An AI server-side real-time enhancement system and data processing method

Abstract

The present invention discloses an AI server-side real-time enhancement system and a data processing method. The AI ​​server-side 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. 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-side 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-side module. The AI ​​server-side module generates a final script file based on the collaborative enhancement information or the matching knowledge and replies to the AI ​​front-end module to solve the defect of insufficient script file generation capability of the original AI server-side module. The present invention enhances the data processing accuracy of the AI ​​self-service system.

Description

Technical Field

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

[0002] Public self-service devices are machines placed in public places that allow citizens to conduct various services. These devices typically integrate multiple functions, such as identity recognition, information query, and service processing, aiming to alleviate queues at automated check-in counters and improve service efficiency. With the emergence of large-scale artificial intelligence (AI), AI can now understand the semantics of conversations with people. By empowering self-service devices with AI, it is now possible for users to pose questions and for AI to complete operations.

[0003] Self-service devices powered by AI assistants can quickly retrofit a large number of existing self-service devices (original self-service modules). However, existing self-service devices have diverse applications and constantly evolving user interfaces. Current AI technology is not yet proficient enough to self-learn and understand new services and interfaces. Alternatively, when the original self-service module's user interface is updated, the AI ​​server-side module lacks the ability to respond and generate script files due to a lack of timely updates. Therefore, when the self-service device's own AI technology fails to meet the requirements for automated service processing, the AI ​​front-end module's script files cannot be restored, rendering the self-service device, equipped with the AI ​​assistant, 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 data processing method, aiming to solve the problems of insufficient script file generation capability and low data processing accuracy of the original AI server of the self-service system in the prior art.

[0006] To achieve the above objectives, the present invention provides an AI server-side real-time enhancement system, which includes:

[0007] An enhanced interface module, a knowledge base module, a collaborative module and a collaborative agent module, wherein the enhanced interface module, the knowledge base module, the collaborative module and the collaborative agent module are connected in sequence;

[0008] 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;

[0009] The knowledge base module is used to extract keywords from the input information, match the input information according to preset matching rules, generate matching knowledge, and send the keywords and the input information to the collaboration module;

[0010] The collaboration module is used to generate routing information based on the keywords and preset collaboration rules, and send the input information to the collaboration agent module according to the routing information;

[0011] The collaborative agent module is configured to generate collaborative enhancement information according to the input information and send the collaborative enhancement information to the collaborative module;

[0012] The collaboration module is further configured to receive the collaboration enhancement information and send the collaboration enhancement information to the knowledge base module;

[0013] The knowledge base module is further configured to process the collaborative enhancement information or the matching knowledge and send the processed information to the enhancement interface module;

[0014] The enhancement interface module is further configured 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.

[0015] 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;

[0016] The keyword extraction unit is used to receive input information from the enhanced interface module, extract keywords from the input information, and send the keywords and the input information to the collaboration module;

[0017] The keyword comparison unit is used to obtain the keyword, match the keyword with the default keyword of the pre-stored knowledge block to obtain a first matching keyword, and obtain first matching knowledge based on the first matching keyword;

[0018] The semantic similarity comparison unit is configured to receive input information from the AI ​​self-service system, perform semantic analysis on the input information to obtain a semantic analysis result, match a keyword whose semantic similarity with the semantic analysis result exceeds a preset threshold from the default keywords of the pre-stored knowledge block as a second matching keyword, and obtain second matching knowledge based on the second matching keyword;

[0019] The knowledge fusion unit is used to fuse the first matching knowledge and the second matching knowledge to generate matching knowledge, and temporarily store the matching knowledge;

[0020] The judgment unit is used to receive the collaborative enhancement information from the collaborative module, judge whether the collaborative enhancement information is empty, and if so, send the matching knowledge to the enhancement interface module; if not, send the collaborative enhancement information to the enhancement interface module.

[0021] Optionally, in the AI ​​server-side real-time enhancement system, the knowledge base module further comprises: a collaborative enhancement information storage unit;

[0022] The collaborative enhancement information storage unit is configured to store the collaborative enhancement information for next use when the collaborative enhancement information is not empty.

[0023] 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;

[0024] The routing information generating unit is configured to receive the keyword, obtain routing information corresponding to the keyword from a preset routing definition table, and send the routing information to the information sending unit;

[0025] 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;

[0026] The feedback unit is configured to receive collaborative enhancement information from the collaborative agent module and send the collaborative enhancement information to the knowledge base module.

[0027] Optionally, in the AI ​​server real-time enhancement system, the collaborative agent module includes: a manual collaborative agent unit and an AI collaborative agent unit;

[0028] The manual collaborative agent unit is configured to receive input information from the collaborative module, send the input information to the online customer service for enhanced processing based on experience operations, receive collaborative enhancement information from the online customer service, and send the collaborative enhancement information from the online customer service to the collaborative module;

[0029] The AI ​​collaborative agent unit is used to receive input information from the collaborative module, process the input information using AI technology, obtain collaborative enhancement information from the AI, and send the collaborative enhancement information from the AI ​​to the collaborative module.

[0030] Optionally, the AI ​​server real-time enhancement system, wherein the AI ​​self-service system includes: an AI front-end module, an APP plug-in module and an AI server module;

[0031] The AI ​​front-end module is used to collect user input content and send the input content to the AI ​​server module;

[0032] The APP plug-in module is used to obtain the screen content of the self-service interface, and package the screen content and send it to the AI ​​server module;

[0033] 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 enhanced interface module;

[0034] 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 based on the collaborative enhancement information or the matching knowledge, and send the script file to the AI ​​front-end module;

[0035] The AI ​​front-end module is also used 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 to automatically execute the self-service function.

[0036] In addition, to achieve the above-mentioned purpose, the present invention also provides a data processing method for an AI server real-time enhancement system, wherein the data processing method includes:

[0037] The enhanced interface module receives input information from the AI ​​self-service system and sends the input information to the knowledge base module;

[0038] 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;

[0039] The collaboration module generates routing information based on the keywords and preset collaboration rules, and sends the input information to the collaboration agent module according to the routing information;

[0040] The collaborative agent module generates collaborative enhancement information according to 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.

[0041] 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 it to the AI ​​self-service system.

[0042] Optionally, the data processing method of the AI ​​server-side real-time enhancement system, wherein the step of extracting keywords from the input information, matching the input information according to preset matching rules, and generating matching knowledge, specifically includes:

[0043] The keyword comparison unit receives input information from the AI ​​self-service system, extracts keywords from the input information, matches the keywords with default keywords of the pre-stored knowledge block to obtain first matching keywords, and obtains first matching knowledge based on the first matching keywords;

[0044] The semantic similarity comparison unit receives input information from the AI ​​self-service system, performs semantic analysis on the input information to obtain a semantic analysis result, matches a keyword whose semantic similarity with the semantic analysis result exceeds a preset threshold from the default keywords of the pre-stored knowledge block as a second matching keyword, and obtains second matching knowledge based on the second matching keyword;

[0045] The knowledge fusion unit fuses the first matching knowledge and the second matching knowledge to generate matching knowledge, and temporarily stores the matching knowledge.

[0046] Optionally, the data processing method of the AI ​​server-side real-time enhancement system, wherein the generating of routing information based on the keywords in combination with preset collaboration rules, and sending the input information to the collaborative agent module based on the routing information, specifically includes:

[0047] The routing information generating unit receives the keyword, obtains routing information corresponding to the keyword in a preset routing definition table, and sends the routing information to the information sending unit;

[0048] The routing information includes a manual collaborative agent path and an AI collaborative agent path;

[0049] When the routing information is a manual collaborative agent path, the information sending unit sends the input information to the manual collaborative agent unit of the collaborative agent module; when the routing information is an AI collaborative agent path, the information sending unit sends the input information to the AI ​​collaborative agent unit of the collaborative agent module.

[0050] Optionally, the data processing method of the AI ​​server-side real-time enhancement system, wherein the processing of the collaborative enhancement information or the matching knowledge and then sending it to the enhancement interface module specifically includes:

[0051] 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.

[0052] 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 next use.

[0053] 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 preset matching rules, generates matching knowledge, and sends the keywords and input information to the collaboration module; the collaboration module generates routing information based on the keywords and preset collaboration rules, and sends the input information to the collaborative agent module according to the routing information; the collaborative agent 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 matching knowledge and sends it to the enhanced interface module, the enhanced interface module receives the collaborative enhancement information or matching knowledge, and sends the collaborative enhancement information or 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

[0054] Figure 1 This is the overall architecture diagram of the AI ​​server real-time enhancement system of the present invention;

[0055] Figure 2 It is a flowchart of a preferred embodiment of the data processing method of the AI ​​server real-time enhancement system of the present invention. DETAILED DESCRIPTION

[0056] In order to make the purpose, technical solutions and advantages of the present invention more clear and distinct, the present invention is further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0057] To solve the problems in the prior art, this embodiment provides an AI server real-time enhancement system, such as Figure 1 As shown, the AI ​​server real-time enhancement system includes: an enhancement interface module, a knowledge base module, a collaboration module and a collaboration agent module.

[0058] Wherein, the enhanced interface module, the knowledge base module, the collaboration module and the collaboration agent module are connected in sequence.

[0059] 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;

[0060] The knowledge base module is used to extract keywords from the input information, match the input information according to preset matching rules, generate matching knowledge, and send the keywords and the input information to the collaboration module;

[0061] The collaboration module is used to generate routing information based on the keywords and preset collaboration rules, and send the input information to the collaboration agent module according to the routing information;

[0062] The collaborative agent module is configured to generate collaborative enhancement information according to the input information and send the collaborative enhancement information to the collaborative module;

[0063] The collaboration module is further configured to receive the collaboration enhancement information and send the collaboration enhancement information to the knowledge base module;

[0064] The knowledge base module is further configured to process the collaborative enhancement information or the matching knowledge and send the processed information to the enhancement interface module;

[0065] The enhancement interface module is further configured 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.

[0066] Furthermore, the present invention also provides a data processing unit included in each module, which is used to perform corresponding data collection and processing operations.

[0067] 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.

[0068] The keyword extraction unit is configured to receive input information from the enhanced interface module, extract keywords from the input information, and send the keywords and the input information to the collaboration module.

[0069] The keyword comparison unit is used to obtain the keyword, match the keyword with the default keyword of the pre-stored knowledge block to obtain a first matching keyword, and obtain first matching knowledge based on the first matching keyword.

[0070] 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 a keyword whose semantic similarity with the semantic analysis result exceeds a preset threshold from the default keywords of the pre-stored knowledge block as a second matching keyword, and obtain second matching knowledge based on the second matching keyword.

[0071] The knowledge fusion unit is used to fuse the first matching knowledge and the second matching knowledge to generate matching knowledge, and temporarily store the matching knowledge.

[0072] The judgment unit is used to receive the collaborative enhancement information from the collaborative module, judge whether the collaborative enhancement information is empty, and if so, send the matching knowledge to the enhancement interface module; if not, send the collaborative enhancement information to the enhancement interface module.

[0073] It is understood that the knowledge base module is used to receive information from the enhanced interface module and match the input information with pre-stored knowledge blocks based on the input information and pre-set matching rules. Pre-stored knowledge blocks that are successfully matched are considered matched knowledge, otherwise the matched knowledge blocks are empty.

[0074] In this embodiment, the pre-set matching rules can be any one of the following or a combination thereof: a database keyword search rule: extracting keywords from the input information and comparing them with the keywords in the pre-stored knowledge block; a match is considered successful if they are identical; or a semantic search rule: using the semantic comparison algorithm, the input information is semantically compared with the semantics of the pre-stored knowledge block for similarity; a match is considered successful if the similarity is greater than or equal to a preset threshold (the threshold ranges from 0 to 1, with a higher threshold indicating a higher degree of semantic match). The semantic comparison algorithm can be a cosine similarity algorithm or other algorithms.

[0075] The knowledge base module is further configured to send the keywords and input information to the collaboration module and wait for a response from the collaboration module. After receiving the collaborative enhancement information from the collaboration module, the knowledge base module processes the information by determining whether the collaborative enhancement information is empty. If so, the matching knowledge is sent to the enhancement interface module. If not, the collaborative enhancement information is sent to the enhancement interface module.

[0076] 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 next use when the collaborative enhancement information is not empty.

[0077] It can be understood that after the knowledge base module receives the collaborative enhancement information replied from the collaborative module, it determines whether the collaborative enhancement information is empty. If the collaborative enhancement information is not empty, the collaborative enhancement information storage unit is directly used to save the collaborative enhancement information. When the same input information is received next time, the collaborative enhancement information in the collaborative enhancement information storage unit is directly called, and the collaborative enhancement information is directly output to the enhancement interface module, and sent to the AI ​​self-service system through the enhancement interface module to improve data processing efficiency.

[0078] Furthermore, the collaboration module includes: a routing information generating unit, an information sending unit and a feedback unit.

[0079] The routing information generating unit is configured to receive the keyword, obtain routing information corresponding to the keyword from a preset routing definition table, and send the routing information to the information sending unit.

[0080] 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.

[0081] The feedback unit is configured to receive collaborative enhancement information from the collaborative agent module and send the collaborative enhancement information to the knowledge base module.

[0082] It can be understood that the collaboration module is used to receive information from the knowledge base module, and in combination with pre-set collaboration rules, send the information as a collaboration request to one or both of the manual collaboration seat and the AI ​​collaboration seat.

[0083] Specifically, the collaborative rules contain a set of operation sequences, called a routing definition table, in the form of keyword-routing information. The keyword can be a noun or phrase, and the routing information includes the manual collaborative agent path and the AI ​​collaborative agent path.

[0084] In this embodiment, the keywords received by the routing information generation unit are compared with the keywords in the routing definition table to obtain routing information corresponding to the keywords, and a collaboration request is sent to the manual collaboration agent or AI collaboration agent corresponding to the keywords according to the routing information corresponding to the keywords; if the received information does not contain any keywords in the routing definition table or the routing definition table does not exist, the collaboration request is not sent to the manual collaboration agent or the AI ​​collaboration agent.

[0085] 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 may be empty or not empty. In this embodiment, the collaborative enhancement information that is not empty is referred to as effective collaborative enhancement information.

[0086] 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 reply information is not blank, then effective collaboration enhancement information is generated based on the information and replied to the knowledge base module.

[0087] If the information sending unit sends a collaboration request to the manual collaboration agent and the AI ​​collaboration agent at the same time, and receives a reply from the manual collaboration agent and the manual agent reply information is not blank, then the effective collaboration enhancement information is directly generated based on the information and replied to the knowledge base module. At the same time, the reply of the AI ​​collaboration agent is ignored (regardless of whether a collaboration request is sent to the AI ​​collaboration agent, that is, the priority of the manual agent is set higher than the priority of the AI ​​collaboration agent).

[0088] Furthermore, the collaborative agent module includes: a manual collaborative agent unit and an AI collaborative agent unit.

[0089] Among them, the manual collaborative agent unit is used to receive input information from the collaborative module, send the input information to the online customer service for enhanced processing based on experience operations, receive collaborative enhancement information from the online customer service, and send the collaborative enhancement information from the online customer service to the collaborative module.

[0090] In this embodiment, the manual collaborative agent unit is used to receive input information from the collaborative module and the routing information is the manual collaborative seat path, and manually generate reply information (i.e., collaborative enhancement information) or blank reply based on experience. Within a preset time period, if there is no manual reply, the agent will automatically reply blank.

[0091] The AI ​​collaborative agent unit is used to receive input information from the collaborative module, process the input information using AI technology, obtain collaborative enhancement information from the AI, and send the collaborative enhancement information from the AI ​​to the collaborative module.

[0092] In this embodiment, the AI ​​collaborative agent unit is configured to receive input from the collaborative module, with routing information indicating an AI collaborative agent path. The AI, which can be an LLM (Large Language Model) or an agent derived from an LLM, processes this information and generates a response. This response information or a blank response is generated. If no AI response is received within a preset time period, the agent will automatically respond with a blank response.

[0093] Furthermore, the AI ​​self-service system includes: an AI front-end module, an APP plug-in module and an AI server module.

[0094] The AI ​​front-end module is used to collect user input content 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 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 enhanced interface module;

[0095] 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 based on the collaborative enhancement information or the matching knowledge, and send the script file to the AI ​​front-end module;

[0096] The AI ​​front-end module is also used 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 to automatically execute the self-service function.

[0097] Based on the AI ​​server real-time enhancement system described in the above embodiment, the present invention also provides a data processing method of the AI ​​server real-time enhancement system, specifically as follows: Figure 2 As shown in , the data processing method of the AI ​​server real-time enhancement system includes the following steps:

[0098] Step S10: The enhanced interface module receives input information from the AI ​​self-service system and sends the input information to the knowledge base module.

[0099] Specifically, it receives user input content from the AI ​​front-end module of the AI ​​self-service system and screen content from the APP plug-in module, generates a response result, and sends the user input content, screen content and response result as input information to the enhanced interface module.

[0100] The enhanced interface module sends the input information to the knowledge base module and receives a response from the knowledge base module. If the response from the knowledge base module is not blank, the enhanced interface module generates an enhanced script based on the information and sends it back to the AI ​​server module. Otherwise, the enhanced script is sent back to the AI ​​server module with a blank response.

[0101] 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.

[0102] Specifically, the knowledge base module receives input information from the AI ​​self-service system, extracts keywords from the input information, matches the keywords with default keywords of pre-stored knowledge blocks to obtain first matching keywords, and obtains first matching knowledge based on the first matching keywords.

[0103] In this embodiment, the keyword comparison unit of the knowledge base module extracts keywords from the input information through database keyword retrieval rules, and compares the keywords with the default keywords of the pre-stored knowledge block. If they are the same, the match is successful, and the first matching keyword is obtained. Knowledge search is performed based on the first matching keyword to obtain the first matching knowledge.

[0104] Furthermore, the semantic similarity comparison unit of the knowledge base module receives input information from the AI ​​self-service system, performs semantic analysis on the input information to obtain a semantic analysis result, matches a keyword whose semantic similarity with the semantic analysis result exceeds a preset threshold from the default keywords of the pre-stored knowledge block as a second matching keyword, and obtains second matching knowledge based on the second matching keyword.

[0105] In this embodiment, the knowledge base module also performs semantic analysis on the input information using semantic search rules to obtain a semantic analysis result. The input information is then compared with the semantics of pre-stored knowledge blocks for similarity. A match is considered successful if the similarity is greater than or equal to a preset threshold (the threshold ranges from 0 to 1, with a higher threshold indicating a higher degree of semantic match). The semantic comparison algorithm may be a cosine similarity algorithm or other algorithms.

[0106] Furthermore, the knowledge fusion unit of the knowledge base module fuses the first matching knowledge and the second matching knowledge to generate matching knowledge, temporarily stores the matching knowledge, and sends the keyword and the input information to the collaboration module.

[0107] It can be understood that this application will fuse the two types of matching knowledge obtained based on the database keyword retrieval rules and semantic retrieval rules, which can obtain the semantics contained in the input information more comprehensively and accurately, 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 collaborative module to obtain collaborative enhancement information. If the collaborative module does not return collaborative enhancement information, the temporarily stored matching knowledge is obtained and sent to the AI ​​self-service system as the final enhanced knowledge.

[0108] Step S30: The collaboration module generates routing information based on the keywords and preset collaboration rules, and sends the input information to the collaboration agent module according to the routing information.

[0109] Specifically, the routing information generation unit receives the keyword, obtains 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 collaborative agent path and an AI collaborative agent path.

[0110] When the routing information is a manual collaborative agent path, the information sending unit sends the input information to the manual collaborative agent unit of the collaborative agent module; when the routing information is an AI collaborative agent path, the information sending unit sends the input information to the AI ​​collaborative agent unit of the collaborative agent module.

[0111] It can be understood that the collaborative rules contain a set of operation sequences, called a routing definition table, in the form of keyword-routing information. The keyword can be a noun or phrase, and the routing information includes the manual collaborative agent path and the AI ​​collaborative agent path.

[0112] In this embodiment, the keywords received by the routing information generation unit are compared with the keywords in the routing definition table to obtain routing information corresponding to the keywords, and a collaboration request is sent to the manual collaboration agent or AI collaboration agent corresponding to the keywords according to the routing information corresponding to the keywords; if the received information does not contain any keywords in the routing definition table or the routing definition table does not exist, the collaboration request is not sent to the manual collaboration agent or the AI ​​collaboration agent.

[0113] Step S40: The collaborative agent module generates collaborative enhancement information according to 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.

[0114] Specifically, the collaborative agent module includes: a manual collaborative agent unit and an AI collaborative agent unit.

[0115] In this embodiment, the manual collaborative agent unit receives input information from the collaborative module and the routing information is the manual collaborative seat path, and manually generates reply information (i.e., collaborative enhancement information) or a blank reply based on experience. Within a preset time period, if there is no manual reply, the agent will automatically reply blank.

[0116] The AI ​​collaborative agent unit receives input information from the collaborative module, processes the input information using AI technology, obtains collaborative enhancement information from the AI, and sends the collaborative enhancement information from the AI ​​to the collaborative module.

[0117] It can be understood that the AI ​​collaborative agent unit is used to receive input information from the collaborative module and the routing information is the AI ​​collaborative seat path, and the AI ​​processes the information to generate a reply. The AI ​​can be an LLM (Large Language Model) or an LLM-derived intelligent agent.

[0118] 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.

[0119] 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 next use.

[0120] Furthermore, the enhancement interface module receives the collaborative enhancement information (the collaborative enhancement information is not empty) or the matching knowledge (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-side 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.

[0121] In summary, the present invention provides an AI server-side real-time enhancement system and a data processing method, wherein the AI ​​server-side real-time enhancement system includes: an enhancement interface module, a knowledge base module, a collaboration module and a collaboration agent module, and the enhancement interface module, the knowledge base module, the collaboration module and the collaboration agent module 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 the screen content, the input information is sent to the AI ​​server-side 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-side module. The AI ​​server-side module can generate a final script file based on the collaborative enhancement information or the matching knowledge and reply it to the AI ​​front-end module to solve the defect of insufficient script file generation capability of the original AI server-side module. The present invention can timely give the latest artificial experience to the AI ​​server-side module, thereby solving user problems, enhancing the data processing accuracy of the AI ​​self-service system, and improving the user experience.

[0122] It should be noted that, in this document, the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or terminal comprising a series of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or terminal. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or terminal comprising the element.

[0123] Of course, those skilled in the art will appreciate that all or part of the processes in the above-described method embodiments can be implemented by instructing related hardware (such as a processor, controller, etc.) through a computer program. The program can be stored in a computer-readable storage medium that can be read by a computer. When executed, the program can include the processes in the above-described method embodiments. The computer-readable storage medium can be a memory, a magnetic disk, an optical disk, etc.

[0124] It should be understood that the application of the present invention is not limited to the above examples. For those skilled in the art, improvements or changes can be made based on the above description. All these improvements and changes should fall within the scope of protection of the claims attached to the present invention.

Claims

1. An AI server-side real-time enhancement system, characterized in that: The AI ​​server real-time enhancement system includes: an enhancement interface module, a knowledge base module, a collaboration module and a collaboration agent module, wherein the enhancement interface module, the knowledge base module, the collaboration module and the collaboration agent module are connected in sequence; 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 preset matching rules, generate matching knowledge, and send the keywords and the input information to the collaboration module; The collaboration module is used to generate routing information based on the keywords and preset collaboration rules, and send the input information to the collaboration agent module according to the routing information; The collaborative agent module is configured to generate collaborative enhancement information according to the input information and send the collaborative enhancement information to the collaborative module; The collaboration module is further configured to receive the collaboration enhancement information and send the collaboration enhancement information to the knowledge base module; The knowledge base module is further configured to process the collaborative enhancement information or the matching knowledge and send the processed information to the enhancement interface module; The enhancement interface module is further configured 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; The knowledge base module includes: Keyword extraction unit, keyword comparison unit, semantic similarity comparison unit, knowledge fusion unit and first judgment unit; The keyword extraction unit is used to receive input information from the enhanced 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 keyword, match the keyword with the default keyword of the pre-stored knowledge block to obtain a first matching keyword, and obtain first matching knowledge based on the first matching keyword; The semantic similarity comparison unit is configured to receive input information from the AI ​​self-service system, perform semantic analysis on the input information to obtain a semantic analysis result, match a keyword whose semantic similarity with the semantic analysis result exceeds a preset threshold from the default keywords of the pre-stored knowledge block as a second matching keyword, and obtain second matching knowledge based on the second matching keyword; The knowledge fusion unit is used to fuse the first matching knowledge and the second matching knowledge to generate matching knowledge, and temporarily store the matching knowledge; The judgment unit is configured to receive the collaborative enhancement information from the collaborative module, judge whether the collaborative enhancement information is empty, and if so, send the matching knowledge to the enhancement interface module; if not, send the collaborative enhancement information to the enhancement interface module; The collaboration module includes: a routing information generating unit, an information sending unit and a feedback unit; The routing information generating unit is configured to receive the keyword, obtain routing information corresponding to the keyword from a preset routing definition table, and send the routing information to the information sending unit; Wherein, the routing definition table is an operation sequence in the form of keyword-routing information; 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 configured to receive collaborative enhancement information from the collaborative agent module and send the collaborative enhancement information to the knowledge base module.

2. The AI ​​server real-time enhancement system according to claim 1, characterized in that: The knowledge base module further includes: a collaborative enhancement information storage unit; The collaborative enhancement information storage unit is configured to store the collaborative enhancement information for next use when the collaborative enhancement information is not empty.

3. The AI ​​server real-time enhancement system according to claim 1, characterized in that: 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 input information from the collaborative module, send the input information to the online customer service for enhanced processing based on experience operations, receive 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 used to receive input information from the collaborative module, process the input information using AI technology, obtain collaborative enhancement information from the AI, and send the collaborative enhancement information from the AI ​​to the collaborative module.

4. The AI ​​server real-time enhancement system according to claim 1, characterized in that: 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 user input content 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 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 enhanced 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 based on 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 also used 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 to automatically execute the self-service function.

5. A data processing method based on the AI ​​server real-time enhancement system according to any one of claims 1 to 4, characterized in that: The data processing method includes: 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 based on the keywords and preset collaboration rules, and sends the input information to the collaboration agent module according to the routing information; The collaborative agent module generates collaborative enhancement information according to 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 it to the AI ​​self-service system.

6. The data processing method of the AI ​​server real-time enhancement system according to claim 5, characterized in that: The step of 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 default keywords of the pre-stored knowledge block to obtain first matching keywords, and obtains first matching knowledge based on the first matching keywords; The semantic similarity comparison unit receives input information from the AI ​​self-service system, performs semantic analysis on the input information to obtain a semantic analysis result, matches a keyword whose semantic similarity with the semantic analysis result exceeds a preset threshold from the default keywords of the pre-stored knowledge block as a second matching keyword, and obtains second matching knowledge based on 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.

7. The data processing method of the AI ​​server real-time enhancement system according to claim 5, characterized in that: The step of generating routing information based on the keyword and a preset collaboration rule, and sending the input information to the collaboration agent module according to the routing information, specifically includes: The routing information generating unit receives the keyword, obtains routing information corresponding to the keyword in a preset routing definition table, and sends the routing information to the information sending unit; The routing information includes a manual collaborative agent path and an AI collaborative agent path; When the routing information is a manual collaborative agent path, the information sending unit sends the input information to the manual collaborative agent unit of the collaborative agent module; when the routing information is an AI collaborative agent path, the information sending unit sends the input information to the AI ​​collaborative agent unit of the collaborative agent module.

8. The data processing method of the AI ​​server real-time enhancement system according to claim 5, characterized in that: The processing of the collaborative enhancement information or the matching knowledge and sending the processed information to the enhancement interface module specifically includes: 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 next use.

Images