AI server real-time enhancement system and data processing method

Abstract

The invention discloses an AI server real-time enhancement system and a data processing method, the AI server real-time enhancement system comprises an enhancement interface module, a knowledge base module, a collaboration module and a collaboration seat module, and the enhancement interface module, the knowledge base module, the collaboration module and the collaboration seat module are connected in sequence. The AI self-service system generates input information according to input content and screen content of a user and then sends the input information to the AI server-side real-time enhancement system, the system supplements and updates the input information and returns final collaborative enhancement information or matching knowledge to the AI server-side module, and the AI server-side module performs collaborative enhancement on the input information. And the AI server module generates a final script file according to the collaborative enhancement information or the matching knowledge and replies the final script file to the AI front-end module, so that the defect that the script file generation capability of the original AI server module is insufficient is overcome, and the data processing precision of the AI self-service system is enhanced.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to an AI server real-time enhancement system and a data processing method. Background Art

[0002] Public self-service devices refer to machine devices placed in public places for citizens to handle various types of business self-service. These devices usually integrate multiple functions such as identity recognition, information query, and business handling, aiming to relieve the queuing pressure at manual windows and improve service efficiency. With the emergence of large models of AI (Artificial Intelligence), AI has been able to complete semantic understanding of conversations with people. By empowering self-service devices with AI, it has become a reality for users to ask questions and for AI to complete the operations of self-service devices.

[0003] Self-service devices based on AI assistants can quickly transform a large number of existing self-service devices (original self-service modules). However, various existing self-service devices have diverse applications, and the user operation interfaces are constantly updated. The existing AI technologies have not yet reached the level of being proficient in business and being able to self-learn and self-understand new services and new operation interfaces. Or when the user interface of the original self-service module is updated, the AI server module lacks the corresponding capabilities to respond due to lack of timely update and thus cannot generate script files. Therefore, when the AI technology of the self-service device itself cannot meet the requirements of automatic business handling, it will result in the inability to reply to the script file of the AI front-end module, thus causing the self-service device with an added AI assistant to be unable to complete its work.

[0004] Therefore, the existing technology still needs to be improved and developed. Summary of the Invention

[0005] The main purpose of the present invention is to provide an AI server real-time enhancement system and a data processing method, aiming to solve the problems of insufficient script file generation ability and low data processing accuracy of the original AI server in the existing self-service system.

[0006] To achieve the above object, the present invention provides an AI server real-time enhancement system, and the AI server real-time enhancement system includes: An enhancement interface module, a knowledge base module, a collaboration module, and a collaboration agent module, which are connected in sequence, the enhancement interface module, the knowledge base module, the collaboration module, and the collaboration agent module; The enhancement interface module is used to receive input information from the AI self-service system and send the input information to the knowledge base module; The knowledge base module is used to extract keywords from the input information, match the input information according to a preset matching rule, generate matching knowledge, and send the keywords and the input information to the collaboration module; The collaboration module is used to generate routing information according to the keywords, and send the input information to the collaboration agent module according to the routing information; The collaboration agent module is used to generate collaborative enhancement information according to the input information, and send the collaborative enhancement information to the collaboration module; The collaboration module is also used to receive the collaborative enhancement information and send the collaborative enhancement information to the knowledge base module; The knowledge base module is also used to process the collaborative enhancement information or the matching knowledge and send it to the enhancement interface module; The enhancement interface module is also used to receive the collaborative enhancement information or the matching knowledge, and send the collaborative enhancement information or the matching knowledge to the AI self-service system.

[0007] Optionally, in the AI server real-time enhancement system, the knowledge base module includes: a keyword extraction unit, a keyword comparison unit, a semantic similarity comparison unit, a knowledge fusion unit, and a first judgment unit; The keyword extraction unit is used to receive the input information from the enhancement interface module, extract keywords from the input information, and send the keywords and the input information to the collaboration module; The keyword comparison unit is used to obtain the keywords, match the keywords with the default keywords of the pre-stored knowledge block to obtain the first matching keywords, and obtain the first matching knowledge according to the first matching keywords; The semantic similarity comparison unit is used to receive the input information from the AI self-service system, perform semantic analysis on the input information to obtain a semantic analysis result, match keywords with a semantic similarity exceeding a preset threshold with the semantic analysis result from the default keywords of the pre-stored knowledge block as the second matching keywords, and obtain the second matching knowledge according to the second matching keywords; The knowledge fusion unit is used to fuse the first matching knowledge and the second matching knowledge, generate matching knowledge, and temporarily store the matching knowledge; The judgment unit is used to receive the collaborative enhancement information from the collaboration module, judge whether the collaborative enhancement information is empty. If the collaborative enhancement information is empty, send the matching knowledge to the enhancement interface module. If the collaborative enhancement information is not empty, send the collaborative enhancement information to the enhancement interface module.

[0008] Optionally, in the AI server real-time enhancement system, the knowledge base module further includes: a collaborative enhancement information storage unit; The collaborative enhancement information storage unit is used to save the collaborative enhancement information for future use when the collaborative enhancement information is not empty.

[0009] Optionally, in the AI server real-time enhancement system, the collaboration module includes: a routing information generation unit, an information sending unit, and a feedback unit; The routing information generation unit is used to receive the keyword, obtain the routing information corresponding to the keyword in a preset routing definition table, and send the routing information to the information sending unit; The information sending unit is used to receive the routing information and the input information, and send the input information to the collaborative agent module according to the routing information; The feedback unit is used to receive the collaborative enhancement information from the collaborative agent module and send the collaborative enhancement information to the knowledge base module.

[0010] Optionally, in the AI server real-time enhancement system, the collaborative agent module includes: a human collaborative agent unit and an AI collaborative agent unit; The human collaborative agent unit is used to receive the input information from the collaboration module, send the input information to an online customer service for enhancement processing based on experience, receive the collaborative enhancement information from the online customer service, and send the collaborative enhancement information from the online customer service to the collaboration module; The AI collaborative agent unit is used to receive the input information from the collaboration module, process the input information using AI technology to obtain the collaborative enhancement information from AI, and send the collaborative enhancement information from AI to the collaboration module.

[0011] Optionally, in the AI server real-time enhancement system, the AI self-service system includes: an AI front-end module, an APP plug-in module, and an AI server module; The AI front-end module is used to collect the input content of the user and send the input content to the AI server module; The APP plug-in module is used to obtain the screen content of the self-service interface, package the screen content, and send it to the AI server module; The AI server module is used to process the input content and the screen content, generate a response result, and send the input content, the screen content, and the response result as input information to the enhancement interface module; The AI server module is further configured to receive the collaborative enhancement information or the matching knowledge from the enhancement interface module, generate a script file according to the collaborative enhancement information or the matching knowledge, and send the script file to the AI front-end module; The AI front-end module is further configured to receive the script file and send the script file to the units corresponding to the APP plug-in module and the AI front-end module respectively to automatically execute the self-service function.

[0012] In addition, to achieve the above object, the present invention also provides a data processing method for an AI server real-time enhancement system, wherein the data processing method includes: The enhancement interface module receives input information from the AI self-service system and sends the input information to the knowledge base module; The knowledge base module extracts keywords from the input information, matches the input information according to a preset matching rule, generates matching knowledge, and sends the keywords and the input information to the collaboration module; The collaboration module generates routing information according to the keywords, and sends the input information to the collaboration agent module according to the routing information; The collaboration agent module generates collaborative enhancement information according to the input information, sends the collaborative enhancement information to the collaboration module, and the collaboration module receives the collaborative enhancement information and sends the collaborative enhancement information to the knowledge base module; The knowledge base module processes the collaborative enhancement information or the matching knowledge and sends it to the enhancement interface module, and the enhancement interface module receives the collaborative enhancement information or the matching knowledge and sends the collaborative enhancement information or the matching knowledge to the AI self-service system.

[0013] Optionally, in the data processing method of the AI server real-time enhancement system, the extracting keywords from the input information, matching the input information according to a preset matching rule, and generating matching knowledge specifically includes: The keyword comparison unit receives input information from the AI self-service system, extracts keywords from the input information, matches the keywords with the default keywords of the pre-stored knowledge block to obtain the first matching keyword, and obtains the first matching knowledge according to the first matching keyword; The semantic similarity comparison unit receives the input information from the AI self-service system, performs semantic analysis on the input information to obtain a semantic analysis result, matches keywords with a semantic similarity exceeding a preset threshold with the semantic analysis result from the default keywords of the pre-stored knowledge chunks as the second matching keywords, and obtains second matching knowledge according to the second matching keywords; The knowledge fusion unit fuses the first matching knowledge and the second matching knowledge, generates matching knowledge, and temporarily stores the matching knowledge.

[0014] Optionally, in the data processing method of the AI server real-time enhancement system, wherein, generating routing information according to the keyword in combination with a preset cooperation rule, and sending the input information to the cooperation seat module according to the routing information, specifically includes: The routing information generation unit receives the keyword, obtains the routing information corresponding to the keyword in a preset routing definition table, and sends the routing information to the information sending unit; Wherein, the routing information includes an artificial cooperation seat path and an AI cooperation seat path; When the routing information is an artificial cooperation seat path, the information sending unit sends the input information to the artificial cooperation seat unit of the cooperation seat module, and when the routing information is an AI cooperation seat path, the information sending unit sends the input information to the AI cooperation seat unit of the cooperation seat module.

[0015] Optionally, in the data processing method of the AI server real-time enhancement system, wherein, processing the cooperation enhancement information or the matching knowledge and then sending it to the enhancement interface module, specifically includes: The knowledge base module determines whether the cooperation enhancement information is empty. If the cooperation enhancement information is empty, the knowledge base module sends the matching knowledge to the enhancement interface module; If the cooperation enhancement information is not empty, the knowledge base module sends the cooperation enhancement information to the enhancement interface module and saves the cooperation enhancement information for future use.

[0016] In the present invention, the enhanced interface module receives input information from the AI self-service system and sends the input information to the knowledge base module; the knowledge base module extracts keywords from the input information, matches the input information according to a preset matching rule, generates matching knowledge, and sends the keywords and the input information to the collaboration module; the collaboration module generates routing information according to the keywords, and according to the routing information, sends the input information to the collaborative agent module; the collaborative agent module generates collaborative enhancement information according to the input information and sends the collaborative enhancement information to the collaboration module, the collaboration module receives the collaborative enhancement information and sends the collaborative enhancement information to the knowledge base module; the knowledge base module processes the collaborative enhancement information or the matching knowledge and sends it to the enhanced interface module, the enhanced interface module receives the collaborative enhancement information or the matching knowledge and sends the collaborative enhancement information or the matching knowledge to the AI self-service system. The present invention enhances the data processing accuracy of the AI self-service system and improves the user experience. Brief Description of the Drawings

[0017] Figure 1 is the overall architecture diagram of the AI server real-time enhancement system of the present invention; Figure 2 is the flowchart of the preferred embodiment of the data processing method of the AI server real-time enhancement system of the present invention. Detailed Embodiments

[0018] To make the objectives, technical solutions and advantages of the present invention clearer and more definite, the following further describes the present invention in detail with reference to the accompanying drawings and by way of examples. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0019] To solve the problems in the prior art, this embodiment provides an AI server real-time enhancement system, as Figure 1 shown, the AI server real-time enhancement system includes: an enhanced interface module, a knowledge base module, a collaboration module, and a collaborative agent module.

[0020] Among them, the enhanced interface module, the knowledge base module, the collaboration module, and the collaborative agent module are connected in sequence.

[0021] The enhanced interface module is used to receive input information from the AI self-service system and send the input information to the knowledge base module; The knowledge base module is used to extract keywords from the input information, match the input information according to a preset matching rule, generate matching knowledge, and send the keywords and the input information to the collaboration module; The collaboration module is used to generate routing information according to the keyword, and combine with a preset collaboration rule, and send the input information to the collaboration agent module according to the routing information; The collaboration agent module is used to generate collaboration enhancement information according to the input information, and send the collaboration enhancement information to the collaboration module; The collaboration module is further used to receive the collaboration enhancement information, and send the collaboration enhancement information to the knowledge base module; The knowledge base module is further used to process the collaboration enhancement information or the matching knowledge and send it to the enhancement interface module; The enhancement interface module is further used to receive the collaboration enhancement information or the matching knowledge, and send the collaboration enhancement information or the matching knowledge to the AI self-service system.

[0022] Furthermore, the present invention also provides a data processing unit included in each module, which is used to execute corresponding data collection and processing operations.

[0023] Specifically, the knowledge base module includes: a keyword extraction unit, a keyword comparison unit, a semantic similarity comparison unit, a knowledge fusion unit, and a first judgment unit.

[0024] Among them, the keyword extraction unit is used to receive the input information from the enhancement interface module, extract keywords from the input information, and send the keywords and the input information to the collaboration module.

[0025] The keyword comparison unit is used to obtain the keywords, match the keywords with the default keywords of the pre-stored knowledge blocks to obtain the first matching keywords, and obtain the first matching knowledge according to the first matching keywords.

[0026] The semantic similarity comparison unit is used to receive the input information from the AI self-service system, perform semantic analysis on the input information to obtain a semantic analysis result, match keywords with a semantic similarity exceeding a preset threshold with the semantic analysis result from the default keywords of the pre-stored knowledge blocks as the second matching keywords, and obtain the second matching knowledge according to the second matching keywords.

[0027] The knowledge fusion unit is used to fuse the first matching knowledge and the second matching knowledge, generate matching knowledge, and temporarily store the matching knowledge.

[0028] The determination unit is configured to receive the collaborative enhancement information from the collaborative module, determine whether the collaborative enhancement information is empty. If the collaborative enhancement information is empty, the matching knowledge is sent to the enhancement interface module. If the collaborative enhancement information is not empty, the collaborative enhancement information is sent to the enhancement interface module.

[0029] It can be understood that the knowledge base module is configured to receive information from the enhancement interface module, and based on the input information and in combination with pre-set matching rules, match the input information with pre-stored knowledge chunks. The pre-stored knowledge chunks with successful matching are the matching knowledge, otherwise the matching knowledge is empty.

[0030] In this embodiment, the pre-set matching rules can be any one or a combination of the following: Database keyword retrieval rule: Extract keywords from the input information and compare them with the keywords of the pre-stored knowledge chunks. If they are the same, the matching is successful; or, semantic retrieval rule: Through the semantic comparison algorithm, compare the semantics of the input information with the semantics of the pre-stored knowledge chunks. If the similarity is greater than or equal to a pre-set threshold (the threshold range is from 0 to 1, and the higher the threshold, the higher the semantic matching degree), the matching is successful. Among them, the semantic comparison algorithm can be the cosine similarity algorithm or other algorithms.

[0031] The knowledge base module is also configured to send the keywords and the input information to the collaborative module and wait for a reply from the collaborative module. After receiving the collaborative enhancement information replied by the collaborative module, it is processed: Determine whether the collaborative enhancement information is empty. If the collaborative enhancement information is empty, the matching knowledge is sent to the enhancement interface module. If the collaborative enhancement information is not empty, the collaborative enhancement information is sent to the enhancement interface module.

[0032] Furthermore, the knowledge base module further includes: a collaborative enhancement information storage unit; the collaborative enhancement information storage unit is configured to save the collaborative enhancement information for future use when the collaborative enhancement information is not empty.

[0033] It can be understood that after the knowledge base module receives the collaborative enhancement information replied by the collaborative module, it determines whether the collaborative enhancement information is empty. If the collaborative enhancement information is not empty, it directly uses the collaborative enhancement information storage unit to save the collaborative enhancement information. When the same input information is received next time, it directly calls the collaborative enhancement information in the collaborative enhancement information storage unit and outputs the collaborative enhancement information directly to the enhancement interface module, and sends it to the AI self-service system through the enhancement interface module to improve data processing efficiency.

[0034] Further, the collaboration module includes: a routing information generation unit, an information sending unit, and a feedback unit.

[0035] Among them, the routing information generation unit is used to receive the keyword, obtain the routing information corresponding to the keyword in a preset routing definition table, and send the routing information to the information sending unit.

[0036] The information sending unit is used to receive the routing information and the input information, and send the input information to the collaborative agent module according to the routing information.

[0037] The feedback unit is used to receive the collaborative enhancement information from the collaborative agent module and send the collaborative enhancement information to the knowledge base module.

[0038] It can be understood that the collaboration module is used to receive information from the knowledge base module, combine preset collaboration rules, and send the information as a collaboration request to one or both of the human collaborative agent and the AI collaborative agent.

[0039] Specifically, a set of operation sequences, called a routing definition table, is included in the collaboration rules, in the form of keyword - routing information. The keyword can be a certain noun or phrase, and the routing information includes the path of the human collaborative agent and the path of the AI collaborative agent.

[0040] In this embodiment, the keyword received by the routing information generation unit is compared with the keywords in the routing definition table to obtain the routing information corresponding to the keyword, and a collaboration request is sent to the corresponding human collaborative agent or AI collaborative agent according to the routing information corresponding to the keyword; if the received information does not contain any keyword in the routing definition table or the routing definition table does not exist, no collaboration request is sent to the human collaborative agent or the AI collaborative agent.

[0041] It can be understood that the feedback unit is used to receive the collaborative enhancement information from the collaborative agent module. The collaborative enhancement information can be empty or not empty. In this embodiment, the non - empty collaborative enhancement information is called valid collaborative enhancement information.

[0042] If the information sending unit only sends a collaboration request to the AI collaborative agent, and receives a reply from the AI collaborative agent and the AI agent's reply information is not blank, then valid collaborative enhancement information is generated based on this information and replied to the knowledge base module.

[0043] If the information sending unit simultaneously sends a collaboration request to both the human collaboration agent and the AI collaboration agent, and receives a response from the human collaboration agent and the response information from the human agent is not blank, then directly generate valid collaboration enhancement information based on this information and reply to the knowledge base module. At the same time, ignore the response from the AI collaboration agent (regardless of whether a collaboration request has been sent to the AI collaboration agent, that is, set the priority of the human agent higher than that of the AI collaboration agent).

[0044] Furthermore, the collaboration agent module includes: a human collaboration agent unit and an AI collaboration agent unit.

[0045] Among them, the human collaboration agent unit is used to receive the input information from the collaboration module, send the input information to the online customer service for enhanced processing based on experience, receive the collaboration enhancement information from the online customer service, and send the collaboration enhancement information from the online customer service to the collaboration module.

[0046] In this embodiment, the human collaboration agent unit is used to receive the input information from the collaboration module with the routing information being the human collaboration agent path, and the human generates a reply message (i.e., collaboration enhancement information) or a blank reply according to experience. If there is no human reply within a preset time period, the agent will also automatically reply with a blank.

[0047] The AI collaboration agent unit is used to receive the input information from the collaboration module, process the input information using AI technology to obtain the collaboration enhancement information from the AI, and send the collaboration enhancement information from the AI to the collaboration module.

[0048] In this embodiment, the AI collaboration agent unit is used to receive the input information from the collaboration module with the routing information being the AI collaboration agent path, and the AI processes this information to generate a reply. This AI can be an LLM (Large Language Model) or an agent derived from an LLM. Generate a reply message or a blank reply. If there is no AI reply within a preset time period, the agent will also automatically reply with a blank.

[0049] Even further, the AI self-service system includes: an AI front-end module, an APP plug-in module, and an AI server module.

[0050] Among them, the AI front-end module is used to collect the input content of the user and send the input content to the AI server module; the APP plug-in module is used to obtain the screen content of the self-service interface and send the screen content to the AI server module after packaging; the AI server module is used to process the input content and the screen content, generate a response result, and send the input content, the screen content, and the response result as input information to the enhancement interface module; The AI server module is further configured to receive the collaborative enhancement information or the matching knowledge from the enhancement interface module, generate a script file according to the collaborative enhancement information or the matching knowledge, and send the script file to the AI front-end module; The AI front-end module is further configured to receive the script file and send the script file to the corresponding units of the APP plug-in module and the AI front-end module respectively to automatically execute the self-service function.

[0051] Based on the AI server real-time enhancement system described in the above embodiments, the present invention further provides a data processing method for the AI server real-time enhancement system, specifically as Figure 2 shown in, the data processing method of the AI server real-time enhancement system includes the following steps: Step S10: The enhancement interface module receives the input information from the AI self-service system and sends the input information to the knowledge base module.

[0052] Specifically, it receives the user input content from the AI front-end module of the AI self-service system and the screen content from the APP plug-in module, generates a response result, and sends the user input content, the screen content, and the response result as input information to the enhancement interface module.

[0053] The enhancement interface module sends the input information to the knowledge base module and receives the reply from the knowledge base module. If the reply information of the knowledge base module is not blank, it generates enhanced script information according to the information and replies to the AI server module; otherwise, it replies blank enhanced script information to the AI server module.

[0054] Step S20: The knowledge base module extracts keywords from the input information, matches the input information according to a preset matching rule, generates matching knowledge, and sends the keywords and the input information to the collaboration module.

[0055] Specifically, the knowledge base module receives the input information from the AI self-service system, extracts keywords from the input information, matches the keywords with the default keywords of the pre-stored knowledge chunks to obtain the first matching keywords, and obtains the first matching knowledge according to the first matching keywords.

[0056] In this embodiment, the keyword comparison unit of the knowledge base module extracts the keywords in the input information through the database keyword retrieval rule, and compares the keywords with the default keywords of the pre-stored knowledge chunks. If they are the same, the matching is successful, and the first matching keywords are obtained. Then, knowledge search is performed according to the first matching keywords to obtain the first matching knowledge.

[0057] Further, the semantic similarity comparison unit of the knowledge base module receives the input information from the AI self-service system, performs semantic analysis on the input information to obtain a semantic analysis result, matches the keywords in the default keywords of the pre-stored knowledge chunks whose semantic similarity with the semantic analysis result exceeds a preset threshold as the second matching keywords, and obtains the second matching knowledge according to the second matching keywords.

[0058] In this embodiment, the knowledge base module also performs semantic analysis on the input information through the semantic retrieval rule to obtain a semantic analysis result, and compares the input information semantically with the semantics of the pre-stored knowledge chunks. If it is greater than or equal to the preset threshold (the threshold range is from 0 to 1, and the higher the threshold, the higher the semantic matching degree), the matching is successful. The semantic comparison algorithm can be the cosine similarity algorithm or other algorithms.

[0059] Furthermore, the knowledge fusion unit of the knowledge base module fuses the first matching knowledge and the second matching knowledge to generate matching knowledge, and temporarily stores the matching knowledge. Then, the keywords and the input information are sent to the collaboration module.

[0060] It can be understood that this application fuses the two types of matching knowledge obtained based on the database keyword retrieval rule and the semantic retrieval rule, can more comprehensively and accurately obtain the semantics contained in the input information, and then generate more accurate matching knowledge and temporarily store the matching knowledge. At the same time, the keywords and the input information are sent to the collaboration module to obtain collaborative enhanced information. If the collaboration module does not return collaborative enhanced information, the temporarily stored matching knowledge is obtained as the final enhanced knowledge and sent to the AI self-service system.

[0061] Step S30: The collaboration module generates routing information according to the keywords, and sends the input information to the collaboration agent module according to the routing information.

[0062] Specifically, the routing information generation unit receives the keyword, obtains the routing information corresponding to the keyword in a preset routing definition table, and sends the routing information to the information sending unit; wherein, the routing information includes a manual collaboration agent path and an AI collaboration agent path.

[0063] When the routing information is a manual collaboration agent path, the information sending unit sends the input information to the manual collaboration agent unit of the collaboration agent module; when the routing information is an AI collaboration agent path, the information sending unit sends the input information to the AI collaboration agent unit of the collaboration agent module.

[0064] It can be understood that the collaboration rules contain a set of operation sequences, called a routing definition table, in the form of keyword - routing information. The keyword can be a certain noun or phrase, and the routing information includes a manual collaboration agent path and an AI collaboration agent path.

[0065] In this embodiment, the keyword received by the routing information generation unit is compared with the keywords in the routing definition table to obtain the routing information corresponding to the keyword, and a collaboration request is sent to the corresponding manual collaboration agent or AI collaboration agent according to the routing information corresponding to the keyword; if the received information does not contain any keyword in the routing definition table or the routing definition table does not exist, no collaboration request is sent to the manual collaboration agent or AI collaboration agent.

[0066] Step S40: The collaboration agent module generates collaboration enhancement information according to the input information, and sends the collaboration enhancement information to the collaboration module. The collaboration module receives the collaboration enhancement information and sends the collaboration enhancement information to the knowledge base module.

[0067] Specifically, the collaboration agent module includes: a manual collaboration agent unit and an AI collaboration agent unit.

[0068] In this embodiment, the manual collaboration agent unit receives the input information from the collaboration module with the routing information being the manual collaboration agent path, and an operator generates a reply message (i.e., collaboration enhancement information) or a blank reply according to experience. If there is no manual reply within a preset time period, the seat will also automatically reply with a blank.

[0069] The AI collaboration agent unit receives the input information from the collaboration module, processes the input information using AI technology to obtain the collaboration enhancement information from the AI, and sends the collaboration enhancement information from the AI to the collaboration module.

[0070] It can be understood that the AI collaborative agent unit is used to receive input information from the collaborative module with the routing information being the AI collaborative path, and the AI processes this information to generate a response. This AI can be an LLM (Large Language Model) or an agent derived from an LLM.

[0071] Step S50: The knowledge base module processes the collaborative enhancement information or the matching knowledge and sends it to the enhancement interface module. The enhancement interface module receives the collaborative enhancement information or the matching knowledge and sends the collaborative enhancement information or the matching knowledge to the AI self-service system.

[0072] Specifically, the knowledge base module determines whether the collaborative enhancement information is empty. If the collaborative enhancement information is empty, the knowledge base module sends the matching knowledge to the enhancement interface module; if the collaborative enhancement information is not empty, the knowledge base module sends the collaborative enhancement information to the enhancement interface module and saves the collaborative enhancement information for future use.

[0073] Further, the enhancement interface module receives the collaborative enhancement information (when the collaborative enhancement information is not empty) or the matching knowledge (when the collaborative enhancement information is empty), and sends the collaborative enhancement information or the matching knowledge to the AI self-service system. The AI server module of the AI self-service system generates an enhanced script file based on the collaborative enhancement information or the matching knowledge, and sends the enhanced script file to the AI front-end module. The AI front-end module receives the enhanced script file and sends the script file to the corresponding units of the APP plug-in module and the AI front-end module to automatically execute the self-service function.

[0074] In summary, the present invention provides an AI server real-time enhancement system and a data processing method. The AI server real-time enhancement system includes: an enhancement interface module, a knowledge base module, a collaboration module, and a collaboration agent module, which are connected in sequence. In the present invention, after the AI self-service system generates input information based on the user's input content and screen content, the input information is sent to the AI server real-time enhancement system. The system supplements and updates the input information, and returns the final collaborative enhancement information or the matching knowledge to the AI server module. The AI server module can generate a final script file based on the collaborative enhancement information or the matching knowledge and reply to the AI front-end module, so as to solve the defect of insufficient script file generation ability of the original AI server module. The present invention can endow the AI server module with the latest manual experience in a timely manner, thereby solving user problems, enhancing the data processing accuracy of the AI self-service system, and improving the user experience.

[0075] It should be noted that in this article, the terms "include", "comprise" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal including a series of elements not only includes those elements, but also includes other elements not expressly listed, or further includes elements inherent to such process, method, article or terminal. Without more limitations, the element defined by the statement "including one..." does not exclude the existence of additional identical elements in the process, method, article or terminal including the element.

[0076] Of course, those of ordinary skill in the art can understand that all or part of the processes of implementing the above method embodiments can be completed by instructing relevant hardware (such as a processor, a controller, etc.) through a computer program. The program can be stored in a computer-readable storage medium readable by a computer. When the program is executed, it can include the processes of the above method embodiments. The computer-readable storage medium can be a memory, a magnetic disk, an optical disk, etc.

[0077] It should be understood that the application of the present invention is not limited to the above examples. For those of ordinary skill in the art, improvements or transformations can be made according to the above description. All such improvements and transformations should fall within the protection scope of the appended claims of the present invention.

Claims

1. An AI server real-time enhancement system, characterized in that, The described AI server real-time enhancement system includes: an enhancement interface module, a knowledge base module, a collaboration module, and a collaboration agent module. The enhancement interface module, the knowledge base module, the collaboration module, and the collaboration agent module are connected in sequence; The enhancement interface module is used to receive input information from the AI self-service system and send the input information to the knowledge base module; The knowledge base module is used to extract keywords from the input information, match the input information according to a preset matching rule, generate matching knowledge, and send the keywords and the input information to the collaboration module; The collaboration module is used to generate routing information according to the keywords, and send the input information to the collaboration agent module according to the routing information; The collaboration agent module is used to generate collaboration enhancement information according to the input information and send the collaboration enhancement information to the collaboration module; The collaboration module is also used to receive the collaboration enhancement information and send the collaboration enhancement information to the knowledge base module; The knowledge base module is also used to process the collaboration enhancement information or the matching knowledge and send it to the enhancement interface module; The enhancement interface module is also used to receive the collaboration enhancement information or the matching knowledge and send the collaboration enhancement information or the matching knowledge to the AI self-service system.

2. The AI server real-time enhancement system according to claim 1, characterized in that The knowledge base module includes: a keyword extraction unit, a keyword comparison unit, a semantic similarity comparison unit, a knowledge fusion unit, and a first judgment unit; The keyword extraction unit is used to receive input information from the enhancement interface module, extract keywords from the input information, and send the keywords and the input information to the collaboration module; The keyword comparison unit is used to obtain the keywords, match the keywords with the default keywords of the pre-stored knowledge block to obtain the first matching keywords, and obtain the first matching knowledge according to the first matching keywords; The semantic similarity comparison unit is used to receive input information from the AI self-service system, perform semantic analysis on the input information to obtain a semantic analysis result, match keywords with a semantic similarity exceeding a preset threshold from the default keywords of the pre-stored knowledge block as the second matching keywords, and obtain the second matching knowledge according to the second matching keywords; The knowledge fusion unit is used to fuse the first matching knowledge and the second matching knowledge, generate matching knowledge, and temporarily store the matching knowledge; The judgment unit is used to receive the collaboration enhancement information from the collaboration module, judge whether the collaboration enhancement information is empty. If the collaboration enhancement information is empty, it sends the matching knowledge to the enhancement interface module. If the collaboration enhancement information is not empty, it sends the collaboration enhancement information to the enhancement interface module.

3. The AI server real-time enhancement system according to claim 2, wherein, The knowledge base module also includes: a collaboration enhancement information storage unit; The collaboration enhancement information storage unit is used to save the collaboration enhancement information for future use when the collaboration enhancement information is not empty.

4. The AI server real-time enhancement system according to claim 1, characterized in that, The collaborative module includes: a routing information generation unit, an information sending unit, and a feedback unit; The routing information generation unit is configured to receive the keyword, obtain the routing information corresponding to the keyword in a preset routing definition table, and send the routing information to the information sending unit; The information sending unit is configured to receive the routing information and the input information, and send the input information to the collaborative agent module according to the routing information; The feedback unit is configured to receive the collaborative enhancement information from the collaborative agent module, and send the collaborative enhancement information to the knowledge base module.

5. The AI server real-time enhancement system according to claim 1, wherein The collaborative agent module includes: a manual collaborative agent unit and an AI collaborative agent unit; The manual collaborative agent unit is configured to receive the input information from the collaborative module, send the input information to an online customer service for enhanced processing based on experience, receive the collaborative enhancement information from the online customer service, and send the collaborative enhancement information from the online customer service to the collaborative module; The AI collaborative agent unit is configured to receive the input information from the collaborative module, process the input information using AI technology to obtain the collaborative enhancement information from AI, and send the collaborative enhancement information from AI to the collaborative module.

6. The AI server real-time enhancement system according to claim 1, wherein, The AI self-service system includes: an AI front-end module, an APP plug-in module, and an AI server module; The AI front-end module is configured to collect the input content of the user, and send the input content to the AI server module; The APP plug-in module is configured to obtain the screen content of the self-service interface, and send the screen content after packaging to the AI server module; The AI server module is configured to process the input content and the screen content, generate a response result, and send the input content, the screen content, and the response result as input information to the enhancement interface module; The AI server module is further configured to receive the collaborative enhancement information or the matching knowledge from the enhancement interface module, generate a script file according to the collaborative enhancement information or the matching knowledge, and send the script file to the AI front-end module; The AI front-end module is further configured to receive the script file, and send the script file to the units corresponding to the APP plug-in module and the AI front-end module respectively to automatically execute the self-service function.

7. A data processing method for an AI server real-time enhancement system according to any one of claims 1-6, characterized in that, The data processing method includes: The enhancement interface module receives the input information from the AI self-service system, and sends the input information to the knowledge base module; The knowledge base module extracts keywords from the input information, matches the input information according to a preset matching rule, generates matching knowledge, and sends the keywords and the input information to the collaborative module; The collaborative module generates routing information according to the keywords, combines with a preset collaborative rule, and sends the input information to the collaborative agent module according to the routing information; The collaborative seat module generates collaborative enhancement information based on the input information, and sends the collaborative enhancement information to the collaborative module. The collaborative module receives the collaborative enhancement information and sends the collaborative enhancement information to the knowledge base module; The knowledge base module processes the collaborative enhancement information or the matching knowledge and sends it to the enhancement interface module. The enhancement interface module receives the collaborative enhancement information or the matching knowledge and sends the collaborative enhancement information or the matching knowledge to the AI self-service system.

8. The data processing method of the AI server real-time enhancement system according to claim 7, characterized in that Extracting keywords from the input information, and matching the input information according to a preset matching rule to generate matching knowledge, specifically including: The keyword comparison unit receives the input information from the AI self-service system, extracts keywords from the input information, matches the keywords with the default keywords of the pre-stored knowledge block to obtain the first matching keyword, and obtains the first matching knowledge according to the first matching keyword; The semantic similarity comparison unit receives the input information from the AI self-service system, performs semantic analysis on the input information to obtain a semantic analysis result, matches keywords with a semantic similarity exceeding a preset threshold from the default keywords of the pre-stored knowledge block as the second matching keyword, and obtains the second matching knowledge according to the second matching keyword; The knowledge fusion unit fuses the first matching knowledge and the second matching knowledge to generate matching knowledge, and temporarily stores the matching knowledge.

9. The data processing method of the AI server real-time enhancement system according to claim 7, characterized in that, Generating routing information according to the keywords, and sending the input information to the collaborative seat module according to the routing information, specifically including: The routing information generation unit receives the keywords, obtains the routing information corresponding to the keywords in the preset routing definition table, and sends the routing information to the information sending unit; Among them, the routing information includes the manual collaborative seat path and the AI collaborative seat path; When the routing information is the manual collaborative seat path, the information sending unit sends the input information to the manual collaborative seat unit of the collaborative seat module. When the routing information is the AI collaborative seat path, the information sending unit sends the input information to the AI collaborative seat unit of the collaborative seat module.

10. The data processing method of the AI server real-time enhancement system according to claim 7, characterized in that, Processing the collaborative enhancement information or the matching knowledge and sending it to the enhancement interface module, specifically including: The knowledge base module determines whether the collaborative enhancement information is empty. If the collaborative enhancement information is empty, the knowledge base module sends the matching knowledge to the enhancement interface module; If the collaborative enhancement information is not empty, the knowledge base module sends the collaborative enhancement information to the enhancement interface module and saves the collaborative enhancement information for future use.

Images