After-sales information processing method and device, electronic equipment and storage medium

Abstract

The invention discloses an after-sales information processing method and device, electronic equipment and a storage medium, and relates to the field of computers, in particular to the technical field of artificial intelligence such as large models, deep learning, agents and intelligent after-sales. The method comprises the steps of obtaining a reference scheme set associated with a target intention of dialogue information under the condition that the dialogue information of the after-sales intention is received, calling a path planning intelligent agent based on an order identifier and the reference scheme set, obtaining candidate scheme information returned by the path planning intelligent agent, calling a scheme auditing intelligent agent based on the candidate scheme information, and obtaining the target intention of the after-sales intention according to the scheme auditing intelligent agent. The method comprises the steps of obtaining target scheme information output by a scheme auditing agent, then generating and displaying an after-sales application form corresponding to an order identifier based on the target scheme information, and finally performing after-sales processing operation on an order corresponding to the order identifier based on the after-sales application form under the condition that a submission instruction for the after-sales application form is received.

Description

After-sales information processing methods, devices, electronic equipment and storage media Technical Field

[0001] This disclosure relates to the field of computer technology, and in particular to the fields of artificial intelligence technology such as large models, deep learning, intelligent agents, and intelligent after-sales service, specifically to after-sales information processing methods, devices, electronic devices, and storage media. Background Technology

[0002] Currently, after-sales systems typically require a significant amount of manual judgment and operation for processing after-sales issues, from selecting a solution to filling out application forms, resulting in low efficiency. Therefore, providing a method for processing after-sales information that improves efficiency is a pressing issue that needs to be addressed. Summary of the Invention

[0003] This disclosure provides an after-sales information processing method, apparatus, electronic device, and storage medium. The specific solution is as follows: According to one aspect of this disclosure, an after-sales information processing method is provided, comprising: upon receiving dialogue information with an after-sales intent, obtaining a set of reference solutions associated with the target intent of the dialogue information, wherein the dialogue information includes an order identifier; based on the order identifier and the set of reference solutions, invoking a path planning agent to obtain candidate solution information returned by the path planning agent; based on the candidate solution information, invoking a solution review agent to obtain target solution information output by the solution review agent; based on the target solution information, generating and displaying an after-sales application form corresponding to the order identifier; upon receiving a submission instruction for the after-sales application form, performing after-sales processing operations on the order corresponding to the order identifier based on the after-sales application form.

[0004] According to another aspect of this disclosure, an after-sales information processing apparatus is provided, comprising: a first acquisition module, configured to acquire a set of reference solutions associated with the target intent of the dialogue information upon receiving dialogue information of an after-sales intent, wherein the dialogue information includes an order identifier; an invocation module, configured to invoke a path planning agent based on the order identifier and the set of reference solutions, and acquire candidate solution information returned by the path planning agent; a second acquisition module, configured to invoke a solution review agent based on the candidate solution information, and acquire target solution information output by the solution review agent; a generation module, configured to generate and display an after-sales application form corresponding to the order identifier based on the target solution information; and a processing module, configured to perform after-sales processing operations on the order corresponding to the order identifier based on the after-sales application form upon receiving a submission instruction for the after-sales application form.

[0005] According to another aspect of this disclosure, an electronic device is provided, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the method described in the above embodiments.

[0006] According to another aspect of this disclosure, a non-transitory computer-readable storage medium is provided storing computer instructions, wherein the computer instructions are used to cause the computer to perform the method described in the above embodiments.

[0007] According to another aspect of this disclosure, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps of the method described in the above embodiments.

[0008] The after-sales information processing method, apparatus, electronic device, and storage medium disclosed herein have the following beneficial effects: First, upon receiving dialogue information with an after-sales intent, a set of reference solutions associated with the target intent of the dialogue information is obtained. Then, based on the order identifier and the set of reference solutions, a path planning agent is invoked to obtain candidate solution information returned by the path planning agent. Based on the candidate solution information, a solution review agent is invoked to obtain the target solution information output by the solution review agent. Subsequently, based on the target solution information, an after-sales application form corresponding to the order identifier is generated and displayed. Finally, upon receiving a submission instruction for the after-sales application form, after-sales processing operations are performed on the order corresponding to the order identifier based on the after-sales application form. Therefore, users do not need to fill out an after-sales application form through cumbersome operations on the after-sales page. Instead, by entering dialogue information with an after-sales intent in any interactive interface, the system can trigger after-sales processing operations for the order, simplifying the user's after-sales application process, saving user time, and improving the accuracy and efficiency of after-sales information processing.

[0009] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this disclosure, nor is it intended to limit the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description

[0010] The accompanying drawings are provided for a better understanding of this solution and do not constitute a limitation of this disclosure. Specifically: Figure 1 is a flowchart illustrating an after-sales information processing method according to an embodiment of this disclosure; Figure 2 is a flowchart illustrating an after-sales information processing method according to another embodiment of this disclosure; Figure 3 is a flowchart illustrating the path planning agent determining candidate solution information in an after-sales information processing method according to another embodiment of this disclosure; Figure 4 is a flowchart illustrating the solution review agent outputting target solution information in an after-sales information processing method according to another embodiment of this disclosure; Figure 5 is a flowchart illustrating an after-sales information processing method according to another embodiment of this disclosure; Figure 6 is a schematic diagram of the system interaction sequence of the after-sales information processing method proposed in this disclosure; Figure 7 is a structural schematic diagram of an after-sales information processing device according to an embodiment of this disclosure; and Figure 8 is a block diagram of an electronic device used to implement the after-sales information processing method of the embodiments of this disclosure. Detailed Implementation

[0011] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0012] This disclosure relates to the fields of artificial intelligence technology, such as large models, deep learning, intelligent agents, and intelligent after-sales service.

[0013] Artificial Intelligence (AI) is a new technological science that studies, develops, and applies theories, methods, technologies, and application systems to simulate, extend, and expand human intelligence.

[0014] Large models, also known as Foundation Models, are models that extract knowledge from hundreds of millions of corpora or images, learn, and then produce large models with hundreds of millions of parameters.

[0015] Deep learning (DL) learns the inherent patterns and hierarchical representations of sample data. The information gained during this learning process greatly aids in interpreting data such as text, images, and sound. The ultimate goal of deep learning is to enable machines to possess analytical and learning capabilities similar to humans, allowing them to recognize data such as text, images, and sound.

[0016] An intelligent agent, also known as a smart agent, is an intelligent entity that autonomously perceives its environment, makes decisions based on goals, and interacts with the environment to achieve specific functions. Intelligent agents perceive changes in the environment (such as through sensors or data input), make judgments and decisions based on their learned knowledge and algorithms, and then execute actions to influence the environment or achieve predetermined goals.

[0017] Intelligent after-sales service refers to the digital and intelligent upgrading of the entire after-sales process after a product or service transaction, relying on digital technologies such as artificial intelligence (AI), big data, the Internet of Things, and robotic process automation (RPA), to build a full-chain intelligent service system that includes data collection, intelligent analysis, accurate decision-making, automated execution, and continuous optimization.

[0018] It should be noted that the acquisition, storage, use, and processing of data and / or information in this disclosed technical solution comply with the relevant provisions of national laws and regulations and do not violate public order and good morals.

[0019] The following description, with reference to the accompanying drawings, outlines an after-sales information processing method, apparatus, electronic device, and storage medium according to embodiments of the present disclosure.

[0020] Figure 1 is a flowchart illustrating an after-sales information processing method according to an embodiment of this disclosure.

[0021] As shown in Figure 1, the after-sales information processing method includes: Step 101, upon receiving dialogue information with an after-sales intent, obtaining a set of reference solutions associated with the target intent of the dialogue information, wherein the dialogue information contains an order identifier.

[0022] It should be noted that the after-sales information processing method proposed in this disclosure can be applied to the after-sales management system and customer service system of e-commerce platforms, covering after-sales scenarios including but not limited to refunds only, returns and refunds, and exchanges, and can be seamlessly adapted to online human customer service, intelligent customer service robots and user self-service after-sales access, etc. This disclosure does not limit it.

[0023] In other words, in some possible implementations, dialogue information regarding users' after-sales intentions can be received through online human customer service, intelligent customer service robots, or user-independent after-sales portals. The specific implementation can be determined according to the actual situation, and this disclosure does not limit it.

[0024] The target intent can be determined based on the user's dialogue information. For example, the target intent could be a refund only, or a return and refund or exchange, etc., and this disclosure does not limit it.

[0025] The reference solution set can be a set of baseline rules used to generate after-sales solutions. Its specific structure and content can be pre-configured according to actual needs. For example, the reference solution set can be set as a multi-level rule set, starting from the first level as the after-sales type (such as refund type), and gradually refining it down to the specific reasons for the after-sales (such as product damage). Taking three levels as an example, it can form a hierarchical relationship of after-sales type → specific after-sales measures → after-sales reasons. Each level sets the parameter constraints required for that level, the functions that must be called, and the list of evidence that must be collected (such as photos of product damage), thereby forming a solution tree with a rigorous structure and a complete chain of evidence. This disclosure does not limit this aspect.

[0026] It should be noted that different objectives require different sets of reference solutions.

[0027] The order identifier can be used to represent an order, and its specific form can be set according to actual needs. For example, the order identifier can be an order number, a tracking number, etc., and this disclosure does not limit it.

[0028] It should be noted that the data structure of the reference scheme set can be set according to actual needs. For example, the reference scheme set can be in JSON format, and this disclosure does not limit this.

[0029] For example, in some possible implementations, when the reference scheme set is in JSON format, the data list corresponding to each level node in the reference scheme set may include the name of the level node (e.g., "Refund Only"), its parent node (e.g., "Refund"), the corresponding use case description, suggestions, execution requirements information, and remarks, etc.

[0030] For example, taking the node "Refund Only" as an example, its corresponding data list can be: {"Node Identifier (key)": "REFUND_MONEY", "Parent Node (parentKey)": "APPLY_REFUND", "Name (name)": "Refund Only", "Usage Scenario Description (description)": "When a user applies for after-sales service, this solution should be used for the following scenarios: the merchant has not shipped the goods; the merchant has shipped the goods but the courier has not delivered them...", "Suggestion (suggestion)": "After-sales type, refund amount, after-sales reason, application instructions", "Execution Requirement Information (execute Require)": "Order Identifier, after-sales type, refund amount, after-sales reason, application instructions, contact information", "Remark (remark)": ""}, this disclosure does not limit this.

[0031] JSON stands for JavaScript Object Notation.

[0032] In this disclosure, after receiving dialogue information from a user with an after-sales intent, the dialogue information can be parsed to determine the target intent of the dialogue information, and then a set of reference solutions associated with the target intent can be obtained to provide a basis for generating an after-sales solution.

[0033] Step 102: Based on the order identifier and reference solution set, invoke the path planning agent to obtain the candidate solution information returned by the path planning agent.

[0034] The path planning agent can be an agent used to generate after-sales solutions. It can generate solution paths with a series of parameter constraints, evidence lists, and function call records based on the rules of the reference solution set. Its specific structure can be set as needed; for example, it can be a large model, or it can be called a "planning agent," "planning agent," etc. This disclosure does not limit it in this way.

[0035] Function calls can be used to obtain context information corresponding to the order identifier, such as order price, status, product, user history information, user rights tags, etc., thereby providing a data foundation for the path planning agent to generate more accurate and adaptable candidate solutions.

[0036] For example, in some possible implementations, when the reference solution set is a hierarchical solution set, the path planning agent can generate solution paths by following the hierarchical logic of the solution set from top to bottom, which is not limited in this disclosure.

[0037] In other words, in some possible implementations, the candidate solution information can be the candidate solution path, which may include parameter constraints, function call records, and evidence lists, etc. This disclosure does not limit this.

[0038] In some possible implementations, when adding after-sales types to the reference solution set, it is only necessary to extend the corresponding nodes, which has a small impact on the system, allows for rapid iteration, and enhances the system's scalability and maintainability.

[0039] It should be noted that the path planning agent can return at least one candidate solution information. That is, the path planning agent can return one candidate solution information or multiple candidate solution information, which can be determined according to the actual situation. This disclosure does not impose any restrictions on this.

[0040] Therefore, after obtaining the set of reference solutions associated with the target intent of the dialogue information, the path planning agent is invoked based on the order identifier and the set of reference solutions contained in the dialogue information. The candidate solution information returned by the path planning agent is obtained, which realizes the after-sales information processing triggered by the user's after-sales intent dialogue information and generates after-sales solutions. This effectively simplifies the user's after-sales application process and improves the efficiency of after-sales processing.

[0041] Step 103: Based on the candidate solution information, invoke the solution review agent to obtain the target solution information output by the solution review agent.

[0042] The solution review agent can act as an independent compliance auditor, objectively verifying the candidate solution information output by the path planning agent. Its review does not rely on the planning process logic, but rather on a reference solution set and function call results, focusing on verifying the legality of the solution path, the completeness of necessary function calls, and the compliance of output fields. Its specific structure can be configured as needed; for example, it can be a large model, or it can be called a "review agent," "auditing agent," etc., which is not limited in this disclosure.

[0043] The compliance of the solution path refers to whether the path in the candidate solution is isomorphic to the parent-child relationship of the nodes defined in the reference solution set. For example, if the solution path is Request Refund → Return and Refund → Refund Reason: Product Damage, is it completely isomorphic to the parent-child relationship of the nodes defined in the corresponding reference solution set? This disclosure does not impose any restrictions on this.

[0044] The completeness of necessary function calls refers to whether all functions that must be called in the candidate solution generation process have been called, and whether the call results satisfy the requirement to proceed to the next level. This disclosure does not impose any limitations on this.

[0045] The compliance of output fields refers to whether the output fields in the candidate solution information, such as the suggested text and the required fields for execution, are strictly derived from the defined content of each node on the hit path through splicing, calculation or conversion, without tampering, missing or redundant information. This disclosure does not impose any restrictions on this.

[0046] The target solution information can be the approved after-sales solution information.

[0047] In other words, after obtaining the candidate solution information returned by the path planning agent, the solution review agent is invoked based on the candidate solution information. This agent verifies the path legality of the candidate solution information, whether all functions that must be called when generating the solution have been called, and the compliance of the output fields, etc. If the verification is successful, the agent will output the candidate solution as the target solution information, thereby realizing the automated and standardized review of after-sales solutions, which greatly reduces the review cost and time.

[0048] In some possible implementations, the solution information output by both the path planning agent and the solution review agent can be in a standardized JSON format, facilitating front-end parsing, rendering, and form pre-filling. Each solution information item is a complete data object containing all information from the solution identifier to the pre-filled form information. This disclosure does not impose any limitations on this.

[0049] Step 104: Based on the target solution information, generate and display the after-sales application form corresponding to the order identifier.

[0050] It should be noted that the format requirements and the content to be filled in for the after-sales application form can be preset as needed, and this disclosure does not impose any restrictions on them.

[0051] In this disclosure, after obtaining the target solution information output by the solution review agent, an after-sales application form corresponding to the order identifier is generated and displayed based on the target solution information so that the user can confirm it. Thus, the user does not need to fill out and generate the after-sales application form through cumbersome operations, but only needs to confirm or supplement a small amount of information to submit the form, which shortens the processing time of each after-sales information and improves the efficiency of after-sales information processing.

[0052] In some possible implementations, when generating the after-sales application form, the target solution information can be automatically converted into pre-filled data according to the after-sales form format requirements and filled in to automatically generate the after-sales application form. This allows users to submit the form with one click after confirmation, simplifying the user's operation and improving the user experience.

[0053] In some possible implementations, the generated after-sales application form can be displayed to the user through the customer service system interface. The specific implementation can be determined according to the actual situation, and this disclosure does not limit it.

[0054] Step 105: Upon receiving a submission instruction for the after-sales application form, perform after-sales processing operations on the order corresponding to the order identifier based on the after-sales application form.

[0055] In this disclosure, after displaying the after-sales application form, upon receiving a submission instruction for the after-sales application form, it can be determined that the user has confirmed the currently generated after-sales application form is correct. At this point, after-sales processing operations can be performed on the order corresponding to the order identifier based on the after-sales application form, thereby improving after-sales efficiency and enhancing user experience.

[0056] In some possible implementations, after the after-sales processing of an order, the after-sales solution, the reasoning process of the path planning agent, all function call records, and the audit report (audit process and audit results, etc.) of the solution review agent can be fully recorded to form a traceable audit chain, ensuring the traceability and reliability of the after-sales solution.

[0057] For example, in some possible implementations, the general application flow of the after-sales information processing method proposed in this disclosure can be as follows: the user submits an after-sales request through the customer service system or self-service page → the system calls the path planning agent to generate personalized candidate solution information → the solution review agent performs independent compliance verification on the solution → the front end displays the approved target solution information and automatically fills the solution information into the corresponding fields of the after-sales application form → the user only needs to check the information and add a small amount of content (such as a more detailed description) to submit with one click. Thus, the user does not need to trigger the system's after-sales processing operation for the order through complex option selection and form filling on a specific page. Instead, the user can trigger the system's after-sales processing operation for the order by entering after-sales intent dialogue in any interactive interface (such as the customer service system interactive interface), realizing a "what you say is what you get" after-sales experience, improving the efficiency of after-sales information processing, and thus enhancing the user experience.

[0058] In some possible implementations, the after-sales processing solution proposed in this disclosure can also be used in different scenarios. For example, in simple scenarios with high determinism and high frequency, after-sales processing can be carried out quickly through a rule engine or a simple model. In complex, ambiguous, or high-value scenarios, a dual-agent system of the after-sales information processing method proposed in this disclosure can be used for in-depth solution planning and review, thereby achieving the best balance between after-sales efficiency and effectiveness. This disclosure does not limit this.

[0059] In this embodiment, upon receiving dialogue information with an after-sales intent, a set of reference solutions associated with the target intent of the dialogue information is first obtained. Then, based on the order identifier and the set of reference solutions, a path planning agent is invoked to obtain candidate solution information returned by the path planning agent. Based on the candidate solution information, a solution review agent is invoked to obtain the target solution information output by the solution review agent. Next, based on the target solution information, an after-sales application form corresponding to the order identifier is generated and displayed. Finally, upon receiving a submission instruction for the after-sales application form, after-sales processing operations are performed on the order corresponding to the order identifier based on the after-sales application form. Therefore, users do not need to fill out an after-sales application form through cumbersome operations on the after-sales page. Instead, by entering dialogue information with an after-sales intent in any interactive interface, the system can trigger after-sales processing operations for the order, simplifying the after-sales application process, saving users time, and improving the accuracy and efficiency of after-sales information processing.

[0060] Figure 2 is a flowchart illustrating an after-sales information processing method according to another embodiment of this disclosure.

[0061] As shown in Figure 2, the after-sales information processing method includes: step 201, when receiving dialogue information with after-sales intent, obtaining a set of reference solutions associated with the target intent of the dialogue information, wherein the dialogue information contains an order identifier.

[0062] In some possible implementations, after receiving user-input dialogue information, intent recognition can be performed first to determine the target intent and its type. Then, if the target intent is an after-sales intent, the reference solution set associated with the target intent is determined based on the association between each reference solution set and the intent. This achieves precise and efficient processing of after-sales information by recognizing the user's dialogue information received through the interactive interface, triggering after-sales service when the user's intent is after-sales, and obtaining the reference solution set associated with the intent. This improves the efficiency and flexibility of after-sales information processing.

[0063] It should be noted that the specific implementation of intent recognition for dialogue information can be set according to actual needs. For example, intent recognition can be performed using rule matching, that is, matching dialogue information with preset keywords, regular expressions, and semantic templates to identify intent. Alternatively, convolutional neural networks can be used to capture key information in dialogue information and match the corresponding intent, etc. This disclosure does not limit this approach.

[0064] It should be noted that the association between each set of reference solutions and the intent can be pre-set; different intents will have different sets of reference solutions associated with them.

[0065] The specific implementation of step 201 can be found in the detailed descriptions of other embodiments in this disclosure, and will not be repeated here.

[0066] Step 202: Call the information retrieval function based on the order identifier to obtain the context information associated with the order identifier.

[0067] The specific structure of the information retrieval function can be set as needed, and this disclosure does not limit it.

[0068] The contextual information may include order information associated with the order identifier, such as order amount, product and status, user rights tags, user history and real-time policies, etc. This disclosure does not limit this information.

[0069] Therefore, by obtaining the context information associated with the order identifier, after-sales solutions can be dynamically generated based on the context information, providing a data foundation for generating more accurate and more adaptable after-sales solutions.

[0070] Step 203: Based on the order identifier, context information, and reference solution set, invoke the path planning agent to obtain the candidate solution information returned by the path planning agent.

[0071] In this disclosure, after obtaining the context information associated with the order identifier, the path planning agent is invoked to perform path planning by combining the context information and generate candidate solution information. This enables the generation of personalized candidate solution information based on the user's order status, historical information, and other information, thereby improving the adaptability and accuracy of the solution.

[0072] Step 204: Based on the candidate solution information, invoke the solution review agent to obtain the target solution information output by the solution review agent.

[0073] Step 205: Based on the target solution information, generate and display the after-sales application form corresponding to the order identifier.

[0074] Step 206: Upon receiving a submission instruction for the after-sales application form, perform after-sales processing operations on the order corresponding to the order identifier based on the after-sales application form.

[0075] The specific implementation of steps 204 to 206 can be found in the detailed descriptions of other embodiments in this disclosure, and will not be repeated here.

[0076] In this embodiment, upon receiving dialogue information with an after-sales intent, a set of reference solutions associated with the target intent of the dialogue information is first obtained. Then, based on the order identifier, an information retrieval function is called to obtain context information associated with the order identifier. Based on the order identifier, context information, and the set of reference solutions, a path planning agent is called to obtain candidate solution information returned by the path planning agent. Subsequently, based on the candidate solution information, a solution review agent is called to obtain the target solution information output by the solution review agent. Based on the target solution information, an after-sales application form corresponding to the order identifier is generated and displayed. Finally, upon receiving a submission instruction for the after-sales application form, after-sales processing operations are performed on the order corresponding to the order identifier based on the after-sales application form. Thus, by inputting dialogue information with an after-sales intent in any interactive interface, the system triggers the invocation of dual agents and combines the context information of the user's order to plan and review after-sales solutions, thereby generating an after-sales application form. This eliminates the need for users to go through complex option selection and form filling to apply for after-sales service. While simplifying the user's after-sales application process, it ensures the accuracy and adaptability of the generated after-sales solution, improves the efficiency and accuracy of after-sales information processing, and ultimately enhances the user experience.

[0077] In this disclosure, after obtaining the reference solution set, order identifier, and context information associated with the target intent of the dialogue information, the path planning agent is invoked to determine the candidate solution information, thereby achieving automated solution generation. The specific implementation of the path planning agent determining the candidate solution information is shown in Figure 3, which is a flowchart illustrating the process of the path planning agent determining candidate solution information in another embodiment of the after-sales information processing method provided by this disclosure. The process of the path planning agent determining the candidate solution information includes: Step 301, performing a pre-verification of the context information based on the verification function associated with the parent node in the reference solution set, and determining the pre-verification result.

[0078] The verification function can be used to verify whether the context information is sufficient to support the current after-sales application, but this disclosure does not limit it.

[0079] It should be noted that the specific structure of the verification function can be set according to actual needs, and this disclosure does not limit it.

[0080] The result of the pre-verification can be either "passed" or "failed," and the specific result can be determined according to the actual situation. This disclosure does not limit this.

[0081] In this disclosure, before generating a solution, the path planning agent can first perform a preliminary verification of the context information based on the verification function associated with the parent node in the reference solution set to confirm its feasibility. This allows for subsequent processing based on the preliminary verification results, effectively avoiding invalid processing and resource waste.

[0082] It should be noted that the structure and content of the prompts for the path planning agent can be set according to actual needs, and this disclosure does not make a current version of them.

[0083] For example, the core instruction of the path planning agent's prompts can be: based on the reference solution set and the functions associated with it, call the function, strictly match the complete path according to the hierarchy, and generate JSON output that can be approved in one pass. It must adhere to the review rules of path validity, field compliance, and standard format, and accurately generate fields such as node identifier (key), name, scenario description, suggestion, execution requirement information, and remark. The remark must include the user scenario, the function called, and the complete matching path. This disclosure does not impose any limitations on this.

[0084] Step 302: If the pre-verification result is that the verification is passed, the context information is matched with the first description information associated with the sub-nodes of the schemes in the scheme set.

[0085] The first descriptive information can be information describing the specific handling measures or solutions of the reference plan, and can be set according to actual needs. For example, taking refund after-sales service as an example, the first descriptive information can be a handling plan such as refund only, return and refund, or exchange, etc., and this disclosure does not limit it.

[0086] In this disclosure, if the pre-verification result is successful, it can be determined that the current context information supports the current after-sales application. At this time, the context information can be matched with the first description information associated with the solution sub-node in the solution set to determine the specific after-sales processing measures.

[0087] Step 303: If the first description information associated with any solution sub-node matches the context information, match the context information with the second description information associated with the cause sub-node under any solution sub-node to determine the target cause sub-node corresponding to any solution sub-node.

[0088] The "Reason" sub-node can be the reason why a user applies for after-sales service. The second description information can be the reason for the after-sales application corresponding to the "Reason" sub-node. For example, it could be that the received goods are partially damaged and require compensation from the merchant, or that the merchant failed to ship the goods, etc. This disclosure does not limit this.

[0089] In this disclosure, when the context information matches a certain sub-node of a scheme in the reference scheme set, in order to generate a complete scheme path, it is also necessary to match the context information with the second description information of the cause sub-node under that scheme sub-node to determine the target cause sub-node corresponding to that scheme sub-node.

[0090] For example, when matching context information with the second description information of the cause sub-node under the solution sub-node to determine the target cause sub-node corresponding to the solution sub-node, the specific problem described by the user in the context information (such as "damaged", "wrong item sent" etc.) can be matched with the second description information of the cause sub-node under the solution sub-node to determine the corresponding target cause sub-node.

[0091] In some possible implementations, when the first description information associated with any solution sub-node matches the context information, the path planning agent can dynamically inject the context information, such as the order status (e.g., received, not shipped), into the processing solution of the corresponding solution sub-node, and solidify the specific parameters required by the solution (e.g., the amount required for a refund, the address required for an exchange, etc.).

[0092] In some possible implementations, after identifying the target cause sub-node, a list of necessary evidence (such as "photos of product damage") can be associated with that cause sub-node and recorded in the solution requirements. This links the after-sales cause with evidence, improving the credibility of the generated solution.

[0093] Step 304: Generate candidate solution information based on any solution sub-node and the target cause sub-node.

[0094] In this disclosure, after determining the solution sub-nodes and target reason sub-nodes for contextual information matching, candidate solution information is generated based on any solution sub-node and the target reason sub-node. By performing hierarchical matching with a reference solution set, candidate solutions are dynamically generated hierarchically, starting from after-sales type (e.g., refund) and gradually refining down to specific reasons (e.g., product damage). This results in highly personalized, adaptable, and accurate candidate solutions, freeing users from complex option selection and improving user experience.

[0095] Step 305: If the pre-verification result is that the verification failed, the information retrieval function is called based on the prompt information in the verification result to obtain supplementary descriptive information.

[0096] The specific content and format of the prompt message can be set according to actual needs. For example, the prompt message can indicate the reason for the failure of the verification, such as missing information in the context and what specific information is missing, etc. This disclosure does not limit this.

[0097] In this disclosure, after performing a preliminary verification of the context information, if the preliminary verification result is a failure, it can be determined that the current context information is insufficient to support the current after-sales application. At this time, based on the prompt information in the verification result, the information retrieval function can be called to obtain supplementary description information, so that after-sales processing can continue, thereby improving the reliability of after-sales processing.

[0098] Step 306: Based on the supplementary description information and context information, return to perform the operation of matching the description information until candidate solution information is generated, or determine that the candidate solution information is empty.

[0099] In this disclosure, after obtaining supplementary description information, the matching operation with the description information can be re-executed based on the supplementary description information and context information. This involves matching the description with the first description information of the solution sub-nodes in the reference solution set, and the second description information of the cause sub-nodes under the matched solution sub-nodes, until candidate solution information is generated. Alternatively, if no relevant solution sub-nodes or cause sub-nodes can be matched when matching description information based on supplementary description information and context information, the candidate solution information can be determined to be empty. Thus, by matching after-sales processing measures with causes based on supplementary description information and context information, candidate solution information is generated when a match is successful, ensuring that the generated candidate solution information meets user needs and improving the accuracy and adaptability of the candidate solution information. If a match fails, it can be determined that the current context information and supplementary description information are insufficient to support after-sales processing. In this case, the candidate solution information can be determined to be empty, thereby improving the reliability and robustness of determining candidate solution information.

[0100] Steps 302 to 304, and steps 305 to 306, are specific implementations of the path planning agent determining candidate solution information under different circumstances. Specifically, steps 302 to 304 are the implementations of the path planning agent determining candidate solution information when the context information passes the pre-verification, and steps 305 to 306 are the implementations of the path planning agent determining candidate solution information when the context information fails the pre-verification.

[0101] In other words, after obtaining the order identifier of the user's current after-sales application, the reference solution set associated with the after-sales intent, and the context information associated with the order identifier, when calling the path planning agent to determine the candidate solution information, step 301 is first executed to perform a pre-verification of the context information. If the context information verification passes, steps 302 to 304 are executed to determine the candidate solution information. In this case, the specific implementation of the path planning agent determining the candidate solution in the after-sales information processing method proposed in this disclosure can be steps 301 to 304.

[0102] If the context information fails the verification, the candidate solution information is determined by executing steps 305 to 306. In this case, the specific implementation of the path planning agent determining the candidate solution in the after-sales information processing method proposed in this disclosure can be steps 301, 305 to 306.

[0103] It should be noted that the specific implementation form of the path planning agent determining the candidate solution is steps 301 to 304, or steps 301, 305 to 306. The specific implementation can be determined according to the actual situation, and this disclosure does not limit it.

[0104] In this embodiment, the context information is first pre-verified based on the verification function associated with the parent node in the reference solution set to determine the pre-verification result. If the pre-verification result is successful, the context information is matched with the first description information associated with the solution child nodes in the solution set. If the first description information associated with any solution child node matches the context information, the context information is matched with the second description information associated with the cause child node under any solution child node to determine the target cause child node corresponding to any solution child node. Then, candidate solution information is generated based on any solution child node and the target cause child node. Alternatively, if the pre-verification result is unsuccessful, an information retrieval function is called based on the prompt information in the verification result to obtain supplementary description information. Based on the supplementary description information and the context information, the operation of matching the description information is returned until candidate solution information is generated, or until the candidate solution information is determined to be empty. Therefore, after obtaining the order identifier, reference solution set, and context information corresponding to the user's after-sales application, the path planning agent performs pre-verification on the context information and performs subsequent processing based on the verification result, improving the reliability and robustness of the path planning agent in determining candidate solution information. If the context information passes validation, the reference solution set is matched level by level based on the context information to determine the corresponding solution sub-nodes and cause sub-nodes. Based on this, candidate solution information is determined, providing a basis for generating personalized after-sales solutions that meet user needs, have high adaptability, and high accuracy, and achieving automated solution generation. If the context information fails validation, supplementary information is obtained by calling a function. Based on the supplementary information and the context information, the process of matching solution sub-nodes and cause sub-nodes in the reference solution set is returned until candidate solution information is generated or it is determined that the candidate solution information is empty, thereby improving the reliability of the path planning agent in determining candidate solutions.

[0105] In this disclosure, after obtaining the candidate solution information returned by the path planning agent, the solution review agent is invoked based on the candidate solution information to obtain the target solution information output by the solution review agent, thereby achieving automated solution review. The specific review implementation of the solution review agent, specifically the process of outputting the target solution information, can be illustrated in Figure 4. Figure 4 is a flowchart illustrating the output of target solution information by the solution review agent in an after-sales information processing method provided in another embodiment of this disclosure. The process of the solution review agent outputting target solution information includes: Step 401, reviewing the candidate solution information based on at least one of the following to determine the target solution information: whether the nodes contained in the candidate solution information belong to the same reference solution set; whether the hierarchy between nodes in the solution path indicated by the candidate solution information satisfies the hierarchical relationship; whether the function calls in the candidate solution information are legal; and whether the content of the candidate solution information is compliant.

[0106] The hierarchical relationship refers to the parent-child relationship among the nodes in the reference scheme set, which can be set according to actual needs. For example, the hierarchical relationship in the reference scheme set can be "After-sales type (parent node) → Specific after-sales measures (scheme child node) → Reason for applying for after-sales service (reason child node)", which is not limited in this disclosure.

[0107] Therefore, by reviewing whether the path described in the solution (such as "refund → return and refund → damaged goods") belongs to the same reference solution set, and whether the hierarchy between nodes in the path satisfies the hierarchical relationship, it is determined whether the solution path is completely isomorphic to the parent-child relationship of the path nodes defined in the reference solution set, thus achieving compliance review of the solution path.

[0108] The validity of a function call can include whether the called function is a function associated with each node on the solution path and the validity of the function call result.

[0109] Whether the content of the candidate solution information is compliant refers to whether the output fields in the solution information, such as the suggested text and the required fields for implementation, are strictly spliced, calculated or converted from the defined content of each node on the solution path, without tampering, missing or redundant information, etc. This disclosure does not impose any restrictions on this.

[0110] Therefore, by conducting multi-dimensional review and verification of the path nodes, function calls, and output content of candidate solution information, the review process becomes quantifiable and verifiable, effectively avoiding inconsistencies caused by subjective judgment and improving the standardization and reliability of the review process.

[0111] It should be noted that the structure and content of the prompts for the solution review agent can be set according to actual needs, and this disclosure does not impose any restrictions on them.

[0112] For example, the core instruction of the solution review agent can be: based on the reference solution set, the execution process record of the path planning agent, and the information of the candidate solutions to be reviewed, a rigorous verification is performed. Review criteria: 1) Whether each node in the candidate solution information belongs to the same reference solution set; 2) Whether the hierarchy between nodes in the solution path indicated by the candidate solution information satisfies the hierarchical relationship; 3) Whether the function calls in the candidate solution information are legal; 4) Whether the content of the candidate solution information is compliant. The final output is a JSON-formatted review conclusion: {"Pass": boolean value, "Reason": ".."}. This disclosure does not impose any limitations on this.

[0113] Step 402: If any candidate solution information satisfies the following conditions, determine any candidate solution information as a target solution information: the nodes contained belong to the same reference solution set; the indicated solution path has a parent-child relationship from the starting node to the ending node; the function called is a function associated with the node in the solution path; the result returned by the called function meets the threshold for entering the next level node; and the contained content is compliant.

[0114] The threshold can be used as a condition for judging the result of a function call when determining whether the function call is legal. It can be set according to actual needs, and this disclosure does not limit it.

[0115] In this disclosure, after reviewing the candidate solution information, the solution review agent can determine that if any candidate solution information meets the above conditions, the solution path of the candidate solution information is isomorphic to the path in the reference solution set, the nodes of the path satisfy the parent-child relationship, the function call is legal, and the solution content is compliant. At this time, the candidate solution information can be determined as the target solution information, thereby ensuring the quality of the output target solution information, effectively intercepting erroneous solutions, and significantly reducing the review cost and time consumption.

[0116] In this embodiment, the solution review agent reviews candidate solution information based on at least one of the following to determine target solution information: whether the nodes contained in the candidate solution information belong to the same reference solution set; whether the hierarchy between nodes in the solution path indicated by the candidate solution information satisfies a hierarchical relationship; whether the function calls in the candidate solution information are legal; and whether the content of the candidate solution information is compliant. Any candidate solution information is determined to be a target solution information if it satisfies the following conditions: the contained nodes belong to the same reference solution set; the indicated solution path has a parent-child relationship from the starting node to the ending node; the called function is associated with a node in the solution path; the returned result of the called function meets the threshold for entering the next level node; and the contained content is compliant. Therefore, by conducting multi-dimensional reviews of the solution path, function calls, and output content of the candidate solution information, the review process is ensured to be quantifiable and verifiable. This achieves automated solution review, reduces review costs, and improves review efficiency and quality.

[0117] Figure 5 is a flowchart illustrating an after-sales information processing method provided in another embodiment of this disclosure.

[0118] As shown in Figure 5, the after-sales information processing method includes: step 501, when receiving dialogue information with after-sales intent, obtaining a set of reference solutions associated with the target intent of the dialogue information, wherein the dialogue information contains an order identifier.

[0119] Step 502: Based on the order identifier and reference solution set, invoke the path planning agent to obtain the candidate solution information returned by the path planning agent.

[0120] Step 503: Invoke the scheme review agent based on the candidate scheme information and obtain the target scheme information output by the scheme review agent.

[0121] The specific implementation of steps 501 to 503 can be found in the detailed descriptions of other embodiments in this disclosure, and will not be repeated here.

[0122] Step 504: When there are multiple target solution information, determine the user rights value corresponding to each target solution information.

[0123] Among them, the user rights value can be a numerical indicator that quantifies the degree to which users obtain the rights and interests from the target plan, such as refund speed, warranty period repair, return shipping costs, etc. It can be used to evaluate the level of protection of users' rights and interests under different target plans.

[0124] It should be noted that the specific implementation form of the user rights value corresponding to each target plan information can be set according to actual needs. For example, the score of each indicator can be determined first based on the pre-set rights assessment benchmark and the weight of each indicator, and then the scores of all indicators can be weighted and summed to determine the final user rights value, etc. This disclosure does not limit this.

[0125] Therefore, when there are multiple target solution information, determining the user rights value corresponding to each target solution information can provide a reference for the display order of the after-sales application form when it is subsequently displayed to the user.

[0126] Step 505: Based on the user's rights value in descending order, generate and display the after-sales application form corresponding to each target solution information.

[0127] In this disclosure, after determining the user rights value corresponding to each target solution information, an after-sales application form corresponding to each target solution information is generated and displayed based on the user rights value from high to low. This allows users to quickly notice after-sales solutions with higher rights values, thereby ensuring that users obtain the best rights and improving the user's after-sales experience.

[0128] Step 506: Upon receiving a submission instruction for the after-sales application form, perform after-sales processing operations on the order corresponding to the order identifier based on the after-sales application form.

[0129] The specific implementation of step 506 can be found in the detailed descriptions of other embodiments in this disclosure, and will not be repeated here.

[0130] In this embodiment, upon receiving dialogue information indicating an after-sales intent, a set of reference solutions associated with the target intent of the dialogue information is first obtained. Then, based on the order identifier and the set of reference solutions, a path planning agent is invoked to obtain candidate solution information returned by the path planning agent. Based on the candidate solution information, a solution review agent is invoked to obtain the target solution information output by the solution review agent. If there are multiple target solutions, the user benefit value corresponding to each target solution is determined. Based on the user benefit values ​​in descending order, an after-sales application form corresponding to each target solution is generated and displayed. Finally, upon receiving a submission instruction for the after-sales application form, after-sales processing operations are performed on the order corresponding to the order identifier based on the after-sales application form. Therefore, when multiple after-sales solutions are generated and reviewed through the path planning agent and the solution review agent, by determining the user benefit value for each after-sales solution and generating and displaying the corresponding point-to-point after-sales application form based on the benefit values ​​in descending order, the process of applying for after-sales service is simplified, allowing users to quickly notice the after-sales solution with the optimal benefit value, thus protecting user rights and improving user experience.

[0131] The following example, with reference to Figure 6, illustrates the system interaction flow of the after-sales information processing method proposed in this disclosure. Figure 6 is a schematic diagram of the system interaction sequence of the after-sales information processing method proposed in this disclosure. The interaction flow shown in Figure 6 is merely an example and is not intended to be limiting.

[0132] Figure 6 illustrates the interaction between the user, customer service system, after-sales coordination agent, path planning agent, solution review agent, and business database. The customer service system, after-sales coordination agent, and business database can all be pre-set according to actual needs, and there are no restrictions here: Step 601, the user submits an after-sales request to the customer service system.

[0133] In this disclosure, users can submit after-sales requests by sending conversation messages to the customer service system.

[0134] Step 602: The customer service system sends the order identifier and user intent to the after-sales coordination agent.

[0135] In this disclosure, after receiving a user's after-sales request, the customer service system can perform intent recognition on the user's after-sales request, i.e., the dialogue information, to determine the user's intent, and then send the order identifier and the user's intent to the after-sales coordination intelligent agent.

[0136] Step 603: The after-sales coordination agent obtains a set of reference solutions associated with the user's intent.

[0137] Step 604: The after-sales coordination agent requests an after-sales solution from the path planning agent.

[0138] Step 605: The path planning agent calls the information retrieval function based on the order identifier to obtain context information.

[0139] Step 606: The information retrieval function retrieves the context information associated with the order identifier from the business database.

[0140] Step 607: The information acquisition function returns context information to the path planning agent.

[0141] Step 608: The after-sales coordination agent sends a set of reference solutions to the path planning agent.

[0142] Step 609: The path planning agent determines candidate solution information based on contextual information and a set of reference solutions.

[0143] Step 610: The path planning agent returns candidate solution information to the after-sales coordination agent.

[0144] Step 611: The after-sales coordination agent sends candidate solution information to the solution review agent for review.

[0145] Step 612: The solution review agent reviews the candidate solution information and determines the target solution.

[0146] Step 613: The solution review agent returns the target solution to the after-sales coordination agent.

[0147] Step 614: The after-sales coordination AI returns the target solution to the customer service system.

[0148] Step 615: The customer service system generates and displays the after-sales application form corresponding to the order identifier to the user based on the target solution information.

[0149] Step 616: The user submits a form to the customer service system.

[0150] Step 617: Based on the submitted form, the customer service system performs after-sales processing on the orders corresponding to the order identifiers.

[0151] The specific implementation of steps 601 to 617 can be referred to the relevant descriptions in the above embodiments of this disclosure, and will not be repeated here.

[0152] Therefore, through the after-sales information processing method proposed in this disclosure, users do not need to go through cumbersome operations and form filling to apply for after-sales service. Instead, they can trigger after-sales service by having a dialogue in the interactive interface, which greatly simplifies user operations, improves the efficiency of after-sales information processing, and enhances the user experience.

[0153] For example, taking the scenario of return and refund - damaged goods as an example, the application of the after-sales information processing method proposed in this disclosure is illustrated: Input: User dialogue information: "I received a broken glass (I bought 10 in total), I want to return it and get a refund. Order number: ×××××××××".

[0154] Processing: The system dynamically identifies the intent, the path planning agent calls the verification function to confirm that the order supports after-sales service, and matches the path: apply for refund → return and refund → refund reason - product damage, calculate the refund amount (total price / 10), and associate the evidence "product damage photo".

[0155] Output and Pre-filling: Generate and display the approved after-sales solution. The front-end automatically pre-fills the form based on this: the "After-sales Type" field automatically selects "Return and Refund".

[0156] The "Refund Amount" field will be automatically calculated and filled with "one-tenth of the total price of the product".

[0157] The "Reason for After-Sales Service" field will automatically select "Product Damage / Packaging Issue".

[0158] The "Application Description" field automatically generates the text "The glass I received was broken...".

[0159] This allows after-sales information processing to be triggered through user-generated after-sales conversations, simplifying the after-sales application process and improving the user experience.

[0160] To implement the above embodiments, this disclosure also proposes an after-sales information processing device.

[0161] Figure 7 is a schematic diagram of the structure of an after-sales information processing device provided in an embodiment of this disclosure.

[0162] As shown in Figure 7, the after-sales information processing device 700 includes: a first acquisition module 701, a calling module 702, a second acquisition module 703, a generation module 704, and a processing module 705.

[0163] The first acquisition module 701 is used to acquire a set of reference solutions associated with the target intent of the dialogue information when receiving dialogue information with an after-sales intent, wherein the dialogue information includes an order identifier; the invocation module 702 is used to invoke a path planning agent based on the order identifier and the set of reference solutions to acquire candidate solution information returned by the path planning agent; the second acquisition module 703 is used to invoke a solution review agent based on the candidate solution information to acquire the target solution information output by the solution review agent; the generation module 704 is used to generate and display an after-sales application form corresponding to the order identifier based on the target solution information; the processing module 705 is used to perform after-sales processing operations on the order corresponding to the order identifier based on the after-sales application form when receiving a submission instruction for the after-sales application form.

[0164] Optionally, the first acquisition module 701 is specifically used to: perform intent recognition on the received dialogue information, determine the target intent of the dialogue information and the type of the target intent; and, if the type of the target intent is after-sales intent, determine the reference scheme set associated with the target intent based on the association relationship between each reference scheme set and the intent.

[0165] Optionally, the aforementioned calling module 702 is specifically used to: call the information acquisition function based on the order identifier to obtain the context information associated with the order identifier; and call the path planning agent based on the order identifier, the context information, and the reference scheme set.

[0166] Optionally, the aforementioned calling module 702 is further configured to: perform pre-verification on the context information based on the verification function associated with the parent node in the reference scheme set, and determine the pre-verification result; if the pre-verification result is a pass verification, match the context information with the first description information associated with the scheme child node in the scheme set; if the first description information associated with any scheme child node matches the context information, match the context information with the second description information associated with the cause child node under any scheme child node, so as to determine the target cause child node corresponding to any scheme child node; and generate candidate scheme information based on any scheme child node and the target cause child node.

[0167] Optionally, the above-mentioned calling module 702 is further configured to: if the pre-verification result is that the verification fails, call the information acquisition function based on the prompt information in the verification result to obtain supplementary description information; and based on the supplementary description information and context information, return to perform the operation of matching the description information until candidate solution information is generated, or determine that the candidate solution information is empty.

[0168] Optionally, the second acquisition module 703 is specifically used to: review the candidate solution information based on at least one of the following to determine the target solution information: whether each node in the candidate solution information belongs to the same reference solution set, whether the hierarchy between nodes in the solution path indicated by the candidate solution information satisfies the hierarchical relationship, whether the function call in the candidate solution information is legal, and whether the content of the candidate solution information is compliant.

[0169] Optionally, the second acquisition module 703 described above is further configured to: determine any candidate solution information as a target solution information if any candidate solution information satisfies the following conditions: the nodes included belong to the same reference solution set; the indicated solution path has a parent-child relationship from the starting node to the ending node; the function called is a function associated with a node in the solution path; the result returned by the called function satisfies the threshold for entering the next level node; and the included content is compliant.

[0170] Optionally, the above-mentioned generation module 704 is specifically used to: determine the user rights value corresponding to each target solution information when there are multiple target solution information; and generate and display the after-sales application form corresponding to each target solution information based on the user rights value in descending order.

[0171] It should be noted that the explanation of the above-mentioned after-sales information processing method embodiment also applies to the after-sales information processing device of this embodiment, so it will not be repeated here.

[0172] In this embodiment, upon receiving dialogue information with an after-sales intent, a set of reference solutions associated with the target intent of the dialogue information is first obtained. Then, based on the order identifier and the set of reference solutions, a path planning agent is invoked to obtain candidate solution information returned by the path planning agent. Based on the candidate solution information, a solution review agent is invoked to obtain the target solution information output by the solution review agent. Next, based on the target solution information, an after-sales application form corresponding to the order identifier is generated and displayed. Finally, upon receiving a submission instruction for the after-sales application form, after-sales processing operations are performed on the order corresponding to the order identifier based on the after-sales application form. Therefore, users do not need to fill out an after-sales application form through cumbersome operations on the after-sales page. Instead, by entering dialogue information with an after-sales intent in any interactive interface, the system can trigger after-sales processing operations for the order, simplifying the after-sales application process, saving users time, and improving the accuracy and efficiency of after-sales information processing.

[0173] According to embodiments of this disclosure, this disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0174] Figure 8 illustrates a schematic block diagram of an example electronic device 800 that can be used to implement embodiments of the present disclosure. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0175] As shown in Figure 8, device 800 includes a computing unit 801, which can perform various appropriate actions and processes based on a computer program stored in ROM (Read-Only Memory) 802 or a computer program loaded from storage unit 808 into RAM (Random Access Memory) 803. RAM 803 can also store various programs and data required for the operation of device 800. The computing unit 801, ROM 802, and RAM 803 are interconnected via bus 804. I / O (Input / Output) interface 805 is also connected to bus 804.

[0176] Multiple components in device 800 are connected to I / O interface 805, including: input unit 806, such as keyboard, mouse, etc.; output unit 807, such as various types of monitors, speakers, etc.; storage unit 808, such as disk, optical disk, etc.; and communication unit 809, such as network card, modem, wireless transceiver, etc. Communication unit 809 allows device 800 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0177] The computing unit 801 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 801 include, but are not limited to, CPUs (Central Processing Units), GPUs (Graphics Processing Units), various special-purpose AI (Artificial Intelligence) computing chips, various computing units running machine learning model algorithms, DSPs (Digital Signal Processors), and any suitable processor, controller, microcontroller, etc. The computing unit 801 performs the various methods and processes described above, such as after-sales information processing methods. For example, in some embodiments, the after-sales information processing method may be implemented as a computer software program tangibly contained in a machine-readable medium, such as storage unit 808. In some embodiments, part or all of the computer program may be loaded and / or installed on device 800 via ROM 802 and / or communication unit 809. When the computer program is loaded into RAM 803 and executed by the computing unit 801, one or more steps of the after-sales information processing method described above may be performed. Alternatively, in other embodiments, the computing unit 801 may be configured to perform an after-sales information processing method by any other suitable means (e.g., by means of firmware).

[0178] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, FPGAs (Field Programmable Gate Arrays), ASICs (Application-Specific Integrated Circuits), ASSPs (Application-Specific Standard Products), SOCs (System-on-Chips), CPLDs (Complex Programmable Logic Devices), computer hardware, firmware, software, and / or combinations thereof. These various implementations may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0179] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0180] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, RAM, ROM, EPROM (Electrically Programmable Read-Only Memory) or flash memory, optical fiber, CD-ROM (Compact Disc Read-Only Memory), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0181] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (Cathode-Ray Tube) or LCD (Liquid Crystal Display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0182] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include LANs (Local Area Networks), WANs (Wide Area Networks), the Internet, and blockchain networks.

[0183] Computer systems can include clients and servers. Clients and servers are generally geographically separated and typically interact via communication networks. The client-server relationship is established by computer programs running on the respective computers and having a client-server relationship with each other. A server can be a cloud server, also known as a cloud computing server or cloud host, a hosting product within the cloud computing service ecosystem, addressing the shortcomings of traditional physical hosts and VPS (Virtual Private Server) services, such as high management difficulty and weak business scalability. Servers can also be servers for distributed systems or servers incorporating blockchain technology.

[0184] According to embodiments of this disclosure, this disclosure also provides a computer program product that, when executed by an instruction processor, performs the after-sales information processing method proposed in the above embodiments of this disclosure.

[0185] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this disclosure can be achieved, and this is not limited herein.

[0186] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A method for processing after-sales information, characterized in that, include: Upon receiving a dialogue message with an after-sales intent, the system retrieves a set of reference solutions associated with the target intent of the dialogue message, wherein the dialogue message contains an order identifier. Based on the order identifier and the set of reference solutions, the system invokes a path planning agent to retrieve candidate solution information returned by the path planning agent. Based on the candidate solution information, the system invokes a solution review agent to retrieve the target solution information output by the solution review agent. Based on the target solution information, the system generates and displays an after-sales application form corresponding to the order identifier. Upon receiving a submission instruction for the after-sales application form, the system performs after-sales processing operations on the order corresponding to the order identifier based on the after-sales application form.

2. The method as described in claim 1, characterized in that, The step of obtaining a set of reference solutions associated with the target intent upon receiving dialogue information with an after-sales intent includes: performing intent recognition on the received dialogue information to determine the target intent of the dialogue information and the type of the target intent; and, if the type of the target intent is an after-sales intent, determining a set of reference solutions associated with the target intent based on the association between each set of reference solutions and the intent.

3. The method as described in claim 1, characterized in that, The step of invoking the path planning agent based on the order identifier and the reference scheme set includes: invoking an information acquisition function based on the order identifier to obtain context information associated with the order identifier; and invoking the path planning agent based on the order identifier, the context information, and the reference scheme set.

4. The method as described in claim 3, characterized in that, The process by which the path planning agent determines candidate solution information includes: performing a preliminary verification on the context information based on the verification function associated with the parent node in the reference solution set, and determining the preliminary verification result; if the preliminary verification result is a pass, matching the context information with the first description information associated with the solution child node in the solution set; if the first description information associated with any solution child node matches the context information, matching the context information with the second description information associated with the cause child node under any solution child node to determine the target cause child node corresponding to any solution child node; and generating the candidate solution information based on the any solution child node and the target cause child node.

5. The method as described in claim 4, characterized in that, After determining the pre-verification result, the method further includes: if the pre-verification result is a failure, calling an information retrieval function based on the prompt information in the verification result to obtain supplementary description information; and based on the supplementary description information and the context information, returning to perform an operation that matches the description information until the candidate solution information is generated, or determining that the candidate solution information is empty.

6. The method according to any one of claims 1-5, characterized in that, The process of the solution review agent outputting target solution information includes: reviewing the candidate solution information based on at least one of the following to determine the target solution information: whether each node in the candidate solution information belongs to the same reference solution set, whether the hierarchy between nodes in the solution path indicated by the candidate solution information satisfies the hierarchical relationship, whether the function call in the candidate solution information is legal, and whether the content of the candidate solution information is compliant.

7. The method as described in claim 6, characterized in that, The method further includes: determining any candidate solution information as a target solution information if any candidate solution information satisfies the following conditions: the nodes included belong to the same reference solution set; the indicated solution path has a parent-child relationship from the starting node to the ending node; the function called is a function associated with the node in the solution path; the result returned by the called function satisfies the threshold for entering the next level node; and the included content is compliant.

8. The method as described in claim 6, characterized in that, The step of generating and displaying the after-sales application form corresponding to the order identifier based on the target solution information includes: when there are multiple target solution information, determining the user rights value corresponding to each target solution information; and generating and displaying the after-sales application form corresponding to each target solution information based on the user rights value in descending order.

9. An after-sales information processing device, characterized in that, include: The first acquisition module is used to acquire a set of reference solutions associated with the target intent of the dialogue information when receiving dialogue information with an after-sales intent, wherein the dialogue information includes an order identifier; the invocation module is used to invoke a path planning agent based on the order identifier and the set of reference solutions, and acquire candidate solution information returned by the path planning agent; the second acquisition module is used to invoke a solution review agent based on the candidate solution information, and acquire target solution information output by the solution review agent; the generation module is used to generate and display an after-sales application form corresponding to the order identifier based on the target solution information; the processing module is used to perform after-sales processing operations on the order corresponding to the order identifier based on the after-sales application form when receiving a submission instruction for the after-sales application form.

10. An electronic device, characterized in that, include: At least one processor; And a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-8.

11. A non-transitory computer-readable storage medium storing computer instructions, characterized in that, The computer instructions are used to cause the computer to perform the method according to any one of claims 1-8.

12. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1-8.

Images