Method and device for processing complaint, storage medium and electronic equipment

Abstract

The embodiment of the invention discloses a complaint processing method and device, a storage medium and electronic equipment, and the method comprises the steps: obtaining a complaint type corresponding to a complaint request of a target user about a target merchant, and obtaining refund prediction information corresponding to the complaint request based on a refund rule model corresponding to the target merchant; inputting the complaint type and the refund prediction information into a negotiation model, obtaining at least one piece of negotiation verbal skill information output by the negotiation model, and carrying out negotiation interaction with the target user according to the at least one piece of negotiation verbal skill information; and obtaining a negotiation result of the negotiation interaction, and if the negotiation result indicates that negotiation is successful, executing a corresponding complaint processing operation for the complaint request according to the negotiation result, so that the complaint processing efficiency can be improved, the labor cost of manual complaint processing can be reduced, and meanwhile, the consistency of refund decisions can be ensured.

Description

Technical Field

[0001] This invention relates to computer technology, and more particularly to a method, apparatus, storage medium, and electronic device for processing complaints. Background Technology

[0002] With the continuous development of electronic payment technology, there are two major pain points in the handling of merchant complaints on current payment platforms: the first is the inefficiency of manual decision-making: merchants need to manually analyze the rationality of each complaint, which takes an average of more than 30 minutes to process; the second is the lack of uniform standards: refund decisions rely on personal experience, and similar complaints may have different handling results. Summary of the Invention

[0003] The purpose of the embodiments in this specification is to provide a method, apparatus, storage medium, and electronic device for processing complaints.

[0004] This specification provides a method for handling complaints, which involves obtaining the complaint type corresponding to a target user's complaint request regarding a target merchant, obtaining refund prediction information corresponding to the complaint request based on a refund rule model corresponding to the target merchant, inputting the complaint type and the refund prediction information into a negotiation model to obtain at least one negotiation script output by the negotiation model, conducting negotiation interaction with the target user based on the at least one negotiation script, obtaining the negotiation result of this negotiation interaction, and if the negotiation result indicates that the negotiation was successful, performing corresponding complaint handling operations on the complaint request based on the negotiation result. This method can improve complaint handling efficiency, reduce the manual cost of handling complaints, and ensure the consistency of refund decisions. The method includes:

[0005] Obtain the complaint type corresponding to the target user's complaint request regarding the target merchant, and based on the refund rule model corresponding to the target merchant, obtain the refund prediction information corresponding to the complaint request;

[0006] Input the complaint type and the refund prediction information into the negotiation model to obtain at least one negotiation script information output by the negotiation model, and conduct negotiation interaction with the target user based on the at least one negotiation script information;

[0007] Obtain the negotiation result of this negotiation interaction. If the negotiation result indicates that the negotiation was successful, perform the corresponding complaint handling operation on the complaint request according to the negotiation result.

[0008] Furthermore, the method also includes:

[0009] The refund rule model corresponding to the target merchant is generated or updated based on the complaint feature information corresponding to the target merchant, wherein the merchant feature information is constructed based on the historical complaint data corresponding to the target merchant.

[0010] Furthermore, the interaction type corresponding to the negotiation interaction includes any one of the following:

[0011] Call interaction;

[0012] Text message interaction;

[0013] Voice message interaction.

[0014] Further, the step of inputting the complaint type and the refund prediction information into the negotiation model to obtain at least one negotiation script output by the negotiation model includes:

[0015] Input the complaint type, the refund prediction information, and the user characteristic information corresponding to the target user into the negotiation model to obtain at least one negotiation script information output by the negotiation model.

[0016] Further, the step of inputting the complaint type, the refund prediction information, and the user characteristic information corresponding to the target user into the negotiation model to obtain at least one negotiation script output by the negotiation model includes:

[0017] Input the complaint type, the refund prediction information, the user characteristic information corresponding to the target user, and the object characteristic information of the associated object corresponding to the complaint request into the negotiation model to obtain at least one negotiation script information output by the negotiation model.

[0018] Furthermore, the step of negotiating with the target user based on the at least one negotiation script includes:

[0019] Based on the current negotiation script information in the at least one negotiation script information, negotiate and interact with the target user to obtain the target user's response information regarding the current negotiation script information;

[0020] Based on the response information, the latest current negotiation script is determined, and the negotiation interaction with the target user continues based on the latest current negotiation script, and so on, until the interaction ends.

[0021] Further, determining the latest current negotiation script information based on the response information includes:

[0022] Based on the response information, determine the target user's current emotional information during the negotiation interaction process;

[0023] Based on the user's current emotional information, determine the latest current negotiation script.

[0024] Furthermore, the refund prediction information includes a refund amount standard. If the current negotiation script information includes a first refund amount, the first refund amount meets the refund amount standard.

[0025] Furthermore, if the current negotiation script does not include a refund amount, the current negotiation script is used to appease the target user;

[0026] The step of determining the latest current negotiation script information based on the user's current emotional information includes:

[0027] If the user's current emotional information meets a preset first condition, the latest current negotiation script information is determined, wherein the latest current negotiation script information includes the first refund amount.

[0028] Furthermore, if the user's current emotional information meets the preset second condition, the latest current negotiation script information includes a second refund amount, the second refund amount meets the refund amount standard, and the second refund amount is greater than the current refund amount corresponding to the historical negotiation script information.

[0029] Furthermore, determining the latest current negotiation script information based on the user's current emotional information includes:

[0030] If the user's current emotional information meets the preset third condition, the second refund amount is determined based on the current refund amount corresponding to the historical negotiation script information;

[0031] Based on the second refund amount, determine the latest current negotiation script information.

[0032] Further, determining the second refund amount based on the current refund amount corresponding to the historical negotiation script information includes:

[0033] The second refund amount is determined based on the current refund amount corresponding to the historical negotiation script information and the user's current emotional information.

[0034] Furthermore, determining the latest current negotiation script information based on the user's current emotional information includes:

[0035] The user's current emotional information, the complaint type, and the refund prediction information are re-inputted into the negotiation model to obtain the latest current negotiation script information.

[0036] Furthermore, obtaining the negotiation result of this negotiation interaction includes:

[0037] The target user's current emotional information is obtained during the negotiation interaction. If the user's current emotional information meets a preset fourth condition, the negotiation result of this negotiation interaction is generated, wherein the negotiation result indicates that the negotiation has been completed.

[0038] Furthermore, the method also includes:

[0039] If the refund prediction information cannot be obtained based on the refund rule model, the target merchant will be prompted to proceed to the manual processing stage regarding the complaint request.

[0040] Furthermore, the method also includes:

[0041] If the negotiation result indicates that the negotiation is unsuccessful, the target merchant will be prompted to proceed to the manual processing stage regarding the complaint request.

[0042] This specification also provides an embodiment of an apparatus for processing complaints, comprising:

[0043] The first acquisition module is used to obtain the complaint type corresponding to the complaint request of the target user about the target merchant, and obtain the refund prediction information corresponding to the complaint request based on the refund rule model corresponding to the target merchant;

[0044] The second acquisition module is used to input the complaint type and the refund prediction information into the negotiation model, obtain at least one negotiation script information output by the negotiation model, and conduct negotiation interaction with the target user based on the at least one negotiation script information;

[0045] The third obtaining module is used to obtain the negotiation result of this negotiation interaction. If the negotiation result indicates that the negotiation was successful, the module performs the corresponding complaint handling operation on the complaint request according to the negotiation result.

[0046] This specification also provides a storage medium storing a computer program adapted to be loaded by a processor and to execute the steps of the method described above.

[0047] This specification also provides an electronic device, including a processor and a memory; wherein the memory stores a computer program adapted to be loaded by the processor and to execute the steps of the method described above.

[0048] This specification also provides a computer program product that stores at least one instruction, characterized in that the at least one instruction, when executed by a processor, implements the steps of the above-described method.

[0049] According to the embodiments of this specification, by obtaining the complaint type corresponding to the target user's complaint request regarding the target merchant, and based on the refund rule model corresponding to the target merchant, refund prediction information corresponding to the complaint request is obtained; the complaint type and the refund prediction information are input into a negotiation model to obtain at least one negotiation script information output by the negotiation model; negotiation interaction is conducted with the target user based on the at least one negotiation script information; the negotiation result of this negotiation interaction is obtained; if the negotiation result indicates that the negotiation has been successful, the corresponding complaint processing operation is performed on the complaint request based on the negotiation result. This can improve the efficiency of complaint processing, reduce the manual cost of handling complaints, and ensure the consistency of refund decisions. Attached Figure Description

[0050] Figure 1 A flowchart illustrating a method for handling complaints provided in an embodiment of this specification;

[0051] Figure 2 A schematic diagram of a device for handling complaints provided in an embodiment of this specification;

[0052] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this specification. Detailed Implementation

[0053] To make the objectives, technical solutions, and advantages of this specification clearer, the technical solutions of this specification will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this specification, and not all of them. Based on the embodiments in this specification, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this specification.

[0054] Please see Figure 1 This is a flowchart illustrating a method for handling complaints provided in an embodiment of this specification. In this embodiment, the method for handling complaints is applied to a device for handling complaints (hereinafter referred to as a "complaint handling device") or an electronic device equipped with a complaint handling device as described in this embodiment. The following will focus on... Figure 1 The process shown is described in detail, and the method for handling complaints may specifically include the following steps:

[0055] S102, obtain the complaint type corresponding to the target user's complaint request about the target merchant, and obtain the refund prediction information corresponding to the complaint request based on the refund rule model corresponding to the target merchant.

[0056] In some embodiments, a target user initiates a complaint request to the payment platform regarding a transaction between themselves and a target merchant. This transaction can be a product transaction (e.g., the target user purchases a product from the target merchant and files a complaint about that product), a service transaction (e.g., the target user purchases a service from the target merchant and files a complaint about that service), or any other form of transaction. This example embodiment does not impose any special limitations on this. In some embodiments, it is necessary to first obtain the complaint type corresponding to the complaint request. This complaint type can be selected by the target user from multiple complaint types, or it can be determined by the payment platform after semantic understanding of the specific complaint content. It should be noted that the above methods of obtaining the complaint type are merely examples and not limitations. Those skilled in the art should understand that any specific method used to obtain the complaint type can be included within the scope of protection of this specification, and this example embodiment does not impose any special limitations on this. In some embodiments, complaint types include, but are not limited to, non-delivery, returns due to quality issues (such as damage or discrepancy with description), discrepancy with description (no return), logistics issues (no goods received, logistics problems, or false delivery), non-performance of service (such as failure to provide services as agreed), minor disputes (such as minor defects), price disputes (such as incorrect pricing or hidden fees), and others (such as merchant fraud). This example embodiment does not impose any special limitations on these types of complaints.

[0057] In some embodiments, the refund rule model is an automated decision-making model trained based on the target merchant's historical complaint data (e.g., complaint data from the past year). By inputting the complaint type corresponding to the complaint request into the refund rule model corresponding to the target merchant, the model can output the refund prediction information corresponding to the complaint request. This refund prediction information includes, but is not limited to, the refund amount (a refund amount of 0 indicates no refund), the refund amount range (e.g., 30%-70%, 50%-100%, etc.), refund attachment conditions (including but not limited to direct order cancellation, return of goods, partial compensation through negotiation or platform ruling, potential liability of the logistics party, judgment based on the degree of non-performance, user acceptance of negotiation, etc.), and the refund period (e.g., refund within 3 days). This example embodiment does not impose any special limitations on these. In some embodiments, the complaint request will first undergo a preliminary review to check whether the complaint request is missing any necessary information. If it is missing, the complaint type corresponding to the complaint request will be obtained only after the target user completes the information, and this complaint type will be input into the refund rule model.

[0058] S104, input the complaint type and the refund prediction information into the negotiation model, obtain at least one negotiation script information output by the negotiation model, and conduct negotiation interaction with the target user based on the at least one negotiation script information.

[0059] In some embodiments, the negotiation model is a trained model used to generate negotiation scripts for use during negotiation interactions with a target user. For example, the negotiation model can be a sequence-to-sequence (Seq2Seq) model based on BERT (Bidirectional Encoder Representations from Transformers). This example embodiment does not specifically limit the specific model structure of the negotiation model. In some embodiments, by inputting the complaint type and refund prediction information into the negotiation model, at least one negotiation script information output by the negotiation model can be obtained. Based on this at least one negotiation script information, automatic negotiation interaction without human intervention can be conducted with the target user. The negotiation script information refers to statements used to automatically negotiate with the target user regarding the complaint request. The type of negotiation interaction includes, but is not limited to, intelligent outbound call interaction, text message interaction, and voice message interaction. This example embodiment does not specifically limit this. In some embodiments, the negotiation script output by the negotiation model may include a specific refund amount. By sending this negotiation script to the target user, the system can determine whether the target user agrees to the refund amount or their level of dissatisfaction with it based on the user's response. Alternatively, the negotiation script may not contain any specific refund amount and may only be used to appease the target user. In some embodiments, the negotiation script may be text-based or audio-based.In some embodiments, the negotiation model may output only one negotiation script at a time, and then send this negotiation script (or, the text-based negotiation script needs to be converted into corresponding audio) to the target user during an automated negotiation interaction without human intervention. Alternatively, the negotiation model may output multiple negotiation scripts at a time, which can be directly sent to the target object. Or, it may send only one negotiation script at a time to the target object, in which case a negotiation script needs to be selected from the multiple negotiation scripts and sent to the target user. After receiving the target user's response regarding the negotiation script, a new negotiation script is selected from the multiple negotiation scripts and sent to the target user, and so on. For example, a negotiation script can be randomly selected from multiple negotiation scripts, or it can be based on... The user profile information of the target user and / or the merchant profile information of the target merchant and / or the attribute information of the goods or services corresponding to the complaint request are used to automatically determine a suitable negotiation script from multiple negotiation scripts. Alternatively, a suitable negotiation script can be automatically determined from multiple negotiation scripts based on the response information. For example, by performing semantic analysis and / or keyword analysis and / or user sentiment analysis on the response information, the most suitable negotiation script for the response information can be automatically determined from multiple negotiation scripts. Alternatively, if the multiple negotiation scripts are arranged in order, the negotiation script that appears first in the multiple negotiation scripts is sent to the target user first. Then, after receiving the user's response information regarding the negotiation script, the negotiation script that appears after the previously sent negotiation script is sent to the target user. In some embodiments, the automated negotiation interaction with the target user without human intervention involves an intelligent negotiation system, such as an AI negotiation interaction system (e.g., an AI telephone system) that interacts with the user through Automatic Speech Recognition (ASR) and Natural Language Processing (NLP).

[0060] S106, obtain the negotiation result of this negotiation interaction. If the negotiation result indicates that the negotiation was successful, perform the corresponding complaint handling operation for the complaint request according to the negotiation result.

[0061] In some embodiments, the negotiation result of this negotiation interaction can be obtained based on whether the target user has performed a confirmation operation during the automatic negotiation interaction. If the user has performed a confirmation operation, a negotiation result indicating successful negotiation is generated. If the user has not performed a confirmation operation within a preset time range, a negotiation result indicating unsuccessful negotiation is generated. For example, the negotiation result of this negotiation interaction can be obtained based on whether the target user has clicked the negotiation completion confirmation button during the automatic negotiation interaction. If the user has clicked the negotiation completion confirmation button, a negotiation result indicating successful negotiation is generated. If the user has not clicked the negotiation completion confirmation button within a preset time range, a negotiation result indicating unsuccessful negotiation is generated. In some embodiments, the success of the negotiation can be automatically determined based on the response information of the target user during the automatic negotiation interaction. If the negotiation is successful, a corresponding negotiation result is generated. For example, by performing semantic analysis and / or keyword analysis and / or user sentiment analysis on the response information, it can be determined whether the user agrees with the current refund amount of the negotiation interaction (e.g., the refund amount in the negotiation script information containing the refund amount most recently sent to the target user during the negotiation interaction). If the user agrees, the negotiation interaction can be determined to be successful. If the success of the negotiation interaction cannot be determined within a preset time range, a negotiation result indicating that the negotiation was unsuccessful is generated. In some embodiments, if the negotiation is successful, the final refund amount of this negotiation interaction is sent back to the automatic refund system in real time. The automatic refund system will automatically execute the corresponding refund operation based on the final refund amount to complete the automatic refund for the complaint request. Otherwise, if the negotiation is unsuccessful, the target merchant can be prompted to enter the manual processing stage for the complaint request. The final refund amount of this negotiation interaction can be determined based on at least one negotiation script information, such as the negotiation script information containing a certain refund amount sent to the target user at the end of this negotiation interaction. Alternatively, the final refund amount can also be determined based on the target user's reply information, such as the reply information from the target user at the end of this negotiation interaction indicating agreement to a certain compensation amount. This example embodiment does not make any special limitations on this. If no negotiation script information containing a refund amount is sent to the target user in this negotiation interaction, the final refund amount of this negotiation interaction can be determined to be 0, that is, no refund will be issued.

[0062] According to the embodiments of this specification, by obtaining the complaint type corresponding to the target user's complaint request regarding the target merchant, and based on the refund rule model corresponding to the target merchant, refund prediction information corresponding to the complaint request is obtained; the complaint type and the refund prediction information are input into a negotiation model to obtain at least one negotiation script information output by the negotiation model; negotiation interaction is conducted with the target user based on the at least one negotiation script information; the negotiation result of this negotiation interaction is obtained; if the negotiation result indicates that the negotiation has been successful, the corresponding complaint processing operation is performed on the complaint request based on the negotiation result. This can improve the efficiency of complaint processing, reduce the manual cost of handling complaints, and ensure the consistency of refund decisions.

[0063] In some embodiments, the method further includes: generating or updating a refund rule model corresponding to the target merchant based on the complaint feature information corresponding to the target merchant, wherein the merchant feature information is constructed based on the historical complaint data corresponding to the target merchant. In some embodiments, the complaint feature information corresponding to the target merchant can be constructed based on the historical complaint data (e.g., complaint data from the past year) corresponding to the target merchant, wherein the complaint feature information includes, but is not limited to, the main complaint types (logistics / quality / service, etc.), historical refund ratio distribution, user value rating, etc. This example embodiment does not make any special limitations on this. The historical complaint data includes, but is not limited to, basic transaction information, complaint reasons (i.e., complaint types), evidence support, platform rule references, negotiation and history records, etc. This example embodiment does not make any special limitations on this. The basic transaction information includes, but is not limited to, the transaction amount - the amount actually paid by the user (the refund limit is usually no more than this amount), the type of goods / services - physical goods, virtual services, offline consumption, etc. (affecting return conditions), and the order status - whether it has been shipped, confirmed receipt, or completed. Evidence may include, but is not limited to, user-provided evidence, chat logs (if the merchant has not fulfilled their promises), product photos / videos (proof of quality issues), logistics information (screenshots of non-delivery or delays), payment vouchers and order details, etc. Platform rules refer to, but are not limited to, the payment platform's refund policy (e.g., unshipped orders usually support full refunds; received goods may require a return and refund (deducting shipping costs); virtual goods or services may not support returns, but partial refunds can be negotiated), merchant promises (e.g., merchants indicate "three times the compensation for counterfeit goods" or "7-day no-reason return policy"), etc. Negotiation and historical records include the results of communication between the two parties (e.g., whether a partial refund agreement has been reached), and historical complaint statistics of the merchant (e.g., if the merchant has been complained about multiple times, the payment platform may favor the user's appeal), etc. This example embodiment does not impose any special limitations on this. In some embodiments, the refund rule model corresponding to the target merchant is trained based on the complaint feature information corresponding to the target merchant, and the refund rule model is generated or updated. The refund rule model is a personalized automated decision-making model trained based on the merchant's historical refund data. In some embodiments, when the number of incremental complaint data corresponding to the target merchant reaches a preset threshold (for example, when the target merchant processes 1,000 complaints), the refund rule model corresponding to the target merchant can be automatically updated based on the incremental complaint data. That is, the refund rule model evolves in real time and is more accurate.

[0064] In some embodiments, the interaction type corresponding to the negotiation interaction includes any of the following: call interaction; text message interaction; and voice message interaction. In some embodiments, the interaction type of the negotiation interaction can be intelligent outbound call interaction, that is, by establishing an audio / video call connection (audio call or video call) with the target user, the negotiation script information is provided to the target user through call audio. If the negotiation script information is in text format, it needs to be converted into call audio format first. In the call connection, the user's reply information is also in call audio format. In some embodiments, the interaction type of the negotiation interaction can also be text message interaction, that is, the negotiation script information is sent to the target user through text messages in the conversation window with the target user. In this case, the user's reply information can be any message format supported by the conversation window (e.g., text message, image message, voice message, video message, etc.). In some embodiments, the interaction type of the negotiation interaction can also be voice message format, that is, the negotiation script information is sent to the target user through voice messages in the conversation window with the target user. In this case, the user's reply information can also be any message format supported by the conversation window.

[0065] In some embodiments, inputting the complaint type and the refund prediction information into the negotiation model to obtain at least one negotiation script information output by the negotiation model includes: inputting the complaint type, the refund prediction information, and the user characteristic information corresponding to the target user into the negotiation model to obtain at least one negotiation script information output by the negotiation model. In some embodiments, by inputting the complaint type, refund prediction information, and the user characteristic information of the target user into the negotiation model, at least one negotiation script information output by the negotiation model can be obtained. This at least one negotiation script information matches the user characteristic information, making it more acceptable or acceptable to the target user, making it easier for the target user to agree to the current refund amount in this negotiation interaction (e.g., the refund amount in the negotiation script information containing the refund amount most recently sent to the target user during this negotiation interaction), making it easier for this negotiation interaction to succeed. The user characteristic information includes, but is not limited to, any characteristic information related to the target user, such as user name, user age, user attributes, identification information of the identity document used to uniquely identify the target user (e.g., ID number), and any other type of user profile information. This example embodiment does not specifically limit this.

[0066] In some embodiments, the step of inputting the complaint type, the refund prediction information, and the user characteristic information corresponding to the target user into the negotiation model to obtain at least one negotiation script information output by the negotiation model includes: inputting the complaint type, the refund prediction information, the user characteristic information corresponding to the target user, and the object characteristic information of the associated object corresponding to the complaint request into the negotiation model to obtain at least one negotiation script information output by the negotiation model. In some embodiments, by inputting the complaint type, refund prediction information, target user's user characteristic information, and object characteristic information of the associated object corresponding to the complaint request into the negotiation model, at least one negotiation script information output by the negotiation model can be obtained. This at least one negotiation script information not only matches the user characteristic information but also matches the object characteristic information, making it more suitable for the associated object corresponding to the complaint request and more easily accepted or recognized by the target user, thus making the negotiation interaction more likely to succeed. The associated object corresponding to the complaint request includes, but is not limited to, product objects (e.g., the target user purchases a product sold by the target merchant and files a complaint against the product) and service objects (e.g., the target user purchases a service provided by the target merchant and files a complaint against the service). The object characteristic information includes, but is not limited to, product name, product type, product price, any attribute information related to the product object, service name, service type, service price, and any attribute information related to the service object. This example embodiment does not impose any special limitations on this.

[0067] In some embodiments, the step of negotiating with the target user based on the at least one negotiation script information includes: negotiating with the target user based on the current negotiation script information in the at least one negotiation script information to obtain the target user's response information regarding the current negotiation script information; determining the latest current negotiation script information based on the response information; continuing to negotiate with the target user based on the latest current negotiation script information, and so on, until the interaction ends. In some embodiments, after the negotiation model outputs at least one negotiation script information, it is necessary to first determine a current negotiation script information from the at least one negotiation script information, and then send the current negotiation script information to the target user in an automated negotiation interaction without human intervention. For example, a negotiation script information can be randomly determined from the at least one negotiation script information as the current negotiation script information. Or, for example, if the at least one negotiation script information is arranged in order, the negotiation script information that appears first in the at least one negotiation script information can be used as the current negotiation script information. In some embodiments, the system obtains the target user's response to the current negotiation script, then determines the latest current negotiation script based on the response, and continues to send the latest current negotiation script to the target user in an automated negotiation interaction without human intervention. The system then obtains the target user's response to the latest current negotiation script, and so on, until the negotiation interaction ends (the negotiation result of the current negotiation interaction has been obtained, i.e., it has been determined whether the negotiation was successful). The latest current negotiation script can be determined from the at least one negotiation script based on the response. For example, by performing semantic analysis and / or keyword analysis on the response, the system automatically determines the negotiation script that best matches the response as the latest current negotiation script. Alternatively, the latest current negotiation script output by the negotiation model can be obtained by re-inputting the complaint type, refund prediction information, and the response into the negotiation model.

[0068] In some embodiments, determining the latest current negotiation script information based on the response information includes: determining the target user's current emotional information during the negotiation interaction based on the response information; and determining the latest current negotiation script information based on the user's current emotional information. In some embodiments, user sentiment analysis can be performed on the target user's response information regarding the negotiation script information sent to them to obtain the target user's current emotional information during this negotiation interaction. For example, large models such as GPT (Generative Pre-trained Transformer) and BERT can be used to identify the user's current emotional information based on the response information using natural language processing (NLP) techniques. This example embodiment does not specifically limit the specific method for identifying the user's current emotional information. In some embodiments, if the response information is in the form of call audio or voice message, it is necessary to first convert the response information into text form using automatic speech recognition (ASR) technology, and then identify the user's current emotional information based on the text-based response information. In some embodiments, the user's current emotional information can be used to indicate whether the target user is satisfied with the current negotiation script (which may include a refund amount, or may not include any refund amount and is only used to appease the target user). Alternatively, the user's current emotional information can also be used to characterize the target user's degree of satisfaction or dissatisfaction with the current negotiation script. The user's current emotional information can be a specific numerical value; for example, a smaller value indicates a higher degree of dissatisfaction. Alternatively, the user's current emotional information can be one of several preset strings, such as "satisfied," "somewhat dissatisfied," or "strongly dissatisfied." This example embodiment does not specifically limit the specific content of the user's current emotional information. In some embodiments, determining the latest current negotiation script based on the user's current emotional information can be achieved by automatically identifying the negotiation script that best matches the user's current emotional information from among at least one negotiation script, or by re-inputting the complaint type, refund prediction information, and the user's current emotional information into the negotiation model to obtain the latest current negotiation script output by the negotiation model.

[0069] In some embodiments, the refund prediction information includes a refund amount standard. If the current negotiation script information includes a first refund amount, the first refund amount must meet the refund amount standard. In some embodiments, the refund prediction information includes a refund amount standard. If the negotiation script information output by the negotiation model includes a first refund amount, the first refund amount must meet the refund amount standard. For example, the refund amount standard may include a specific numerical value. In this case, the refund amount standard is used to indicate that the first refund amount in the negotiation script information must be less than or equal to that numerical value. Alternatively, the refund amount standard may also include an amount range. In this case, the refund amount standard is used to indicate that the first refund amount in the negotiation script information must be within that amount range. Or, for example, the refund amount standard may also be used to indicate no refund. In this case, the negotiation script information output by the negotiation model will not include any refund amount.

[0070] In some embodiments, if the current negotiation script information does not include a refund amount, the current negotiation script information is used to appease the target user; wherein, determining the latest current negotiation script information based on the user's current emotional information includes: if the user's current emotional information meets a preset first condition, determining the latest current negotiation script information, wherein the latest current negotiation script information includes the first refund amount. In some embodiments, if the current negotiation script information does not include any refund amount, i.e., the current negotiation script information is only used to appease the target user, then after obtaining the target user's current emotional information, if the user's current emotional information meets a preset first condition, the determined latest current negotiation script information needs to include a first refund amount that meets the refund amount standard, wherein the first condition may be that the user's current emotional information is one of at least one preset string, or the first condition may also be that the user's current emotional information is within a preset numerical range, or the user's current emotional information is less than or equal to a first preset value, or the user's current emotional information is greater than or equal to a second preset value. This example embodiment does not specifically limit the content of the first condition.

[0071] In some embodiments, if the user's current emotional information meets a preset second condition, the latest current negotiation script information includes a second refund amount, the second refund amount meets the refund amount standard, and the second refund amount is greater than the current refund amount corresponding to the historical negotiation script information. In some embodiments, if the current negotiation script information includes a second refund amount that meets the refund amount standard, then after obtaining the target user's current emotional information, if the user's current emotional information meets the preset second condition, the determined latest current negotiation script information needs to include the second refund amount, and the second refund amount needs to be greater than the current refund amount corresponding to the historical negotiation script information. The second condition can be that the user's current emotional information is one of at least one preset string, or the second condition can be that the user's current emotional information is within a preset value range, or that the user's current emotional information is less than or equal to a first preset value, or that the user's current emotional information is greater than or equal to a second preset value. This example embodiment does not specifically limit the content of the second condition. Historical negotiation script information refers to the negotiation script information sent to the target user in this negotiation interaction, and the current refund amount corresponding to the historical negotiation script information refers to the refund amount in the negotiation script information containing the refund amount most recently sent to the target user during this negotiation interaction.

[0072] In some embodiments, determining the latest current negotiation script information based on the user's current emotional information includes: if the user's current emotional information meets a preset third condition, determining the second refund amount based on the current refund amount corresponding to the historical negotiation script information; and determining the latest current negotiation script information based on the second refund amount. In some embodiments, if the user's current emotional information meets the preset third condition, the second refund amount needs to be determined based on the current refund amount corresponding to the historical negotiation script information. The second refund amount needs to be greater than the current refund amount, and the second refund amount still needs to meet the refund amount standard. For example, the second refund amount can be obtained by adding a preset value to the current refund amount, or by multiplying the current refund amount by a preset coefficient greater than 1. This example embodiment does not specifically limit this. The third condition can be that the user's current emotional information is one of at least one preset string, or the third condition can be that the user's current emotional information is within a preset value range, or that the user's current emotional information is less than or equal to a first preset value, or that the user's current emotional information is greater than or equal to a second preset value. This example embodiment does not specifically limit the specific content of the third condition.

[0073] In some embodiments, determining the second refund amount based on the current refund amount corresponding to the historical negotiation script information includes: determining the second refund amount based on the current refund amount corresponding to the historical negotiation script information and the user's current emotional information. In some embodiments, the target user's level of dissatisfaction with the current refund amount can be obtained based on the user's current emotional information, and then the second refund amount can be determined based on the current refund amount and the level of dissatisfaction. For example, the higher the level of dissatisfaction, the higher the determined second refund amount; the lower the level of dissatisfaction, the lower the determined second refund amount. For example, the second refund amount can be obtained by adding a preset amount to the current refund amount, where the amount of the amount increase is determined based on the level of dissatisfaction; the higher the level of dissatisfaction, the larger the amount of the amount increase; the lower the level of dissatisfaction, the smaller the amount of the amount increase. Another example is to multiply the current refund amount by a preset coefficient greater than 1 to obtain the second refund amount, where the preset coefficient is determined based on the level of dissatisfaction; the higher the level of dissatisfaction, the larger the preset coefficient; the lower the level of dissatisfaction, the smaller the preset coefficient. This example embodiment does not impose any special limitations on this.

[0074] In some embodiments, determining the latest current negotiation script information based on the user's current emotional information includes: re-inputting the user's current emotional information, the complaint type, and the refund prediction information into the negotiation model to obtain the latest current negotiation script information. In some embodiments, the user's current emotional information, the complaint type, and the refund prediction information can be re-inputted into the negotiation model to obtain the latest negotiation script information output by the negotiation model, and this negotiation script information can be sent to the target user as the latest current negotiation script information.

[0075] In some embodiments, obtaining the negotiation result of the current negotiation interaction includes: acquiring the target user's current emotional information during the negotiation interaction; if the user's current emotional information meets a preset fourth condition, generating the negotiation result of the current negotiation interaction, wherein the negotiation result indicates that the negotiation has been completed. In some embodiments, user emotion analysis can be performed on the target user's response information regarding the negotiation script information sent to them to obtain the target user's current emotional information during the current negotiation interaction. If the user's current emotional information meets a preset fourth condition, it can be determined that the current negotiation interaction has been completed, and a negotiation result indicating that the negotiation has been completed is generated. The fourth condition may be that the user's current emotional information is one of at least one preset string, or the fourth condition may be that the user's current emotional information is within a preset numerical range, or the user's current emotional information is less than or equal to a first preset value, or the user's current emotional information is greater than or equal to a second preset value. This example embodiment does not specifically limit the content of the fourth condition. In some embodiments, if it cannot be determined whether the current negotiation interaction has been completed based on the user's current emotional information within a preset time range, a negotiation result indicating that the negotiation was unsuccessful is generated.

[0076] In some embodiments, the method further includes: if the refund prediction information cannot be obtained based on the refund rule model, prompting the target merchant to enter the manual processing stage regarding the complaint request. In some embodiments, if the refund rule model fails to output the refund prediction information corresponding to the complaint request after the complaint type corresponding to the complaint request is input into the refund rule model corresponding to the target merchant, then the target merchant can be directly prompted to enter the manual processing stage regarding the complaint request, without having to enter the automated negotiation interaction with the target user without human intervention.

[0077] In some embodiments, the method further includes: if the negotiation result indicates that the negotiation was unsuccessful, prompting the target merchant to proceed to the manual processing stage regarding the complaint request. In some embodiments, if no negotiation result indicating successful negotiation is obtained within a preset time range, i.e., if it cannot be determined whether the negotiation interaction was successful within the preset time range, a negotiation result indicating unsuccessful negotiation is generated, and the target merchant is prompted to proceed to the manual processing stage regarding the complaint request.

[0078] Figure 2This is a schematic diagram of a device for handling complaints provided in an embodiment of this specification. This device (hereinafter referred to as "complaint handling device 1") can be implemented as all or part of an electronic device through software, hardware, or a combination of both. According to some embodiments, the complaint handling device 1 includes a first obtaining module 11, a second obtaining module 12, and a third obtaining module 13.

[0079] The first acquisition module 11 is used to obtain the complaint type corresponding to the complaint request of the target user about the target merchant, and obtain the refund prediction information corresponding to the complaint request based on the refund rule model corresponding to the target merchant.

[0080] The second acquisition module 12 is used to input the complaint type and the refund prediction information into the negotiation model, obtain at least one negotiation script information output by the negotiation model, and conduct negotiation interaction with the target user based on the at least one negotiation script information;

[0081] The third obtaining module 13 is used to obtain the negotiation result of this negotiation interaction. If the negotiation result indicates that the negotiation was successful, the module performs the corresponding complaint handling operation on the complaint request according to the negotiation result.

[0082] In some embodiments, the complaint processing device 1 is further configured to: generate or update a refund rule model corresponding to the target merchant based on the complaint feature information corresponding to the target merchant, wherein the merchant feature information is constructed based on the historical complaint data corresponding to the target merchant.

[0083] In some embodiments, the interaction type corresponding to the negotiation interaction includes any of the following: call interaction; text message interaction; voice message interaction.

[0084] In some embodiments, the step of inputting the complaint type and the refund prediction information into the negotiation model to obtain at least one negotiation script information output by the negotiation model includes: inputting the complaint type, the refund prediction information, and the user characteristic information corresponding to the target user into the negotiation model to obtain at least one negotiation script information output by the negotiation model.

[0085] In some embodiments, the step of inputting the complaint type, the refund prediction information, and the user characteristic information corresponding to the target user into the negotiation model to obtain at least one negotiation script information output by the negotiation model includes: inputting the complaint type, the refund prediction information, the user characteristic information corresponding to the target user, and the object characteristic information of the associated object corresponding to the complaint request into the negotiation model to obtain at least one negotiation script information output by the negotiation model.

[0086] In some embodiments, the step of negotiating and interacting with the target user based on the at least one negotiation script information includes: negotiating and interacting with the target user based on the current negotiation script information in the at least one negotiation script information to obtain the target user's response information regarding the current negotiation script information; determining the latest current negotiation script information based on the response information; continuing to negotiate and interact with the target user based on the latest current negotiation script information, and so on, until the interaction ends.

[0087] In some embodiments, determining the latest current negotiation script information based on the response information includes: determining the target user's current emotional information during the negotiation interaction based on the response information; and determining the latest current negotiation script information based on the user's current emotional information.

[0088] In some embodiments, the refund prediction information includes a refund amount standard, and if the current negotiation script information includes a first refund amount, the first refund amount meets the refund amount standard.

[0089] In some embodiments, if the current negotiation script information does not include the refund amount, the current negotiation script information is used to appease the target user; wherein, determining the latest current negotiation script information based on the user's current emotional information includes: determining the latest current negotiation script information if the user's current emotional information meets a preset first condition, wherein the latest current negotiation script information includes the first refund amount.

[0090] In some embodiments, if the user's current emotional information meets a preset second condition, the latest current negotiation script information includes a second refund amount, the second refund amount meets the refund amount standard, and the second refund amount is greater than the current refund amount corresponding to the historical negotiation script information.

[0091] In some embodiments, determining the latest current negotiation script information based on the user's current emotional information includes: if the user's current emotional information meets a preset third condition, determining the second refund amount based on the current refund amount corresponding to the historical negotiation script information; and determining the latest current negotiation script information based on the second refund amount.

[0092] In some embodiments, determining the second refund amount based on the current refund amount corresponding to the historical negotiation script information includes: determining the second refund amount based on the current refund amount corresponding to the historical negotiation script information and the user's current emotional information.

[0093] In some embodiments, determining the latest current negotiation script information based on the user's current emotional information includes: re-inputting the user's current emotional information, the complaint type, and the refund prediction information into the negotiation model to obtain the latest current negotiation script information.

[0094] In some embodiments, obtaining the negotiation result of this negotiation interaction includes: acquiring the target user's current emotional information during the negotiation interaction; if the user's current emotional information meets a preset fourth condition, generating the negotiation result of this negotiation interaction, wherein the negotiation result indicates that the negotiation has been completed.

[0095] In some embodiments, the complaint processing device 1 is further configured to: if the refund prediction information cannot be obtained based on the refund rule model, prompt the target merchant to enter the manual processing stage regarding the complaint request.

[0096] In some embodiments, the complaint handling device 1 is further configured to: if the negotiation result indicates that the negotiation is unsuccessful, prompt the target merchant to enter the manual processing stage regarding the complaint request.

[0097] The above-described apparatus embodiments correspond to the aforementioned method embodiments. For detailed descriptions, please refer to the description in the method embodiments section; further details will not be repeated here. The apparatus embodiments are derived from the corresponding method embodiments and have the same technical effects. For detailed descriptions, please refer to the corresponding method embodiments.

[0098] This specification also provides a computer storage medium storing a computer program thereon, which, when executed by a processor, implements the method described in this specification.

[0099] This specification also provides a computer program product that stores at least one instruction, which is loaded by the processor and executes the method described in this specification embodiment.

[0100] This specification also provides an electronic device, including a processor and a memory; wherein the memory stores a computer program adapted to be loaded by the processor and execute the method described in the embodiments of this specification.

[0101] The embodiments in this specification also provide Figure 3 The diagram shows the structure of the electronic device. Figure 3 At the hardware level, the electronic device includes a processor, internal bus, network interface, memory, and non-volatile memory, and may also include other hardware required for business operations. The processor reads the corresponding computer program from the non-volatile memory into memory and then runs it to implement the above method.

[0102] The systems, devices, modules, or units described in the above embodiments can be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, a computer can be, for example, a personal computer, laptop computer, cellular phone, camera phone, smartphone, personal digital assistant, media player, navigation device, email device, game console, tablet computer, wearable device, or any combination of these devices.

[0103] Those skilled in the art will understand that embodiments of this specification can be provided as methods, systems, or computer program products. Therefore, this specification may take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this specification may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0104] This specification is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this specification. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a machine for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0105] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0106] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1The steps of the function specified in one or more boxes.

[0107] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0108] This specification can be described in the general context of computer-executable instructions that are executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform a specific task or implement a specific abstract data type. This specification can also be practiced in distributed computing environments, where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0109] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to interchangeably. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments.

[0110] The above description is merely an embodiment of this specification and is not intended to limit this specification. Various modifications and variations can be made to this specification by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this specification should be included within the scope of the claims of this specification.

Claims

1. A method for handling complaints, comprising: Obtain the complaint type corresponding to the target user's complaint request regarding the target merchant, and based on the refund rule model corresponding to the target merchant, obtain the refund prediction information corresponding to the complaint request; Input the complaint type and the refund prediction information into the negotiation model to obtain at least one negotiation script information output by the negotiation model, and conduct negotiation interaction with the target user based on the at least one negotiation script information; Obtain the negotiation result of this negotiation interaction. If the negotiation result indicates that the negotiation was successful, perform the corresponding complaint handling operation on the complaint request according to the negotiation result.

2. The method according to claim 1, further comprising: The refund rule model corresponding to the target merchant is generated or updated based on the complaint feature information corresponding to the target merchant, wherein the merchant feature information is constructed based on the historical complaint data corresponding to the target merchant.

3. The method according to claim 1, wherein the interaction type corresponding to the negotiation interaction includes any one of the following: Call interaction; Text message interaction; Voice message interaction.

4. The method according to claim 1, wherein inputting the complaint type and the refund prediction information into the negotiation model to obtain at least one negotiation script information output by the negotiation model includes: Input the complaint type, the refund prediction information, and the user characteristic information corresponding to the target user into the negotiation model to obtain at least one negotiation script information output by the negotiation model.

5. The method according to claim 4, wherein inputting the complaint type, the refund prediction information, and the user characteristic information corresponding to the target user into the negotiation model to obtain at least one negotiation script information output by the negotiation model includes: Input the complaint type, the refund prediction information, the user characteristic information corresponding to the target user, and the object characteristic information of the associated object corresponding to the complaint request into the negotiation model to obtain at least one negotiation script information output by the negotiation model.

6. The method according to claim 1, wherein the step of negotiating with the target user based on the at least one negotiation script information comprises: Based on the current negotiation script information in the at least one negotiation script information, negotiate and interact with the target user to obtain the target user's response information regarding the current negotiation script information; Based on the response information, the latest current negotiation script is determined, and the negotiation interaction with the target user continues based on the latest current negotiation script, and so on, until the interaction ends.

7. The method according to claim 6, wherein determining the latest current negotiation script information based on the response information includes: Based on the response information, determine the target user's current emotional information during the negotiation interaction process; Based on the user's current emotional information, determine the latest current negotiation script.

8. The method according to claim 7, wherein the refund prediction information includes a refund amount standard, and if the current negotiation script information includes a first refund amount, the first refund amount conforms to the refund amount standard.

9. The method according to claim 8, wherein if the current negotiation script information does not include a refund amount, the current negotiation script information is used to appease the target user; in, The step of determining the latest current negotiation script information based on the user's current emotional information includes: If the user's current emotional information meets a preset first condition, the latest current negotiation script information is determined, wherein the latest current negotiation script information includes the first refund amount.

10. The method according to claim 8, wherein if the user's current emotional information meets a preset second condition, the latest current negotiation script information includes a second refund amount, the second refund amount meets the refund amount standard, and the second refund amount is greater than the current refund amount corresponding to the historical negotiation script information.

11. The method according to claim 10, wherein determining the latest current negotiation script information based on the user's current emotional information includes: If the user's current emotional information meets the preset third condition, the second refund amount is determined based on the current refund amount corresponding to the historical negotiation script information; Based on the second refund amount, determine the latest current negotiation script information.

12. The method according to claim 11, wherein determining the second refund amount based on the current refund amount corresponding to the historical negotiation script information includes: The second refund amount is determined based on the current refund amount corresponding to the historical negotiation script information and the user's current emotional information.

13. The method according to claim 7, wherein determining the latest current negotiation script information based on the user's current emotional information includes: The user's current emotional information, the complaint type, and the refund prediction information are re-inputted into the negotiation model to obtain the latest current negotiation script information.

14. The method according to claim 1, wherein obtaining the negotiation result of this negotiation interaction includes: The target user's current emotional information is obtained during the negotiation interaction. If the user's current emotional information meets a preset fourth condition, the negotiation result of this negotiation interaction is generated, wherein the negotiation result indicates that the negotiation has been completed.

15. The method according to claim 1, further comprising: If the refund prediction information cannot be obtained based on the refund rule model, the target merchant will be prompted to proceed to the manual processing stage regarding the complaint request.

16. The method according to claim 1, further comprising: If the negotiation result indicates that the negotiation is unsuccessful, the target merchant will be prompted to proceed to the manual processing stage regarding the complaint request.

17. An apparatus for processing complaints, comprising: The first acquisition module is used to obtain the complaint type corresponding to the complaint request of the target user about the target merchant, and obtain the refund prediction information corresponding to the complaint request based on the refund rule model corresponding to the target merchant; The second acquisition module is used to input the complaint type and the refund prediction information into the negotiation model, obtain at least one negotiation script information output by the negotiation model, and conduct negotiation interaction with the target user based on the at least one negotiation script information; The third obtaining module is used to obtain the negotiation result of this negotiation interaction. If the negotiation result indicates that the negotiation was successful, the module performs the corresponding complaint handling operation on the complaint request according to the negotiation result.

18. A storage medium having a computer program stored thereon, 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 to 16.

19. An electronic device, characterized in that, include: A processor and a memory; wherein the memory stores a computer program adapted to be loaded by the processor and to execute the steps of the method as claimed in any one of claims 1 to 16.

20. A computer program product having at least one instruction stored thereon, characterized in that, When the at least one instruction is executed by the processor, it implements the steps of the method according to any one of claims 1 to 16.

Images