FRAM, ROM, PROM, EPROM, EEPROM, Flash Memory, magnetic surface memory, optical disc, or CD-ROM, etc.

[0211] In the example embodiments, the embodiments of the present application also provide a computer program product including a computer program executable by the processor 1502 in the question and answer processing device 1500 to complete the steps of the question and answer processing method described in the foregoing embodiments of the present application.

[0212] It should be noted that "first", "second", "third", etc. are used to distinguish similar objects, and do not necessarily describe a specific order or sequence.

[0213] In addition, the technical solutions described in the embodiments of the present application can be combined arbitrarily without conflict.

[0214] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto, any person skilled in the art can easily think of changes or replacements within the technical range disclosed in the present application, which should be covered in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. An intelligent question answering system, characterized in that, The system comprises a scheduling module, an intelligent hub module and a task execution module, execution states and execution details of the scheduling module, the intelligent hub module and the task execution module are visualized to an end user throughout the process, and The scheduling module is configured to receive a first request sent by the end user, the first request being used to request a reply to a target problem to be processed, and the first request comprising the target problem to be processed. The scheduling module is further configured to analyze the target problem to be processed to obtain intention analysis information. The intelligent hub module is configured to determine a planning task category based on the intention analysis information, and determine an executable capability matching the planning task category. The task execution module is configured to execute the executable capability matching the planning task category to obtain reply information for the target problem to be processed.

2. The system of claim 1, wherein, The intelligent hub module comprises a task planning module and a capability processing module, and The task planning module is configured to convert the intention analysis information into a planning task category related to the intention analysis information. The capability processing module is configured to determine the executable capability matching the planning task category based on the planning task category related to the intention analysis information.

3. The system of claim 2, wherein, The capability processing module comprises a capability screening module and an intelligent transaction module, and The capability screening module is configured to screen the planning task category related to the intention analysis information to obtain a sub-task category of a candidate capability. The intelligent transaction module is configured to determine the executable capability matching the planning task category based on the sub-task category of the candidate capability.

4. The system of claim 3, wherein, The intelligent transaction module is specifically configured to: determine an intelligent transaction strategy; perform intelligent transaction processing on the sub-task category of the candidate capability based on the intelligent transaction strategy to obtain the executable capability matching the planning task category.

5. The system of claim 2, wherein, The intelligent hub module further comprises a result summarizing module, and The result summarizing module is configured to summarize the planning task category related to the intention analysis information, the executable capability matching the planning task category and the reply information for the target problem to be processed respectively to obtain corresponding summary results.

6. The system of claim 1, wherein, The system further comprises a dialogue management module, and The dialogue management module is configured to save the reply information for the target problem to be processed and realize interaction with the end user.

7. The system of claim 6, wherein, The dialogue management module is specifically configured to: in a case where the executable capability matching the planning task category meets a set condition, acquire a target capability selected by the end user, the target capability and the executable capability matching the planning task category being strongly associated with the planning task category; change the executable capability matching the planning task category into the target capability; and The task execution module is further configured to execute the target capability to obtain the reply information for the target problem to be processed.

8. The system of claim 1, wherein, The system further comprises a capability center module, and The capability center module is configured to register various types of capabilities and aggregate capability information of the various types of capabilities to obtain aggregated capability information.

9. The system of claim 1, wherein, The scheduling module is further configured to determine an information carrier type of the intent analysis information. The system further includes an object storage module, and The object storage module is configured to store the information carrier type of the intent analysis information.

10. A question and answer processing method characterized by, The method includes: receiving a first request sent by a terminal user, the first request being used to request a reply to a target problem to be processed, the first request including the target problem to be processed; analyzing the target problem to be processed to obtain intent analysis information; determining a planning task category based on the intent analysis information, and determining executable capabilities matching the planning task category; executing the executable capabilities matching the planning task category to obtain reply information for the target problem to be processed.

11. The method of claim 10, wherein, The determination of the planning task category based on the intent analysis information includes: performing task category conversion on the intent analysis information to obtain a planning task category related to the intent analysis information; The determination of the executable capabilities matching the planning task category includes: determining the executable capabilities matching the planning task category based on the planning task category related to the intent analysis information.

12. The method of claim 11, wherein, The determination of the executable capabilities matching the planning task category based on the planning task category related to the intent analysis information includes: performing capability screening on the planning task category related to the intent analysis information to obtain a sub-task category of a candidate capability; determining the executable capabilities matching the planning task category based on the sub-task category of the candidate capability.

13. The method of claim 12, wherein, The determination of the executable capabilities matching the planning task category based on the sub-task category of the candidate capability includes: determining an intelligent transaction strategy; performing intelligent transaction processing on the sub-task category of the candidate capability based on the intelligent transaction strategy to obtain the executable capabilities matching the planning task category.

14. The method of claim 11, wherein, The method further includes: summarizing the planning task category related to the intent analysis information, the executable capabilities matching the planning task category, and the reply information for the target problem to be processed respectively to obtain corresponding summary results.

15. The method of claim 10, wherein, After the execution of the executable capabilities matching the planning task category to obtain the reply information for the target problem to be processed, the method further includes: saving the reply information for the target problem to be processed and realizing interaction with the terminal user.

16. The method of claim 15, wherein, The realization of the interaction with the terminal user includes: in a case where the executable capabilities matching the planning task category meet a set condition, obtaining a target capability selected by the terminal user, the target capability and the executable capabilities matching the planning task category being strongly associated with the planning task category; changing the executable capabilities matching the planning task category to the target capability; and Performing the target capability to obtain reply information for the target problem to be processed.

17. The method of claim 10, wherein, The method further comprises: Registering each type of capability and collecting capability information of the each type of capability to obtain collected capability information.

18. The method of claim 10, wherein, The method further comprises: Determining an information carrier type of the intent analysis information; Storing the information carrier type of the intent analysis information.

19. A question and answer processing apparatus characterized by comprising: The apparatus comprises: A receiving unit configured to receive a first request sent by a terminal user, the first request being used to request a reply to a target problem to be processed, the first request comprising the target problem to be processed; An analysis unit configured to analyze the target problem to be processed to obtain intent analysis information; A first determining unit configured to determine a planning task type based on the intent analysis information; A second determining unit configured to determine an executable capability matching the planning task type; An executing unit configured to perform the executable capability matching the planning task type to obtain reply information for the target problem to be processed.

20. A question and answer processing device, characterized by comprise: a processor and a memory for storing a computer program capable of running on the processor; wherein the processor is configured to run the computer program to perform the steps of the method according to any one of claims 10 to 18.

21. A storage medium having stored thereon a computer program, characterized in that The computer program is executed by the processor to implement the steps of the method according to any one of claims 10 to 18.

22. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to implement the steps of the method according to any one of claims 10 to 18. The computer program is executed by the processor to implement the steps of the method according to any one of claims 10 to 18.

Images