Information feedback method, apparatus and device

Abstract

The application discloses an information feedback method, device and equipment, and the method comprises the following steps: receiving user input information; if the user input information meets the first condition, searching based on the user input information through an artificial instruction library; if the search result shows that the artificial instruction library does not have historical instructions matched with the user input information, issuing an information acquisition instruction; receiving artificial guidance information in response to the information acquisition instruction; and generating feedback information for feeding back the user input information according to the artificial guidance information.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, and in particular to an information feedback method, apparatus, and device. Background Technology

[0002] Human-machine collaboration has wide applications in various industries. Taking intelligent customer service as an example, transferring to a human agent is one of the simplest methods of human-machine collaboration. Currently, the known methods of transferring to a human agent are mainly as follows: (1) Through the human agent identification module in the intelligent customer service or when the user provides negative feedback, the user's problem is directly transferred to a human agent for processing; (2) For user problems, when the intelligent customer service system cannot accurately determine the user's intent, the human agent manually selects an intent for the intelligent customer service system, and then the intelligent customer service system continues to process it. It heavily relies on human agent resources. Summary of the Invention

[0003] This application provides an information feedback method, apparatus, and device.

[0004] According to a first aspect of this application, an information feedback method is provided, the method comprising: receiving user input information; if the user input information satisfies the first condition, performing a search based on the user input information using a manual instruction library; if the search result indicates that there is no historical instruction in the manual instruction library that matches the user input information, issuing an information retrieval instruction; receiving manual guidance information in response to the information retrieval instruction; and generating feedback information for providing feedback on the user input information based on the manual guidance information.

[0005] According to one embodiment of this application, generating feedback information for responding to user input information based on the human guidance information includes: performing natural language understanding on the human guidance information to obtain an information understanding result, wherein the information understanding result is used to characterize the associated information required to respond to user input information; and generating the feedback information based on the information understanding result.

[0006] According to one embodiment of this application, the method further includes: updating the artificial instruction library based on the information understanding results.

[0007] According to one embodiment of this application, the retrieval based on the user input information through a manual instruction database includes: matching the user input information with multiple historical user input information in the manual instruction database; wherein, the manual instruction database includes multiple historical instructions and the correspondence between the historical instructions and the corresponding historical user input information.

[0008] According to one embodiment of this application, the retrieval based on the user input information through a manual instruction library includes: obtaining context information of the user input information; matching the context information with historical context information corresponding to multiple historical instructions; wherein, the manual instruction library includes multiple historical instructions and the correspondence between the historical instructions and the context information of the corresponding historical user input information.

[0009] According to one embodiment of this application, the method further includes: if the search result shows that the manual instruction library has historical instructions that match the user input information, providing feedback to the user input information based on the historical instructions that match the user input information.

[0010] According to one embodiment of this application, the user input information satisfying the first condition includes at least one of the following: the confidence level of the pre-feedback information used to provide feedback on the user input information is less than a set confidence threshold; a processing instruction is received indicating a request for manual processing of the user input information; and feedback reminder information is received on the user input information, the feedback reminder information indicating that manual processing of the user input information is required.

[0011] According to one embodiment of this application, if the confidence level of the pre-feedback information used to provide feedback to the user input information is less than a set confidence threshold, the following operations are performed to determine this: obtaining context information of the user input information; determining the confidence level of the pre-feedback information based on the context information; and determining whether the confidence level is less than the set confidence threshold.

[0012] According to a second aspect of this application, an information feedback device is also provided, the device comprising: a first receiving module for receiving user input information; a retrieval module for performing a retrieval based on the user input information through a manual instruction library when the user input information meets a first condition; an information acquisition module for issuing an information acquisition instruction if the retrieval result shows that there is no historical instruction in the manual instruction library that matches the user input information; a second receiving module for receiving manual guidance information in response to the information acquisition instruction; and a generation module for generating feedback information for responding to the user input information based on the manual guidance information.

[0013] According to a third aspect of this application, an apparatus is provided, the apparatus comprising at least one processor, and at least one memory and a bus connected to the processor; wherein the processor and the memory communicate with each other via the bus; the processor is used to invoke program instructions in the memory to execute the above-described information feedback method.

[0014] It should be understood that the teachings of this application are not required to achieve all the beneficial effects described above, but rather that a specific technical solution can achieve a specific technical effect, and other embodiments of this application can also achieve beneficial effects not mentioned above. Attached Figure Description

[0015] The above and other objects, features, and advantages of exemplary embodiments of this application will become readily apparent from the following detailed description taken in conjunction with the accompanying drawings. Several embodiments of this application are illustrated in the drawings by way of example and not limitation, in which:

[0016] In the accompanying drawings, the same or corresponding reference numerals indicate the same or corresponding parts.

[0017] Figure 1 A schematic diagram illustrating the implementation flow of an information feedback method according to an embodiment of this application is shown;

[0018] Figure 2 A schematic diagram illustrating the implementation flow of another embodiment of the information feedback method of this application is shown;

[0019] Figure 3a This illustration shows an application scenario diagram of a specific application example of the information feedback method according to an embodiment of this application;

[0020] Figure 3b It shows Figure 3a This diagram illustrates the comparison between the application scenario of this specific application example and existing technologies.

[0021] Figure 4 The diagram illustrates the implementation flow of a specific application example of the information feedback method according to an embodiment of this application;

[0022] Figure 5 A schematic diagram of the composition structure of the information feedback device according to an embodiment of this application is shown;

[0023] Figure 6 A schematic diagram of the composition structure of the device according to an embodiment of this application is shown. Detailed Implementation

[0024] The principles and spirit of this application will now be described with reference to several exemplary embodiments. It should be understood that these embodiments are provided merely to enable those skilled in the art to better understand and implement this application, and are not intended to limit the scope of this application in any way. Rather, these embodiments are provided to make this application more thorough and complete, and to fully convey the scope of this application to those skilled in the art.

[0025] The technical solution of this application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0026] Figure 1A schematic diagram illustrating the implementation flow of an information feedback method according to an embodiment of this application is shown.

[0027] refer to Figure 1 The information feedback method in this application embodiment includes at least the following operation flow: Operation 101, receiving user input information; Operation 102, if the user input information meets a first condition, performing a search based on the user input information through a manual instruction library; Operation 103, if the search results show that there is no historical instruction in the manual instruction library that matches the user input information, issuing an information retrieval instruction; Operation 104, receiving manual guidance information in response to the information retrieval instruction; Operation 105, generating feedback information for providing feedback to the user input information based on the manual guidance information.

[0028] This application provides an information feedback method. If the user input information meets a first condition, the user input information is retrieved through a manual instruction database, avoiding direct transfer to a human agent and saving agent resources. When the manual instruction database does not have a matching instruction, feedback information for the user input information is generated by obtaining human guidance information, effectively handling new problems included in the user input information.

[0029] In operation 101, user input information is received.

[0030] In this embodiment of the application, the user input information can be questions raised by the user, such as: 1. I am an online celebrity and can do free push notifications for you. Can you provide me with your latest mobile phone? 2. I want to take the senior engineer exam. Do you offer such training?

[0031] In another embodiment of this application, the user input information may also be an instruction such as "transfer to human agent". Here, "transfer to human agent" may be text input information or a pre-configured virtual button or physical button.

[0032] In operation 102, if the user input information meets the first condition, a search is performed based on the user input information through the manual instruction library.

[0033] In this embodiment of the application, the user input information satisfying the first condition may include at least one of the following: the confidence level of the pre-feedback information used to provide feedback on the user input information is less than a set confidence threshold; a processing instruction is received indicating a request for manual processing of the user input information; or a feedback reminder is received for the user input information, the feedback reminder indicating that manual processing of the user input information is required.

[0034] In this embodiment of the application, if the confidence level of the pre-feedback information used to provide feedback to user input information is less than a set confidence threshold, the following operation can be used to determine this: obtain the context information of the user input information, determine the confidence level of the pre-feedback information based on the context information, and determine whether the confidence level is less than the set confidence threshold.

[0035] Here, an illustrative example of the application scenario of this embodiment is provided: a user inputs information through an intelligent customer service system and communicates with the user through the system. If the system automatically provides feedback on the user's input, whether the feedback meets the user's expectations depends primarily on whether the confidence level of the pre-feedback information obtained by the system based on multiple natural language understanding models already completed within the system reaches a set value. If the confidence level of the pre-feedback information is less than the set confidence threshold, it indicates that the system's automatic feedback on the user's input cannot meet the user's expectations.

[0036] In this embodiment of the application, if the user input information represents the first round of interaction between the user and the system, then the user input information here only includes the following content. If the user has already had one or more rounds of interaction with the system, then the contextual information of the user input information can be obtained here. Based on the contextual information, the confidence level of the pre-feedback information is determined, which can be done using a confidence level determination method commonly used in the field of natural language understanding. The confidence threshold set here can be set according to actual needs, and this application does not impose specific limitations on it.

[0037] In this embodiment of the application, user input information can be matched with multiple historical user input information in a manual command database to achieve retrieval based on user input information. The manual command database includes multiple historical commands and the correspondence between historical commands and their corresponding historical user input information.

[0038] Specifically, referring to the exemplary application scenarios mentioned above, user interaction with the system can be multiple interactions between a single user and the system, or multiple interactions between multiple users and the system. The manual instruction database can record the information processing steps transitioning from user to system interaction, recording a one-to-one correspondence between historical user input information and historical instructions for that historical user input information in a mapping or other form. Thus, the manual instruction database can include the correspondence between multiple historical user input information and their corresponding historical instructions. To ensure the effectiveness of the manual instruction database while effectively conserving storage resources, methods such as least-used methods can be used to delete historical user input information in the manual instruction database that has been retrieved less frequently than a set frequency within a set time period, deleting the corresponding historical instructions along with the historical user input information. Alternatively, methods such as most recently used methods can be used to delete historical user input information and corresponding historical instructions entered before a set time period from the manual instruction database.

[0039] In operation 103, if the search results show that there is no historical instruction in the manual instruction library that matches the user input information, an information retrieval instruction is issued.

[0040] In this embodiment of the application, the result of matching user input information with multiple historical user input information in the manual instruction library can include: the manual instruction library does not contain any historical instructions that match the user input information, and the manual instruction library does not contain any historical instructions that match the user input information. If the manual instruction library does not contain any historical instructions that match the user input information, then an information retrieval instruction can be issued to obtain human guidance information.

[0041] In operation 104, manual guidance information is received in response to the information retrieval instruction.

[0042] In this embodiment of the application, the manual guidance information can be the information needed to provide feedback on user input. For example, if a user inputs, "I'm an online celebrity, I can promote your products for free, can you provide me with your latest mobile phone?", natural language understanding can determine that the following information is needed to provide feedback: confirming whether the user is indeed an online celebrity, which can be determined by obtaining information such as the user's number of followers; and the standard for providing the latest mobile phone, which can be determined by whether the user's number of followers exceeds a set threshold. For example, the final generated manual guidance information could be: "Based on the dialogue, obtain the user's current number of followers. If the number of followers is greater than 5000, the user's promotion request can be accepted, and the user's contact information can be left; otherwise, the user is politely refused."

[0043] In operation 105, feedback information is generated based on the manual guidance information to provide feedback on the user's input.

[0044] In this embodiment of the application, feedback information for responding to user input can be generated based on human guidance information through the following operations: performing natural language understanding on the human guidance information to obtain information understanding results, and generating feedback information based on the information understanding results. The information understanding results are used to characterize the relevant information needed to provide feedback to the user input.

[0045] For example, a user inputs, "I'm an online influencer, I can do free promotions for you, can you provide me with your latest mobile phone?". The human guidance message could be, "Based on the dialogue, obtain the user's current number of followers. If the number of followers is greater than 5000, accept the user's promotion request and leave their contact information; otherwise, politely decline the user." Natural language understanding of this human guidance message can yield the following results: 1. It's necessary to confirm whether the user is indeed an online influencer, specifically by obtaining information such as the user's number of followers; and specifically, whether the user's number of followers exceeds a set threshold. 2. After obtaining the number of followers, it's necessary to determine the content of the next round of interaction based on the relationship between the number of followers and the set follower threshold, where the follower threshold is set according to the standard of providing the latest mobile phone. For example: if the number of followers is greater than the threshold, accept the user's promotion request and leave their contact information; otherwise, politely decline. All of the above information understanding results can be considered as the relevant information needed to represent the feedback to the user's input.

[0046] Based on the information understanding results obtained through natural language understanding of the human guidance information, the first feedback message generated could be "How many followers do you have?". At this point, the system can obtain the user's actual number of followers and make a further judgment based on that number. In this way, the system can acquire the ability to handle multi-round interaction processes in new and complex scenarios, instead of relying on continuous human interaction with the user.

[0047] Furthermore, in another embodiment of this application, the manual instruction library is updated based on the information understanding results.

[0048] In this embodiment of the application, if the search results show that there are historical instructions in the manual instruction library that match the user input information, then the user input information is fed back based on the historical instructions that match the user input information.

[0049] In this way, the human command database is constantly updated. For similar scenarios, users can directly search through the human command database, effectively avoiding the problems of large workload for human processing and individual differences such as language styles among different human agents. This significantly reduces the number of times users need to be transferred to human agents and significantly improves the user experience in terms of response speed and the targeted feedback of user input.

[0050] Figure 2 A schematic diagram illustrating the implementation flow of another embodiment of the information feedback method of this application is shown.

[0051] refer to Figure 2 The information feedback method in this application includes at least the following operation steps:

[0052] Operation 201: Receive user input information.

[0053] Operation 202: If the user input information meets the first condition, obtain the context information of the user input information.

[0054] Operation 203 involves retrieving information based on user input from a manual command database.

[0055] In this embodiment of the application, the manual instruction library includes multiple historical instructions and the correspondence between the historical instructions and the historical context information of the corresponding historical user input information.

[0056] Here, the context information of the user input can be obtained, and the obtained context information can be matched with the historical context information corresponding to multiple historical instructions.

[0057] Operation 204: If the search results show that there is no historical instruction in the manual instruction library that matches the user input information, issue an information retrieval instruction.

[0058] Operation 205: Receive manual guidance information in response to the information retrieval instruction.

[0059] Operation 206: Based on the manual guidance information, generate feedback information to respond to the user's input.

[0060] The specific implementation process of operations 201, 204 to 206 is as follows: Figure 1 The specific implementation processes of operations 101 to 105 in the illustrated embodiments are similar and will not be described again here.

[0061] Figure 3a The illustration shows a schematic diagram of an application scenario for a specific application example of the information feedback method according to an embodiment of this application.

[0062] refer to Figure 3a In this specific application example, the information feedback method is applied to an intelligent customer service system. The interactive interface diagram 301 of the intelligent customer service system illustrates the multi-round interaction process between the user and the system. The intelligent customer service system 302 simply illustrates the system's processing of user input information. When a human agent 303 needs human guidance information, they input it into the intelligent customer service system. The system's processing of user input information and the process of human agent 303 inputting human guidance information into the system will be discussed below in conjunction with... Figure 4 A detailed explanation will not be provided here.

[0063] Figure 3b It shows Figure 3a This diagram illustrates the comparison between the application scenario of this specific application example and existing technologies.

[0064] refer to Figure 3b As can be seen, in the existing technology, user 304 needs to interact not only with the intelligent customer service system 302, but also with a human agent 303 when encountering problems that the intelligent customer service system cannot solve. However, in the specific application example of the information feedback method in this application, the user only needs to interact with the intelligent customer service system. When encountering problems that the intelligent customer service system cannot solve, the user obtains information from the human agent 303 through the intelligent customer service system for natural language understanding and model training, reducing direct interaction between the user and the human agent 303 and saving human agent resources. Thus, all information obtained by the user has been processed by the intelligent customer service system, ensuring the consistency of interaction language style and other interaction forms. Furthermore, the intelligent customer service system continuously updates its human instruction library, quickly and effectively handling similar complex problems and significantly improving the user experience.

[0065] Figure 4 The diagram illustrates the implementation flow of a specific application example of the information feedback method according to an embodiment of this application.

[0066] refer to Figure 4 In this specific application example of this application, the information feedback method may include at least the following process:

[0067] Operation 401: Enter the current round of dialogue.

[0068] Operation 402: Receive user input information—a new round of user input question (Query).

[0069] Operation 403: Obtain context information for user input.

[0070] The contextual information of the user input here may include the interaction information of the previous round or multiple rounds before the current round, as well as the pre-feedback information obtained by the system after understanding the user input based on the existing natural language understanding function.

[0071] Operation 404 requires inputting the Query and Context into a scoring model that assigns a confidence score to the pre-feedback information. Here, any confidence scoring model commonly used in the field of natural language understanding can be used; this application does not impose any specific limitations on it.

[0072] Operation 405 calls the scoring model to obtain the confidence level of the pre-feedback information.

[0073] Operation 406: Determine if the confidence level is greater than the set confidence threshold. If yes, proceed to operation 407; otherwise, proceed to operation 408.

[0074] Operation 407: Based on the current Query, Context, and existing configuration of the intelligent customer service system, output the feedback information for the current round.

[0075] Operation 408, based on the current Query and Context, performs a search using a manual command library. The search results can also be scored based on the confidence score commonly used in the field of natural language understanding.

[0076] Operation 409: Determine if the score of the search result is greater than the set matching threshold. If yes, proceed to operation 412; otherwise, proceed to operation 410.

[0077] Operation 410 pushes the current Query and Context to a human agent.

[0078] Operation 411: Obtain new human guidance information.

[0079] Operation 412: Based on the current Query, Context, and human guidance information, call the natural language understanding module of the intelligent customer service system to perform natural language understanding and obtain feedback information on the user input.

[0080] Operation 413: Update the manual instruction database.

[0081] Operation 414: The current round of dialogue processing has ended.

[0082] The information feedback method in this application receives user input information. If the user input information meets a first condition, it searches a manual instruction database based on the user input information. If the search results show that there is no historical instruction in the manual instruction database that matches the user input information, it issues an information retrieval instruction, receives manual guidance information in response to the information retrieval instruction, and generates feedback information to provide feedback on the user input information based on the manual guidance information. Therefore, if the user input information meets the first condition, it searches the user input information through the manual instruction database, effectively avoiding direct transfer to a human. When there is no matching instruction in the manual instruction database, it generates feedback information for the user input information by retrieving manual guidance information, effectively handling new issues included in the user input information.

[0083] Similarly, based on the information feedback method described above, this application embodiment also provides a computer-readable storage medium storing a program that, when executed by a processor, causes the processor to perform at least the following operation steps: operation 101, receiving user input information; operation 102, if the user input information meets a first condition, performing a search based on the user input information using a manual instruction library; operation 103, if the search results show that there is no historical instruction in the manual instruction library that matches the user input information, issuing an information retrieval instruction; operation 104, receiving manual guidance information in response to the information retrieval instruction; operation 105, generating feedback information for responding to the user input information based on the manual guidance information.

[0084] Furthermore, based on the information feedback method described above, embodiments of this application also provide an information feedback device, such as... Figure 5 The device 50 includes: a first receiving module 501 for receiving user input information; a retrieval module 502 for performing a retrieval based on the user input information through a manual instruction library when the user input information meets a first condition; an information acquisition module 503 for issuing an information acquisition instruction if the retrieval result shows that there is no historical instruction in the manual instruction library that matches the user input information; a second receiving module 504 for receiving manual guidance information in response to the information acquisition instruction; and a generation module 505 for generating feedback information for responding to the user input information based on the manual guidance information.

[0085] In this embodiment of the application, the generation module 505 includes: an understanding submodule, used to perform natural language understanding on the human guidance information to obtain an information understanding result, the information understanding result being used to characterize the related information required to provide feedback to the user input information; and a generation submodule, used to generate feedback information based on the information understanding result.

[0086] In this embodiment of the present application, the device 50 further includes an update module for updating the manual instruction library based on the information understanding results.

[0087] In this embodiment of the application, the retrieval module 502 includes: a first matching submodule, used to match user input information with multiple historical user input information in a manual instruction library; wherein, the manual instruction library includes multiple historical instructions and the correspondence between historical instructions and corresponding historical user input information.

[0088] In this embodiment of the application, the retrieval module 502 includes: an information acquisition submodule for acquiring context information of user input information; and a second matching submodule for matching the context information with historical context information corresponding to multiple historical instructions; wherein, the manual instruction library includes multiple historical instructions and the correspondence between historical instructions and the context information of corresponding historical user input information.

[0089] In this embodiment of the application, the device 50 further includes a feedback module, configured to provide feedback to the user input information based on the historical instruction that matches the user input information if the search result indicates that there is a historical instruction in the manual instruction library that matches the user input information.

[0090] In this embodiment of the application, the user input information satisfying the first condition includes at least one of the following: the confidence level of the pre-feedback information used to provide feedback on the user input information is less than a set confidence threshold; a processing instruction is received indicating a request for manual processing of the user input information; or a feedback reminder is received for the user input information, the feedback reminder indicating that manual processing of the user input information is required.

[0091] In this embodiment of the application, if the confidence level of the pre-feedback information used to provide feedback to user input information is less than a set confidence threshold, the following operations are used to determine this: obtain context information of the user input information; determine the confidence level of the pre-feedback information based on the context information; and determine whether the confidence level is less than the set confidence threshold.

[0092] Furthermore, based on the information feedback method described above, embodiments of this application also provide a device, such as... Figure 6 The device 60 includes: at least one processor 601, and at least one memory 602 and a bus 603 connected to the processor 601; wherein the processor 601 and the memory 602 communicate with each other through the bus 603; the processor 601 is used to call program instructions in the memory 602 to execute the above-mentioned information feedback method.

[0093] It should be noted here that the above description of the embodiments of the information feedback device and equipment is consistent with the foregoing Figures 1 to 4 The method embodiments shown are described similarly and have the same characteristics as described above. Figures 1 to 4 The beneficial effects of the methods illustrated are similar and will not be repeated here. For technical details not disclosed in the embodiments of the information feedback device and equipment in this application, please refer to the foregoing description of this application. Figures 1 to 4 The method embodiments shown are for understanding purposes only and will not be described in detail here for the sake of brevity.

[0094] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0095] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods, such as: multiple units or components can be combined, or integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the various components shown or discussed can be through some interfaces, and the indirect coupling or communication connection between devices or units can be electrical, mechanical, or other forms.

[0096] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units. They may be located in one place or distributed across multiple network units. Some or all of the units may be selected to achieve the purpose of this embodiment according to actual needs.

[0097] In addition, each functional unit in the various embodiments of this application can be integrated into one processing unit, or each unit can be a separate unit, or two or more units can be integrated into one unit; the integrated unit can be implemented in hardware or in the form of hardware plus software functional units.

[0098] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media that can store program code, such as mobile storage devices, read-only memory (ROM), magnetic disks, or optical disks.

[0099] Alternatively, if the integrated units described above are implemented as software functional modules and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of this application, or the parts that contribute to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, ROMs, magnetic disks, or optical disks.

[0100] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. An information feedback method, the method comprising: Receive user input information; If the user input information meets the first condition, a search is performed based on the user input information using a manual instruction database. If the search results show that there is no historical instruction in the manual instruction library that matches the user input information, an information retrieval instruction is issued. Receive human guidance information in response to the information acquisition instruction; Natural language understanding is performed on the human guidance information to obtain information understanding results, which are used to characterize the relevant information that needs to be obtained to provide feedback on the user input information; Based on the information received, the results are understood, and feedback information is generated.

2. The method according to claim 1, further comprising: The manual instruction library is updated based on the information understanding results.

3. The method according to claim 1, wherein the retrieval based on the user input information using a manual command database includes: The user input information is matched with multiple historical user input information from the manual instruction database; The manual instruction library includes multiple historical instructions and the correspondence between the historical instructions and the corresponding historical user input information.

4. The method according to claim 1, wherein the retrieval based on the user input information using a manual command database includes: Obtain the context information of the user input; The context information is matched with the historical context information corresponding to multiple historical instructions; The manual instruction library includes multiple historical instructions and the correspondence between the historical instructions and the context information of the corresponding historical user input information.

5. The method according to claim 1, further comprising: If the search results show that there are historical instructions in the manual instruction library that match the user input information, then the user input information is fed back based on the historical instructions that match the user input information.

6. The method according to claim 1, wherein the user input information satisfies at least one of the following conditions: The confidence level of the pre-feedback information used to provide feedback on the user input is less than a set confidence threshold; Received a processing instruction indicating a request for manual processing of the user input information; Upon receiving feedback notification information regarding the user input, the feedback notification information indicates that the user input information requires manual processing.

7. The method according to claim 6, wherein the confidence level of the pre-feedback information used to provide feedback on the user input information is less than a set confidence threshold, is determined by the following operation: Obtain the context information of the user input; Based on the context information, determine the confidence level of the pre-feedback information; Determine whether the confidence level is less than the set confidence threshold.

8. An information feedback device, the device comprising: The first receiving module is used to receive user input information; The retrieval module is used to retrieve information based on the user input information by means of a manual instruction library, provided that the user input information meets the first condition. The information acquisition module is used to issue an information acquisition command if the search results show that there is no historical command in the manual command library that matches the user input information. The second receiving module is used to receive manual guidance information in response to the information acquisition instruction; The generation module is used to generate feedback information based on the manual guidance information to provide feedback on the user input information; The generation module includes: an understanding submodule, used to perform natural language understanding on the human guidance information to obtain an information understanding result, wherein the information understanding result is used to characterize the associated information required to provide feedback on the user input information; A generation submodule is used to understand the results based on the information and generate the feedback information.

9. An apparatus, the apparatus comprising at least one processor, and at least one memory and a bus connected to the processor; wherein, The processor and the memory communicate with each other via the bus; The processor is used to invoke program instructions in the memory to execute the information feedback method according to any one of claims 1-7.

Images