Customer service scheduling method, system, terminal device and storage medium

Abstract

The present invention relates to the field of financial technology, and discloses a method, system, terminal device, and computer storage medium for scheduling customer service. The customer service scheduling method obtains a business request to be served and extracts request parameters from the business request; obtains the target service processing logic from a preset distributed configuration center according to the request parameters, wherein the distributed configuration center is configured with the service processing logic of all business products; dynamically loads each business module code file according to the target service processing logic; and executes the business module code file to schedule the customer service corresponding to the business request. The present invention solves the technical problem that the scheduling business development and operation and maintenance process in the traditional way is time-consuming and cannot meet the business side's needs for rapid response to customer service scheduling.

Description

Technical Field

[0001] The present invention relates to the technical field of financial technology (Fintech), and in particular to a scheduling method, system, terminal equipment and computer storage medium for customer service. Background Art

[0002] With the development of computer technology, more and more technologies are being applied in the financial field. Traditional finance is gradually transforming into financial technology. However, due to the security, real-time and stability requirements of the financial industry, higher requirements are also placed on technology.

[0003] At present, the scheduling of providing corresponding customer services for different business products is carried out in the form of a combination of hard-coded redis (remote dictionary service, an open source log-type, "Key-Value" database written in ANSI C language, supporting the network, and can be based on memory or persistent, and providing APIs (Application Programming Interface, application program interface) in multiple languages) message queues and state machines. In this way, not only is it necessary to carry out the development, testing and online deployment process of a full set of service processing logic for each different business product, but even when the difference between the two business products is small, it is still necessary to develop separate service processing logic for the two business products and perform test deployment. In addition, in the later operation and maintenance process, if the business side needs to adjust the service processing logic and thus modify the corresponding parameters, it is necessary to rewrite the code and test it. In this way, the existing customer service scheduling method that uses a combination of hard-coded redis message queues and state machines is time-consuming in the scheduling business development and operation and maintenance process, and cannot meet the needs of the business side to quickly respond to customer service scheduling. Summary of the invention

[0004] The main purpose of the present invention is to provide a customer service scheduling method, system, terminal device and computer storage medium, aiming to solve the technical problems that the existing customer service scheduling method that uses a hard-coded redis message queue combined with a state machine is time-consuming in the scheduling business development and operation and maintenance process, and cannot meet the business side's technical problems of quickly responding to customer service scheduling needs.

[0005] To achieve the above object, the present invention provides a method for scheduling customer service, the method comprising:

[0006] Obtaining a service request to be serviced and extracting request parameters from the service request;

[0007] Obtaining the target service processing logic from a preset distributed configuration center according to the request parameters, wherein the distributed configuration center is configured with the service processing logic of all business products;

[0008] Dynamically load each business module code file according to the target service processing logic;

[0009] The business module code file is executed to schedule customer service corresponding to the business request.

[0010] In addition, to achieve the above-mentioned purpose, the present invention also provides a customer service scheduling system, the customer service scheduling system comprising:

[0011] An acquisition module, used to acquire a business request to be served and extract request parameters from the business request, and acquire a target service processing logic from a preset distributed configuration center according to the request parameters, wherein the distributed configuration center is configured with the service processing logic of all business products;

[0012] A code dynamic loading module, used to dynamically load each business module code file according to the target service processing logic;

[0013] The customer service scheduling module is used to execute the business module code file to schedule the customer service corresponding to the business request.

[0014] Among them, each functional module of the customer service scheduling system of the present invention implements the steps of the customer service scheduling method described above during operation.

[0015] In addition, to achieve the above-mentioned purpose, the present invention also provides a terminal device, which includes: a memory, a processor, and a customer service scheduling program stored on the memory and executable on the processor, and the customer service scheduling program, when executed by the processor, implements the steps of the customer service scheduling method described above.

[0016] In addition, to achieve the above objectives, the present invention also provides a computer storage medium, on which a customer service scheduling program is stored, and when the customer service scheduling program is executed by a processor, the steps of the customer service scheduling method described above are implemented.

[0017] In addition, to achieve the above-mentioned purpose, the present invention also provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, the steps of the scheduling method for customer service as described above are implemented.

[0018] The present invention provides a scheduling method, system, terminal device, computer storage medium and computer program product for customer service, which obtains a business request to be served and extracts request parameters from the business request; obtains target service processing logic from a preset distributed configuration center according to the request parameters, wherein the distributed configuration center is configured with the service processing logic of all business products; dynamically loads each business module code file according to the target service processing logic; and executes the business module code file to schedule the customer service corresponding to the business request.

[0019] In the process of scheduling customer services for business requests initiated by customers for any business products, the present invention receives business requests initiated by any customers and extracts request parameters from the business requests through a terminal device configured as a scheduling system. Then, based on the request parameters, the terminal device obtains the target service processing logic of the business product corresponding to the service request initiated by the customer from a distributed configuration center that is pre-configured with the service processing logics of all business products. Then, the terminal device dynamically loads the business module code file required to process the service request according to the target service processing logic. Finally, the terminal device executes the dynamically loaded business module code file to schedule the customer service that matches the business request to perform the corresponding service operation.

[0020] Compared with the traditional method of combining hard-coded redis message queues with state machines for customer service scheduling, the present invention configures the service processing logic of all business products in a distributed configuration center, so that when part or all of the service processing logic of a business product needs to be updated, or the service processing logic of a business product needs to be superimposed with the same logic of another business product, the configuration file corresponding to the service processing logic can be modified directly in the distributed configuration center, that is, the service processing logic can be dynamically adjusted without the need for complex processes such as redevelopment, testing and online deployment of the entire scheduling system, thereby ensuring the responsiveness of customer service scheduling for business requests and solving the technical problems that the development and operation and maintenance processes of scheduling business in the traditional method are time-consuming and cannot meet the needs of the business side to quickly respond to customer service scheduling. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 A schematic diagram of the device structure of the hardware operating environment of the terminal device involved in the embodiment of the present invention;

[0022] Figure 2 A flow chart of a first embodiment of a scheduling method for serving customers of the present invention;

[0023] Figure 3 A schematic diagram of a system framework involved in an embodiment of a scheduling method for serving customers of the present invention;

[0024] Figure 4 The protocol code for dynamically adapting the service processing logic involved in the first embodiment of the scheduling method for serving customers of the present invention;

[0025] Figure 5 A protocol code for dynamically adapting a customer service matching model according to an embodiment of a scheduling method for customer service of the present invention;

[0026] Figure 6 Another schematic diagram of a system framework involved in an embodiment of a scheduling method for serving customers of the present invention;

[0027] Figure 7 A schematic diagram of another system framework involved in an embodiment of a scheduling method for serving customers of the present invention;

[0028] Figure 8 A schematic diagram of the functional modules of an embodiment of a scheduling system for serving customers of the present invention.

[0029] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0030] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0031] Reference Figure 1 , Figure 1 The present invention is a schematic diagram of the device structure of the hardware operating environment of the terminal device involved in the embodiment of the present invention.

[0032] The terminal device of the embodiment of the present invention may be a client device configured as a scheduling system to perform customer service scheduling, and the terminal device may be a smart phone, a PC (Personal Computer), a tablet computer, a portable computer, etc.

[0033] like Figure 1As shown, the terminal device may include: a processor 1001, such as a CPU, a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. Among them, the communication bus 1002 is used to realize the connection and communication between these components. The user interface 1003 may include a display screen (Display), an input unit such as a keyboard (Keyboard), and the optional user interface 1003 may also include a standard wired interface and a wireless interface. The network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a Wi-Fi interface). The memory 1005 may be a high-speed RAM memory, or it may be a stable memory (non-volatile memory), such as a disk memory. The memory 1005 may optionally be a storage device independent of the aforementioned processor 1001.

[0034] Those skilled in the art will understand that Figure 1 The terminal device structure shown in the figure does not constitute a limitation on the terminal device, and may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently.

[0035] like Figure 1 As shown, the memory 1005 as a computer storage medium may include an operating system, a network communication module, a user interface module, and a scheduling program for customer service.

[0036] exist Figure 1 In the terminal shown, the network interface 1004 is mainly used to connect to the background server and communicate data with the background server; the user interface 1003 is mainly used to connect to the client and communicate data with the client; and the processor 1001 can be used to call the customer service scheduling program stored in the memory 1005, and execute the following embodiments of the customer service scheduling method.

[0037] Based on the above hardware structure, various embodiments of the scheduling method for customer service of the present invention are proposed.

[0038] It should be noted that, at present, the scheduling of providing corresponding customer services for different business products is carried out in the form of a combination of a hard-coded redis (remote dictionary service, an open source log-type, "Key-Value" database written in ANSI C language, supporting the network, and can be based on memory or persistent, and providing APIs (Application Programming Interface) in multiple languages) message queue and a state machine. In this way, not only is it necessary to carry out the development, testing and online deployment process of a full set of service processing logic for each of the different business products, but even when the difference between the two business products is small, it is still necessary to develop separate service processing logic for the two business products and perform test deployment. In addition, in the later operation and maintenance process, if the business side needs to adjust the service processing logic to modify the corresponding parameters, it is necessary to rewrite the code and test it. In this way, the existing customer service scheduling method that uses a hard-coded redis message queue combined with a state machine is time-consuming in the scheduling business development and operation and maintenance process, and cannot meet the needs of the business side to quickly respond to customer service scheduling.

[0039] In view of the above phenomenon, the present invention provides a method for scheduling customer service. Figure 2 , Figure 2 A flow chart of a first embodiment of a scheduling method for serving customers of the present invention. In this embodiment, the scheduling method for serving customers is applied to the terminal device configured as a scheduling system. The scheduling method for serving customers of the present invention includes:

[0040] Step S10, obtaining a service request to be serviced and extracting request parameters from the service request;

[0041] The terminal device configured as the dispatching system first receives the service request initiated by the customer for the service product through the unified gateway interface, and then parses the received service request to extract the request parameters encapsulated in the service request.

[0042] It should be noted that, in this embodiment, any service request received by the terminal device includes at least: the customer identification of the customer initiating the service request, the product identification of the service product targeted by the customer, and the question type of the question corresponding to the service request that requires the customer service to provide an answer.

[0043] Specifically, for example, please refer to Figure 3The system framework shown in the figure, in which the business system represents various products that require customer service (referred to as customer service) to provide corresponding service support, and the system framework also includes a terminal device configured as a scheduling system, a gateway interface provided to customers for inputting request parameters to initiate business requests, a distributed configuration center configured with the service processing logic of all business products, a file storage space for storing specific business module code files, a model storage space for storing customer service matching models, and a feature library for storing customer features and customer service features.

[0044] Under this system framework, all business requests initiated by customers for any business products are input by calling the unified gateway interface: clientGateWay. When the customer inputs the request parameters based on the gateway interface: clientGateWay to initiate a business request, he or she needs to at least enter his or her own customer identification - customer number, the product identification of the business product - product number, and the type of question that needs to be answered by the customer service (typed in an input box with a drop-down option). Therefore, after the terminal device receives the business request encapsulated with the user number, product number and question type through the gateway interface: clientGateWay, it parses the business request based on any mature instruction parsing method to extract the request parameters: customer number, product number and question type.

[0045] Step S20, obtaining the target service processing logic from a preset distributed configuration center according to the request parameters, wherein the distributed configuration center is configured with the service processing logic of all business products;

[0046] It should be noted that, in the present embodiment, the distributed configuration center in the above-mentioned system architecture is configured with the service processing logic of all business products respectively, by abstracting the service processing logic common to all business products (for example, user information query, user product query and user authority verification, etc.) into business module files, and then dynamically configuring the business module files by the distributed configuration center, thereby dynamically loading the code module corresponding to the service processing logic. In addition, since the service processing logic common to business products has been configured in the distributed configuration center, when it is necessary to add the service processing logic of a new business product in the future, it is only necessary to configure the service processing logic of the new service product, except for the common service processing logic. In this way, the configuration efficiency of the service processing logic of the business product can be effectively improved, thereby further ensuring the response enthusiasm of customer service scheduling for service requests initiated by business products.

[0047] It should be understood that, based on different design requirements of actual applications, in different feasible implementations, the distributed configuration center is of course also configured with any other parameter configurations required for scheduling customer service. The scheduling method for customer service of the present invention only limits all parameter configurations used to schedule customer service to be configured in the distributed configuration center, and the scheduling method for customer service of the present invention does not make specific limitations on the specific content of the parameter configuration.

[0048] After the terminal device receives a business request initiated by a customer that requires support from the customer service and extracts the request parameters encapsulated in the business request, it adapts the target service processing logic of the business product to which the business request initiated by the customer belongs based on the request parameters in the service processing logic of all business products configured in the distributed configuration center.

[0049] Specifically, for example, please refer to Figure 3 The system framework shown in the figure, the distributed configuration center is pre-configured as shown in the attached Figure 4 The protocol code shown in FIG. 1 specifies the service processing logic of two business products, product A and product B. In the protocol code, productCode refers to the number of the business product. Each business product has a different number (such as "A" and "B"). productBiz refers to the specific service processing logic of the business product. Among them, bizCode refers to the number of the business module code. The terminal configured as the scheduling system relies on this number to match and load the corresponding code block. bizName refers to the business name of the code block. bizOrder refers to the execution order of the business module code: 1 is the first module executed, 2 is the module executed after 1, and so on. bizError has two enumeration values, namely break and continue. Among them, break means interrupting the program and returning directly when a system exception occurs. Continue means that when an exception occurs in the business module, the judgment of the code module can be skipped and the processing of the next code module can be directly performed without directly interrupting the thread to return an error code.

[0050] In this way, after the terminal device receives the business request initiated by the customer and extracts the request parameters, assuming that the customer number in the request parameters is the customer ID: xxx…3 (x is any positive integer greater than or equal to 0) and the product number is A, the terminal device first pulls the service processing logic of all products from the distributed configuration center, and then matches the product number A with the value "A" of the productCode field. When they are determined to be equal, the corresponding productBiz field is read as the target service processing logic of product A to which the business request initiated by the current customer belongs.

[0051] It should be noted that, in this embodiment, the target service processing logic—productBiz is an array, and each object in the array is a set of business logic. The terminal device first sorts according to the size value of bizOrder, and arranges the code modules to be executed in order from small to large. For example, according to the target service processing logic of product A—productBiz, the terminal device arranges the business module codes to be executed in order according to the size value of bizOrder corresponding to the number: checkUserId, complaintHistory, customerComplaintHistory...

[0052] Step S30, dynamically loading each business module code file according to the target service processing logic;

[0053] After the terminal device obtains the target service processing logic from the distributed configuration center according to the request parameters, it begins to process the business request initiated by the customer according to the target service processing logic. In the process of processing the business request, it dynamically loads each business module code file in sequence according to the number of the business module code in the target service processing logic.

[0054] It should be noted that, in this embodiment, the business module code file is an abstraction of some common service processing logics (for example, user information query, user product query, and user authority verification, etc.). The terminal device configured as the scheduling system abstracts these common service processing logics into configuration files and configures them in the distributed configuration center, so that it can rely on the spi mechanism (Spi mechanism refers to Service Provider Interface: it actually originates from the service provider framework (Service Provider Framework), which is a mechanism that separates the service interface from the service implementation to achieve decoupling and greatly improves the scalability of the program. Introducing a service provider means introducing the implementer of the spi interface, and obtaining the specific implementation class through local registration discovery, which is easily pluggable) to dynamically load the code module.

[0055] Specifically, for example, please refer to Figure 3The system framework shown is configured as a terminal device of the scheduling system that matches the product number A with the value "A" of the productCode field, thereby reading the corresponding productBiz field as the target service processing logic of product A to which the business request initiated by the current customer belongs, and according to the target service processing logic of product A—productBiz, the business module codes to be executed in sequence according to the size values ​​of bizOrder are: checkUserId, complaintHistory, customerComplaintHistory..., that is, according to the business module code numbers: checkUserId, complaintHistory, customerComplaintHistory..., in the file storage space where the specific business module code files are stored in the distributed configuration center, the business module code files matching the business module code numbers are dynamically loaded into the system (such as a java system) application, and the business module code files are executed in sequence.

[0056] It should be noted that, in this embodiment, the terminal device configured as the scheduling system dynamically loads the code file by adopting the Java SPI mechanism. Specifically, for example, the terminal device configured as the scheduling system implements dynamic loading of the business module code file according to the configuration file pre-added by the staff for controlling the loading of the business module code file through the custom programming of the interface and the preset strategy mode by the staff. In addition, the terminal device configured as the scheduling system also implements service search through the tool class "java.util.ServiceLoader" provided in JDK (Java Development Kit (JDK) is a software development kit for Java developers.).

[0057] Furthermore, in this embodiment, the custom programming of the interface in the terminal device configured as the scheduling system is obtained by encoding based on an interface-oriented programming method.

[0058] It should be noted that since the abstract code modules often have many different implementation schemes, such as the common log module scheme, XML parsing module, JDBC driver module, etc., the staff, in the object-oriented design concept, encodes based on the interface-oriented programming method on the interface in the terminal device configured as the scheduling system to achieve the connection between the various code modules, rather than directly hard-coding the implementation class.

[0059] Furthermore, in this embodiment, the terminal device configured as the scheduling system realizes the discovery of the service request initiated by the client based on the SPI technology.

[0060] It should be noted that, in this embodiment, SPI technology is a mechanism provided in JAVA for finding a service implementation class for a certain interface.

[0061] Specifically, for example, the terminal device configured as the scheduling system adopts the IOC concept similar to that in the Spring framework (i.e., the control of program loading and assembly is moved outside the program), and through the agreement in the JAVA SPI mechanism, when the service provider (such as a new log component) provides a certain implementation of the service interface, a file named after the service interface is created in the META-INF / services / directory of the jar package, and the fully qualified class name of the specific implementation class that implements the service interface is filled in the file. In this way, when the external program assembles the code module, the specific implementation class path can be found through the configuration file in the META-INF / services / directory of the jar package, so that the module can be loaded and instantiated through the reflection mechanism, thereby completing the injection of the module.

[0062] Step S40: executing the business module code file to schedule customer service corresponding to the business request.

[0063] After dynamically loading each business module code file, the terminal device executes the business module file, and when executing the module code file for allocating customer service, it further adapts and schedules the corresponding customer service based on the customer identifier in the request parameter to provide service support for the business request.

[0064] It should be noted that, in this embodiment, please refer to Figure 3 The system framework shown in Figure 4 The protocol code shown configures the terminal device of the scheduling system. When executing the module code files corresponding to the code module numbers checkUserId, complaintHistory and / or customerComplaintHistory..., if an exception occurs, it is directly determined according to the bizError configured in the protocol code whether to skip the current exception and continue execution or to directly return the exception information at the current processing node.

[0065] Further, in a feasible embodiment, after the above step S40, the method for scheduling customer service of the present invention may further include:

[0066] Step S50, obtaining a modification request for the target service processing logic;

[0067] After the terminal device dispatches the customer service to provide corresponding service support for the business request initiated by the customer, it further obtains a modification request initiated by the customer based on his own needs to modify the target service processing logic of the business product to which the business request belongs.

[0068] It should be noted that in this embodiment, after the customer service provides the corresponding service support to the customer, the customer may raise the demand to delete the contact history with the customer service or other demands, so the terminal device configured as the scheduling system generates a modification request for modifying the target service processing logic of the business product to which the business request belongs based on the encapsulation of the demand. It should be understood that based on the different design needs of actual applications, in different feasible implementation methods, the customer can of course also raise other demands during or after the customer service provides the corresponding service support, and the scheduling method of the customer service of the present invention is not limited to the specific type of the modification request.

[0069] Step S60: modify the target service processing logic in the distributed configuration center according to the modification request to obtain a new service processing logic.

[0070] After receiving the modification request initiated by the customer, the terminal device modifies the target service processing logic of the business product to which the business request initiated by the customer belongs, which has been configured by the distributed configuration center, according to the customer requirements encapsulated in the modification request, and thereby obtains the new service processing logic of the business product after the modification.

[0071] Specifically, for example, after the terminal device configured as the scheduling system schedules the customer service to provide corresponding service support for the business request initiated by the customer for business product A, the customer and / or the development staff of business product A submits a modification request to delete the contact history between the customer and the customer service. The terminal device receives the modification request and directly modifies the parameter configuration related to the service processing logic of business product A on the distributed configuration center, and removes the id of the contact history module code, so that the new service processing logic of the modified business product A is changed from the original "first call the user authority verification code module, then perform the user transaction record query code module, then perform the user contact history query code module, and finally enter the customer service allocation code module" to: first call the user authority verification code module, then perform the user transaction record query code module, and finally enter the direct customer service allocation code module.

[0072] The embodiment of the present invention provides a scheduling method for customer service, wherein a terminal device configured as a scheduling system first receives a service request initiated by a customer for a service product through a unified gateway interface, and then parses the received service request to extract the request parameters encapsulated in the service request; after receiving the service request initiated by the customer and requiring support from the customer service and extracting the request parameters encapsulated in the service request, the terminal device adapts the target service processing logic of the service product to which the service request initiated by the customer belongs in the service processing logics of all the service products configured in the distributed configuration center according to the request parameters; after the terminal device obtains the target service processing logic from the distributed configuration center according to the request parameters, it starts to process the service request initiated by the customer according to the target service processing logic, and in the process of processing the service request, dynamically loads each service module code file in sequence according to the number of the service module code in the target service processing logic; after dynamically loading each service module code file, the terminal device executes the service module file, and when executing the module code file for allocating customer service, further adapts and schedules the corresponding customer service based on the customer identifier in the request parameter and the preset customer service matching model to provide service support for the service request.

[0073] In addition, after the terminal device dispatches the customer service to provide corresponding service support for the business request initiated by the customer, it further obtains the modification request initiated by the customer based on the customer's own needs to modify the target service processing logic of the business product to which the business request initiated by the customer belongs; after obtaining the modification request initiated by the customer, the terminal device modifies the target service processing logic of the business product to which the business request initiated by the customer belongs, which has been configured by the distributed configuration center, in accordance with the customer's needs encapsulated in the modification request, and thereby obtains the new service processing logic of the business product after the modification.

[0074] Compared with the traditional method of combining hard-coded RADIUS message queue with state machine for customer service scheduling, the present invention configures the service processing logic of all business products in a distributed configuration center, so that when part or all of the service processing logic of a business product needs to be updated, or the service processing logic of a business product needs to be superimposed with the same logic of another business product, the configuration file corresponding to the service processing logic can be modified directly in the distributed configuration center, that is, the service processing logic can be dynamically adjusted without the need for complex processes such as redevelopment, testing and online deployment of the entire scheduling system, thereby ensuring the responsiveness of customer service scheduling for business requests and solving the technical problems that the development and operation and maintenance processes of scheduling business in the traditional method are time-consuming and cannot meet the needs of the business side to quickly respond to customer service scheduling.

[0075] Further, based on the above first embodiment, a second embodiment of the method for scheduling customer service of the present invention is proposed. The main difference between this embodiment and the above first embodiment is that, in this embodiment, each of the business module code files includes a module code file for allocating customer service, and the above step S40 may include:

[0076] Step S401, executing the module code file for allocating customer service to load a preset customer service matching model, wherein the customer service matching model is obtained by deep learning training based on customer characteristics and customer service characteristics;

[0077] Step S402: Adapt the customer service corresponding to the business request using the customer service matching model, and schedule the customer service.

[0078] It should be noted that in this embodiment, the customer service matching model is a deep learning algorithm model based on tensorFlow, and the calculation of each customer service matching model is based on the above Figure 3 In the system framework shown, multiple feature values ​​in the feature library (including feature values ​​of customer features and feature values ​​of customer service features) are calculated. In addition, the mathematical model of each model is slightly different and the focus of each model is also different. Different business products also provide customer service scheduling services for business requests initiated by customers for business products by configuring different models.

[0079] In addition, the number of customer service matching models is greater than one, e.g. Figure 3 Within the system framework shown, the customer service matching model can specifically include: customer service response rate model, customer service satisfaction rate model, customer service professionalism model, customer service waiting time model, user contact history model, user behavior model, user information model and user waiting time model. It should be understood that based on the different design needs of actual applications, in different feasible implementation methods, within the system framework where the client terminal configured as the scheduling system is located, the customer service matching model may of course also include other algorithm models not listed here. The scheduling method for customer service of the present invention is not limited to the specific type of the customer service matching model.

[0080] After dynamically loading each business module code file, the terminal device executes the business module file, and when executing the module code file for allocating customer service, it further adapts and schedules the corresponding customer service based on the customer identifier in the request parameters and the preset customer service matching model to provide service support for the business request.

[0081] Furthermore, in a feasible embodiment, the above step S401 may include:

[0082] Step S4011, adapting the algorithm model number in the module code file for allocating customer service according to the product identification;

[0083] Step S4012, determining the target customer service matching model corresponding to the algorithm model number in each of the customer service matching models;

[0084] During the process of dynamically loading various business module code files, when the terminal device executes the module code file for allocating customer services in the various business module code files, it first uses the product identifier extracted from the request parameters by parsing the business request initiated by the customer, adapts the algorithm model number in the module code file, and determines the target customer service matching model corresponding to the algorithm model number among multiple customer service matching models.

[0085] Specifically, for example, please refer to Figure 3 The system framework shown in the figure, the distributed configuration center is also pre-configured as shown in the attached Figure 5 As shown in the protocol code, ProductCode represents the product number, learningModel represents the algorithm model number, and the protocol code stipulates that the business product—productA calls skillModel (customer service skill model), and productB calls userInfoModel user information model. modelConfig refers to the configuration information of the algorithm model, because the same business product and the same algorithm model can run multiple versions at the same time (this can quickly benchmark the effect of the algorithm model), modelVersion refers to the version of the algorithm model, and flow refers to the traffic using this version, and the total traffic is 100.

[0086] Specifically, for example, please refer to the attached Figure 3 In the system framework shown in the figure, after the terminal device configured as the scheduling system processes the business module - querySupportCustomer, the overall customer service scheduling process reaches the algorithm protocol matching stage (i.e., using the customer service matching model to adapt to the customer service that provides service support for the business request initiated by the customer). Similarly, the terminal device will pull the latest algorithm module logic protocol configured on the above distributed configuration center (i.e., the above Figure 5The protocol code shown in the figure) is then matched with productA through the value of the productCode field. If they are equal, the value corresponding to the learningModel field is read, which represents the id (algorithm model number) of the corresponding algorithm model. Then, when the terminal device reads the value corresponding to the learningModel field as skillModel, it determines that the target customer service matching model currently used for corresponding customer service adaptation for business requests of business product A is the customer service skill model.

[0087] Step S4013, calculating the model version number according to the customer identifier;

[0088] Step S4014, loading the model of the target version pointed to by the model version number from among the candidate versions of the target customer service matching model.

[0089] After the terminal device determines the target customer service matching model currently used to adapt the corresponding customer service to the business request initiated by the customer, it further uses the user identifier in the request parameter extracted by parsing the business request to calculate the model version number, so as to load the target version of the model pointed to by the model version number among the various candidate versions of the target customer service matching model to adapt the corresponding customer service.

[0090] Specifically, for example, please refer to Figure 3 The system framework shown in the figure, the distributed configuration center is also pre-configured as shown in the attached Figure 5The protocol code shown is configured as a terminal device of the scheduling system. After determining that the target customer service matching model currently used for corresponding customer service adaptation for the business request of business product A is a customer service skill model, it continues to read the value under the ModelConfig field in the latest algorithm module logic protocol. Among them, ModelVersion represents the version number of the skillModel algorithm model (customer service skill model) corresponding to the productA product. There are three versions, which are 1.1.2, 1.1.3, and 1.1.4 respectively. The three versions of the skillModel algorithm model have a flow rate of 30% for version 1.1.2 and 30% for version 1.1.3. , the traffic corresponding to version 1.1.4 is 40%. The terminal device determines the target version of the model by the customer identifier - the remainder of the customer number divided by 100. That is, if the remainder of the customer number divided by 100 is between 0 and 29, the skillModel algorithm model of version 1.1.2 is assigned as the target version model to be executed to adapt to customer service; if the remainder is between 30 and 59, the skillModel algorithm model of version 1.1.3 is assigned as the target version model to be executed to adapt to customer service; or, if the remainder is between 60 and 99, the skillModel algorithm model of version 1.1.4 is assigned as the target version model to be executed to adapt to customer service.

[0091] It should be noted that, in this embodiment, an algorithm model id (algorithm model number) has many algorithm versions. The corresponding algorithm model can be loaded into the memory of the terminal device through the algorithm model id and the algorithm model version number. Figure 5 In the protocol code shown, the value of the flow field represents the traffic owned by the specific algorithm model.

[0092] Furthermore, in another feasible embodiment, the above step S402 may include:

[0093] Step S4021, if it is detected that the customer service matching model fails, or it is detected that the offset of the model calculation result output by the customer service matching model exceeds a preset deviation threshold, a new customer service matching model is dynamically loaded to adapt the customer service corresponding to the business request.

[0094] It should be noted that, in this embodiment, the preset deviation threshold is the maximum allowable deviation between the result output by the customer service matching model after calculation and the optimal solution.

[0095] When the terminal device determines the customer service matching model for adapting the corresponding customer service to the business request initiated by the customer, and in the process of using the customer service matching model to adapt the customer service, if it detects that the customer service matching model cannot execute the calculation process and thus determines that the model has failed, the terminal device can immediately redetermine a new customer adaptation model, and load and execute the new customer service matching model to adapt the customer service. Alternatively, if the terminal device detects that the model calculation result output by the customer service matching model based on the input parameter (the characteristic value of the customer feature) is greater than or equal to the maximum allowable deviation, the terminal device also immediately redetermines a new customer adaptation model, and loads and executes the new customer service matching model to adapt the customer service.

[0096] It should be noted that, in this embodiment, please refer to Figure 3 The system framework shown in the figure, in which the feature library can be a feature library corresponding to each customer and customer service (customer service), and the feature library contains indicators of each feature value of each customer and each customer service. The terminal device configured as the scheduling system uses the historical records of each customer and customer service pairing in the past as data samples to train the algorithm model. The input value is the feature values ​​of the corresponding customer and the feature values ​​of the corresponding customer service. The output value is the score between the customer and the customer service. Then, a weighted calculation is performed on the basis of each score item to obtain the output value y. When the weighted calculation is performed, the weight values ​​are different for different models. For example, due to the different emphases of the user satisfaction model and the customer service response rate model, in the user satisfaction model, the customer's waiting time weight and the customer service idle time weight are lower than those in the customer service response rate model, but the customer's customer service professional ability score and the customer service attitude score are higher than the weights of the two scores in the customer service response rate model.

[0097] In addition, different algorithm models have different model input parameters, and the same input parameters have different weights in different models (for example, in the customer service response rate model, the input feature values ​​include customer service professional ability characteristics, customer service call service duration characteristics, customer service age, place of origin and other basic information characteristics, customer basic information characteristics, customer problem type characteristics, customer service waiting time characteristics, customer waiting time characteristics, and customer urgency level characteristics. Although the user response rate model also includes the customer's basic information characteristics, customer service waiting time characteristics and customer waiting time characteristics, the weights of the customer's basic information characteristics, customer service waiting time characteristics and customer waiting time characteristics in the two algorithm models are inconsistent).

[0098] Based on this, the scheduling method of customer service of the present invention constructs a linear equation: y=W T *X=θ T*X processes the input data of the customer service matching model, where W represents the weight value, and X represents the value after multiplying the corresponding eigenvalue by the weight value specified by the algorithm model, minus the mean of the eigenvalue and then dividing it by the variance of the eigenvalue. According to the corresponding relationship between X1, X2, X3...Xn and Y, the error function is: Cost = (actual value - predicted value) 2 =(y pre - y) 2 , Cost is the error value. With the data training of TensorFlow, the value of cost will become smaller and smaller, thus reaching an optimal solution. Then, the optimal solution of each model is compared to find the optimal model (customer service matching model) for each business product to adapt to customer service.

[0099] In another feasible embodiment, in addition to building a customer service matching model based on the above-mentioned deep school framework training to adapt and schedule customer service for users, the terminal device configured as a scheduling system can also directly schedule customer service based on any mature weight data formula.

[0100] Specifically, for example, please refer to Figure 6 In the system framework shown, during the process of dynamically loading various business module code files, when executing the module code files for allocating customer services in the various business module code files, the terminal device first parses the business request initiated by the customer to extract the product identifier and the customer identifier in the request parameters, and then extracts the corresponding customer characteristics and the customer service characteristics associated with the business product pointed to by the product identifier from the characteristics storing the customer characteristics and customer service characteristics according to the product identifier and the customer identifier, thereby directly substituting the customer characteristics and the customer service characteristics into any mature weight calculation formula for calculation to determine and schedule the optimal customer service suitable for the customer.

[0101] In this embodiment, during the process of dynamically loading various business module code files through the terminal device, when executing the module code file for allocating customer service in the various business module code files, the product identifier in the request parameter extracted by parsing the business request initiated by the customer is first used to adapt the algorithm model number in the module code file, and the target customer service matching model corresponding to the algorithm model number among multiple customer service matching models is determined; after determining the target customer service matching model currently used to adapt the corresponding customer service to the business request initiated by the customer, the terminal device further uses the user identifier in the request parameter extracted by parsing the business request to calculate the model version number, so as to load the target version pointed to by the model version number among the various candidate versions of the target customer service matching model. model to adapt the corresponding customer service; and when the terminal device determines the customer service matching model for adapting the corresponding customer service for the business request initiated by the customer, and in the process of using the customer service matching model to adapt the customer service, if it is detected that the customer service matching model cannot execute the calculation process and thus determines that the model has failed, the terminal device can immediately redetermine a new customer adaptation model, and load and execute the new customer service matching model to adapt the customer service, or, if the terminal device detects that the model calculation result output by the customer service matching model based on the input parameter (the characteristic value of the customer feature) is greater than or equal to the maximum allowable deviation, the terminal device also immediately redetermines a new customer adaptation model, and loads and executes the new customer service matching model to adapt the customer service.

[0102] In this way, the customer service scheduling method of the present invention realizes the matching and scheduling between users and customer services according to the customer service matching model trained by the deep learning framework. Compared with the traditional customer service scheduling method that directly assigns the customer service with the longest idle time to the user with the longest waiting time, the customer service scheduling method of the present invention calculates the weights according to the deep learning framework, thereby differentially assigning customer service to different users, thereby making the degree of adaptation between customer service and users higher, and customer service can provide services to users more accurately and efficiently, which effectively improves the accuracy and efficiency of customer service scheduling and improves the user experience.

[0103] Further, based on the above-mentioned first and second embodiments, a third embodiment of a method for scheduling customer service of the present invention is proposed. The main difference between this embodiment and the above-mentioned first and second embodiments is that the method for scheduling customer service of the present invention may further include:

[0104] Step A, obtaining the service processing logic of each of the business products from the distributed configuration center, and storing each of the service processing logic in each local node machine;

[0105] Before the terminal device configured as a scheduling system performs the adaptation operation of the target service processing logic according to the request parameters extracted from the business request initiated by the customer, it can also pull the service processing logic of all business products from the distributed configuration center to each local node machine for storage in advance. In this way, when the terminal device performs the adaptation operation of the target service processing logic, the corresponding node machine can directly adapt it among all the service processing logics stored locally.

[0106] Specifically, for example, please refer to Figure 3 The system framework shown is configured such that after each power-on, the terminal device of the scheduling system pulls the service processing logic of all products from the distributed configuration center based on the communication connection established with the distributed configuration center, and then stores all the service processing logic in association in the local storage space (running memory or solid-state storage medium) of each local node machine.

[0107] Furthermore, in a feasible embodiment, in the above step A, after the step of "storing each of the service processing logics in each local node machine", the scheduling method for customer service of the present invention may further include:

[0108] Step B, receiving a data update notification from the distributed configuration center, and updating each of the service processing logics stored in each of the node machines according to the data update notification;

[0109] The terminal device configured as the scheduling system receives in real time the data update notification issued by the distributed configuration center. The data update notification is generated by the distributed configuration center after modifying the service processing logic of the local configuration based on the dynamic effectiveness mechanism. After receiving the data update notification, the terminal device updates the service processing logic stored on each local node machine according to the modification content carried in the data update notification.

[0110] Specifically, for example, after the distributed configuration center receives the configuration parameters modified and adjusted by the staff for the service processing logic of a certain business product C that has been configured, thereby updating the service processing logic of the business product C to obtain a new service processing logic, the distributed configuration center immediately encapsulates the configuration parameters modified and adjusted for the service processing logic of the business product C as a real-time data update notification and sends it to the terminal device configured as the scheduling system. Therefore, after receiving the data update notification, the terminal device parses the configuration parameters and modifies and adjusts the original service processing logic of the business product C stored on each local node machine according to the configuration parameters, thereby obtaining the new service processing logic of the business product C.

[0111] It should be noted that, in this embodiment, the dynamic validation mechanism of the above-mentioned distributed configuration center can be specifically implemented based on Http long polling request and Spring extension mechanism, that is, during the startup process of Spring container, the distributed configuration center registers the Bean containing the "${…}" placeholder and the "@Value" annotation in the parameter to the registry defined in the distributed configuration center framework through the customized BeanPostProcessor and BeanFactoryPostProcessor (both are extension interfaces provided by Spring IOC container), and then continuously obtains the configuration information of the server through Http long polling. Once the configuration changes, the corresponding Bean is found according to the Key of the changed configuration, and then the properties of the Bean are modified, thereby realizing the dynamic validation of the configuration.

[0112] Step C: obtaining updated new service processing logic from the distributed configuration center according to a preset time period, and using the new service processing logic to update each service processing logic stored in each node machine.

[0113] It should be noted that, in the present embodiment, the preset time period is a time period for a terminal device configured as a scheduling system to actively and repeatedly pull service processing logic from a distributed configuration center. It should be understood that, based on different design requirements of actual applications, in different feasible implementations, the time period can of course be set to different sizes. The scheduling method for customer service of the present invention does not limit the specific size of the time period.

[0114] In addition to passively receiving data update notifications generated by the distributed configuration center based on modified service processing logic, the terminal device configured as a scheduling system also actively pulls the new service processing logic obtained by the distributed configuration center after modifying the original service processing logic according to a pre-set preset time cycle, and then uses the new service processing logic to update the service processing logic stored on each local node machine of the terminal device.

[0115] Specifically, for example, a terminal device configured as a scheduling system takes 24 hours as a time cycle, and pulls a new service processing logic from the distributed configuration center every 24 hours, and repeats this cycle. Therefore, after the distributed configuration center receives the configuration parameters modified and adjusted by the staff for the service processing logic of a certain business product C that has been configured within a certain 24-hour time cycle, and updates the service processing logic of the business product C to obtain a new service processing logic, the terminal device of the scheduling system actively initiates a "detect new service processing logic and return" service request to the distributed configuration center when the current system time reaches the beginning of the next time cycle, so that the distributed configuration center receives the service request and detects the new service processing logic of the business product C locally, and feeds the new service processing logic back to the terminal device, so that after receiving the new service processing logic, the terminal device replaces the original service processing logic of the business product C stored on each local node machine to implement the update processing of the service processing logic.

[0116] It should be noted that, in this embodiment, the terminal device configured as the scheduling system passively receives data update notifications issued by the distributed configuration center, and actively pulls new service processing logic from the distributed configuration center, which can be executed synchronously or selectively. The scheduling method of customer service of the present invention does not specifically limit the execution method.

[0117] Further, in another feasible embodiment, in addition to configuring the service processing logic corresponding to the business product through the distributed configuration center, the scheduling method for customer service of the present invention can also directly configure the service processing logic locally. The scheduling method for customer service of the present invention can also include:

[0118] Step D: configuring and generating the service processing logic of all the business products in each of the local node machines.

[0119] In addition to pulling the service processing logic of all business products from the distributed configuration center, the terminal device configured as the scheduling system can also directly configure the service processing logic of all business products that is the same as that of the distributed configuration center on each local node machine. Therefore, when the terminal device is performing the adaptation operation of the target service processing logic, the corresponding node machine can directly adapt the entire service processing logic stored locally.

[0120] Specifically, for example, please refer to Figure 7 The system framework shown in the figure is configured such that the terminal device of the scheduling system receives the configuration parameters input by the staff before receiving the service request initiated by the customer and extracting the request parameters. Figure 4 The protocol code shown generates corresponding configuration files directly in the local storage space (solid storage medium) of each local node machine for storage, that is, the service processing logic of all business products is configured in the local storage space of each local node machine.

[0121] Further, in a feasible embodiment, after obtaining the service request to be served and extracting the request parameters from the service request in the above step S10, the scheduling method for customer service of the present invention may further include:

[0122] Step E: adapting the target service processing logic in each of the service processing logics stored in each of the node machines according to the request parameters.

[0123] Specifically, for example, after a terminal device configured as a scheduling system pulls the service processing logic of all business products from a distributed configuration center and stores all the service processing logic in the local storage space of each local node machine, if the terminal device receives a business request initiated by a customer and extracts the request parameters, any local node machine directly matches the product number contained in the request parameters with the value of the productCode field in all the service processing logics stored in the local storage space, and when it is determined that the product number is equal to the value, reads the corresponding productBiz field in all the service processing logics as the target service processing logic for the product to which the business request initiated by the current customer belongs.

[0124] Alternatively, after the terminal device receives the business request initiated by the customer and extracts the request parameters, any local node machine matches the product number contained in the request parameters with all the service processing logic stored in the local storage space, thereby reading the target service processing logic of the product to which the business request initiated by the current customer belongs.

[0125] In this embodiment, before performing the adaptation operation of the target service processing logic according to the request parameters extracted from the business request initiated by the customer, the terminal device configured as the scheduling system can also pull the service processing logic of all business products from the distributed configuration center to each local node machine for storage in advance. In this way, when the terminal device performs the adaptation operation of the target service processing logic, the corresponding node machine can directly perform the adaptation among all the service processing logics stored locally. Alternatively, in addition to pulling the service processing logic of all business products from the distributed configuration center, the terminal device configured as the scheduling system can also directly configure the service processing logic of all business products that is the same as that of the distributed configuration center on each local node machine. Therefore, when the terminal device performs the adaptation operation of the target service processing logic, the corresponding node machine can directly perform the adaptation among all the service processing logics stored locally.

[0126] Also, by configuring the terminal device as a scheduling system to receive in real time the data update notification issued by the distributed configuration center, the data update notification is generated by the distributed configuration center after modifying the service processing logic of the local configuration based on the dynamic effectiveness mechanism. After receiving the data update notification, the terminal device will update the service processing logic stored on each local node machine according to the modification content carried in the data update notification; or, by configuring the terminal device as a scheduling system, in addition to passively receiving the data update notification generated by the distributed configuration center based on the modification of the service processing logic, the terminal device also actively pulls the new service processing logic obtained by the distributed configuration center after modifying the original service processing logic from the distributed configuration center according to a pre-set preset time period cycle, and then uses the new service processing logic to update the service processing logic stored on each local node machine of the terminal device.

[0127] Compared with the traditional customer service scheduling method, the present invention stores or directly configures the service processing logic used for scheduling to each node of the scheduling system. That is, when the terminal device configured as the scheduling system pulls the configuration related to the service processing logic in the distributed configuration center, it will save a copy of the configuration data locally on each node machine, and also perform real-time update operations on the configuration data. Therefore, when the distributed configuration center is unavailable or the configuration center cannot be set up separately due to cost, the terminal device can also directly read the local configuration data and perform customer service scheduling operations normally, further improving the stability and overall efficiency of customer service scheduling.

[0128] Furthermore, the present invention also provides a customer service scheduling system, which is applied to a message client device. Figure 8 , Figure 8A schematic diagram of the functional modules of an embodiment of a scheduling system for serving customers of the present invention. Figure 8 As shown, the scheduling system for customer service of the present invention includes:

[0129] The acquisition module 10 is used to acquire the service request to be served and extract the request parameters from the service request, and acquire the target service processing logic from the preset distributed configuration center according to the request parameters, wherein the distributed configuration center is configured with the service processing logic of all business products;

[0130] A code dynamic loading module 20, used to dynamically load each business module code file according to the target service processing logic;

[0131] The customer service scheduling module 30 is used to execute the business module code file to schedule the customer service corresponding to the business request.

[0132] Furthermore, each of the business module code files includes a module code file for allocating customer service, and the customer service scheduling module 30 includes:

[0133] A model loading unit, used to execute the module code file for allocating customer service to load a preset customer service matching model, wherein the customer service matching model is obtained by deep learning training based on customer characteristics and customer service characteristics;

[0134] The customer service adaptation unit is used to adapt the customer service corresponding to the business request by using the customer service matching model and schedule the customer service.

[0135] Furthermore, the request parameters include: a customer identifier of a customer initiating the service request and a product identifier of the service product; the number of customer service matching models is greater than one; and the model loading unit includes:

[0136] An adaptor unit, for adapting an algorithm model number in a module code file for allocating customer service according to the product identification;

[0137] A determination subunit, used to determine a target customer service matching model corresponding to the algorithm model number in each of the customer service matching models;

[0138] A calculation subunit, used for calculating a model version number according to the customer identifier;

[0139] The loading subunit is used to load the model of the target version pointed to by the model version number from each candidate version of the target customer service matching model.

[0140] Furthermore, the customer service adaptation unit is also used to dynamically load a new customer service matching model to adapt the customer service corresponding to the business request if it detects that the customer service matching model fails, or detects that the offset of the model calculation result output by the customer service matching model exceeds a preset deviation threshold.

[0141] Furthermore, the scheduling system for customer service of the present invention further includes:

[0142] A logic storage module is used to obtain the service processing logic of each of the business products from the distributed configuration center, and store each of the service processing logic in each local node machine;

[0143] A logic generation module, used to configure and generate the service processing logic of all the business products in each of the local node machines;

[0144] The acquisition module of the scheduling system for customer service of the present invention is also used to adapt the target service processing logic in each of the service processing logics stored in each of the node machines according to the request parameters.

[0145] Furthermore, the scheduling system for customer service of the present invention further includes:

[0146] A logic update module is used to receive data update notifications from the distributed configuration center, and update the service processing logic stored in each of the node machines according to the data update notifications; and to obtain updated new service processing logic from the distributed configuration center according to a preset time period, and use the new service processing logic to update the service processing logic stored in each of the node machines.

[0147] Furthermore, the logic update module is also used to obtain a modification request for the target service processing logic; and to modify the target service processing logic in the distributed configuration center according to the modification request to obtain a new service processing logic.

[0148] Among them, the functional implementation of each module in the above-mentioned customer service scheduling system corresponds to the various steps in the above-mentioned customer service scheduling method embodiment, and its functions and implementation processes will not be repeated here one by one.

[0149] The present invention also provides a computer storage medium storing a customer service scheduling program, which, when executed by a processor, implements the steps of the customer service scheduling method described in any of the above embodiments.

[0150] The specific embodiments of the computer storage medium of the present invention are basically the same as the embodiments of the scheduling method for customer service described above, and will not be described in detail here.

[0151] The present invention also provides a computer program product, which includes a computer program. When the computer program is executed by a processor, the steps of the customer service scheduling method described in any of the above embodiments are implemented.

[0152] The specific embodiments of the computer storage medium of the present invention are basically the same as the embodiments of the scheduling method for customer service described above, and will not be described in detail here.

[0153] The above are only preferred embodiments of the present invention, and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for scheduling customer service, characterized in that: The customer service dispatching method includes: Obtaining a business request to be serviced and extracting request parameters from the business request; the request parameters include: customer number, product number and problem type; Obtaining the target service processing logic from a preset distributed configuration center according to the request parameters, wherein the distributed configuration center is configured with the service processing logic of all business products, that is, all parameter configurations for scheduling customer service are configured in the distributed configuration center; Dynamically loading each business module code file according to the target service processing logic, specifically: dynamically loading each business module code file in sequence according to the number of the business module code in the target service processing logic; Executing the service module code file to schedule customer service corresponding to the service request; Among them, each of the business module code files includes a module code file for allocating customer service, The step of executing the business module code file to schedule the customer service corresponding to the business request includes: Execute the module code file for allocating customer service to load a preset customer service matching model, wherein the customer service matching model is obtained by deep learning training based on customer characteristics and customer service characteristics; Adapting the customer service corresponding to the business request using the customer service matching model, and scheduling the customer service; Wherein, the request parameters include: the customer identification of the customer initiating the service request and the product identification of the service product, and the number of the customer service matching models is greater than one, The step of executing the module code file for allocating customer service to load a preset customer service matching model includes: Adapting the algorithm model number in the module code file for the allocated customer service according to the product identification; Determine a target customer service matching model corresponding to the algorithm model number in each of the customer service matching models; Calculate the model version number according to the customer identifier; Among the candidate versions of the target customer service matching model, the model of the target version pointed to by the model version number is loaded.

2. The method for dispatching customer service according to claim 1, characterized in that: The step of using the customer service matching model to adapt the customer service corresponding to the business request includes: If a failure of the customer service matching model is detected, or if an offset of a model calculation result output by the customer service matching model is detected to exceed a preset deviation threshold, a new customer service matching model is dynamically loaded to adapt to the customer service corresponding to the business request.

3. The method for dispatching customer service according to claim 1, characterized in that: The customer service scheduling method further includes: Obtain the service processing logic of all the business products from the distributed configuration center, and store the service processing logic in each local node machine; or, Configuring and generating the service processing logic of all the business products in each of the local node machines; After the step of obtaining the service request to be served and extracting request parameters from the service request, the method further includes: Adapt the target service processing logic in each of the service processing logics stored in each of the node machines according to the request parameters.

4. The method for dispatching customer service according to claim 3, characterized in that: After the step of storing each of the service processing logics in each local node machine, the method further includes: receiving a data update notification from the distributed configuration center, and updating each of the service processing logics stored in each of the node machines according to the data update notification; and / or, The updated new service processing logic is obtained from the distributed configuration center according to a preset time period, and the new service processing logic is used to update each service processing logic stored in each node machine.

5. The method for dispatching customer service according to claim 1, characterized in that: After the step of executing the service module code file to schedule the customer service corresponding to the service request, the method further includes: Obtaining a modification request for the target service processing logic; The target service processing logic is modified in the distributed configuration center according to the modification request to obtain a new service processing logic.

6. A customer service dispatching system, characterized in that: The dispatch system for customer service includes: An acquisition module is used to acquire a business request to be served and extract request parameters from the business request, wherein the request parameters include: customer number, product number and problem type; and, according to the request parameters, acquire a target service processing logic from a preset distributed configuration center, wherein the distributed configuration center is configured with the service processing logic of all business products, that is, all parameter configurations for scheduling customer service are configured in the distributed configuration center; A code dynamic loading module, used to dynamically load each business module code file according to the target service processing logic, specifically: dynamically load each business module code file in sequence according to the business module code number in the target service processing logic; A customer service scheduling module, used to execute the business module code file to schedule customer service corresponding to the business request; Among them, each of the business module code files includes a module code file for allocating customer service, The step of executing the business module code file to schedule the customer service corresponding to the business request includes: Execute the module code file for allocating customer service to load a preset customer service matching model, wherein the customer service matching model is obtained by deep learning training based on customer characteristics and customer service characteristics; Adapting the customer service corresponding to the business request using the customer service matching model, and scheduling the customer service; Wherein, the request parameters include: the customer identification of the customer initiating the service request and the product identification of the service product, and the number of the customer service matching models is greater than one, The step of executing the module code file for allocating customer service to load a preset customer service matching model includes: Adapting the algorithm model number in the module code file for the allocated customer service according to the product identification; Determine a target customer service matching model corresponding to the algorithm model number in each of the customer service matching models; Calculate the model version number according to the customer identifier; Among the candidate versions of the target customer service matching model, the model of the target version pointed to by the model version number is loaded.

7. A terminal device, characterized in that: The terminal device includes: a memory, a processor, and a customer service scheduling program stored in the memory and executable on the processor. When the customer service scheduling program is executed by the processor, the steps of the customer service scheduling method as described in any one of claims 1 to 5 are implemented.

8. A computer storage medium, characterized in that: The computer storage medium stores a customer service scheduling program, and when the customer service scheduling program is executed by the processor, the steps of the customer service scheduling method according to any one of claims 1 to 5 are implemented.

Images