Work order processing method and device, storage medium and program product

Abstract

The embodiment of the invention provides a work order processing method and device, a storage medium and a program product. In the work order processing method, for a to-be-processed target work order, target knowledge information required for processing the target work order can be generated through a first model, and work order detail information corresponding to the target work order is analyzed through a second model to obtain summary information corresponding to the target work order. The target knowledge information and the summary information can be displayed on a page, so that the target work order can be processed according to the summary information and the knowledge information. Wherein the summary information comprises the appeal information and / or the solution information corresponding to the appeal, so that the customer service staff can quickly know the user appeal and dispute point in the target work order. The knowledge information can help the customer service staff to quickly obtain information required for solving disputes, reduce the time for searching and thinking, and improve the work order processing efficiency.

Description

Technical Field

[0001] This application relates to the field of Internet technologies, and in particular, to a work order processing method, device, storage medium, and program product. Background Art

[0002] A work order is an information carrier for managing and tracking customer service requests or problem reports. In an e-commerce scenario, an e-commerce platform can create a work order for a customer when the customer makes a service request or encounters a dispute. For example, when a customer has needs such as order query, product return and exchange, complaint handling, rights protection, etc., the e-commerce platform can create corresponding work orders for the above needs respectively. Each work order can contain relevant information about the request or problem, such as customer contact information, problem description, historical interaction records, and processing status, etc. The customer service of the e-commerce platform can process these work orders to handle the customer's requests or problems.

[0003] In some existing solutions, the e-commerce platform can present a large amount of information related to work orders. The customer service needs to extract effective information from the large amount of information related to work orders and process the work orders according to the extracted effective information. This existing work order processing method has low efficiency and there is a need to propose a new solution. Summary of the Invention

[0004] Multiple aspects of this application provide a work order processing method, device, storage medium, and program product to improve work order processing efficiency.

[0005] An embodiment of this application provides a work order processing method, including: displaying a first page, where the first page includes a target work order to be processed, and the target work order is associated with target knowledge information generated by a first model for processing the target work order; invoking a second model to analyze work order detail information corresponding to the target work order to obtain summary information corresponding to the target work order, where the summary information includes: request information and / or solution information corresponding to the request; displaying the target knowledge information and the summary information on the first page or a second page for processing the target work order according to the summary information and the target knowledge information, and the second page is different from the first page.

[0006] An embodiment of this application further provides an electronic device, including: a memory and a processor; the memory is used to store one or more computer instructions; the processor is used to execute the one or more computer instructions to: execute the steps in the method provided by the embodiment of this application.

[0007] An embodiment of this application further provides a computer-readable storage medium storing a computer program, and when the computer program is executed by a processor, it can implement the steps in the method provided by the embodiment of this application.

[0008] An embodiment of this application also provides a computer program product, including: computer programs / instructions, which can implement the steps in the method provided by the embodiment of this application when executed by a processor.

[0009] In an embodiment of this application, for a target work order to be processed, target knowledge information required to process the target work order can be generated through a first model, and the work order details information corresponding to the target work order can be analyzed through a second model to obtain summary information corresponding to the target work order. The target knowledge information and the summary information can be displayed on a page for processing the target work order according to the summary information and the knowledge information. Among them, the summary information includes: claim information and / or solution information corresponding to the claim, which can enable customer service personnel to quickly understand the user's claims and dispute points in the target work order. The knowledge information can help customer service personnel quickly obtain the information required to resolve disputes, reduce the time spent on searching and thinking, and improve the processing efficiency of work orders. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] The drawings described herein are used to provide a further understanding of this application, and constitute a part of this application. The illustrative embodiments of this application and their descriptions are used to explain this application and do not constitute an improper limitation of this application. In the drawings:

[0011] Figure 1 It is a schematic flow chart of a work order processing method provided by an exemplary embodiment of this application;

[0012] Figure 2 It is a schematic diagram of a second page provided by an exemplary embodiment of this application;

[0013] Figure 3 It is a schematic flow chart of obtaining knowledge information provided by an exemplary embodiment of this application;

[0014] Figure 4 It is a schematic flow chart of obtaining summary information provided by an exemplary embodiment of this application;

[0015] Figure 5 It is a schematic flow chart of obtaining emotion classification information provided by an exemplary embodiment of this application;

[0016] Figure 6a It is a schematic flow chart of a clue rule configuration interface provided by an exemplary embodiment of this application;

[0017] Figure 6b It is a schematic flow chart of a clue rule configuration interface provided by another exemplary embodiment of this application;

[0018] Figure 6c It is a schematic flow chart of a clue rule configuration interface provided by yet another exemplary embodiment of this application;

[0019] Figure 7 Schematic architecture diagram of the work order processing method provided by an exemplary embodiment of the present application;

[0020] Figure 8 Schematic structural diagram of an electronic device provided by an exemplary embodiment of the present application. Detailed implementation manners

[0021] To make the objectives, technical solutions and advantages of the present application clearer, the technical solutions of the present application will be clearly and completely described below in conjunction with the specific embodiments of the present application and the corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0022] The terms used in the embodiments of the present invention are only for the purpose of describing specific embodiments, and are not intended to limit the present invention. The singular forms "a", "the" and "said" used in the embodiments of the present invention and the appended claims are also intended to include the plural forms, unless the context clearly indicates otherwise. "Plural" generally includes at least two, but does not exclude the case of including at least one.

[0023] It should be understood that the term "and / or" used herein is only a kind of association relationship describing associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " herein generally represents an "or" relationship between the associated objects before and after.

[0024] It should also be noted that the terms "comprise", "include" or any other variant thereof are intended to cover non-exclusive inclusion, so that a product or system including a series of elements not only includes those elements, but also includes other elements not explicitly listed, or further includes elements inherent to such product or system. Without further limitations, the element defined by the statement "including one..." does not exclude the existence of other identical elements in the product or system including the element.

[0025] In some work order processing solutions, an e-commerce platform can present a large amount of information related to work orders. Customer service needs to extract effective information from the large amount of information related to work orders and process the work orders based on the extracted effective information. This work order processing method has low efficiency. In view of the above technical problems, in some embodiments of the present application, a solution is provided. The technical solutions provided by the embodiments of the present application are described in detail below with reference to the drawings.

[0026] Figure 1It is a schematic flowchart of a work order processing method provided by an exemplary embodiment of the present application. The method may include the following Figure 1 shown steps:

[0027] Step 101: Display a first page, where the first page includes a target work order to be processed, and the target work order is associated with target knowledge information generated by a first model for processing the target work order.

[0028] Step 102: Invoke a second model to analyze the work order detail information corresponding to the target work order, and obtain summary information corresponding to the target work order. The summary information includes: appeal information and / or solution information corresponding to the appeal.

[0029] Step 103: Display the target knowledge information and the summary information on the first page or a second page for processing the target work order according to the summary information and the target knowledge information. The second page is different from the first page.

[0030] This embodiment can be executed by a work order processing tool. In some embodiments, the work order processing tool can be implemented as a client program running on a terminal device, and the terminal device can be implemented as a desktop computer, a laptop computer, or a smart phone, etc. In other embodiments, the work order processing tool can be implemented as a system composed of a front-end program and a back-end program. The front-end program can include a client program or a web page. In this system, the front-end program can run on the above terminal device, and the back-end program can run on a server device, and the server device can include, but is not limited to: conventional servers, cloud servers, or server arrays and other server devices. In the work order processing scenario, the terminal device usually refers to the terminal device used by customer service personnel for processing work orders.

[0031] In this embodiment, a work order refers to an information carrier for managing and tracking customer service requests or problem reports. In the e-commerce scenario, an e-commerce platform can create a work order for a customer when the customer (including a buyer or a seller) makes a service appeal or encounters a dispute. In step 101, the work order processing tool can display a first page through the terminal device. The first page can be a work order task management page. The work order task management page can display a list of work orders to be processed and the processing status of any work order for the customer service personnel to view, and can provide an operation entry for the customer service personnel to process the work order. When the customer service personnel has logged in to their personal account, the work orders to be processed displayed on the work order task management page refer to the work orders assigned to the customer service personnel. In this embodiment, any work order to be processed displayed on the first page will be used as an example to exemplarily illustrate the work order processing method. Any work order to be processed on the first page is described as a target work order.

[0032] The target work order is associated with target knowledge information, which is used to provide the knowledge required for processing the target work order, so as to provide reference and support for the work order processing operation. The target knowledge information may include: operation guides required for work order processing, historical cases, product knowledge information corresponding to the target work order, etc. In this embodiment, the target work order associated knowledge information is generated by the first model. Among them, the first model can be a model for knowledge retrieval, or can be a large language model for content generation, which is not limited in this embodiment.

[0033] If the first model is a model for knowledge retrieval, the first model can retrieve the knowledge information related to the target work order from a large amount of existing data and information by means of keyword matching or semantic analysis, querying databases, knowledge bases or search engines, etc. If the first model is a large language model, the first model can retrieve the implicit knowledge representation matching the target work order from the large amount of implicit knowledge representations absorbed and stored during its training process, and generate explicit target knowledge information according to the retrieved implicit knowledge representation. The target knowledge information is used to enable customer service personnel to quickly obtain the information required to solve problems, reduce the time spent on searching and thinking, and improve the processing efficiency of work orders.

[0034] In step 102, the work order processing tool can call the second model to analyze the work order details information corresponding to the target work order to obtain the summary information corresponding to the target work order. Optionally, the operation of calling the second model to analyze the work order details information corresponding to the target work order can be triggered by the work order processing tool under specified operations, or can be actively executed by the work order processing tool, which is not limited in this embodiment.

[0035] The work order details information refers to the set of information and data used to describe the creation background of the target work order, and is used to comprehensively display the context status of the target work order. In this embodiment, the work order details information may include, but is not limited to, at least one of the order information corresponding to the target work order, transaction remarks, processing records, communication records between the buyer and the seller, communication records with online customer service, communication records with hotline customer service, voice communication records, action records, and instant messaging remarks. Among them, the action record refers to the record of relevant operations initiated by both parties in dispute due to the dispute, such as the record of applying for after-sales service, the record of applying for the intervention of the artificial customer service, the record of urging the customer service to follow up, and so on.

[0036] Among them, the second model can be a large language model with the ability to summarize abstracts. The work order processing tool can provide the work order detail information to the large language model and construct prompts to prompt the large language model to perform the task of generating an abstract, so that the large language model analyzes the work order detail information to obtain the abstract information. In this embodiment, the abstract information may at least include: claim information and / or solution information corresponding to the claim. Among them, the claim information is information used to describe the claims or assertions of users with disputes. For example, in an e-commerce scenario, the claim information may include at least one of the claim information of the buyer and the claim information of the seller. The solution information corresponding to the claim refers to the solution provided by the respondent user for the claim, or the solution expected by the complainant for the respondent to give. When constructing the prompt, the work order processing tool can add different task types to the prompt, that is, the summary task of the claim information and the summary task of the solution corresponding to the claim, and can specify the task content. For example, the task content can be specifically described as: according to the input work order detail information, analyze the claims of the buyer, the claims of the seller, and the solutions corresponding to each claim.

[0037] After obtaining the abstract information based on the above implementation, in step 103, the work order processing tool can display the knowledge information and the abstract information through the terminal device for processing the target work order according to the abstract information and the knowledge information. Among them, the target knowledge information and the abstract information can be displayed on the first page or the second page, and the second page and the first page can be different pages. In some embodiments, the work order processing tool can respond to a specified operation, jump from the first page for displaying the target work order to the second page, and display the knowledge information and the abstract information on the second page. As Figure 2 shown, the second page can be the processing page of the target work order. This page can display the information card corresponding to the intelligent diagnosis. This information card can display abstract information such as order information, buyer's claim, buyer's emotion, seller's solution, platform solution, etc., and can also display product knowledge for the reference of customer service staff.

[0038] In this embodiment, for the target work order to be processed, the target knowledge information required for processing the target work order can be generated through the first model, and the work order detail information corresponding to the target work order can be analyzed through the second model to obtain the abstract information corresponding to the target work order. The target knowledge information and the abstract information can be displayed on the page for processing the target work order according to the abstract information and the knowledge information. Among them, the abstract information includes: claim information and / or solution information corresponding to the claim, which can enable customer service staff to quickly understand the user's claims and dispute points in the target work order. The knowledge information can help customer service staff quickly obtain the information required to resolve disputes, reduce the time spent on searching and thinking, and improve the processing efficiency of work orders.

[0039] In the above and following embodiments of the present application, a large language model refers to a natural language processing (NLP) model that has been trained on a large scale. Large language models are typically built based on deep learning techniques and trained on large-scale training datasets, enabling them to exhibit powerful performance when processing natural language tasks. Given a piece of text (i.e., context), the large language model attempts to predict the most likely next word. The number of parameters of the large language model is greater than a set threshold, which is usually on the order of millions or billions. During the training process, these parameters are continuously adjusted and optimized based on the difference between the prediction results and the actual results of the large language model to improve the prediction accuracy of the large language model. In some embodiments, the large language model typically adopts an advanced neural network architecture, such as the Transformer architecture, to build the model structure, which enables the large language model to capture complex patterns in the text and handle long-range dependencies. After sufficient training, the large language model has powerful generation capabilities and can produce coherent and contextually appropriate text content based on a given prompt.

[0040] In this embodiment, a large language model pre-trained on a large number of general datasets can be used as the base model of the first model. To adapt to the specific application scenario of the present application, that is, for generating the target knowledge information required for processing work orders, the pre-trained large language model can be fine-tuned on a dataset formed by a large number of work orders and the knowledge information required for processing work orders, so that the fine-tuned large language model is more suitable for performing the task of generating work order knowledge information. Similarly, a large language model pre-trained on a large number of general datasets can be used as the base model of the second model. To adapt to the specific application scenario of the present application, that is, for summarizing work order details information, the pre-trained large language model can be fine-tuned on a dataset formed by a large number of work order details information and their summaries, so that the fine-tuned large language model is more suitable for performing the task of summarizing work order details information.

[0041] In some alternative embodiments, the target knowledge information is pre-generated before the target work order is displayed on the first page, so that when the customer service staff processes the target work order, there is no need to wait for the first model to return the knowledge information.

[0042] Optionally, before presenting the target work order on the first page, the work order processing tool can respond to the assignment operation of the target work order and identify the target question to be answered from the work order details information corresponding to the target work order. Optionally, the target question can be various questions associated with the target work order, which can be raised by different users involved in the target work order. For example, in an e-commerce scenario, the target question can be a question raised by a buyer regarding a product, a question raised by a buyer / seller regarding a solution, etc. After identifying the target question, the work order processing tool can call the first model to answer the target question and obtain the target knowledge information associated with the target work order. Optionally, the target knowledge information associated with the target work order can be stored in the cache of a specified database. When a selection operation on the target work order is detected, the work order processing tool can query the target knowledge information associated with the target work order from the cache of the specified database and present the queried target knowledge information.

[0043] In this implementation, the assignment operation of the target work order is used to trigger the pre-generation operation of the target knowledge information, enabling the customer service staff to view the knowledge information in a timely manner when processing the target work order, without having to wait for the first model to return the target knowledge information when processing the target work order, thus improving the processing efficiency of the work order.

[0044] In an e-commerce scenario, the target work order is usually created when a dispute occurs between a buyer and a seller regarding a product order, and the work order details information may include information for identifying the product category corresponding to the target work order. The product category refers to the category formed by classifying products according to certain principles, such as classifying products according to rules such as the use, process, brand, target consumer group, etc. of the products. Optionally, when the work order processing tool identifies the target question to be answered from the work order details information corresponding to the target work order, it can identify the product category corresponding to the target work order according to the work order details information corresponding to the target work order. For example, it can obtain the order information from the work order details information, identify the product name from the order information, and query the product category corresponding to the product name in the product category table as the product category corresponding to the target work order.

[0045] The work order processing tool can determine whether the product category corresponding to the target work order is within the set category range. Among them, the set category range includes product categories with relatively high professionalism, and customer service personnel need to have a relatively high level of knowledge when processing work orders for such products. For example, the technology of mobile phones and digital products is updated rapidly, involving multiple aspects such as hardware specifications, software compatibility, and operating systems, with relatively high professionalism. For example, medical devices and health care products are directly related to the health of users, with relatively high professionalism. Another example is that the identification of the authenticity of metal or gem products relies on complex criteria and has relatively high professionalism. In some alternative implementation manners, multiple product categories with relatively high professionalism can be preset within the category range. If the product category corresponding to the target work order hits any one of the preset product categories, it is determined that the product category corresponding to the target work order is within the set category range. In some other alternative embodiments, the work order processing tool can call a large language model, identify the product category corresponding to the target work order according to the work order details information corresponding to the target work order, and determine whether the product category corresponding to the target work order is within the set category range. This embodiment is not limited.

[0046] If the product category is not within the set category range, the work order processing tool may not perform the pre-generation operation of knowledge information. If the product category is within the set category range, the work order processing tool can identify the question to be answered from the work order details information corresponding to the target work order. The work order processing tool can determine whether the identified question to be answered is a question in a specific professional field. Among them, a question in a specific professional field refers to a question that needs professional knowledge and skills to solve due to factors such as product characteristics, technical requirements, market norms, or cultural backgrounds. These questions usually go beyond the common sense of ordinary consumers and involve deeper technical details, industry standards, laws and regulations, and consumer protection, etc. For example, for product categories such as food, cosmetics, and metal jewelry, the safety and applicability of raw materials and ingredients are professional questions. For product categories such as digital cameras and mobile phones, technical indicators such as pixel count, processor frequency, and memory specifications are professional questions.

[0047] In some alternative embodiments, for each product category, at least one tag corresponding to a professional field may be preset. After the work order processing tool identifies the question to be answered and determines the product category corresponding to the target work order, it may determine whether the question to be answered matches any of the tags corresponding to the professional fields under the product category. If it matches any of the tags corresponding to the professional fields under the product category, it may be determined that the question to be answered is a question in the professional field under the product category. If it does not match any of the tags corresponding to at least one professional field under the product category, it may be determined that the question to be answered is not a question in the professional field under the product category. In some other alternative embodiments, the work order processing tool may call a large language model to determine whether the question to be answered is a question in a professional field, and this embodiment is not limited.

[0048] If the question to be answered identified from the target work order is a question in a specific professional field, the work order processing tool may not perform the pre-generation operation of knowledge information. If the question to be answered identified from the target work order is a question in a specific professional field, the work order processing tool may use the question to be answered as the target question to be answered. After obtaining the target question based on the above implementation, the work order processing tool may call the first model to answer the target question, obtain the answer result, and use the answer result as the target knowledge information associated with the target work order.

[0049] In some embodiments, the work order processing tool may obtain the answer result of the target question in an asynchronous callback manner. Specifically, the work order processing tool may register a listener or callback method with the first model. After providing the target question to the first model, the work order processing tool may continue to perform the processing operations of other questions or other work orders without waiting for the first model to return the processing result. After the first model generates the answer result of the target question, it may return the answer result to the work order processing tool in the form of a callback message through the listener or callback method registered by the work order processing tool. Thus, even if the time taken by the first model to answer the target question is relatively long, it will not block the processing logic of the work order processing tool, and for example, it can improve the processing efficiency of the work order processing tool.

[0050] It should be noted that before pre-generating the target knowledge information associated with the target work order, the work order processing tool may add a distributed lock to the target work order to prevent the risk of concurrently pre-generating knowledge information for the same work order. Additionally, in some possible scenarios, due to network reasons or model reasons, the work order processing tool fails to successfully obtain the pre-generated target knowledge information. In this case, when the work order processing tool detects a selection operation on the target work order, it may perform the above steps of calling the first model to pre-generate knowledge information again.

[0051] The following will be combined with Figure 3A further exemplary description is provided for the alternative implementation manners of the work order processing tool to obtain the target knowledge information associated with the target work order.

[0052] As Figure 3 shown, the work order processing tool can respond to the work order assignment message, filter the work order assignment message. If the type of the work order corresponding to the work order assignment message is a dispute work order, the work order assignment message is added to the group message. For any work order corresponding to the work order assignment message in the group message, the work order processing tool can query the cache in the database to determine whether the knowledge information corresponding to the work order is hit. If the knowledge information corresponding to the work order is hit, the hit knowledge information is used as the target knowledge information and directly returned. If the knowledge information corresponding to the work order is not hit, the algorithm link for pre-generating knowledge information is executed.

[0053] As Figure 3 shown, in the algorithm link for pre-generating knowledge information, a distributed lock can be added to the work order, and the work order details information required for pre-generating knowledge information is constructed as the algorithm input parameter. Then, according to the algorithm input parameter, the operation of filtering the commodity category is performed to determine whether the commodity category corresponding to the work order is the specified commodity category. If it is the specified commodity category, the problem title in the algorithm input parameter can be obtained, and the large language model is called through the specified interface to determine whether the problem title is a professional field problem. If it is a professional field problem, the large language model is called through the specified interface to obtain the problem answer. After obtaining the problem answer, the distributed lock can be released. As Figure 3 shown, the answer information generated by the large language model can be added to the message queue, and the work order processing tool can read the answer information from the message queue by subscribing to the message and write it into the cache of the database. When the customer service staff selects a work order, the cache in the database can be queried to determine whether the knowledge information corresponding to the work order is hit. If the knowledge information corresponding to the work order is hit, the target knowledge information is directly returned. If the knowledge information corresponding to the work order is not hit, the algorithm link for pre-generating knowledge information can be repeatedly called, and this call operation is executed asynchronously without waiting for the result of the algorithm link to be returned.

[0054] Based on this implementation manner, the work order processing tool can call the first model to obtain knowledge information when it is determined that the commodity category corresponding to the target work order is within the set category range and the problem identified from the work order details information is a problem in a specific professional field, reducing the number of calls to the first model and thus reducing the cost of obtaining knowledge information.

[0055] In some alternative embodiments, the operation of the work order processing tool calling the second model can be triggered by the selection operation of the target work order. When the work order processing tool detects the selection operation of the target work order, it can call the second model to analyze the work order details information corresponding to the target work order to obtain the summary information corresponding to the target work order.

[0056] Optionally, the operation of selecting a target work order can be a user operation or a system operation. Among them, a user operation refers to an operation directly triggered by the interaction operation of a customer service staff. The interaction operation of the user can be issued through the human-machine interface provided by the terminal device, and the human-machine interface can include at least one of: command line interface (CLI), graphical user interface (GUI), text user interface (TUI), voice user interface (VUI), touch user interface (Touch UI), gesture user interface (Gestural UI), and peripheral interfaces such as keyboards and mice.

[0057] Among them, a system operation refers to an operation automatically generated by an operating system or other software components. System operations can include but are not limited to: timer arrival operations, operations for logical or state changes within an application (such as completion of a specified task), operations for receiving specific communication messages from other processes / applications, application programming interface (API) call operations, etc. For example, in some scenarios, the target work order is configured with a waiting time. After the target work order is assigned, a countdown operation for the waiting time can be started. When the countdown ends, the work order processing tool can actively initiate an operation to select the target work order. Another example is that in some scenarios, when there are multiple work orders to be processed, the multiple work orders are arranged in sequence. After the previous work order is processed, the work order processing tool can actively initiate an operation to select the next work order, and so on.

[0058] Optionally, when the work order processing tool analyzes the work order detail information corresponding to the target work order by invoking the second model, it can call a specified service object according to the identifier of the target work order, so that the service object queries the description information of at least one dimension of the target work order to obtain the work order detail information corresponding to the target work order. Among them, the service object is obtained by encapsulating a series of methods, and these methods are used to obtain information from different data sources or data domains. For example, as Figure 4 shown, in some embodiments, the acquisition methods corresponding to the information of at least one dimension can be encapsulated in the service object, and the information of the at least one dimension can include but is not limited to: order information, transaction remarks, processing records, communication records between buyers and sellers, communication records with online customer service, voice communication records, and instant messaging remarks. Furthermore, the service object can obtain the corresponding information from the data sources or data domains corresponding to different information in parallel, so as to facilitate the unified integration of the parameters required by the work order processing tool, so that the work order processing tool can use these data.

[0059] As Figure 4As shown, after obtaining the work order details information, the work order processing tool may pre-process the work order details information, and call the second model to analyze the pre-processed work order details information to obtain the demand information and / or the solution information corresponding to the demand. Optionally, pre-processing the work order details information may include: converting the format of the work order details information to convert the work order details information into a standardized interface input parameter that conforms to the preset interface format, thereby facilitating the analysis of the work order details information by calling the second model through the preset interface. Among them, the call of the work order processing tool to the second model may be a synchronous call, that is, after the work order processing tool initiates a call operation to the second model through the preset interface, it can wait for the analysis result returned by the second model. After obtaining the analysis result returned by the second model, the work order processing tool may write the analysis result to the specified cache space for query.

[0060] Based on this implementation, the work order processing tool uses the service object to obtain description information of different dimensions in parallel, which can effectively reduce the waiting time and improve the efficiency of calling the second model, thereby facilitating high-performance response to the selection operation of the target work order.

[0061] In some optional embodiments, the work order processing tool may also obtain and display the emotion classification information corresponding to the target work order. The emotion classification information refers to the information obtained by classifying the emotion of at least one user associated with the target work order. In the e-commerce scenario, the emotion classification information corresponding to a target work order may include: the emotion classification information of the buyer user and / or the emotion classification information of the seller user. For example, the buyer's emotion information may include: three emotions such as calm, fluctuating and excited in a progressive level. The emotion classification information can facilitate the determination of the processing priority of the target work order, and can assist customer service personnel in selecting a more reasonable work order processing method.

[0062] Optionally, the emotion classification information corresponding to each work order is stored in a specified storage space. For example, in some scenarios, the identification information of the work order and the emotion classification information corresponding to the work order can be stored in a database in the form of key-value pairs for querying the corresponding emotion classification information according to the identification information of the work order. The identification information of the work order is used to uniquely identify and track each work order, such as the work order number. For the target work order, when the work order processing tool detects a selection operation on the target work order, it can query the specified storage space according to the identification information of the target work order to obtain the emotion classification information corresponding to the target work order, and display the queried emotion classification information for processing the target work order according to the summary information, knowledge information, and emotion classification information. In some embodiments, the emotion classification information corresponding to the target work order can be displayed on the second page together with the target knowledge information and summary information associated with the work order for the customer service staff to view uniformly. In some other alternative embodiments, the emotion classification information corresponding to the target work order can be displayed on other pages other than the first page and the second page, and this embodiment is not limited.

[0063] Based on this implementation, the work order processing tool can actively provide the customer service staff with the emotion classification information corresponding to the target work order, thereby facilitating the customer service staff to make decisions on the processing priority and processing method of the target work order and improving the work order processing efficiency.

[0064] Optionally, the emotion classification information corresponding to the target work order is obtained before the target work order is displayed on the first page. Before the work order processing tool displays the target work order on the first page, it can call the third model to analyze the work order details information corresponding to the target work order to obtain the emotion classification information corresponding to the target work order, and store the emotion information corresponding to the target work order in the specified storage space. Optionally, the work order details information can at least include the voice conversation information between different user roles associated with the target work order, and the voice conversation information includes the tones of different user roles to facilitate the analysis of the emotional states of users in different roles.

[0065] In some possible application scenarios, the emotion classification information corresponding to the target work order is obtained in the quality inspection link of the target work order. As Figure 5 shown, before allocating the target work order, the work order processing tool can configure a quality inspection task for the target work order and use the quality inspection platform to execute the quality inspection task. When the quality inspection platform executes the quality inspection task of the target work order, it can provide the work order details information corresponding to the target work order to the third model, as Figure 5The shown emotion recognition model enables the third model to classify the emotions of different user roles involved in the target work order according to the work order details information. The quality inspection platform can call the third model in an asynchronous callback manner. After the third model obtains the emotion classification information corresponding to the target work order, it can store the emotion classification information in the message queue. The server in the work order processing tool can read the emotion classification information from the message queue and store the identification information of the target work order and its corresponding emotion classification information in a specified database in the form of a cache for query.

[0066] In this embodiment, the third model can be an algorithm model based on traditional machine learning or deep learning. For example, the third model can be a machine learning model based on the Naive Bayes algorithm, Support Vector Machine (SVM), or Logistic Regression. Or, it can be a deep learning model based on the Recurrent Neural Network (RNN) or Convolutional Neural Network. This embodiment does not make any restrictions. Taking the third model implemented as a Convolutional Neural Network as an example, when the third model analyzes the work order details information corresponding to the target work order, it can perform word segmentation on the text corresponding to the work order details information and map each word obtained by word segmentation into a vector space of a fixed length to form a sentence-level matrix representation. In the convolutional layer of the third model, multiple filters can be used to scan the entire sentence matrix to generate local feature maps. In the pooling layer of the third model, pooling operations can be used to reduce the feature dimensions and retain important features. In the fully connected layer of the third model, further feature extraction can be performed on the feature vectors that have undergone convolution and pooling to obtain high-level abstract features. In the output layer of the third model, the emotion category probability distribution can be output according to the high-level abstract features. In some optional embodiments, the third model can also be the large language model described in the foregoing embodiments. This embodiment does not make any restrictions.

[0067] Based on this implementation method, in the quality inspection link before the target work order is processed, calling the third model to obtain the emotion classification information corresponding to the target work order is beneficial for pre-classifying the processing priority of the target work order. In addition, based on the third model, the emotion characteristics in the work order details information can be captured more accurately, so as to accurately identify the emotion classification information corresponding to the target work order and provide more data support for the processing operation of the target work order.

[0068] In some alternative embodiments, the work order processing tool may further provide the dispute clue information required to process the target work order, so as to provide more comprehensive information support for the work order processing operation and guide the work order processing operation to a certain extent. Among them, the dispute clue information refers to the key data that helps to understand the dispute background and customer intention corresponding to the target work order, as well as the relevant data that helps to accelerate the settlement of disputes. The dispute clue information can be screened from the dispute voucher data corresponding to the target work order, or can be generated according to the detail data corresponding to the target work order, which is not limited in this embodiment. The following will give an exemplary description of the alternative implementation methods for obtaining the dispute clue information.

[0069] Optionally, the work order processing tool may obtain the dispute voucher data corresponding to the target work order. The dispute voucher data may include at least one of a purchase and sale view, a product view, and a review view. Different views are used to present the dispute event from different perspectives to help customer service personnel make more reasonable decisions. Among them, the purchase and sale view may include the interaction history between the buyer and the seller and the specific situation of the current dispute, including the communication records between the two parties, transaction details, payment information, the refund amount of the buyer, the dispute amount of the buyer / seller, whether the buyer is a malicious buyer, etc. The product view is used to describe the information about the characteristics, specifications, quality status, etc. of the product itself, and is used to evaluate whether the product meets the expected standards and whether there are manufacturing defects, risks or other problems. The review view is used to describe the overall satisfaction of the customer with the product and service, so as to identify potential problems. In some scenarios, an interface for obtaining the dispute voucher data can be customized and developed, and the interface can be configured as a service object, and the service object can obtain the dispute voucher data corresponding to the target work order from multiple data domains provided by the e-commerce platform.

[0070] After obtaining the dispute voucher data corresponding to the target work order, the work order processing tool may match the dispute voucher data with the conditions in the preset clue rules to obtain partial detail data that hits the target rules in the clue rules. Among them, the clue rules contain a set of predefined conditions for automatically screening out the dispute clue information that meets the conditions by means of condition matching. Optionally, in the e-commerce application scenario, the clue rules may include: clue rules from the buyer dimension, clue rules from the buyer dimension, and clue rules from the product dimension.

[0071] The target rule is the rule in the preset clue rules that is hit. During the rule matching process, one clue rule may match multiple dispute vouchers, and one dispute voucher may also match multiple clue rules. In this embodiment, different types of rules can be set, and the reliability of different types of rules is different. For example, in some embodiments, accurate clue rules, suspected clue rules, and negotiation clue rules can be set.

[0072] Among them, the accurate clue rule refers to a rule that triggers corresponding actions based on clear and definite conditions. For example, an accurate clue rule can be: if the buyer meets specific conditions (such as the buyer being a member, or the buyer's purchase date being within 7 days), then automatically approve a full refund; if the logistics system feedbacks that the package has been delivered, then immediately update the order status to "completed". The suspected clue rule is a rule that identifies potential problems or abnormal situations based on a certain probability or fuzzy logic. This type of rule is applicable to situations where action decisions cannot be made based on a single condition, but can be made through a comprehensive consideration of multiple factors. For example, a suspected clue rule can be: mark suspicious transactions according to the user's behavior patterns (such as frequently changing addresses, abnormally high-value orders), and prompt for additional security checks; for product batches that have received multiple reports of the same product defect, issue a warning and recommend suspending sales. Among them, the negotiation clue rule refers to a rule that requires multi-party collaboration or further communication to make action decisions. For example, a negotiation clue rule can be: if the number of only-refund requests from the buyer is greater than 3 times, then a refund needs to be negotiated with the seller; if the refund amount involved in the order is greater than the set threshold, then a refund needs to be negotiated with the seller.

[0073] When the dispute voucher data hits different types of rules, the corresponding clue types of the dispute voucher data are different. The work order processing tool can obtain different types of dispute clue information according to the dispute voucher data and the type of the target rule it hits. For example, if a certain dispute voucher data hits an accurate clue rule, the work order processing tool can determine that this dispute voucher data is an accurate clue. If a certain dispute voucher data hits a suspected clue rule, the work order processing tool can determine that this dispute voucher data is a suspected clue. After obtaining different types of dispute clue information, the work order processing tool can display different types of dispute clue information for processing the target work order according to different types of dispute clue information. Optionally, different types of dispute clue information can be displayed together with knowledge information and summary information on the second page, or on other pages, which is not limited in this embodiment.

[0074] Based on this implementation method, obtaining the dispute clue information in the dispute voucher data through rule matching is beneficial to providing intuitive guidance information to customer service personnel, thereby facilitating the improvement of work order processing efficiency. In addition, classifying the clue information can intuitively display the importance and uses of different clues, facilitating customer service personnel to quickly locate the required information without having to manually re-classify multiple types of clue information. In some scenarios, customer service personnel can quickly process work orders according to accurate clues, quickly identify potential risks according to suspected clues, and can conduct multi-party communication according to negotiation clues, greatly improving the work order processing efficiency.

[0075] In some alternative embodiments, any rule in the preset clue rules includes conclusion information and / or action information corresponding to the conditions. For example, any rule in the clue rules can be described as a triple including conditions, conclusions, and actions. In the triple corresponding to the rule, the condition is the prerequisite for triggering this rule, the conclusion is the result or inference obtained when the condition is met, and the action is the specific operation or behavior that needs to be executed when the condition is met. For example, a certain rule is: if the number of only-refund requests from the buyer is greater than 3 times (condition), then the direct refund processing method is unavailable (conclusion), and a refund can be made after negotiating with the seller (action). Another example, a rule is: if the buyer hits any malicious label in the malicious label list (condition), then it is considered that the buyer has a risk of forged vouchers and thus the buyer's vouchers are unavailable (conclusion), and the buyer can be requested to verify the vouchers on the specified page (action).

[0076] For any of the above different types of dispute clue information, the work order processing tool can obtain the work order processing guidance information corresponding to the dispute clue information according to the conclusion information and / or action information in the target rule hit by the dispute clue information, and display the work order processing guidance information together with the dispute clue information. Among them, the work order processing guidance information is used to provide directional suggestions for customer service staff to handle work orders, so as to improve the work order processing efficiency. In some alternative embodiments, the work order processing tool can directly use the conclusion information and / or action information in the target rule hit by the dispute clue information as the work order processing guidance information corresponding to the dispute clue information. In other alternative embodiments, the work order processing tool can call a large language model to generate the work order processing guidance information corresponding to the dispute clue information according to the conclusion information and / or action information in the target rule hit by the dispute clue information.

[0077] Based on this implementation method, guidance information required to handle the target work order can be provided to customer service staff to reduce the risk of poor work order processing effects caused by customer service staff having less experience or relying too much on experience. In some alternative embodiments, the clue rules can be set according to the relatively new version of the dispute handling specification to automatically apply and promote the relatively new version of the dispute handling specification.

[0078] Optionally, the work order processing guidance information can be bound with action points. When the action points are triggered, other detailed information associated with the work order processing guidance information can be displayed. For example, in some alternative embodiments, the work order processing tool can respond to the trigger operation of the work order processing guidance information, jump to the third page, and display the dispute voucher data corresponding to the dispute clue information and / or the dispute handling specification information on the third page. Among them, the dispute voucher data corresponding to the clue information refers to at least one of the purchase view, product view, and evaluation view corresponding to the clue information, so as to more specifically present the dispute event from different perspectives. Among them, the dispute handling specification information is a preset systematic and standardized process and rule, which is used to ensure that the dispute cases can be handled reasonably and fairly.

[0079] Based on this implementation manner, when the work order processing guidance information is given, more detailed voucher information or specification information related to the work order processing guidance information can be provided to the customer service staff, which is convenient for the customer service staff to process the target work order more efficiently and reasonably.

[0080] In some alternative embodiments, the clue rules can be flexibly configured on the configuration interface. As Figure 6a shown, on the rule configuration interface, the clue rules for the buyer dimension, the clue rules for the buyer dimension, and the clue rules for the product dimension can be configured respectively. When configuring the clue rules, the accurate rules, suspected rules, and negotiation rules can be set respectively. Taking Figure 6a the clue rule configuration interface for the buyer dimension shown as an example, the conclusion of the rule is "buyer vouchers are unavailable", the conditions can be configured as: meeting any label in the malicious label list, the type can be configured as an accurate clue, and the corresponding clue action point can be configured. The clue action point is used to configure the action information corresponding to the clue rule to provide the guidance information required for the customer service staff to process the target work order. As Figure 6a shown, the "AND" icon is displayed in the condition configuration area of the clue rule. When the icon is triggered, a new condition input box can be added to configure multiple conditions, and the multiple conditions satisfy the "AND" logical relationship.

[0081] For example, the accurate clue rules for the buyer dimension can be configured as:

[0082] Buyer Rule 1 = {Condition: The number of buyer proofs > 0; Conclusion: The buyer has provided {the number of buyer proofs} vouchers; Action: Clue action point type: view, clue action point target: xxx}, as Figure 6b shown. On the Figure 6b configuration interface shown, other conditions or condition groups can also be added, which will not be illustrated here.

[0083] Buyer Rule 2 = {Condition: Hit any one of the specified list of tags; Conclusion: The compensation scheme based on the principle of taking the higher value cannot be used}.

[0084] The clue rules of the suspected type in the buyer dimension can be configured as:

[0085] Buyer Rule 3 = {Condition: Hit any one of the specified list of tags; Conclusion: Conduct a suspected case filing}.

[0086] The clue rules of the negotiation type in the buyer dimension can be configured as:

[0087] Buyer Rule 4 = {Condition: The number of orders paid by the buyer in 90 days, the refund rate in 90 days, the goods status, and the first-level commodity category respectively meet their corresponding conditions; Conclusion: The buyer has a positive contribution to the seller and is a core buyer of the store}. As Figure 6c shown, indicators such as the number of orders paid by the buyer in 90 days, the refund rate in 90 days, the goods status, and the first-level commodity category can be added on the configuration interface, and the conditions that each of the above indicators needs to meet can be configured respectively.

[0088] For example, the clue rules of the accurate type in the seller dimension can be configured as:

[0089] Seller Rule 1 = {Condition: The number of evidences provided by the seller > 0; Conclusion: The seller has provided {the number of evidences provided by the buyer} vouchers; Action: Clue action point type: view, clue action point target: xxx}.

[0090] The clue rules of the negotiation type in the seller dimension can be configured as:

[0091] Seller Rule 4 = {Condition: The number of penalties for the seller in the same category in the recent 90 days > the number of penalties for the same category in the recent 90 days; Conclusion: The number of penalties for the same commodity > 2 times}.

[0092] Another example, the clue rules of the accurate type in the commodity dimension can be configured as:

[0093] Commodity Rule 1 = {Condition: The refund amount is greater than the set threshold; Conclusion: Involves a large amount of refund; Action: Clue action point type: view, clue action point target: xxxx}.

[0094] Commodity Rule 2 = {Condition: The commodity hits the in-sale tag or the after-sale tag; Conclusion: The commodity is in the in-sale / after-sale refund state}.

[0095] Commodity Rule 3 = {Condition: The commodity hits the tag that supports 7 days after activation or use; Conclusion: Supports 7 days after activation or use}.

[0096] The clue rules of the suspected type in the commodity dimension can be configured as:

[0097] Commodity Rule 4 = {Condition: The commodity hits high / low risk labels; Conclusion: The commodity is a high / low risk commodity}.

[0098] The negotiable clue rules in terms of commodities can be configured as follows:

[0099] Commodity Rule 5 = {Condition: The number of negative reviews and the negative review rate of the commodity meet their respective corresponding thresholds; Conclusion: The negative review rate of the commodity is relatively high}.

[0100] Commodity Rule 6 = {Condition: The quantity of quality-related refunds and the quality-related refund rate of the commodity meet their respective corresponding thresholds; Conclusion: The negative review rate of the commodity caused by this quality is relatively high}.

[0101] Commodity Rule 7 = {Condition: The quantity of business responsibility disputes in the most recent 90 days, the ratio of business responsibility disputes in the most recent 90 days, and the refund reasons of the commodity meet their respective conditions; Conclusion: The quantity of disputes caused by the merchant for the commodity is relatively high.

[0102] Based on the above implementation manners, different-dimension and different-type clue rules can be flexibly configured on the configuration interface according to requirements, so as to meet the rule configuration requirements of various different scenarios.

[0103] Figure 7 Taking the processing of dispute work orders in the e-commerce scenario as an example, the overall architecture of the work order processing tool provided by the embodiments of the present application is schematically shown as Figure 7 shown. The auxiliary information provided by the work order processing tool to the customer service staff mainly includes: basic information, commodity knowledge, and key clues. Among them, the basic information includes: summary of the buyer's appeal, buyer's emotion level, and summary of the seller's solution. The key clues include: buyer's clues, seller's clues, and commodity clues, etc.

[0104] As Figure 7 shown, the work order processing tool can synchronously call a preset algorithm link to obtain basic information. Specifically, in this algorithm link, a service object can be called to obtain input parameter information in parallel. The input parameter information includes: order information, records of in / hotline services, buyer / seller chat records, transaction messages, action records, instant messaging messages, and other work order details information. The above work order details information can be obtained from at least one data domain among the order domain, dispute domain, refund domain, work order domain, transaction domain, lockdown domain, and code domain as Figure 7 shown. After obtaining the input parameter information, parameter processing operations such as formatting or factoring can be performed on the input parameter information to make the input parameter information meet the format requirements of the specified interface. Then, through this specified interface, the processed input parameters can be input into the large language model so that the large language model generates and returns the summary of the buyer's appeal, the buyer's emotion, and the summary of the seller's solution. In some other embodiments, the work order processing tool can also asynchronously call the above algorithm link to obtain the buyer's emotion and cache the buyer's emotion in the database, which is not shown.

[0105] As Figure 7 shown, the work order processing tool can asynchronously call the knowledge pre-generation link to obtain the target knowledge information corresponding to the work order. In the knowledge pre-generation link, the work order processing tool can call a large language model through a specified interface to generate product knowledge information. The product knowledge information generated by the large language model is stored in a message queue, and the server in the work order processing tool writes the product knowledge information stored in the message queue into the cache of the database for use.

[0106] As Figure 7 shown, the acquisition of key clues depends on the rule engine. The rule engine can batch execute clue matching operations according to pre-configured clue rules. The clue rules can include clue rules from the buyer dimension, seller dimension, and product dimension. The rule engine can perform rule matching on the dispute voucher data of the work order according to the above-mentioned multiple different dimension clue rules, and display the dispute voucher data that hits the clue rules as key clues.

[0107] Based on the above embodiments, the work order can be intelligently diagnosed, so as to assist customer service personnel in quickly obtaining the information required to resolve disputes, reducing the time spent on searching and thinking, and improving the processing efficiency of work orders.

[0108] It should be noted that the execution subject of each step of the method provided in the above embodiments can be the same device, or the method can also be executed by different devices as the execution subject. For example, the execution subject of steps 101 to 104 can be device A; another example is that the execution subject of steps 101 and 102 can be device A, and the execution subject of step 103 can be device B; and so on.

[0109] In addition, in some processes described in the above embodiments and the accompanying drawings, there are multiple operations that appear in a specific order, but it should be clearly understood that these operations can be executed not in the order in which they appear in this article or in parallel. The operation numbers such as 101 and 102 are only used to distinguish different operations, and the numbers themselves do not represent any execution order. In addition, these processes can include more or fewer operations, and these operations can be executed in sequence or in parallel. It should be noted that the descriptions such as "first" and "second" in this article are used to distinguish different messages, devices, modules, etc., and do not represent a sequence, nor do they limit that "first" and "second" are of different types.

[0110] In addition, it should be noted that in the case where the embodiments of the present application involve user interaction operations or triggering operations, the user interaction operations or triggering operations involved in the embodiments of the present application include, but are not limited to: interaction operations in various ways such as touch operations, gesture operations, voice operations, head movement operations, and eye movement operations; among them, touch operations include, but are not limited to: click operations, double-click operations, long-press operations, swipe operations, pinch operations, or mouse hover operations, etc. Swipe operations include, but are not limited to: straight-line swipes, curved swipes, etc.

[0111] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data that have been authorized by the user or fully authorized by all parties. And the collection, use, and processing of relevant data need to comply with the relevant laws, regulations, and standards of relevant countries and regions, and corresponding operation entrances are provided for users to choose to authorize or refuse.

[0112] Figure 8 Schematically shows a schematic structural diagram of an electronic device provided by an exemplary embodiment of the present application. This electronic device is applicable to the work order processing method provided in the foregoing embodiment. As Figure 8 shown, the electronic device includes: a memory 801, a processor 802, and a display component 803.

[0113] The memory 801 is used to store computer programs and can be configured to store various other data to support operations on the electronic device. Examples of these data include instructions for any application program or method for operating on the electronic device.

[0114] The processor 802 is coupled to the memory 801 and is used to execute the computer program in the memory 801 for: controlling the display component 803 to display a first page, the first page including a target work order to be processed, and the target work order is associated with target knowledge information required for processing the target work order generated by a first model; calling a second model to analyze the work order detail information corresponding to the target work order to obtain summary information corresponding to the target work order, the summary information including: appeal information and / or solution information corresponding to the appeal; controlling the display component 803 to display the target knowledge information and the summary information on the first page or a second page for processing the target work order according to the summary information and the target knowledge information, and the second page is different from the first page.

[0115] Optionally, the processor 802 is further configured to: before presenting the target work order on the first page, in response to an assignment operation for the target work order, identify a target question to be answered from the work order details information corresponding to the target work order; call the first model to answer the target question to obtain the target knowledge information.

[0116] Optionally, when the processor 802 identifies a target question to be answered from the work order details information corresponding to the target work order, it is specifically configured to: identify the product category corresponding to the target work order according to the work order details information corresponding to the target work order; if the product category is within a set category range, identify the question to be answered from the work order details information corresponding to the target work order; if the question to be answered is a question in a specific professional field, use the question to be answered as the target question to be answered.

[0117] Optionally, when the processor 802 calls the second model to analyze the work order details information corresponding to the target work order to obtain the summary information corresponding to the target work order, it is specifically configured to: when detecting a selection operation for the target work order, call the second model to analyze the work order details information corresponding to the target work order to obtain the summary information corresponding to the target work order.

[0118] Optionally, when the processor 802 calls the second model to analyze the work order details information corresponding to the target work order to obtain the summary information corresponding to the target work order, it is specifically configured to: call a specified service object according to the identifier of the target work order, so that the service object queries the description information of at least one dimension of the target work order to obtain the work order details information corresponding to the target work order; preprocess the work order details information, and call the second model to analyze the preprocessed work order details information to obtain the appeal information and / or the solution information corresponding to the appeal.

[0119] Optionally, the processor 802 is further configured to: when detecting a selection operation for the target work order, query a specified storage space according to the identifier information of the target work order to obtain the emotion classification information corresponding to the target work order, where the specified storage space is used to store the emotion classification information corresponding to different work orders; present the emotion classification information for processing the target work order according to the emotion classification information.

[0120] Optionally, before presenting the target work order on the first page, the processor 802 is further configured to: call a third model to analyze the work order details information corresponding to the target work order to obtain the emotion classification information corresponding to the target work order; store the emotion information corresponding to the target work order in the specified storage space.

[0121] Optionally, the processor 802 is further configured to: obtain the dispute voucher data corresponding to the target work order; match the dispute voucher data with the conditions in the preset clue rules to obtain partial detail data that hits the target rules in the clue rules; obtain different types of dispute clue information according to the partial detail data and the type of the target rules it hits; display the different types of dispute clue information for processing the target work order according to the different types of dispute clue information.

[0122] Optionally, in the clue rules, any rule includes conclusion information and / or action information corresponding to the conditions; the processor 802 is further configured to: for any dispute clue information in the different types of dispute clue information, obtain the work order processing guidance information corresponding to the dispute clue information according to the conclusion information and / or action information in the target rule hit by the dispute clue information; display the work order processing guidance information together with the dispute clue information.

[0123] Optionally, the processor 802 is further configured to: in response to a trigger operation on the work order processing guidance information, jump to a third page; display the dispute voucher data and / or dispute handling specification information corresponding to the dispute clue information on the third page.

[0124] Further, as Figure 8 shown, the electronic device further includes: other components such as a power supply component 804, a communication component 805, and an audio component 806. Figure 8 Only some components are schematically shown, which does not mean that the electronic device only includes Figure 8 the components shown. Figure 8 In, the components within the dashed box are optional components, not mandatory components, and can be determined according to the product form of the electronic device. The electronic device in this embodiment can be implemented as a terminal device such as a desktop computer, a laptop computer, a smart phone, or an IOT device, or can also be a server device such as a conventional server, a cloud server, or a server array. If the electronic device in this embodiment is implemented as a terminal device such as a desktop computer, a laptop computer, or a smart phone, it can include Figure 8 the components within the dashed box; if the electronic device in this embodiment is implemented as a server device such as a conventional server, a cloud server, or a server array, it may not include Figure 8 the components within the dashed box.

[0125] Among them, the memory 801 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0126] Among them, the communication component 805 is configured to facilitate communication between the device where the communication component is located and other devices in a wired or wireless manner. The device where the communication component is located can access a wireless network based on a communication standard, such as 2G (such as Global System for Mobile Communications (GSM), etc.), 3G (such as Wideband Code Division Multiple Access (WCDMA)), 4G (such as Long Term Evolution (LTE), etc.), 4G+ (such as LTE-Advanced (LTE-A), etc.) or 5G (5th Generation Mobile Communication Technology), or a combination thereof. In an exemplary embodiment, the communication component receives a broadcast signal or broadcast-related information from an external broadcast management system via a broadcast channel.

[0127] Among them, the power supply component 804 is used to provide power for various components of the device where the power supply component is located. The power supply component may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power for the device where the power supply component is located.

[0128] The display component includes a screen, which may include a Liquid Crystal Display (LCD) and a Touch panel (TP). If the screen includes a touch panel, the screen can be implemented as a touch screen to receive input signals from a user. The touch panel includes one or more touch sensors to sense touches, swipes, and gestures on the touch panel. The touch sensors can sense not only the boundaries of touch or swipe actions but also detect the duration and pressure associated with the touch or swipe operations.

[0129] The audio component can be configured to output and / or input audio signals. For example, the audio component includes a microphone (MIC), which is configured to receive external audio signals when the device where the audio component is located is in an operating mode, such as a call mode, a recording mode, and a voice recognition mode. The received audio signals can be further stored in the memory or sent via the communication component. In some embodiments, the audio component further includes a speaker for outputting audio signals.

[0130] In this embodiment, for the target work order to be processed, the target knowledge information required to process the target work order can be generated by the first model, and the work order details information corresponding to the target work order can be analyzed by the second model to obtain the summary information corresponding to the target work order. The target knowledge information and the summary information can be displayed on the page for processing the target work order based on the summary information and the knowledge information. Among them, the summary information includes: appeal information and / or solution information corresponding to the appeal, which can enable customer service staff to quickly understand the user's appeal and dispute points in the target work order. The knowledge information can help customer service staff quickly obtain the information required to resolve disputes, reduce the time spent on searching and thinking, and improve the processing efficiency of work orders.

[0131] Accordingly, an embodiment of the present application further provides a computer-readable storage medium storing a computer program, which when executed can implement each step executable by an electronic device in the above method embodiment. Among them, the computer-readable storage medium can be implemented by volatile or non-volatile or a combination thereof, and can be removable or non-removable. Examples of computer-readable storage media include, but are not limited to, phase-change random access memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette tapes, magnetic disk storage or other magnetic storage devices or any other non-transmission medium.

[0132] An embodiment of the present application further provides a computer program product, including: computer programs / instructions, which when executed by a processor can implement the steps in the method provided by the embodiment of the present application. It should be understood that each process or a combination of multiple processes in the above method flow can be implemented by computer programs or instructions. In addition, these computer programs or instructions can be applied to the processors of general-purpose computers, special-purpose computers, embedded processors or other programmable data processing devices, so that the processors of general-purpose computers, special-purpose computers, embedded processors or other programmable data processing devices can be used as devices to implement the corresponding functions in the above method embodiment.

[0133] It should also be noted that the term "including", "comprising" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, product or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or further includes elements inherent to such process, method, product or device. Without further limitation, an element defined by the statement "including one..." does not exclude the existence of another identical element in the process, method, product or device including the element.

[0134] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, various changes and modifications can be made to the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included within the scope of the claims of the present application.

Claims

1. A work order processing method, characterized in that: include: Displaying a first page, wherein the first page includes a target work order to be processed, and the target work order is associated with target knowledge information required for processing the target work order generated by a first model; Calling the second model to analyze the work order details information corresponding to the target work order to obtain summary information corresponding to the target work order, wherein the summary information includes: demand information and / or solution information corresponding to the demand; The target knowledge information and the summary information are displayed on the first page or the second page so as to process the target work order according to the summary information and the target knowledge information, and the second page is different from the first page.

2. The method according to claim 1, characterized in that Also includes: Before displaying the target work order on the first page, in response to an assignment operation on the target work order, identifying a target problem to be answered from work order detail information corresponding to the target work order; The first model is called to answer the target question to obtain the target knowledge information.

3. The method according to claim 2, characterized in that Identifying a target question to be answered from the work order details information corresponding to the target work order, including: Identify the commodity category corresponding to the target work order according to the work order details information corresponding to the target work order; If the product category is within the set category range, identifying the question to be answered from the work order details information corresponding to the target work order; If the question to be answered is a question in a specific professional field, the question to be answered will be used as the target question to be answered.

4. The method according to claim 1, characterized in that The second model is called to analyze the work order details information corresponding to the target work order to obtain summary information corresponding to the target work order, including: When a selection operation on the target work order is detected, the second model is called to analyze the work order detail information corresponding to the target work order to obtain summary information corresponding to the target work order.

5. The method according to claim 4, characterized in that The second model is called to analyze the work order details information corresponding to the target work order to obtain summary information corresponding to the target work order, including: Calling a specified service object according to the identifier of the target work order, so that the service object queries the description information of at least one dimension of the target work order to obtain the work order details information corresponding to the target work order; The work order detail information is preprocessed, and the second model is called to analyze the preprocessed work order detail information to obtain the demand information and / or solution information corresponding to the demand.

6. The method according to claim 1, characterized in that Also includes: When a selection operation on the target work order is detected, a designated storage space is queried according to the identification information of the target work order to obtain the emotion classification information corresponding to the target work order, and the designated storage space is used to store the emotion classification information corresponding to different work orders; The emotion classification information is displayed so as to process the target work order according to the emotion classification information.

7. The method according to claim 6, characterized in that Before displaying the target work order on the first page, the method further includes: Calling a third model to analyze the work order details information corresponding to the target work order to obtain the emotion classification information corresponding to the target work order; The emotion information corresponding to the target work order is stored in the designated storage space.

8. The method according to any one of claims 1 to 7, characterized in that: Also includes: Obtain dispute voucher data corresponding to the target work order; Matching the dispute voucher data with the conditions in the preset clue rules to obtain partial detail data that hits the target rules in the clue rules; Obtaining different types of dispute clue information according to the partial detail data and the type of target rule it hits; The different types of dispute clue information are displayed so as to process the target work order according to the different types of dispute clue information.

9. The method according to claim 8, characterized in that In the clue rules, any rule contains conclusion information and / or action information corresponding to a condition; the method further includes: For any dispute clue information among the different types of dispute clue information, obtaining work order processing guidance information corresponding to the dispute clue information according to conclusion information and / or action information in the target rule hit by the dispute clue information; The work order processing guidance information is displayed together with the dispute clue information.

10. The method according to claim 9, characterized in that Also includes: In response to a triggering operation on the work order processing guidance information, jumping to a third page; The dispute voucher data and / or dispute handling specification information corresponding to the dispute clue information are displayed on the third page.

11. An electronic device, characterized in that: include: Memory and processor; The memory is used to store one or more computer instructions; The processor is configured to execute the one or more computer instructions to: perform the steps in the method according to any one of claims 1-10.

12. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps in the method described in any one of claims 1 to 10 can be implemented.

13. A computer program product, characterized in that include: A computer program / instruction, which, when executed by a processor, can implement the steps of the method according to any one of claims 1 to 10.

Images