Customer service scheduling processing method and device

Abstract

The embodiment of the invention provides a customer service scheduling processing method and device.The customer service scheduling processing method comprises the steps that after user service data are obtained, the scheduling category is determined according to the flow state and the data category of the user service data, and the scheduling category is determined according to the data category of the user service data; inputting the user service data and the user reference data into a portrait generation model corresponding to the scheduling category for portrait generation to obtain user portrait data, and determining candidate customer service in a customer service list according to a customer service determination strategy corresponding to the scheduling category; and finally, according to the user portrait data and the customer service portrait data of the candidate customer service, customer service allocation processing is carried out on the user, so that customer service scheduling processing is realized.

Description

Technical Field

[0001] This document relates to the field of data processing technology, and in particular to a customer service scheduling processing method and device. Background Art

[0002] With the development of Internet technology, more and more users perceive and participate in services online. In the process of service perception and service participation, in order to enhance users' awareness of services, customer service can be set up to interact with users during the service perception or service participation process, answer questions, recommend services to users, etc.; however, there is a mismatch between the number of users who need to interact with services and the number of customer service staff, resulting in a situation where there are no available customer service staff to allocate to users or there are many available customer service staff but not enough users to interact with them. How to achieve effective customer service scheduling is a focus of increasing concern for users and service providers. Summary of the invention

[0003] One or more embodiments of the present specification provide a customer service scheduling processing method, including: obtaining user service data, and determining a scheduling category according to a traffic state and a data category of the user service data. Determining a portrait generation model according to the scheduling category, and inputting the user service data and user benchmark data into the portrait generation model to generate a portrait, thereby obtaining user portrait data. Determining candidate customer service in a customer service list according to a customer service determination strategy corresponding to the scheduling category. Performing customer service allocation processing on the user according to the user portrait data and the customer service portrait data of the candidate customer service.

[0004] One or more embodiments of the present specification provide a customer service scheduling processing device, including: a data acquisition module, configured to acquire user service data, and determine a scheduling category according to a traffic status and a data category of the user service data. A portrait generation module, configured to determine a portrait generation model according to the scheduling category, and input the user service data and user benchmark data into the portrait generation model to generate a portrait, thereby obtaining user portrait data. A candidate customer service determination module, configured to determine a candidate customer service in a customer service list according to a customer service determination strategy corresponding to the scheduling category. A customer service scheduling processing module, configured to perform customer service allocation processing on a user according to the user portrait data and the customer service portrait data of the candidate customer service.

[0005] One or more embodiments of the present specification provide a customer service scheduling processing device, including: a processor; and a memory configured to store computer executable instructions, wherein when the computer executable instructions are executed, the processor: obtains user service data, and determines a scheduling category according to a traffic state and a data category of the user service data. Determine a portrait generation model according to the scheduling category, and input the user service data and user benchmark data into the portrait generation model to generate a portrait, thereby obtaining user portrait data. Determine candidate customer service in a customer service list according to a customer service determination strategy corresponding to the scheduling category. Perform customer service allocation processing on the user according to the user portrait data and the customer service portrait data of the candidate customer service.

[0006] One or more embodiments of the present specification provide a computer-readable storage medium for storing computer-executable instructions, which implement the following process when executed: obtaining user service data, and determining a scheduling category based on the traffic status and the data category of the user service data. Determining a portrait generation model based on the scheduling category, and inputting the user service data and user benchmark data into the portrait generation model for portrait generation to obtain user portrait data. Determining candidate customer service in a customer service list based on a customer service determination strategy corresponding to the scheduling category. Performing customer service allocation processing on the user based on the user portrait data and the customer service portrait data of the candidate customer service. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] In order to more clearly illustrate one or more embodiments of the present specification or the technical solutions in the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present specification. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative labor. Figure 1 A schematic diagram of an implementation environment of a customer service scheduling processing method provided in one or more embodiments of this specification; Figure 2 A processing flow chart of a customer service scheduling processing method provided in one or more embodiments of this specification; Figure 3 A schematic diagram of a customer service dispatching system provided for one or more embodiments of this specification; Figure 4 A processing flow chart of a customer service scheduling processing method for a dialogue scheduling scenario for a security project provided by one or more embodiments of this specification; Figure 5 A processing flow chart of a customer service scheduling processing method for an offline scheduling scenario of a security project provided by one or more embodiments of this specification; Figure 6 A processing flow chart of a customer service scheduling processing method for a real-time scheduling scenario of a security project provided by one or more embodiments of this specification; Figure 7 A schematic diagram of an embodiment of a customer service dispatch processing device provided in one or more embodiments of this specification; Figure 8 A schematic diagram of the structure of a customer service scheduling processing device provided in one or more embodiments of this specification. DETAILED DESCRIPTION

[0008] In order to enable those skilled in the art to better understand the technical solutions in one or more embodiments of this specification, the technical solutions in one or more embodiments of this specification will be clearly and completely described below in conjunction with the drawings in one or more embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of them. Based on one or more embodiments of this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this document.

[0009] The customer service dispatch processing method provided in one or more embodiments of this specification can be applied to the implementation environment of customer service allocation of services, referring to Figure 1 , the implementation environment at least includes: a server 101, a customer service terminal 102 and a user terminal 103; The server 101 may be a single server, or a server cluster consisting of several servers, or one or more cloud servers in a cloud computing platform; used for customer service scheduling processing; The customer service terminal 102 may specifically be a mobile phone, a personal computer, a tablet computer, an e-book reader, a device for information interaction based on VR (Virtual Reality), a vehicle terminal, an IoT device, a wearable smart device, a laptop computer, a desktop computer, etc.; and is used for service interaction with the user terminal 103; The user terminal 103 can specifically be a mobile phone, a personal computer, a tablet computer, an e-book reader, a device for information interaction based on VR (Virtual Reality), a vehicle-mounted terminal, an IoT device, a wearable smart device, a laptop computer and a desktop computer, etc.; it is used to submit user service data to the server 101 and / or interact with the customer service terminal 102 for service.

[0010] In addition, the implementation environment may also include a portrait generation model 104 and an allocation model 105; wherein the portrait generation model 104 is used to generate a portrait according to user service data and user benchmark data; the allocation model 105 is used to perform customer service allocation processing according to user portrait data and customer service portrait data; In this implementation environment, after the server 101 obtains the user service data, it determines the scheduling category according to the current traffic status and the data category of the user service data, determines the portrait generation model 104 according to the scheduling category, inputs the user service data and the user benchmark data into the portrait generation model 104 for portrait generation, and obtains the user portrait data; then, according to the customer service determination strategy corresponding to the scheduling category, the candidate customer service is determined in the customer service list, and finally, through the allocation model 105, the customer service allocation processing is performed on the user according to the user portrait data and the customer service portrait data of the candidate customer service, so that the interactive customer service assigned to the user is through the customer service terminal 102, and the user performs service interaction through the user terminal 103; in this way, the customer service allocation processing is performed based on the user portrait data and the customer service portrait data, so as to improve the effectiveness and accuracy of the allocation results.

[0011] It should be noted that, considering that the user service data, user benchmark data, customer service benchmark data and other related data involved in this specification may belong to the privacy of the user or customer service to a certain extent, if you want to collect the user's user service data, user benchmark data and customer service benchmark data of the customer service and other related data, you can obtain the user's authorization before collecting the data, so that the operation of collecting data complies with relevant data management regulations. For example, data authorization can be performed when the user is currently submitting user service data to the server, and data authorization can also be performed when the user submits user service data to the server for the first time; the specific method of data authorization can be to send a data authorization reminder to the user, and the user can obtain data collection authorization after confirming the reminder through an instruction, or the method of data authorization can also be to obtain data collection authorization by signing a data authorization agreement; the data authorization of customer service is similar, and this embodiment will not be repeated here.

[0012] One or more embodiments of a customer service scheduling processing method provided in this specification are as follows: Reference Figure 2 The customer service scheduling processing method provided in this embodiment specifically includes steps S202 to S208.

[0013] Step S202: acquiring user service data, and determining a scheduling category according to a traffic state and a data category of the user service data.

[0014] The user service data in this embodiment includes user conversation data submitted by the user for conversation interaction; for example, user conversation data submitted by the user through a subroutine in an application or a third-party application for conversation interaction; taking the guarantee project subroutine in a third-party payment application as an example, the user enters "Hello, may I ask what are the vehicle guarantee items?" through the message module in the guarantee project subroutine and submits it; the user service data obtained by the server is the user conversation data "Hello, may I ask what are the vehicle guarantee items?" It should be noted that the server in this embodiment can be a server corresponding to the guarantee project subroutine or a server corresponding to the third-party payment application, and this embodiment does not limit this.

[0015] In addition to the user conversation data, the user service data in this embodiment may also be user description data describing the relevant circumstances of the user, such as description data of the user's personal, economic, health, and family dimensions.

[0016] Corresponding to the two types of user service data, namely user dialogue data and / or user description data, the data categories of the user service data include dialogue data categories and / or description data categories.

[0017] The flow state includes a state characterizing the amount of current access flow; specifically, if the current access flow is less than the flow threshold, the flow state is determined to be the first flow state, and if the current access flow is greater than or equal to the flow threshold, the flow state is determined to be the second flow state. The first flow state includes a low flow state with less flow; the second flow state includes a high flow state with more flow.

[0018] In addition to directly determining the traffic status based on the current access traffic, the traffic status can also be determined based on the customer service status data in the customer service list; specifically, if the number of customer service staff in the customer service list whose customer service status is in service status is less than the quantity threshold, the traffic status is determined to be the first traffic status; if the number of customer service staff in the customer service list whose customer service status is in service status is greater than or equal to the quantity threshold, the traffic status is determined to be the second traffic status; or, the available status duration of the customer service in the customer service list is counted, and if the available status duration is greater than or equal to the duration threshold, the traffic status is determined to be the first traffic status; if the available status duration is less than the duration threshold, the traffic status is determined to be the second traffic status; wherein, the available status duration may include the sum of the status durations of each customer service in the customer service list in the available status. The above describes the process of determining the flow status. In the specific implementation process, any one of the above methods can be used to determine the flow status, or at least two of the above methods can be combined to determine the flow status. Other methods can also be used to determine the flow status, which can be configured according to actual needs. This embodiment does not limit this. Taking the combination of access traffic and the number of customer services as an example, in the process of determining the flow status, it is detected whether the access traffic is less than the traffic threshold, and whether the number of customer services in the customer service list whose customer service status is service status is less than the number threshold. If so, the flow status is determined to be the first flow status; if one of the detection results is no, the flow status is determined to be the second flow status.

[0019] In this embodiment, there are different service interactions, such as a conversation interaction with a user after obtaining user conversation data, or a recommendation interaction for making service recommendations to a user; different service interactions have certain differences in the customer service scheduling process. In this embodiment, the scheduling category includes a category for performing customer service scheduling processing.

[0020] In specific implementation, after obtaining the user service data, the scheduling category is determined according to the traffic status and the data category of the user service data; in the recommendation interaction process to the user, since it is not initiated by the user, the user service data includes user description data, and due to the difference between the user conversation data and the user description data, that is, the different scenarios of conversation interaction and service recommendation, after obtaining the user service data, the traffic status, that is, the current traffic status, is determined, and there are also differences in the process of determining the scheduling category according to the traffic status and the data category of the user service data.

[0021] In an optional implementation manner provided by this embodiment, in the process of determining the scheduling category according to the traffic state and the data category of the user service data, if the data category is the conversation data category, the traffic state is determined according to the customer service state data of each customer service in the customer service list, and if the traffic state is the first traffic state, the scheduling category is determined as the first conversation scheduling category, and if the traffic state is the second traffic state, the scheduling category is determined as the second conversation scheduling category; In an optional implementation provided by this embodiment, if the data category is a descriptive data category, the traffic state is determined based on the access traffic. If the traffic state is the first traffic state, the scheduling category is determined as the first recommended scheduling category. If the traffic state is the second traffic state, the scheduling category is determined as the second recommended scheduling category.

[0022] Specifically, if the user service data is user conversation data and the traffic state is the first traffic state, the scheduling category is determined to be the first conversation scheduling category; if the user service data is user conversation data and the traffic state is the second traffic state, the scheduling category is determined to be the second conversation scheduling category; if the user service data is user description data and the traffic state is the first traffic state, the scheduling category is determined to be the first recommended scheduling category; if the user service data is user description data and the traffic state is the second traffic state, the scheduling category is determined to be the second recommended scheduling category.

[0023] Step S204, determining a portrait generation model according to the scheduling category, and inputting the user service data and user benchmark data into the portrait generation model to generate a portrait, thereby obtaining user portrait data.

[0024] In the above process, after obtaining the user service data, the scheduling category is first determined according to the traffic status and the data category of the user service data; in this step, the portrait generation model is determined according to the scheduling category, and the user portrait is generated through the portrait generation model to obtain the user portrait data.

[0025] The user portrait data in this embodiment includes data characterizing the user obtained by characterizing the user's purchasing power, willingness, and mentality; optionally, the user portrait data in this embodiment includes at least one of the following: demand data, recommended services, user ratings, purchasing power data, purchasing willingness data, purchasing mentality data.

[0026] The user benchmark data includes relevant data of the user maintained by the server, such as historical behavior data of the user, age and / or job data of the user, etc.

[0027] During specific implementation, in order to improve the effectiveness and accuracy of the portrait generation and improve the degree of representation of the user by the generated user portrait data, in this embodiment, the portrait generation model is first determined according to the scheduling category, and then the user service data and user benchmark data are input into the portrait generation model to generate the portrait and obtain the user portrait data; the following specifically describes the determination of the portrait generation model under different scheduling categories and the portrait generation process.

[0028] (1) First dialogue scheduling category In an optional implementation manner provided by this embodiment, if the scheduling category is the first dialogue scheduling category, in the process of determining a portrait generation model according to the scheduling category, and inputting the user service data and the user benchmark data into the portrait generation model to generate a portrait, and obtaining the user portrait data, the following operations are performed: According to the first dialogue scheduling category, determining a large language model, a recommendation model and / or a rating model as a portrait generation model; The user service data and the first prompt text are input into the large language model for demand identification to obtain dialogue demands, the dialogue demands and / or historical behavior data are input into the recommendation model for service recommendation processing to obtain recommended services, and / or the user basic data, dialogue demands and / or recommended services are input into the rating model for rating processing to obtain user ratings.

[0029] Optionally, the first prompt text includes a prompt text prompting the large language model to perform demand identification; for example, the first prompt text is "The user conversation data is: [data], please identify the user's service interaction needs based on the user conversation data". The large language model includes LLM (Large Language Model); in addition, the large language model described in this embodiment can also be a pre-trained natural language model. The large language model can adopt a foundation model (Foundation Models) or a pretrained model (Pretrained Models). The specific architecture of the large language model can be a neural network architecture, a Transform architecture, or other architectures using a large number of parameters. In the specific execution process, the large language model can directly adopt a foundation model or a pretrained model, and can also fine-tune the foundation model or the pretrained model on the basis of the foundation model or the pretrained model for tasks related to the conversation data such as demand identification and service recommendation processing (Supervised Fine-Tuning, SFT), so as to obtain a large language model that can perform tasks related to the conversation data such as demand identification and service recommendation processing.

[0030] The recommendation model and rating model may be deep learning models. The conversation requirements include the user's interaction intention and / or the user's recommended service category; for example, whether the user needs health protection items or property protection items more; the recommended services include specific services recommended to the user; for example, health protection item A is recommended to the user. The user rating includes the user rating determined based on the purchasing power, willingness and mind after identifying the user's purchasing power, willingness and mind.

[0031] The historical behavior data includes data related to the user's historical participation in services, such as the user's historical purchase behavior data for insurance items; or the historical behavior data can also be historical conversation data of the user's historical conversations. It should be noted that if the historical behavior data includes historical conversation data, the historical conversation data and the recommendation prompt text are input into the large language model for service recommendation processing to obtain the recommended service; that is, for the data of the conversation data category, the large language model is used for processing, the data of the conversation data category and the corresponding prompt text are input into the large language model for corresponding processing, and for the data of the description data category, the corresponding deep learning model is used for corresponding processing, and this embodiment will not be repeated here.

[0032] Specifically, if the scheduling category is the first dialogue scheduling category, it means that the current traffic state is the first traffic state, that is, the access traffic is relatively small. In this case, the manual customer service can be determined as the interactive customer service that interacts with the user in conversation. In order to improve the accuracy and effectiveness of the obtained user portrait data, according to the first dialogue scheduling category, the user dialogue data and the demand identification prompt text are input into the large language model for demand identification to obtain dialogue needs. The dialogue needs and historical behavior data are input into the recommendation model for service recommendation processing to obtain recommended services. The user basic data, dialogue needs and recommended services are input into the rating model for rating processing to obtain user ratings. The dialogue needs, recommended services and user ratings are used as user portrait data.

[0033] It should be noted that in addition to conversation needs, recommended services and user ratings, user basic data can also be input into the risk identification model for risk identification to obtain risk identification results, which are also used as user portrait data; historical conversation data and interaction preference prompt text can also be input into the large language model to generate interaction preferences, obtain interaction preferences, and use interaction preferences as user portrait data. Optionally, the interaction preference includes interaction mode and / or interaction feedback time; wherein the interaction mode is, for example, telephone mode or online chat mode; the interaction feedback time is, for example, the reply speed of the message; in addition, it can also include interaction preferences of other dimensions, which are not limited in this embodiment.

[0034] (2) Second dialogue scheduling category In an optional implementation manner provided by this embodiment, if the scheduling category is the second dialogue scheduling category, in the process of determining a portrait generation model according to the scheduling category, and inputting the user service data and the user benchmark data into the portrait generation model to generate a portrait, and obtaining the user portrait data, the following operations are performed: According to the second dialogue scheduling category, the user service data and the second prompt text are input into the large language model to evaluate the dialogue index, so as to obtain the dialogue index; If the dialogue indicators meet the manual customer service interaction conditions, the user service data and the user benchmark data are input into the portrait generation model corresponding to the first dialogue scheduling category to generate a portrait and obtain user portrait data.

[0035] Optionally, the second prompt text includes prompt text prompting the large language model to perform dialogue indicator evaluation; for example, the second prompt text is "The user dialogue data is: [data], please evaluate the urgency of the user"; the dialogue indicators include evaluation indicators that characterize the user's urgency, potential conversion rate and / or type; wherein the type is, for example, complaint, consultation and / or suggestion.

[0036] Specifically, if the scheduling category is the first dialogue scheduling category, it means that the current traffic state is the second traffic state, that is, the access traffic is relatively large. In this case, if each user is assigned to customer service processing, it may cause access traffic accumulation, which not only causes data processing pressure on the server, but also affects the user's interactive mentality to a certain extent. Based on this, in this embodiment, if the scheduling category is the second dialogue scheduling category, the user dialogue data and the dialogue indicator evaluation prompt text are first input into the large language model for dialogue indicator evaluation to obtain the dialogue indicator. If the dialogue indicator is greater than the indicator threshold, it is determined that the dialogue indicator meets the manual customer service interaction condition, and the portrait recognition model corresponding to the first dialogue scheduling category is determined as the portrait recognition model corresponding to the second dialogue scheduling category, and the user service data and user benchmark data are input into the portrait generation model corresponding to the first dialogue scheduling type for portrait generation to obtain user portrait data. The specific portrait generation process is similar to the portrait generation process corresponding to the first dialogue scheduling type mentioned above, and the above-mentioned related content can be referred to, and this embodiment will not be repeated here.

[0037] If the conversation index is less than or equal to the index threshold, it is determined that the conversation index does not meet the manual customer service interaction conditions, and the machine customer service is determined as the interactive customer service for conversational interaction with the user, so as to have a conversation with the user through the machine customer service. In this way, when there is a lot of access traffic, machine customer service is introduced, and the appropriate customer service is assigned to the user according to the user's conversation index, which not only reduces the pressure on manual customer service, but also shortens the user's interaction waiting time and improves the service experience.

[0038] Furthermore, when a machine customer service is assigned to a user, the user may also submit a manual interaction request. In an optional implementation manner provided in this embodiment, the user is marked as a manual waiting state based on the manual interaction request submitted by the user during a conversation with the machine customer service. If it is detected that there is a customer service representative with an available status in the customer service list, the customer service representative is determined as an interactive customer service representative for conversational interaction with the user, so that the interactive customer service representative can conduct conversational interaction with the user.

[0039] During the specific execution process, if the user is marked as being in a manual waiting state due to heavy access traffic, and manual customer service is not assigned to the user for a long time, it may lead to user loss and affect user mentality. In this embodiment, it is also possible to interact with the user through the "customer service grabbing order" mode; optionally, if an interaction request from the target customer service to the user in the manual waiting state is detected, the target customer service is determined as an interactive customer service who interacts with the user in a conversation, and the manual waiting state mark of the user is cancelled. After the interaction is over, the target customer service can also be marked for interaction to issue an interaction record to the target customer service, thereby improving the customer service mentality and enabling the customer service to actively participate, thereby achieving effective allocation of users and improving the user's interactive experience.

[0040] (3) Recommended scheduling category In an optional implementation manner provided by this embodiment, regardless of whether it is the first recommended scheduling category or the second recommended scheduling category, since the user service data is user description data, the demand identification model, recommendation model and / or rating model are determined to be the portrait generation model; the user service data is input into the demand identification model for demand identification to obtain recommended demand, the recommended demand and historical behavior data are input into the recommendation model for service recommendation processing to obtain recommended service, and / or the user basic data, recommended demand and recommended service are input into the rating model for rating processing to obtain user rating.

[0041] It should be noted that the relevant processing process is similar to the relevant processing process under the above-mentioned dialogue scheduling category, and will not be repeated in this embodiment.

[0042] In addition, steps S202 to S204 may also be replaced by obtaining user service data, determining a portrait generation model according to the traffic status and / or the data category of the user service data, and inputting the user service data and the user benchmark data into the portrait generation model for portrait generation, obtaining user portrait data, and forming a new implementation method with one or more other processing steps provided in this embodiment. Optionally, the user service data may also be input into the portrait generation model for portrait generation; in this case, the user service data may include the user benchmark data.

[0043] Alternatively, steps S202 to S204 may also be replaced by obtaining user service data, determining a portrait generation model according to the traffic state, inputting the user service data into the portrait generation model for portrait generation, obtaining user portrait data, and forming a new implementation method with one or more processing steps provided in this embodiment. Optionally, if the traffic state is the first traffic state, the large language model recommendation model and / or rating model is determined as the portrait generation model; if the traffic state is the second traffic state, the large language model is determined as the portrait generation model; specifically, in the second traffic state, a conversation index evaluation is first performed, and if the conversation index is greater than the index threshold, the portrait is generated according to the portrait generation model corresponding to the first traffic state; the specific process can be referred to above, and this embodiment will not be repeated here.

[0044] Step S206: Determine candidate customer services in the customer service list according to the customer service determination strategy corresponding to the scheduling category.

[0045] In the above process, a user portrait is generated to obtain user portrait data. In this step, candidate customer services are determined in the customer service list according to the customer service determination strategy corresponding to the scheduling category.

[0046] The customer service list includes a customer service list for managing online customer service created based on relevant data of online customer service in the current time period. The customer service list may only record manual customer service, or may record online manual customer service and machine customer service, which is not limited in this embodiment. Optionally, the customer service list records the customer service identification, customer service status, customer service portrait data, actual service flow and / or predicted service flow of each customer service.

[0047] In specific implementation, when a customer service representative goes online, he needs to pay a fee to the customer service representative. If there are many online customer service representatives but less access traffic, the customer service representative will be idle for a long time. Therefore, in this embodiment, the access traffic can be predicted in combination with historical traffic data and time data. The online customer service representative can be determined based on the predicted access traffic and the predicted service traffic of the customer service representative to achieve "customer service scheduling based on supply and demand prediction" and avoid economic losses caused by a large number of online customer service representatives being idle. In an optional implementation provided in this embodiment, the customer service list can be created in the following way: Input the time data and historical traffic data into the traffic prediction model to predict the access traffic and obtain the predicted access traffic within the time interval; Inputting the historical service flow and qualification data of each customer service into the flow prediction model to perform service flow prediction, and obtaining the predicted service flow of each customer service within the time interval; According to the predicted access traffic and the predicted service traffic, online customer services within the time interval are determined, and the customer service list is created based on the customer service identification and customer service portrait data of the online customer services.

[0048] Specifically, on the one hand, the access traffic can be predicted based on the current time data and historical traffic data to obtain the predicted access traffic within the time interval, thereby realizing the prediction of "demand"; that is, the current time data and historical traffic data can be input into the traffic prediction model to predict the access traffic, and obtain the predicted access traffic within the time interval corresponding to the current time data; on the other hand, the service traffic can be predicted based on the historical service traffic and qualification data of each customer service, and the predicted service traffic of each customer service within the time interval can be obtained, thereby realizing the prediction of "supply"; that is, the historical service traffic and qualification data of each customer service can be input into the traffic prediction model to predict the service traffic, and obtain the predicted service traffic of each customer service within the time interval; on the basis of the supply and demand prediction, the online customer service within the time interval can be determined based on the predicted access traffic and the predicted service traffic; the difference between the sum of the predicted service traffic of the online customer service and the predicted access traffic is made less than the preset threshold, thereby realizing more reasonable scheduling, and finally creating a customer service list based on the customer service identification and customer service portrait data of the online customer service.

[0049] In the specific implementation process, time data is introduced to strengthen the supply and demand situation at different times. In addition to time data, weather data, seasonal data, etc. can also be introduced to improve the comprehensiveness and flexibility of supply and demand forecasts. For example, access traffic forecasts are made based on current time data, weather data, and historical traffic data to obtain the predicted access traffic within a time interval.

[0050] It should be noted that the customer service list can be created before the start of service interaction with users every day, or it can be created before the start of each week for each day of the week, or it can be created before the start of each week for each day of the week. This embodiment does not limit this.

[0051] On the basis of the above-mentioned predicted access traffic, if the predicted access traffic is less than the predicted threshold, it means that the access traffic in the time interval is less. In this case, it is inconvenient to manage the customer service if fewer customer service staff are arranged online, and it is also impossible to correct the deviation caused by the prediction; therefore, the online customer service can be determined according to the benchmark customer service threshold and a customer service list can be created; in order to avoid a large number of customer service staff in the customer service list being available, users who recommend services can be screened from historical visiting users to recommend services to users. In an optional implementation provided by this embodiment, if the predicted access traffic is less than the predicted threshold, historical visiting users are read, and users who recommend services are screened from historical visiting users based on user portrait features of historical visiting users. Further, the user service data of the user is obtained, and the scheduling category is determined based on the traffic status and the data category of the user service data. Correspondingly, the determined scheduling category is the first recommended scheduling category.

[0052] The above describes the creation of a customer service list and the determination of service recommendation users based on supply and demand forecasts.

[0053] In the process of determining candidate customer service in the customer service list according to the customer service determination strategy corresponding to the scheduling category, in the first optional implementation mode provided by this embodiment, according to the dialogue scheduling category, the customer service in the customer service list with the customer service status as available is read as the candidate customer service; if the reading result is not empty, the following step S208 is executed to perform customer service allocation processing to the user according to the user portrait data and the customer service portrait data of the candidate customer service; if the reading result is empty, the updated customer service is read from the customer service library, the updated customer service is updated to the customer service list, and the customer service allocation processing is performed to the user according to the user portrait data and the customer service portrait data of the updated customer service. Optionally, the dialogue scheduling category includes the first dialogue scheduling category and / or the second dialogue scheduling category.

[0054] Specifically, if the user has a need for manual dialogue, the updated customer service outside the customer service list in the customer service library can be put online for service interaction with the user.

[0055] In the second optional implementation manner provided by this embodiment, if the scheduling category is the first recommended scheduling category, the customer service in the customer service list is determined as the candidate customer service; if the scheduling category is the second recommended scheduling category, the customer service in the customer service list with an available status is read as the candidate customer service.

[0056] Specifically, if the scheduling category is the first recommended scheduling category, it means that the access traffic is relatively small, and all customer service staff in the customer service list can be identified as candidate customer service staff, so that users can be assigned to each customer service staff, so that service recommendations can be made based on user portrait data when the customer service staff is available; if the scheduling category is the second recommended scheduling category, it means that the access traffic is relatively large. If all customer service staff are identified as candidate customer service staff, it may cause the customer service staff to handle both more conversation interactions and more service recommendations, causing customer service fatigue and affecting service results. Therefore, the customer service staff in the available state in the customer service list can be identified as candidate customer service staff.

[0057] That is, in an offline state, user allocation of service recommendations to customer services can be made based on time data, historical traffic data, the customer service's historical service traffic and qualification data, which is the processing process under the first recommendation scheduling category; and in real time, user allocation of service recommendations can be made to customer services that have been available for a long time, that is, idle customer services, thereby maximizing the utilization rate of customer services and improving users' perception of services and service conversion through service recommendations.

[0058] Step S208: Allocate customer service to the user based on the user portrait data and the customer service portrait data of the candidate customer service.

[0059] In the above steps, user portrait data and candidate customer service personnel were obtained. In this step, customer service assignment is performed to the user based on the user portrait data and the customer service portrait data of the candidate customer service personnel. In this way, customer service assignment is performed based on the user portrait data and the customer service portrait data, thereby improving the effectiveness of the interactive customer service assigned to the user and thereby improving the quality of service interaction.

[0060] In an optional implementation manner provided in this embodiment, customer service portrait data is generated in the following manner: The qualification data of the candidate user is determined based on the service registration data of the candidate customer service, the historical conversation data and the third prompt text of the candidate user are input into the large language model for service label identification to obtain the service label, and / or the customer service basic data, qualification data and / or service label are input into the rating model for rating processing to obtain the customer service rating.

[0061] Optionally, the third prompt text includes a service logo identification prompt text.

[0062] In addition to the above-mentioned service labels, qualification data and customer service ratings, the customer service portrait data can also input the customer service's historical predicted traffic and historical actual traffic into the completion calculation model for completion calculation, and use the obtained customer service completion degree as one of the customer service portrait data.

[0063] In specific implementation, the user profile data and the customer service profile data of the candidate customer service can be input into the allocation model for allocation processing to obtain the interactive customer service assigned to the user. Optionally, the allocation model can be allocated in the following manner: According to the user rating in the user portrait data, select the first candidate customer service whose customer service rating matches the user rating from the candidate customer service; Screening out a second candidate customer service representative whose service tag matches the user requirement from the first candidate customer service representatives; optionally, the user requirement includes a conversation requirement and / or a recommendation requirement; An interactive customer service representative assigned to the user is determined from among the second candidate customer service representatives according to the saturation of each second candidate customer service representative.

[0064] Optionally, the saturation can be calculated based on the actual service flow and predicted service flow of the customer service; for example, the quotient of the actual service flow and the predicted service flow is used as the saturation. Specifically, the second candidate customer service with low saturation can be determined as the interactive customer service assigned to the user.

[0065] It should be noted that customer service allocation processing can be performed on users based on user portrait data and customer service data that at least includes customer service portrait data; that is, the user portrait data and customer service data that at least includes customer service portrait data are input into the allocation model for allocation processing to obtain interactive customer service allocated to the user. Optionally, the customer service data also includes the actual service flow and / or predicted service flow of the customer service. It should also be noted that the specific allocation processing process of the above allocation model is merely illustrative, and allocation processing can also be performed according to actual needs, which is not limited in this embodiment.

[0066] In order to improve the effectiveness of the customer service list, in an optional implementation manner provided by the present embodiment, after customer service is assigned to the user, the customer service status of the interactive customer service in the customer service list that performs service interaction with the user is updated to the service status, and if an instruction to end the interaction with the service interaction is detected, the actual service flow of the interactive customer service in the customer service list is updated; in order to avoid the customer service performing too many service interactions and affecting the service quality, in an optional implementation manner provided by the present embodiment, if it is detected that the actual service flow of any customer service in the customer service list is greater than the predicted service flow, any customer service in the customer service list is deleted, thereby improving the customer service's enthusiasm for service interaction.

[0067] In addition, in this embodiment, in addition to using different portrait generation models for portrait generation according to different scheduling categories, the processing model can also be configured according to different time periods, different users, and different service scenarios. For example, after obtaining user service data and determining the scheduling category according to the traffic status and the data category of the user service data, the portrait generation model is determined according to the scheduling category and time data, or the portrait generation model is determined according to the scheduling category and service scenario. Other models, such as allocation models, can be configured according to different time periods, different users, and different service scenarios. The corresponding model can be read during specific use, and this embodiment will not be repeated here.

[0068] The customer service scheduling processing method provided in this embodiment can be implemented through a customer service scheduling system. The performance of the customer service scheduling system under different conditions can be detected by simulating the actual operating environment. For example, the processing capability of the customer service scheduling system under high concurrency can be evaluated, different models and strategies can be verified, various failure scenarios can be simulated, and the fault tolerance and recovery speed of the system can be improved. In addition, indicators such as system response time, CPU (Central Processing Unit) utilization rate and / or memory utilization rate can be set, and thresholds can be set for each indicator. When the indicator exceeds the threshold, an alarm can be processed to achieve system management.

[0069] To sum up, the one or more customer service scheduling processing methods provided in this embodiment, after obtaining the user service data, perform customer service allocation processing on the user according to the user portrait data of the user and the customer service portrait data of the candidate customer service. Specifically, after obtaining the user service data, first determine the portrait generation strategy according to the traffic status and the data category of the user service data, generate the portrait according to the portrait generation model corresponding to the portrait generation strategy, obtain the user portrait data, and then determine the candidate customer service in the customer service list, and finally input the user portrait data and the customer service portrait data of the candidate customer service into the allocation model to perform customer service allocation processing on the user, and obtain the interactive customer service who interacts with the user for service. In this way, not only the accuracy of the characterization of users and customer service is improved, but also customer service allocation processing is performed on users based on the user portrait data and the customer service portrait data, thereby improving the allocation efficiency and allocation effect, and improving the matching degree between the interactive customer service and the user, thereby improving the user's interactive mentality.

[0070] The following takes the application of a customer service scheduling processing method provided by this embodiment in a dialogue scheduling scenario of a security project as an example to further illustrate the customer service scheduling processing method provided by this embodiment. Figure 3 and Figure 4 The customer service scheduling processing method applied to the dialogue scheduling scenario of the security project specifically includes the following steps.

[0071] Step S402, obtaining user conversation data and determining the traffic state according to the access traffic.

[0072] Step S404, if the traffic state is a low traffic state, the user conversation data and the user benchmark data are input into the portrait generation model to generate a portrait, obtain the user portrait data, and execute steps S414 to S416.

[0073] Step S406: If the traffic state is a high traffic state, the user conversation data and the conversation index evaluation prompt text are input into the large language model to perform conversation index evaluation to obtain the conversation index.

[0074] Step S408: If the dialogue indicator does not meet the manual customer service interaction condition, the machine customer service is determined as the interactive customer service for dialogue interaction with the user.

[0075] Step S410: If a manual interaction request submitted during the dialogue interaction between the user and the machine customer service is detected, steps S414 to S416 are executed.

[0076] If a manual interaction request submitted during the conversational interaction between the user and the machine customer service is detected, a candidate manual customer service is determined in the manual customer service list. If the candidate manual customer service read is empty, the user can be published to a publishing space that can be confirmed by the manual customer service. If an interactive request from the target manual customer service to the user is detected, the target manual customer service can be determined as the interactive customer service for the conversational interaction with the user, and after detecting that the conversational interaction between the target manual customer service and the user is completed, an interactive incentive is issued to the target manual customer service. Figure 3 As shown, in the traffic control module, users can be assigned to manual customer service through the order grabbing mode, and interactive incentives can be provided to the manual customer service who receives the order; In addition, a reservation callback function can also be configured during the traffic scheduling process. If the user submits a manual interaction request, but there is no available candidate manual customer service in the manual customer service list, a reservation callback can be made, such as recommending an available appointment time. According to the callback appointment time selected by the user in the recommended available appointment time, the interactive customer service assigned to the user will call the user back to interact with the user. Optionally, the callback can be in the form of an in-application message or a call, which is not limited in this embodiment.

[0077] Step S412: If the dialogue indicators meet the manual customer service interaction conditions, determine the portrait generation model, and input the user dialogue data and user benchmark data into the portrait generation model to generate a portrait, thereby obtaining user portrait data.

[0078] For example, Figure 3 The customer service dispatching system shown generates a user portrait through a user characterization module to obtain user portrait data. Optionally, the user portrait data includes user ratings, user purchasing power for insurance items, willingness to purchase insurance items, and / or mindset for insurance items.

[0079] Step S414: Determine a candidate human customer service representative in the human customer service representative list.

[0080] Step S416: Input the user portrait data and the customer service portrait data of the candidate human customer service into the allocation model for allocation processing to obtain an interactive customer service who interacts with the user in dialogue.

[0081] For example, Figure 3 As shown, the customer service portrait module can be used to generate a portrait of the manual customer service to obtain customer service portrait data. Optionally, the customer service portrait data includes customer service ratings, qualification data of manual customer service, service tags and / or service completion. After obtaining the user portrait data and customer service portrait data, in the real-time scheduling process, customer service allocation is performed according to the user portrait data and customer service portrait data.

[0082] Optionally, the traffic status is used to determine whether to directly assign human customer service to the user, or to first evaluate the user's conversation indicators and then determine whether to assign human customer service or machine customer service to the user based on the conversation indicators. That is, the traffic control module implements traffic control under high traffic conditions by diverting users.

[0083] It should be noted that any one of steps S402 to S416 or any combination of multiple steps can be combined with any one of steps S202 to S208 to form a new implementation method according to the needs of implementation deployment; in addition, according to the needs of actual deployment, any one or multiple technical features can be selected from steps S402 to S416 and combined with any one or multiple technical features provided by steps S202 to S208 to form a new implementation method; or, any one or multiple technical features in steps S402 to S416 can be replaced with any one or multiple technical features provided by steps S202 to S208 to form a new implementation method according to the needs of actual deployment, which will not be repeated here.

[0084] The following takes the application of a customer service scheduling processing method provided by this embodiment in an offline scheduling scenario of a security project as an example to further illustrate the customer service scheduling processing method provided by this embodiment. Figure 3 and Figure 5 The customer service scheduling processing method applied to the offline scheduling scenario of the guarantee project specifically includes the following steps.

[0085] Step S502: input the time data and historical traffic data into the traffic prediction model to predict the access traffic and obtain the predicted access traffic within the time interval.

[0086] Step S504: input the historical service flow and qualification data of each manual customer service into the flow prediction model to perform service flow prediction, and obtain the predicted service flow of each manual customer service within the time interval.

[0087] For example, Figure 3 As shown, supply and demand forecasting is performed through the supply and demand forecasting module to obtain supply-forecast service flow and demand-forecast access flow.

[0088] Step S506: determining online manual customer service within the time interval according to the predicted access traffic and the predicted service traffic of each manual customer service.

[0089] Optionally, by determining online manual customer service based on predicted access traffic and predicted service traffic, the flexibility in coping with different traffic conditions is improved.

[0090] For example, Figure 3As shown, in the real-time scheduling module, supply and demand are matched based on the predicted access traffic and predicted service traffic, so that the sum of the predicted service traffic of the determined online manual customer service is more matched with the predicted access traffic; or the difference between the sum of the predicted service traffic and the predicted access traffic is made smaller than the difference threshold, thereby realizing the scheduling of manual customer service within the time interval and ensuring that effective services can be provided even during peak hours.

[0091] Step S508: If it is detected that the predicted access traffic is less than the predicted threshold, read the historical access users.

[0092] Step S510, determine a portrait generation model, input user benchmark data of each historical visiting user into the portrait generation model to generate a portrait, and obtain user portrait data of each historical visiting user.

[0093] Step S512: input the customer service benchmark data of each online manual customer service into the portrait generation model to generate a portrait, and obtain the customer service data of each online manual customer service.

[0094] Step S514, input the user portrait data of each historical visiting user and the customer service data of each online manual customer service into the allocation model for allocation processing, and obtain the historical visiting user for service recommendation allocated to each online manual customer service.

[0095] Optionally, the number of historical access users allocated to each online manual customer service may be one or more.

[0096] It should be noted that any one of steps S502 to S514 or any combination of multiple steps can be combined with any one of steps S202 to S208 to form a new implementation method according to the needs of implementation deployment; in addition, according to the needs of actual deployment, any one or multiple technical features can be selected from steps S502 to S514 and combined with any one or multiple technical features provided by steps S202 to S208 to form a new implementation method; or, any one or multiple technical features in steps S502 to S514 can be replaced with any one or multiple technical features provided by steps S202 to S208 to form a new implementation method according to the needs of actual deployment, which will not be elaborated here.

[0097] The following takes the application of a customer service scheduling processing method provided by this embodiment in a real-time scheduling scenario of a security project as an example to further illustrate the customer service scheduling processing method provided by this embodiment. Figure 3 and Figure 6 , a customer service scheduling processing method applied to the security project scenario specifically includes the following steps.

[0098] Step S602, determining the traffic state according to the access traffic.

[0099] Step S604: If the traffic state is a low traffic state, the historical service data of the historically accessed user is read, and a portrait generation model is determined according to the data category of the historical service data.

[0100] Step S606: Input the historical service data into the portrait generation model to generate a portrait, and obtain the user portrait data of each historical visiting user.

[0101] Step S608, determining secondary interaction users from historical visiting users based on the user ratings included in the user portrait data.

[0102] Step S610: read the candidate human customer service personnel in the human customer service list that are in an available state.

[0103] Step S612: Input the user portrait data of each secondary interaction user and the customer service portrait data of each candidate human customer service into the allocation model for allocation processing, and obtain the secondary interaction user for secondary interaction allocated to each candidate human customer service.

[0104] Optionally, the situation of manual customer service is analyzed through actual access traffic, and secondary interaction users who are suitable for secondary interaction are recommended to manual customer service under low traffic conditions, so as to make service recommendations according to the recommended services in the user portrait data, thereby improving customer service utilization and user conversion.

[0105] It should be noted that any one of steps S602 to S612 or any combination of multiple steps can be combined with any one of steps S202 to S208 to form a new implementation method according to the needs of implementation deployment; in addition, according to the needs of actual deployment, any one or multiple technical features can be selected from steps S602 to S612 and combined with any one or multiple technical features provided by steps S202 to S208 to form a new implementation method; or, any one or multiple technical features in steps S602 to S612 can be replaced with any one or multiple technical features provided by steps S202 to S208 to form a new implementation method according to the needs of actual deployment, which will not be repeated here.

[0106] An embodiment of a customer service dispatch processing device provided in this specification is as follows: In the above-mentioned embodiment, a customer service scheduling processing method is provided, and correspondingly, a customer service scheduling processing device is also provided, which is described below with reference to the accompanying drawings.

[0107] Reference Figure 7, which shows a schematic diagram of an embodiment of a customer service scheduling processing device provided in this embodiment.

[0108] Since the device embodiment corresponds to the method embodiment, the description is relatively simple, and the relevant parts can refer to the corresponding description of the method embodiment provided above. The device embodiment described below is only illustrative.

[0109] This embodiment provides a customer service scheduling processing device, the device comprising: The data acquisition module 702 is configured to acquire user service data and determine a scheduling category according to a traffic state and a data category of the user service data; A portrait generation module 704 is configured to determine a portrait generation model according to the scheduling category, and input the user service data and the user benchmark data into the portrait generation model to generate a portrait, thereby obtaining user portrait data; A candidate customer service determination module 706 is configured to determine a candidate customer service in the customer service list according to a customer service determination strategy corresponding to the scheduling category; The customer service scheduling processing module 708 is configured to perform customer service allocation processing on the user based on the user portrait data and the customer service portrait data of the candidate customer service.

[0110] An embodiment of a customer service dispatch processing device provided in this specification is as follows: Corresponding to the customer service scheduling processing method described above, based on the same technical concept, one or more embodiments of this specification also provide a customer service scheduling processing device, which is used to execute the customer service scheduling processing method provided above. Figure 8 A schematic diagram of the structure of a customer service scheduling processing device provided in one or more embodiments of this specification.

[0111] This embodiment provides a customer service scheduling processing device, including: like Figure 8As shown, the customer service dispatch processing device may have relatively large differences due to different configurations or performances, and may include one or more processors 801 and memory 802, and the memory 802 may store one or more storage applications or data. Among them, the memory 802 may be a short-term storage or a persistent storage. The application stored in the memory 802 may include one or more modules (not shown in the figure), and each module may include a series of computer executable instructions in the customer service dispatch processing device. Furthermore, the processor 801 may be configured to communicate with the memory 802, and execute a series of computer executable instructions in the memory 802 on the customer service dispatch processing device. The customer service dispatch processing device may also include one or more power supplies 803, one or more wired or wireless network interfaces 804, one or more input / output interfaces 805, one or more keyboards 806, etc.

[0112] In a specific embodiment, the customer service dispatch processing device includes a memory and one or more programs, wherein the one or more programs are stored in the memory, and the one or more programs may include one or more modules, and each module may include a series of computer executable instructions in the customer service dispatch processing device, and the one or more programs are configured to be executed by one or more processors, including computer executable instructions for performing the following: Acquire user service data, and determine a scheduling category according to a traffic state and a data category of the user service data; Determine a portrait generation model according to the scheduling category, and input the user service data and user benchmark data into the portrait generation model to generate a portrait, thereby obtaining user portrait data; Determine candidate customer services in the customer service list according to the customer service determination strategy corresponding to the scheduling category; Customer service is assigned to the user based on the user portrait data and the customer service portrait data of the candidate customer service.

[0113] An embodiment of a computer-readable storage medium provided in this specification is as follows: Corresponding to the customer service scheduling processing method described above, based on the same technical concept, one or more embodiments of this specification also provide a computer-readable storage medium.

[0114] The computer-readable storage medium provided in this embodiment is used to store computer-executable instructions, and the computer-executable instructions implement the following process when executed: Acquire user service data, and determine a scheduling category according to a traffic state and a data category of the user service data; Determine a portrait generation model according to the scheduling category, and input the user service data and user benchmark data into the portrait generation model to generate a portrait, thereby obtaining user portrait data; Determine candidate customer services in the customer service list according to the customer service determination strategy corresponding to the scheduling category; Customer service is assigned to the user based on the user portrait data and the customer service portrait data of the candidate customer service.

[0115] It should be noted that an embodiment of a computer-readable storage medium in this specification and an embodiment of a customer service scheduling processing method in this specification are based on the same inventive concept, so the specific implementation of this embodiment can refer to the implementation of the aforementioned corresponding method, and the repeated parts will not be repeated.

[0116] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referenced to each other. Each embodiment focuses on the differences from other embodiments. For example, the device embodiment, equipment embodiment and computer-readable storage medium embodiment are similar to the method embodiment, so the description is relatively simple. To read the relevant contents in the device embodiment, equipment embodiment and computer-readable storage medium embodiment, please refer to the partial description of the method embodiment.

[0117] The above is a description of a specific embodiment of the specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0118] In the 1930s, improvements to a technology could be clearly distinguished as hardware improvements (for example, improvements to circuit structures such as diodes, transistors, switches, etc.) or software improvements (improvements to method flows). However, with the development of technology, many of today's method flow improvements can be regarded as direct improvements to hardware circuit structures. Designers almost always obtain the corresponding hardware circuit structure by programming the improved method flow into the hardware circuit. Therefore, it cannot be said that an improvement in a method flow cannot be implemented using hardware entity modules. For example, a programmable logic device (PLD) (such as a field programmable gate array (FPGA)) is such an integrated circuit whose logical function is determined by the user's programming of the device. Designers can "integrate" a digital system on a PLD by programming it themselves, without having to ask chip manufacturers to design and produce dedicated integrated circuit chips. Moreover, nowadays, instead of manually making integrated circuit chips, this kind of programming is mostly implemented by "logic compiler" software, which is similar to the software compiler used when developing and writing programs, and the original code before compilation must also be written in a specific programming language, which is called hardware description language (HDL). There is not only one kind of HDL, but many kinds, such as ABEL (Advanced Boolean Expression Language), AHDL (Altera Hardware Description Language), Confluence, CUPL (Cornell University Programming Language), HDCal, JHDL (Java Hardware Description Language), Lava, Lola, MyHDL, PALASM, RHDL (Ruby Hardware Description Language), etc. The most commonly used ones are VHDL (Very-High-Speed ​​Integrated Circuit Hardware Description Language) and Verilog. Those skilled in the art should also know that it is only necessary to program the method flow slightly in the above-mentioned hardware description languages ​​and program it into the integrated circuit, and then it is easy to obtain the hardware circuit that implements the logic method flow.

[0119] The controller may be implemented in any suitable manner, for example, the controller may take the form of a microprocessor or processor and a computer-readable medium storing a computer-readable program code (e.g., software or firmware) executable by the (micro)processor, a logic gate, a switch, an application-specific integrated circuit (ASIC), a programmable logic controller, and an embedded microcontroller, examples of which include but are not limited to the following microcontrollers: ARC 625D, Atmel AT91SAM, Microchip PIC18F26K20, and Silicone Labs C8051F320, and the memory controller may also be implemented as part of the control logic of the memory. It is also known to those skilled in the art that, in addition to implementing the controller in a purely computer-readable program code manner, the controller may be implemented in the form of a logic gate, a switch, an application-specific integrated circuit, a programmable logic controller, and an embedded microcontroller by logically programming the method steps. Therefore, such a controller may be considered as a hardware component, and the devices for implementing various functions included therein may also be considered as structures within the hardware component. Or even, the devices for implementing various functions may be considered as both software modules for implementing the method and structures within the hardware component.

[0120] The systems, devices, modules or units described in the above embodiments may be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, the computer may be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

[0121] For the convenience of description, the above devices are described in terms of functions and are divided into various units. Of course, when implementing the embodiments of this specification, the functions of each unit can be implemented in the same or multiple software and / or hardware.

[0122] It should be understood by those skilled in the art that one or more embodiments of the present specification may be provided as a method, system or computer program product. Therefore, one or more embodiments of the present specification may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present specification may take the form of a computer program product implemented on one or more computer-readable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0123] This specification is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of this specification. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0124] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0125] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0126] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0127] Memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.

[0128] Computer-readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer-readable instructions, data structures, program modules or other data. Examples of computer-readable storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include transitory media such as modulated data signals and carrier waves.

[0129] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of further restrictions, the elements defined by the sentence "includes at least one ..." do not exclude the presence of other identical elements in the process, method, commodity or device including the elements.

[0130] One or more embodiments of the present specification may be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data categories. One or more embodiments of the present specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules may be located in local and remote computer storage media, including storage devices.

[0131] The above description is only an embodiment of this document and is not intended to limit this document. For those skilled in the art, this document may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of this document should be included in the scope of the claims of this document.

Claims

1. A customer service dispatch processing method, comprising: Acquire user service data, and determine a scheduling category according to a traffic state and a data category of the user service data; Determine a portrait generation model according to the scheduling category, and input the user service data and user benchmark data into the portrait generation model to generate a portrait, thereby obtaining user portrait data; Determine candidate customer services in the customer service list according to the customer service determination strategy corresponding to the scheduling category; Customer service is assigned to the user based on the user portrait data and the customer service portrait data of the candidate customer service.

2. The customer service scheduling processing method according to claim 1, wherein determining the scheduling category according to the traffic status and the data category of the user service data comprises: If the data category is a conversation data category, determining the flow status according to the customer service status data of each customer service in the customer service list; If the traffic state is the first traffic state, determining the scheduling category as the first dialogue scheduling category; If the traffic state is the second traffic state, the scheduling category is determined as the second dialogue scheduling category.

3. The customer service scheduling processing method according to claim 2, wherein determining a portrait generation model according to the scheduling category, and inputting the user service data and user benchmark data into the portrait generation model to generate a portrait, and obtaining user portrait data, comprises: According to the first dialogue scheduling category, determining a large language model, a recommendation model and / or a rating model as the portrait generation model; The user service data and the first prompt text are input into the large language model for demand identification to obtain dialogue demands, the dialogue demands and / or historical behavior data are input into the recommendation model for service recommendation processing to obtain recommended services, and / or the user basic data, the dialogue demands and / or the recommended services are input into the rating model for rating processing to obtain user ratings.

4. The customer service dispatch processing method according to claim 2, wherein determining a portrait generation model according to the dispatch category, and inputting the user service data and user benchmark data into the portrait generation model to generate a portrait, thereby obtaining user portrait data, comprises: According to the second dialogue scheduling category, inputting the user service data and the second prompt text into a large language model to perform dialogue index evaluation to obtain a dialogue index; If the dialogue indicator meets the manual customer service interaction condition, the user service data and the user benchmark data are input into the portrait generation model corresponding to the first dialogue scheduling category to generate a portrait, so as to obtain user portrait data.

5. The customer service dispatch processing method according to claim 4, wherein the user service data and the second prompt text are input into a large language model for dialogue index evaluation, and after the dialogue index operation is performed, the method further comprises: If the conversation indicator does not meet the manual customer service interaction condition, the machine customer service is determined as the interactive customer service for conversation interaction with the user, so as to conduct conversation interaction with the user through the machine customer service.

6. The customer service scheduling processing method according to claim 5, wherein if the dialogue indicator does not meet the manual customer service interaction condition, the machine customer service is determined as the interactive customer service for dialogue interaction with the user, and after the dialogue interaction operation is performed through the machine customer service and the user, it also includes: According to a manual interaction request submitted by the user during the conversation with the machine customer service, marking the user as being in a manual waiting state; If it is detected that there is a customer service representative whose customer service status is available in the customer service list, the customer service representative is determined as an interactive customer service representative for dialog interaction with the user.

7. The customer service scheduling processing method according to claim 6, after the operation of marking the user as a manual waiting state according to the manual response request submitted by the user during the conversation with the machine customer service is performed, it also includes: If an interaction request of the target customer service to the user is detected, the target customer service is determined as an interactive customer service for dialog interaction with the user, and the manual waiting state mark of the user is cancelled; The target customer service is interactively marked to issue an interactive incentive to the target customer service.

8. The customer service scheduling processing method according to claim 2, wherein the step of determining candidate customer services in the customer service list according to the customer service determination strategy corresponding to the scheduling category comprises: According to the first dialogue scheduling category or the second dialogue scheduling category, a customer service whose customer service status is available in the customer service list is read as the candidate customer service.

9. The customer service scheduling processing method according to claim 8, after the sub-step of reading a customer service in the customer service list whose customer service status is available as the candidate customer service is executed according to the first dialogue scheduling category or the second dialogue scheduling category, further comprising: If the read result is empty, read and update the customer service from the customer service database; The updated customer service is updated to the customer service list, and customer service is assigned to the user based on the user portrait data and the customer service portrait data of the updated customer service.

10. The customer service scheduling processing method according to claim 1, wherein determining the scheduling category according to the traffic status and the data category of the user service data comprises: If the data category is a descriptive data category, determining a flow state according to the access flow; If the traffic state is the first traffic state, determining the scheduling category as a first recommended scheduling category; If the traffic state is the second traffic state, the scheduling category is determined as the second recommended scheduling category.

11. The customer service dispatch processing method according to claim 10, wherein determining a portrait generation model according to the dispatch category, and inputting the user service data and user benchmark data into the portrait generation model to generate a portrait, and obtaining user portrait data comprises: Determining, according to the first recommended scheduling category or the second recommended scheduling category, a demand identification model, a recommendation model and / or a rating model as the portrait generation model; The user service data is input into the demand identification model for demand identification to obtain recommended demand, the recommended demand and historical behavior data are input into the recommendation model for service recommendation processing to obtain recommended service, and / or the user basic data, the recommended demand and / or the recommended service are input into the rating model for rating processing to obtain user rating.

12. The customer service scheduling processing method according to claim 10, wherein the step of determining candidate customer services in the customer service list according to the customer service determination strategy corresponding to the scheduling category comprises: If the scheduling category is the first recommended scheduling category, determining the customer service in the customer service list as a candidate customer service; If the scheduling category is the second recommended scheduling category, a customer service whose customer service status is available in the customer service list is read as the candidate customer service.

13. The customer service scheduling processing method according to claim 1, wherein the customer service list is created in the following manner: Input the time data and historical traffic data into the traffic prediction model to predict the access traffic and obtain the predicted access traffic within the time interval; Inputting the historical service flow and qualification data of each customer service into the flow prediction model to perform service flow prediction, and obtaining the predicted service flow of each customer service within the time interval; According to the predicted access traffic and the predicted service traffic, online customer services within the time interval are determined, and the customer service list is created based on the customer service identification and customer service portrait data of the online customer services.

14. The customer service scheduling processing method according to claim 13, after the step of determining the online customer service within the time interval according to the predicted access traffic and the predicted service traffic, and creating the customer service list based on the customer service identification and customer service portrait data of the online customer service, further comprising: If the predicted access traffic is less than the predicted threshold, read the historical access users; According to the user portrait features of the historical visiting users, users who make service recommendations are screened from among the historical visiting users.

15. According to the customer service scheduling processing method of claim 1, the customer service portrait data is generated in the following manner: The qualification data of the candidate customer service is determined based on the service registration data of the candidate customer service, the historical conversation data and the third prompt text of the candidate user are input into the large language model for service label identification to obtain the service label, and / or the customer service basic data, the qualification data and / or the service label are input into the rating model for rating processing to obtain the customer service rating.

16. The customer service dispatch processing method according to claim 1, after the step of assigning a customer service to a user according to the user portrait data and the customer service portrait data of the candidate customer service is executed, further comprising: Updating the customer service status of the interactive customer service in the customer service list that performs service interaction with the user to the service status; If an interaction end instruction for the service interaction is detected, the actual service flow of the interactive customer service in the customer service list is updated.

17. The customer service dispatch processing method according to claim 1, further comprising: If it is detected that the actual service flow of any customer service in the customer service list is greater than the predicted service flow, any customer service in the customer service list is deleted.

18. A customer service dispatch processing device, comprising: A data acquisition module is configured to acquire user service data and determine a scheduling category according to a traffic state and a data category of the user service data; A portrait generation module is configured to determine a portrait generation model according to the scheduling category, and input the user service data and the user benchmark data into the portrait generation model to generate a portrait, thereby obtaining user portrait data; A candidate customer service determination module is configured to determine a candidate customer service in a customer service list according to a customer service determination strategy corresponding to the scheduling category; The customer service scheduling processing module is configured to perform customer service allocation processing on the user based on the user portrait data and the customer service portrait data of the candidate customer service.

19. A customer service dispatch processing device, comprising: processor; and a memory configured to store computer executable instructions that, when executed, cause the processor to: Acquire user service data, and determine a scheduling category according to a traffic state and a data category of the user service data; Determine a portrait generation model according to the scheduling category, and input the user service data and user benchmark data into the portrait generation model to generate a portrait, thereby obtaining user portrait data; Determine candidate customer services in the customer service list according to the customer service determination strategy corresponding to the scheduling category; Customer service is assigned to the user based on the user portrait data and the customer service portrait data of the candidate customer service.

20. A computer-readable storage medium for storing computer-executable instructions, wherein the computer-executable instructions implement the steps of the method of claim 1 when executed.

Images