Complaint processing method and device

Abstract

The invention provides a complaint processing method and device, relates to the technical field of communication, and aims to accurately determine a complaint problem and a solution for the problem, effectively improve the efficiency of customer service in solving a complaint work order and improve the network awareness experience of a user. The method comprises the following steps: after a customer service accesses a complaint call of a user, acquiring call data of the complaint call in real time; processing the call data by adopting a natural language processing (NLP) technology to obtain key information of the complaint call; according to the key information and a remote dictionary service Redis knowledge base, determining a user complaint problem and a first solution corresponding to the complaint problem, the complaint problem being a complaint problem associated with the key information; and / or, according to the key information and the language model, determining a complaint problem of the user and a second solution corresponding to the complaint problem.

Description

Technical Field

[0001] This application relates to the field of communication technologies, and in particular, to a method and apparatus for handling complaints. Background Art

[0002] With the rapid development of operator services, the number of complaint work orders generated by users based on operator services is also increasing, bringing huge challenges to operator customer service. Facing a large number of complaint work orders, operator customer service may need to repeatedly answer the same question, or may not be able to quickly determine a solution to the problem based on the user's call content.

[0003] The current main means to support customer service work include intelligent customer service, speech libraries, problem explanation libraries, etc. However, these methods require customer service to retrieve keywords based on the call content and then obtain corresponding speech or solutions. Therefore, this method results in low efficiency of customer service in handling complaint work orders and poor user perception experience. Summary of the Invention

[0004] This application provides a method and apparatus for handling complaints, which can accurately determine the complaint problem and the corresponding solution for the problem, effectively improve the efficiency of customer service in handling complaint work orders, and enhance the user's network perception experience.

[0005] To achieve the above object, this application adopts the following technical solutions:

[0006] In a first aspect, this application provides a method for handling complaints. The method includes: after connecting to the user's complaint call, real-time obtaining the call data of the complaint call; using natural language processing (NLP) technology to process the call data to obtain the key information of the complaint call; determining the user's complaint problem and the corresponding first solution for the complaint problem according to the key information and a remote dictionary service (Redis) knowledge base, where the complaint problem is the complaint problem associated with the key information; and / or determining the user's complaint problem and the corresponding second solution for the complaint problem according to the key information and a language model.

[0007] Based on the above technical solution, after receiving a user's complaint call, the call data of the complaint call is obtained in real time, and NLP technology is used to process the call data to obtain the key information of the complaint call. Then, according to the key information, the language model, and the Redis knowledge base, the user's complaint problem and the corresponding solution to the complaint problem are determined. Since the call data is obtained in real time, correspondingly, the first solution corresponding to the complaint problem can be determined in real time, and / or the user's complaint problem and the second solution corresponding to the complaint problem can be determined in real time. In addition, since the second solution and / or the first solution can be determined, the solution to the complaint problem determined is more comprehensive and accurate. The solution of this application does not require customer service personnel to manually retrieve the solution to the complaint problem, effectively improving the efficiency of customer service in solving complaint work orders, and thus can improve the user's network perception experience.

[0008] A possible implementation manner is to determine the user's complaint problem and the first solution corresponding to the complaint problem according to the key information and the remote dictionary service Redis knowledge base, including: tagging the key information with target tags, and retrieving the target encoding corresponding to the target tags in the Redis knowledge base; according to the target encoding, determining the user's complaint problem and the first solution corresponding to the complaint problem.

[0009] A possible implementation manner is to determine the user's complaint problem and the second solution corresponding to the complaint problem according to the key information and the language model, including: inputting the key information into the language model, where the language model is determined based on the Bert model; outputting the user's complaint problem and the second solution corresponding to the complaint problem.

[0010] A possible implementation manner is that the method further includes: obtaining sample complaint work orders, where the sample complaint work orders include sample complaint problems and the corresponding solutions to the sample complaint problems; using NLP technology to process the sample complaint work orders, and based on the fuzzy clustering algorithm, performing clustering analysis on the sample complaint problems to obtain multiple sample key information corresponding to the sample complaint problems and the solutions corresponding to each sample key information; tagging the multiple sample key information and encoding the tags to obtain the encodings corresponding to the multiple tags; determining the Redis knowledge base, where the Redis knowledge base includes the encodings corresponding to the multiple tags and the solutions corresponding to each encoding.

[0011] A possible implementation manner is that the method further includes: based on the Bert model, using the multiple sample key information as the input and the solutions corresponding to each sample key information as the output, training to obtain a language model, where the language model is used to determine the user's complaint problem and the second solution corresponding to the complaint problem.

[0012] Second aspect, the present application provides a complaint processing device, which includes an acquisition module and a processing module; the acquisition module is used to acquire the call data of the complaint call in real time after the customer service connects to the user's complaint call. The processing module is used to process the call data by using natural language processing (NLP) technology to obtain the key information of the complaint call; the processing module is further used to determine the user's complaint problem and the corresponding first solution for the complaint problem according to the key information and the remote dictionary service (Redis) knowledge base, where the complaint problem is the complaint problem associated with the key information; and / or, the processing module is further used to determine the user's complaint problem and the corresponding second solution for the complaint problem according to the key information and the language model.

[0013] Third aspect, the present application provides a complaint processing device, which includes a processor coupled to a memory, and the processor is used to run a computer program or instruction to implement the method described in the first aspect and any possible implementation manner of the first aspect.

[0014] Fourth aspect, the present application provides a computer-readable storage medium, in which instructions are stored, and when the instructions run on the device, the computer is caused to execute the method described in the first aspect and any possible implementation manner of the first aspect.

[0015] Fifth aspect, the embodiments of the present application provide a computer program product containing instructions, and when the computer program product runs on the computer, the computer is caused to execute the method described in the first aspect and any possible implementation manner of the first aspect.

[0016] Sixth aspect, the embodiments of the present application provide a chip, which includes a processor coupled to a memory, and the processor is used to run a computer program or instruction to implement the method described in the first aspect and any possible implementation manner of the first aspect.

[0017] Among them, the technical effects brought by any possible implementation manner in the second aspect to the sixth aspect can be referred to the technical effects brought by the first aspect or different possible implementation manners in the first aspect, which will not be elaborated here. Description of the Drawings

[0018] Figure 1 It is a schematic diagram of the complaint processing process provided by the embodiments of the present application;

[0019] Figure 2 It is a flowchart of a complaint processing method provided by the embodiments of the present application;

[0020] Figure 3 It is a schematic diagram of the verification of the Redis knowledge base provided by the embodiments of the present application;

[0021] Figure 4 Schematic diagram of the trained language model provided by the embodiment of the present application;

[0022] Figure 5 Schematic diagram of the structure of a complaint processing device provided by the embodiment of the present application;

[0023] Figure 6 Schematic diagram of the structure of another complaint processing device provided by the embodiment of the present application;

[0024] Figure 7 Schematic diagram of the structure of a chip provided by the embodiment of the present application. Detailed implementation manners

[0025] The following describes in detail the complaint processing method and device provided by the embodiment of the present application with reference to the accompanying drawings.

[0026] The term "and / or" in this article is only a description of the association relationship of 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.

[0027] The terms "first" and "second" in the description and drawings of the present application are used to distinguish different objects or different processes for the same object, rather than to describe a specific order of the objects.

[0028] In addition, the terms "including" and "having" and any variations thereof mentioned in the description of the present application are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not limited to the listed steps or units, but optionally further includes other steps or units not listed, or optionally further includes other steps or units inherent to these processes, methods, products, or devices.

[0029] It should be noted that in the embodiments of the present application, words such as "exemplary" or "for example" are used to represent examples, illustrations, or explanations. Any embodiment or design solution described as "exemplary" or "for example" in the embodiments of the present application should not be construed as being more preferred or more advantageous than other embodiments or design solutions. Rather, the use of words such as "exemplary" or "for example" is intended to present relevant concepts in a specific manner.

[0030] It is recorded in the background technology that currently, when facing a user's complaint, a customer service needs to retrieve keywords based on the call content and then obtain corresponding conversation scripts or solutions. However, this method results in low efficiency in resolving complaint work orders and poor user perception experience.

[0031] To solve the above problems, an embodiment of the present application provides a method for handling complaints. In this method, after connecting to the user's complaint call, the call data of the complaint call can be obtained in real time; the natural language processing (NLP) technology is used to process the call data to obtain the key information of the complaint call; according to the key information and the language model, the user's complaint problem and the corresponding second solution to the complaint problem are determined; and / or, according to the key information and the remote dictionary service Redis knowledge base, the user's complaint problem and the corresponding first solution to the complaint problem are determined. In this way, the problem of the complaint and the solution to the problem can be accurately determined, the efficiency of the customer service in solving the complaint work order can be effectively improved, and the user's network perception experience can be enhanced.

[0032] It can be understood that the above method can be executed by any device with computing capabilities. For example, it can be executed by a server, a computer, a computing device, etc. (hereinafter referred to as the complaint handling device), or can be executed by a module in the complaint handling device, such as a chip, a circuit, a chip system, a logical node, or a logical module. Among them, the server includes but is not limited to: tower server, blade server, rack server, physical server, virtual host, virtual private server (VPS), cloud server, home server, enterprise server, etc.

[0033] An example is as Figure 1 shown. The user initiates a complaint call. After the customer service connects to the user's complaint call, the complaint handling device obtains the call data of the complaint call in real time, uses the natural language processing (NLP) technology to process the call data to obtain the key information of the complaint call, determines the user's complaint problem and the corresponding first solution to the complaint problem according to the key information and the remote dictionary service Redis knowledge base, and / or determines the user's complaint problem and the corresponding second solution to the complaint problem according to the key information and the language model, and presents the determined complaint problem and the solution corresponding to the complaint problem to the customer service.

[0034] Optionally, the complaint handling device can be configured with a display interface, which can be used to display the determined user's complaint problem and the corresponding second solution to the complaint problem, and / or the determined user's complaint problem and the corresponding first solution to the complaint problem, without limitation.

[0035] The following describes in detail the method for handling complaints provided by the embodiments of the present application with reference to the accompanying drawings.

[0036] As Figure 2 shown, a method for handling complaints provided by an embodiment of the present application includes the following steps:

[0037] S201. After the customer service answers the user's complaint call, obtain the call data of the complaint call in real time.

[0038] Among them, the call data of the complaint call may include the user's voice data during the complaint call, and may also include the customer service's voice data during the complaint call.

[0039] S202. Use natural language processing (NLP) technology to process the call data to obtain the key information of the complaint call.

[0040] Among them, the key information includes the keywords of the complaint problems or reasons presented in the context of the complaint call.

[0041] Exemplarily, taking the complaint call about the billing dispute of the phone number 186******** as an example, the complaint processing device uses NLP technology to process the call data (for example, word segmentation processing or semantic analysis processing), and the key information obtained for this complaint call includes billing rules, billing differences, and billing problems.

[0042] S203. Determine the user's complaint problem and the corresponding first solution to the complaint problem according to the key information and the Remote Dictionary Server (Redis) knowledge base.

[0043] Among them, the complaint problem is the complaint problem associated with the key information. For example, if the key information includes billing rules, the complaint problem is a billing problem. Another example, if the key information includes call interruption, the complaint problem is a network problem or a signal quality problem.

[0044] In the embodiment of the present application, the Redis knowledge base includes the encodings corresponding to the multiple tags and the solutions corresponding to each encoding.

[0045] A possible implementation manner is that the complaint processing device tags the key information with a target tag and retrieves the target encoding corresponding to the target tag in the Redis knowledge base; according to the target encoding, determine the user's complaint problem and the corresponding first solution to the complaint problem.

[0046] Exemplarily, taking the key information including charging rules, charging differences, and charging problems, where the target code corresponding to target label 1 is 1, the target code corresponding to target label 2 is 2, and the target code corresponding to target label 3 is 3 as an example, the complaint processing device tags the charging rules with target label 1, the charging differences with target label 2, and the charging problems with target label 3, and retrieves the target codes corresponding to the target labels in the Redis knowledge base; according to target code 1, target code 2, and target code 3, it determines that the user's complaint problem is "charging dispute" and the first solution corresponding to "charging dispute" is to view the historical prompt information sent by the operator to the user terminal. For example, as of * month * day * minute, you still have * minutes left in your domestic voice package.

[0047] A possible implementation method is that the complaint processing device obtains sample complaint work orders, processes the sample complaint work orders using NLP technology, and based on the fuzzy clustering algorithm, conducts cluster analysis on the sample complaint problems to obtain multiple sample key information corresponding to the sample complaint problems and the solutions corresponding to each sample key information; tags the multiple sample key information and encodes the tags to obtain the encodings corresponding to the multiple tags; determines the Redis knowledge base. Among them, the sample complaint work order includes the sample complaint problem and the solution corresponding to the sample complaint problem.

[0048] It can be understood that by using the fuzzy clustering algorithm to classify and process the sample complaint work orders, the complaint processing device can make more full and effective use of the sample complaint work orders, explore the value of the sample complaint work orders, reduce manual classification processing, and provide great help for the construction of the knowledge base. In addition, traditional means such as charts and experience are time-consuming and laborious to process sample complaint work orders with a high degree of discrete distribution, and there are problems of a large amount of repetitive labor and repetitive analysis. The labor cost is extremely high and the efficiency is low. Therefore, the fuzzy clustering algorithm can process abnormal sample complaint work orders or sample complaint work orders with more noise, improve the stability and reliability of the cluster analysis, and can also truly restore the modeling of the real world, with good interpretability and understandability. The fuzzy clustering algorithm can obtain the membership degree of each sample point to all class centers by optimizing the objective function, so as to determine the class belonging of the sample point to achieve the purpose of automatically classifying the samples.

[0049] Exemplarily, the complaint processing device can obtain sample complaint work orders from a work order system, a problem ledger, a production problem plan, etc., format and standardize the complaint work orders, and eliminate abnormal complaint work orders to remove noise, thus obtaining sample complaint work orders. Then, NLP technology is used to process the sample complaint work orders, and based on the fuzzy clustering algorithm, cluster analysis is performed on the sample complaint problems to obtain multiple sample key information corresponding to the sample complaint problems and the solutions corresponding to each sample key information. Each sample key information corresponds to a category of complaint problems. Then, the Redis knowledge base is determined.

[0050] It can be understood that the complaint processing device constructs a knowledge base engine with Redis as the core, tags multiple sample key information in the sample complaint work orders, encodes the tags to obtain the encodings corresponding to multiple tags to construct the Key values, and associates the key values with the solutions corresponding to the sample complaint problems, that is, the value corresponding to each key value is the solution corresponding to the sample complaint problem, thereby determining the Redis knowledge base.

[0051] Furthermore, the Redis knowledge base can be iteratively updated in real time based on the newly added complaint calls of users. Traditional knowledge bases have problems such as unclear keywords and inaccurate matching, especially for personnel with poor business capabilities, which are likely to cause off-topic answers. Compared with traditional knowledge bases, the Redis knowledge base is more accurate. The Redis knowledge base can support customer service personnel to associate complaint problems and the corresponding solutions based on users' complaint calls, enabling customer service to quickly and accurately answer users' questions.

[0052] In the embodiments of the present application, the complaint processing device can fully verify and test the Redis knowledge base based on the knowledge graph to ensure the accuracy of complaint problem classification and tagging. If there are deviations, the threshold is adjusted in a timely manner for repair. This solves the problems of data association, data semantics, and data intelligence in the Redis knowledge base, and provides rich industry knowledge models and knowledge reasoning components to realize the convergence, fusion, reasoning, and complex operations of data and knowledge.

[0053] Next, taking Figure 3 as an example, the verification and testing process of the Redis knowledge base based on the knowledge graph will be specifically introduced. In Figure 3Among them, the complaint processing device can perform attribute extraction, relationship extraction, or entity extraction on system pop-up reminder data, complaint work order data, and operation data to obtain key information, and perform noise processing on the key information and the corresponding solutions (also known as standard phrases) based on the knowledge graph, and determine weights (also known as quality calculation) for different solutions corresponding to the key information, and perform knowledge fusion and knowledge reasoning on other knowledge bases of the billing system, business system, accounting system, and customer service system to verify the key information and the corresponding solutions in the Redis knowledge base, so as to obtain an accurate Redis knowledge base.

[0054] S204. Determine the user's complaint problem and the corresponding second solution for the complaint problem according to the key information and the language model.

[0055] A possible implementation method is that the complaint processing device inputs the key information into the language model and outputs the user's complaint problem and the corresponding second solution for the complaint problem. Among them, the language model is determined based on the Bert model. The language model is used to determine the user's complaint problem and the corresponding second solution for the complaint problem.

[0056] Exemplarily, taking the key information including billing rules, billing differences, and billing problems as an example, the complaint processing device inputs the billing rules, billing differences, and billing problems into the language model, and outputs that the user's complaint problem is "billing dispute", and the corresponding second solution for "billing dispute" is to output the user's package situation and the usage situation of the package. For example, the package you subscribed to is *** yuan for *** minutes of calls and ** GB of traffic, the billing rule is ** tiered billing, the minimum guarantee fee is ** yuan, your consumption situation is **, where the effective method is ***, and there is a situation of combined billing with **.

[0057] A possible implementation method is that the complaint processing device trains a language model based on the Bert model, using multiple sample key information as input and the corresponding solutions for each sample key information as output.

[0058] Exemplarily, as Figure 4 shown, the complaint processing device selects multiple sample key information from the Redis knowledge base as the input feature vector of the Bert model based on the Bert model, and uses the corresponding solutions for each sample key information as output to train the language model. Optionally, the complaint processing device can also combine a convolutional neural network (CNN) to determine the output complaint problem and the solution for the complaint problem.

[0059] It can be understood that the complaint processing device can iteratively train and optimize the maturity and stability of the language model through the newly added complaint work orders. The language model originates from deep learning and has an absolute advantage in the application of the NLP field. By learning a large number of various patterns and structures of key information data of multiple samples, the language model finally aims to understand and generate human language and outputs solutions corresponding to the key information of each sample.

[0060] It can be understood that there is no sequence preference for the execution order of S203 and S204, and the present application does not limit the execution order of S203 and S204.

[0061] Furthermore, when the complaint processing device determines the user's complaint problem and the first solution and / or the second solution corresponding to the complaint problem, it displays the determined user's complaint problem and the first solution and / or the second solution (the first speech and / or the second speech) corresponding to the complaint problem to the customer service.

[0062] Based on Figure 2 According to the method shown above, after the complaint processing device accesses the user's complaint call, it can obtain the call data of the complaint call in real time, process the call data using NLP technology to obtain the key information of the complaint call, and determine the user's complaint problem and the solution corresponding to the complaint problem according to the key information, the language model, and the Redis knowledge base. Since the call data is obtained in real time, correspondingly, the user's complaint problem and the second solution corresponding to the complaint problem can be determined in real time, and / or the first solution corresponding to the complaint problem can be determined in real time. In addition, since the second solution and / or the first solution can be determined, the solution determined for the complaint problem is more comprehensive and accurate. The solution of the present application does not require customer service personnel to manually retrieve and determine the solution to the complaint problem, effectively improving the efficiency of customer service in solving complaint work orders, and thus can improve the user's network perception experience.

[0063] As Figure 5 As shown in the figure, it is a schematic structural diagram of a complaint processing device 50 provided by an embodiment of the present application. The complaint processing device 50 includes an acquisition module 501 and a processing module 502.

[0064] The acquisition module 501 is used to obtain the call data of the complaint call in real time after the customer service accesses the user's complaint call. For example: the acquisition module 501 is used to execute the above S201.

[0065] The processing module 502 is used to process the call data using natural language processing NLP technology to obtain the key information of the complaint call. For example: the processing module 502 is used to execute the above S202.

[0066] The processing module 502 is further configured to determine the user's complaint problem and a first solution corresponding to the complaint problem according to the key information and the Redis knowledge base. The complaint problem is the complaint problem associated with the key information. For example, the processing module 502 is used to execute S203 described above.

[0067] The processing module 502 is further configured to determine the user's complaint problem and a second solution corresponding to the complaint problem according to the key information and the language model. For example, the processing module 502 is used to execute S204 described above.

[0068] In a possible design, the processing module 502 is specifically configured to label the key information with a target label and retrieve a target code corresponding to the target label in the Redis knowledge base. The processing module 502 is further specifically configured to determine the user's complaint problem and a first solution corresponding to the complaint problem according to the target code.

[0069] In a possible design, the processing module 502 is further specifically configured to input the key information into the language model, and the language model is determined based on the Bert model. The processing module 502 is further specifically configured to output the user's complaint problem and a second solution corresponding to the complaint problem.

[0070] In a possible design, the acquisition module 501 is further configured to acquire sample complaint work orders, and the sample complaint work orders include sample complaint problems and solutions corresponding to the sample complaint problems. The processing module 502 is further configured to process the sample complaint work orders by using NLP technology, and perform clustering analysis on the sample complaint problems based on a fuzzy clustering algorithm to obtain multiple sample key information corresponding to the sample complaint problems and solutions corresponding to each sample key information. The processing module 502 is further configured to label the multiple sample key information and encode the labels to obtain codes corresponding to the multiple labels. The processing module 502 is further configured to determine the Redis knowledge base, and the Redis knowledge base includes the codes corresponding to the multiple labels and the solutions corresponding to each code.

[0071] In a possible design, the processing module 502 is further configured to use the multiple sample key information as input and the solutions corresponding to each sample key information as output based on the Bert model to train the language model, and the language model is used to determine the user's complaint problem and a second solution corresponding to the complaint problem.

[0072] It can be understood that the above-mentioned complaint processing device can also be implemented by hardware. For example, when implemented by hardware, the acquisition module 501 in the embodiments of the present application can be integrated on a communication interface, and the processing module 502 can be integrated on a processor. Another example is that when implemented by hardware, both the acquisition module 501 and the processing module 502 in the embodiments of the present application are integrated on the processor. The structure of the hardware can be as Figure 6 shown.

[0073] Figure 6 FIG. shows a schematic diagram of a possible hardware structure of the complaint processing device involved in the above embodiments. The complaint processing device includes: a processor 602. Optionally, the complaint processing device further includes: a communication interface 603, a memory 601, and a bus 604.

[0074] The processor 602 is used to control and manage the operations of the complaint processing device. For example, it executes the steps performed by the above-mentioned processing module 502, and / or is used to execute other processes of the technologies described herein. Optionally, the processor 602 can also execute the steps performed by the above-mentioned acquisition module 501. The above-mentioned processor 602 can be a central processing unit, a general-purpose processor, a digital signal processor, an application-specific integrated circuit, a field-programmable gate array, or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute various exemplary logical blocks, modules, and circuits described in connection with the disclosure of the present application. The processor can also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc.

[0075] The communication interface 603 is used to support the communication between the complaint processing device and other network entities. For example, it executes the steps performed by the above-mentioned acquisition module 501.

[0076] The memory 601 is used to store the program code and data of the complaint processing device. For example, the memory 601 can be the memory in the complaint processing device, etc. The memory can include volatile memory, such as random access memory; the memory can also include non-volatile memory, such as read-only memory, flash memory, hard disk, or solid-state drive; the memory can also include a combination of the above types of memory.

[0077] The bus 604 can be an Extended Industry Standard Architecture (EISA) bus, etc. The bus 604 can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 6It is represented only by a thick line, but it does not mean that there is only one bus or one type of bus.

[0078] Figure 7 It is a schematic structural diagram of a chip 70 provided by an embodiment of the present application. The chip 70 includes one or more than two (including two) processors 701. Optionally, the chip 70 further includes a communication interface 703, a bus 702, and a memory 704.

[0079] Among them, the above-mentioned processor 701 can implement or execute various exemplary logic blocks, units, and circuits described in connection with the disclosure of the present application. The processor can be a central processing unit, a general-purpose processor, a digital signal processor, an application-specific integrated circuit, a field-programmable gate array, or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute various exemplary logic blocks, units, and circuits described in connection with the disclosure of the present application. The processor can also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc.

[0080] The memory 704 can include a read-only memory and a random access memory, and provide operation instructions and data to the processor 701. A part of the memory 704 can also include a non-volatile random access memory (NVRAM).

[0081] In some embodiments, the memory 704 stores the following elements, execution modules, or data structures, or subsets thereof, or extended sets thereof.

[0082] In the embodiments of the present application, by calling the operation instructions stored in the memory 704 (the operation instructions can be stored in the operating system), corresponding operations are executed.

[0083] The memory 704 can include a volatile memory, such as a random access memory; the memory can also include a non-volatile memory, such as a read-only memory, a flash memory, a hard disk, or a solid-state drive; the memory can also include a combination of the above types of memories.

[0084] The bus 702 can be an Extended Industry Standard Architecture (EISA) bus, etc. The bus 702 can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 7 it is represented only by a single line, but it does not mean that there is only one bus or one type of bus.

[0085] From the description of the above embodiments, those skilled in the art can clearly understand that for the convenience and brevity of description, only the division of the above functional modules is used as an example. In actual applications, the above functions can be allocated to different functional modules according to needs, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. The specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0086] The embodiment of the present application provides a computer program product including instructions. When the computer program product runs on a computer, the computer is caused to execute the method in the foregoing method embodiment.

[0087] The embodiment of the present application further provides a computer-readable storage medium. Instructions are stored in the computer-readable storage medium. When the instructions run on a computer, the computer is caused to execute the method in the method flow shown in the foregoing method embodiment.

[0088] Among them, the computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples (non-exhaustive list) of the computer-readable storage medium include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a register, a hard disk, an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above, or any other form of computer-readable storage medium well-known in the art. An exemplary storage medium is coupled to the processor so that the processor can read information from the storage medium and write information to the storage medium. Of course, the storage medium can also be a component of the processor. The processor and the storage medium can be located in an application specific integrated circuit (ASIC). In the embodiment of the present application, the computer-readable storage medium can be any tangible medium that contains or stores a program, and the program can be used by or combined with an instruction execution system, apparatus, or device.

[0089] Since the complaint processing device, computer-readable storage medium, and computer program product in the embodiments of the present application can be applied to the above method, the technical effects that can be obtained can also refer to the method embodiments above, and are not described in detail in the embodiments of the present application.

[0090] In several embodiments provided by the present application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are only illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces, and the indirect couplings or communication connections of the devices or units can be in electrical, mechanical, or other forms.

[0091] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they can be located in one place, or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0092] In addition, in each embodiment of the present application, the functional units can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit.

[0093] The above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions within the technical scope disclosed in the present application should be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for handling complaints, characterized in that, the method includes: After the customer service connects to the user's complaint call, real-time obtain the call data of the complaint call; Use natural language processing NLP technology to process the call data to obtain the key information of the complaint call; According to the key information and the remote dictionary service Redis knowledge base, determine the user's complaint problem and the corresponding first solution for the complaint problem, where the complaint problem is the complaint problem associated with the key information; and / or, according to the key information and the language model, determine the user's complaint problem and the corresponding second solution for the complaint problem.

2. The method according to claim 1, characterized in that, The determining the user's complaint problem and the corresponding first solution according to the key information and the remote dictionary service Redis knowledge base includes: Label the key information with a target label, and retrieve the target encoding corresponding to the target label in the Redis knowledge base; According to the target encoding, determine the user's complaint problem and the corresponding first solution for the complaint problem.

3. The method according to claim 1, characterized in that, The determining the user's complaint problem and the corresponding second solution according to the key information and the language model includes: Input the key information into the language model, and the language model is determined based on the Bert model; Output the user's complaint problem and the corresponding second solution for the complaint problem.

4. The method according to any one of claims 1-3, characterized in that, the method further includes: Obtain sample complaint work orders, where the sample complaint work orders include sample complaint problems and the corresponding solutions for the sample complaint problems; Use NLP technology to process the sample complaint work orders, and based on the fuzzy clustering algorithm, perform clustering analysis on the sample complaint problems to obtain multiple sample key information corresponding to the sample complaint problems and the solutions corresponding to each sample key information; Label the multiple sample key information and encode the labels to obtain the encodings corresponding to the multiple labels; Determine the Redis knowledge base, where the Redis knowledge base includes the encodings corresponding to the multiple labels and the solutions corresponding to each encoding.

5. The method according to claim 4, characterized in that, the method further includes: Based on the Bert model, use the multiple sample key information as input and the solutions corresponding to each sample key information as output to train the language model, and the language model is used to determine the user's complaint problem and the corresponding second solution for the complaint problem.

6. A complaint handling device, characterized in that, the device includes: an acquisition module and a processing module; The acquisition module is used to, after the customer service connects to the user's complaint call, real-time obtain the call data of the complaint call; The processing module is used to process the call data by using natural language processing (NLP) technology to obtain the key information of the complaint call; The processing module is further used to determine the user's complaint problem and the corresponding first solution for the complaint problem according to the key information and the Redis knowledge base. The complaint problem is the complaint problem associated with the key information; and / or, The processing module is further used to determine the user's complaint problem and the corresponding second solution for the complaint problem according to the key information and the language model.

7. The device according to claim 6, wherein, The processing module is specifically used to label the key information with a target label and retrieve the target code corresponding to the target label in the Redis knowledge base; The processing module is further specifically used to determine the user's complaint problem and the corresponding first solution for the complaint problem according to the target code.

8. The device according to claim 6, wherein, The processing module is further specifically used to input the key information into the language model, and the language model is determined based on the Bert model; The processing module is further specifically used to output the user's complaint problem and the corresponding second solution for the complaint problem.

9. The device according to any one of claims 6-8, wherein, The acquisition module is further used to acquire sample complaint work orders, and the sample complaint work orders include sample complaint problems and the corresponding solutions for the sample complaint problems; The processing module is further used to process the sample complaint work orders by using NLP technology and perform clustering analysis on the sample complaint problems based on a fuzzy clustering algorithm to obtain multiple sample key information corresponding to the sample complaint problems and the corresponding solutions for each sample key information; The processing module is further used to label the multiple sample key information and encode the labels to obtain the codes corresponding to the multiple labels; The processing module is further used to determine the Redis knowledge base, and the Redis knowledge base includes the codes corresponding to the multiple labels and the corresponding solutions for each code.

10. The device according to claim 9, wherein, The processing module is further used to train the language model by using the multiple sample key information as input and the corresponding solution for each sample key information as output based on the Bert model. The language model is used to determine the user's complaint problem and the corresponding second solution for the complaint problem.

11. A device for processing complaints, wherein, comprises: a processor; The processor is coupled with a memory, and the memory is used to store programs or instructions. When the programs or instructions are executed by the processor, the device for processing complaints executes the method according to any one of claims 1 to 5.

12. A computer-readable storage medium, in which instructions are stored, wherein, When the computer executes the instructions, the computer performs the method of any one of claims 1 to 5.

Images