Customer service resource scheduling method and device, equipment, storage medium and product

Abstract

The invention discloses a customer service resource scheduling method and device, equipment, a storage medium and a product, and relates to the technical field, the method comprises the following steps: combining pre-generated operation index labels based on a preset label combination logic to obtain a resource matching strategy; constructing a skill queue-customer service resource relationship according to the resource matching strategy; and performing customer service resource scheduling based on the skill queue-customer service resource relationship and a preset reinforcement learning DQN scheduling model, through the above scheme, combining the operation index labels based on the specific combinatorial logic to form the corresponding resource matching strategy, and further constructing the corresponding relationship between the skill queue and the customer service resources, thereby improving the resource matching efficiency. And finally, scheduling the customer service resources based on the deep learning model and the corresponding relationship between each skill queue and the customer service resources, thereby realizing accurate and efficient scheduling of the telephone customer service resources, improving the accuracy and efficiency of scheduling of the customer service resources, and optimizing resource allocation.

Description

Technical Field

[0001] The present application relates to the technical field of resource scheduling, and in particular to a customer service resource scheduling method, device, equipment, storage medium and product. Background Art

[0002] With the rapid development of information technology, telephone customer service system has become an important channel for enterprises to communicate with customers. In the modern business environment, telephone customer service is not only a service provider, but also a key point for enterprises to obtain customer feedback and maintain customer relationships. Effective telephone customer service resource management is of great significance to improving customer satisfaction, reducing operating costs, and enhancing the market competitiveness of enterprises.

[0003] Current telephone customer service management systems generally rely on traditional manual scheduling and dispatching methods. Customer service personnel are assigned to different service queues based on their respective skills to respond to customer phone inquiries. In this model, customer service personnel scheduling and skill allocation are usually based on historical data and management experience, lacking real-time data analysis and flexible adjustment mechanisms.

[0004] The above contents are only used to assist in understanding the technical solution of the present application and do not constitute an admission that the above contents are prior art. Summary of the invention

[0005] The main purpose of this application is to provide a customer service resource scheduling method, device, equipment, storage medium and computer program product, aiming to improve customer service resource scheduling efficiency and optimize resource allocation.

[0006] To achieve the above objectives, the present application proposes a customer service resource scheduling method, wherein the method comprises:

[0007] Combine the pre-generated operation indicator labels based on the preset label combination logic to obtain the resource matching strategy;

[0008] Building a skill queue-customer service resource relationship according to the resource matching strategy;

[0009] Customer service resource scheduling is performed based on the skill queue-customer service resource relationship and the preset reinforcement learning DQN scheduling model.

[0010] In one embodiment, before the step of combining the pre-generated operation indicator tags based on the preset tag combination logic to obtain the resource matching strategy, the step further includes:

[0011] Collect customer service operation data generated by the telephone customer service system through the preset stream processing engine;

[0012] The customer service operation data is analyzed and calculated through a preset streaming processing framework to generate operation indicator labels.

[0013] In one embodiment, the step of combining the operation indicator tags based on a preset tag combination logic to obtain a resource matching strategy includes:

[0014] Determine the target scenarios for customer service resource allocation;

[0015] Determine several target operation indicator labels according to the customer service resource allocation target scenario;

[0016] Determine the target tag combination logic based on the scenario characteristics of the customer service resource allocation target scenario;

[0017] The operation indicator tags are combined based on the target tag combination logic to obtain a resource matching strategy.

[0018] In one embodiment, the step of scheduling customer service resources based on the skill queue-customer service resource relationship and a preset reinforcement learning DQN scheduling model includes:

[0019] Generate a customer service resource allocation queue based on the skill queue-customer service resource relationship;

[0020] Performing customer service resource scheduling according to the customer service resource allocation queue;

[0021] Obtaining system feedback data in the customer service resource scheduling process through a preset DQN scheduling model;

[0022] The customer service resource scheduling is adjusted based on the system feedback data.

[0023] In one embodiment, the steps of constructing the DQN scheduling model include:

[0024] Cleaning the customer service operation data and the previously acquired historical customer service operation data for outliers, and extracting standardized feature data;

[0025] Constructing a model training set, a model validation set, and a model test set based on the standardized feature data;

[0026] Performing model training on the pre-built original scheduling model through the model training set to obtain a DQN scheduling training model;

[0027] Adjust the target hyperparameters of the DQN scheduling training model through the model verification set to obtain a DQN scheduling adjustment model;

[0028] After the performance of the DQN scheduling adjustment model is evaluated through the model test set, a DQN scheduling model is obtained.

[0029] In one embodiment, after the step of adjusting the customer service resource scheduling based on the system feedback data, the step further includes:

[0030] Generate target condition adjustment parameters through the DQN scheduling model;

[0031] The parameters are adjusted according to the target conditions, the customer service resource allocation queue is updated, and the steps are performed: the system feedback data in the customer service resource scheduling process is obtained through the DQN scheduling model.

[0032] In addition, to achieve the above purpose, the present application also proposes a customer service resource scheduling device, the customer service resource scheduling device comprising:

[0033] A combination module is used to combine pre-generated operation indicator labels based on a preset label combination logic to obtain a resource matching strategy;

[0034] A construction module, used to construct a skill queue-customer service resource relationship according to the resource matching strategy;

[0035] The scheduling module is used to schedule customer service resources based on the skill queue-customer service resource relationship and a preset reinforcement learning DQN scheduling model.

[0036] In addition, to achieve the above-mentioned purpose, the present application also proposes a customer service resource scheduling device, which includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, and the computer program is configured to implement the steps of the customer service resource scheduling method described above.

[0037] In addition, to achieve the above-mentioned purpose, the present application also proposes a storage medium, which is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by the processor, the steps of the customer service resource scheduling method described above are implemented.

[0038] In addition, to achieve the above-mentioned purpose, the present application also provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, the steps of the customer service resource scheduling method described above are implemented.

[0039] One or more technical solutions proposed in the present application obtain a resource matching strategy by combining pre-generated operation indicator labels based on a preset label combination logic; construct a skill queue-customer service resource relationship according to the resource matching strategy; and perform customer service resource scheduling based on the skill queue-customer service resource relationship and a preset reinforcement learning DQN scheduling model. Through the above solution, the operation indicator labels are combined based on a specific combination logic to form a corresponding resource matching strategy, and the correspondence between the skill queue and the customer service resources is further constructed. Finally, the customer service resources are scheduled based on the deep learning model and the correspondence between each skill queue and the customer service resources, thereby realizing accurate and efficient scheduling of telephone customer service resources, thereby improving the accuracy and efficiency of customer service resource scheduling and optimizing resource allocation. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0042] Figure 1 A flowchart of the first embodiment of the customer service resource scheduling method of this application is provided;

[0043] Figure 2 A schematic diagram of the system architecture of a customer service resource scheduling system corresponding to the customer service resource scheduling method of this application;

[0044] Figure 3 This is a schematic diagram of the customer service resource scheduling method label construction and application process for this application;

[0045] Figure 4 This is a flow chart of the flow processing and storage process of the customer service resource scheduling method label for this application;

[0046] Figure 5 This is a schematic diagram of the target tag combination process of the customer service resource scheduling method for this application;

[0047] Figure 6 A flowchart of the second embodiment of the customer service resource scheduling method of this application is provided;

[0048] Figure 7 This is a schematic diagram of the module structure of the customer service resource scheduling device according to an embodiment of the present application;

[0049] Figure 8Schematic diagram of the device structure of the hardware operating environment involved in the customer service resource scheduling method in the embodiment of the present application.

[0050] The purpose, features and advantages of this application will be further described in conjunction with the embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0051] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application.

[0052] In order to better understand the technical solution of the present application, a detailed description will be given below in conjunction with the accompanying drawings and specific implementation methods.

[0053] The main solution of the embodiment of the present application is to propose a real-time intelligent scheduling system for telephone customer service resources based on deep learning and expert rules. The system first collects key operational data of the telephone customer service system, such as customer service scheduling, business queue number, service connection rate and customer satisfaction rate. These data are abstracted into labels, and label rules and their value ranges are defined based on business needs and expert experience.

[0054] The label data is combined using logical operators such as AND, OR, NOT, and more complex tree combinations to form matching rules. These rules are used to filter the skill queues and corresponding customer service personnel that need to be adjusted. Through the real-time data processing system, the label data is continuously monitored and updated, and the allocation of skill queues and customer service personnel is dynamically matched and adjusted according to the predefined rule combination. A big data streaming processing system is designed, and Apache Flink is used as the streaming processing engine to collect and process the operational data of the telephone customer service system in real time. At the same time, Apache Spark is integrated for offline batch analysis to process historical data to generate long-term statistical labels. The calculated label values ​​are stored in the Redis database to support the real-time operation requirements of the system.

[0055] The Deep Q Network (DQN) model is introduced to optimize skill scheduling. The DQN model predicts changes in the customer service queue by learning historical and real-time data, and adjusts conditional parameters such as the thresholds of the queue number and service connection rate. Model training involves data preprocessing, feature extraction, model training, hyperparameter adjustment, and performance evaluation.

[0056] The trained DQN model is integrated into the real-time scheduling system. The model provides suggestions for adjusting conditional parameters. The scheduling system selects the optimal adjustment strategy based on the Q value output by DQN and applies it to actual operations. The system continuously monitors the effect of the adjustment and feeds the data back to the model to update and optimize the decision-making strategy, forming a closed-loop improvement mechanism that enables the system to dynamically adapt to changes in business needs.

[0057] This structured and intelligent approach can improve the real-time performance, flexibility and resource utilization efficiency of the telephone customer service system.

[0058] Since the current telephone customer service management system generally relies on traditional manual scheduling and dispatching methods, customer service personnel are assigned to different service queues according to their respective skills to respond to customer telephone inquiries. In this model, the scheduling and skill allocation of customer service personnel are usually based on historical data and the experience of managers, lacking real-time data analysis and flexible adjustment mechanisms.

[0059] Although existing telephone customer service management systems can provide basic customer service functions, they have some obvious defects. First, the real-time performance of manual operations is insufficient, and operational data often lags behind, which cannot reflect the actual needs of the current system in a timely manner, resulting in an inability to quickly respond to business changes and affecting overall service efficiency. Secondly, the existing system lacks structured rule precipitation, and the rules for skill adjustment mainly rely on manual experience, which limits the flexibility and scalability of the system. In addition, in the face of rapid changes and sudden scenarios, it is difficult for existing systems to respond effectively because most of them rely on prediction models based on historical data and lack real-time adjustment capabilities. Finally, the existing system lacks a closed-loop improvement mechanism, and the adjusted effect data often does not receive systematic feedback and analysis, making it difficult to continuously optimize and improve service strategies. These problems limit the performance of the telephone customer service system in terms of real-time performance, rule structuring, ability to respond to rapid changes, and closed-loop improvement mechanisms. An innovative solution is urgently needed to overcome these limitations.

[0060] The present application provides a solution, which combines pre-generated operation indicator labels based on a preset label combination logic to obtain a resource matching strategy; constructs a skill queue-customer service resource relationship according to the resource matching strategy; and performs customer service resource scheduling based on the skill queue-customer service resource relationship and a preset reinforcement learning DQN scheduling model. Through the above solution, operation indicator labels are generated based on customer service operation data, and based on the strategy and deep learning model formed by the labels, adaptive telephone customer service resource scheduling is dynamically implemented, which improves the accuracy and efficiency of scheduling, thereby improving the efficiency of customer service resource scheduling and optimizing resource allocation.

[0061] It should be noted that the execution subject of this embodiment may be a computing service device with customer service resource scheduling, network communication and program running functions, such as a tablet computer, a personal computer, a mobile phone, etc., or an electronic device capable of realizing the above functions, a customer service resource scheduling system, etc. The following takes the customer service resource scheduling system as an example to illustrate this embodiment and the following embodiments.

[0062] Based on this, the embodiment of the present application provides a customer service resource scheduling method, referring to Figure 1 , Figure 1 This is a flowchart of the first embodiment of the customer service resource scheduling method of the present application.

[0063] In this embodiment, the customer service resource scheduling method includes steps S1000 to S3000:

[0064] Step S1000: combining pre-generated operation indicator tags based on a preset tag combination logic to obtain a resource matching strategy;

[0065] In this embodiment, considering the continuous growth of business volume and the diversification of customer needs, how to efficiently manage and dispatch customer service resources has become the key to improving service quality and efficiency. To address this challenge, this solution proposes a real-time intelligent dispatching system based on deep learning and expert rules, which aims to optimize the allocation of customer service resources through real-time data analysis and intelligent decision-making. The system collects and processes customer service operation data, generates operation indicator labels, and provides decision support for subsequent resource matching and dispatching. Figure 2 , Figure 2 This is a schematic diagram of the system architecture of the customer service resource scheduling system corresponding to the customer service resource scheduling method of this application. This embodiment proposes a real-time intelligent scheduling system architecture for telephone customer service resources based on deep learning and expert rules. This architecture forms a complete intelligent scheduling solution from the underlying data source to the functional modules of the application layer, and then to the different customer service channels of the traffic layer.

[0066] At the data source layer, the system collects and stores various data from the customer service system through a variety of databases and message queue systems, such as HBASE, ClickHouse, MySQL, Kafka, and Elasticsearch. These data include customer interaction records, customer service behavior data, and operation logs, providing a basis for the system's data analysis and decision-making.

[0067] The data analysis layer is responsible for processing and analyzing the collected data. The system first processes the data in real time and offline through Flink real-time jobs and Spark offline jobs to generate operational indicator labels. These labels transform raw data into useful information that can be used for decision-making through methods such as rule calculation, dynamic functions, and frequency control. For example, the system can generate a "peak pressure" label based on the frequency of customer calls and the response time of customer service personnel, indicating that the current system is facing high service pressure.

[0068] The application layer includes various functional modules for system operation and intelligent scheduling. The policy management module allows administrators to configure and adjust scheduling policies, while the skill scheduling module dynamically matches skill queues and customer service resources according to preset rules and policies. The intelligent scheduling engine is the core of the system. It uses technical means such as rule calculation and behavior tracking to analyze and adjust the allocation of customer service resources in real time to meet the ever-changing customer needs.

[0069] The traffic layer includes different customer service channels handled by the system, including IVR self-service channels, manual telephone customer service, intelligent customer service channels, and text service channels. These channels guide customer requests to the corresponding processing queues through traffic access and distribution modules. The intelligent scheduling engine intelligently schedules these requests based on real-time data and the preset DQN scheduling model to ensure that every customer can receive timely and effective service.

[0070] In addition, the system also includes evaluation feedback and data analysis modules, which continuously optimize and adjust the scheduling strategy by collecting feedback data of scheduling results. Through this closed-loop improvement mechanism, the system can continuously learn and adapt to improve the accuracy and efficiency of scheduling. In short, the system architecture in this embodiment realizes real-time intelligent scheduling of telephone customer service resources through real-time data processing, intelligent scheduling engine and closed-loop feedback mechanism. This architecture not only improves the response speed and service quality of the customer service system, but also enhances the system's ability to respond to rapidly changing and sudden scenarios through the combination of deep learning and expert rules.

[0071] It should be noted that the resource matching strategy refers to the rules or strategies for determining how to allocate customer service resources based on the current operational indicator tags. This process involves determining the target scenarios for customer service resource allocation, such as the resource allocation measures that need to be taken during high-load periods or low satisfaction situations.

[0072] In addition, it should be noted that after determining the target operation indicator labels, the system will determine the label combination logic based on these labels and scenario characteristics. For example, if the labels "high load" and "low satisfaction" appear at the same time, the system may combine these two labels and generate a resource matching strategy to increase the number of customer service staff.

[0073] For example, in a specific implementation, the system first identifies that the current operation scenario is a high-load period, and then determines that additional customer service personnel are needed to handle a large number of customer calls based on historical data and expert experience. The system matches the "high load" label with the "increase customer service" strategy to obtain a resource matching strategy.

[0074] In a feasible implementation manner, before step S1000, steps S1100 to S1200 may be included:

[0075] Step S1100: collecting customer service operation data generated by the telephone customer service system through a preset stream processing engine;

[0076] Step S1200: Analyze and calculate the customer service operation data through a preset streaming processing framework to generate an operation indicator label.

[0077] It should be noted that in this embodiment, the customer service operation data includes various information captured in real time from the customer service system, such as customer call frequency, customer service staff working status, service response time, etc. These data are the key to evaluating the current customer service operation status and predicting future trends.

[0078] In addition, it should be noted that the operation indicator labels are obtained by analyzing and processing the original data, and they can describe the current operation status more intuitively and specifically. For example, the system may generate a label "peak hours" to indicate that the number of customer calls in the current time period has increased sharply and additional customer service support is required.

[0079] For example, in a specific implementation, the system first collects real-time data generated by the telephone customer service system through a preset stream processing engine, such as Apache Flink. Then, the system analyzes and calculates these data through a preset stream processing framework to generate operational indicator labels that reflect the current operating status, such as "high load" and "low satisfaction". These labels will serve as the basis for formulating resource matching strategies, helping the system to adjust customer service resources in real time to cope with changing customer needs.

[0080] Stream processing engines such as Apache Flink are used to collect real-time data generated by the telephone customer service system. These data include key indicators such as the frequency of customer calls, customer service response time, and service connection rate. These data streams are dynamic, continuous, and updated in real time with each customer interaction. Figure 3 , Figure 3 This is a schematic diagram of the label construction and application process of the customer service resource scheduling method of this application. In this embodiment, the system first formulates a scheduling strategy at the policy decision layer, which may be based on real-time operational data and historical scheduling results. Then, at the label data processing layer, the system conducts an in-depth analysis of the operational indicator labels to ensure that these labels can accurately reflect the current operational status and customer needs. The agent label adaptation layer and the line label adaptation layer match the appropriate customer service resources to the correct service queue based on these labels and business needs.

[0081] For example, the system may identify that a skill queue is currently facing a high load, and through the analysis of the line label adaptation layer, it is determined that customer service personnel with specific skills need to be added to meet this challenge. At the same time, the agent label adaptation layer ensures that the assigned customer service personnel have the corresponding skills and experience. Through this refined resource matching, the system can improve response speed and service quality, while optimizing the utilization efficiency of customer service resources.

[0082] In addition, it should be noted that the stream processing framework analyzes and calculates the collected real-time data to generate operational indicator labels. These labels are to transform complex data into an easy-to-understand form, such as simplifying a series of complex interaction data into labels such as "peak hours" or "low satisfaction" to quickly identify key problems and needs in the current service.

[0083] In addition, it should be noted that, referring to Figure 4 , Figure 4 This is a schematic diagram of the flow processing and storage process of the customer service resource scheduling method label of this application. In this embodiment, a big data streaming processing system (such as Apache Flink) is designed, and Apache Flink is used as a streaming processing engine to collect and process the operational data of the telephone customer service system in real time. Through the streaming processing framework, the data is analyzed and calculated in real time to generate label values. The integrated offline batch processing system Apache Spark is used for offline batch processing analysis, and historical data is processed regularly to generate long-term statistical labels. The batch processing results are combined with the real-time calculation results to provide comprehensive data support. The calculated label values ​​are stored in the Redis database to ensure high concurrency and low latency query performance. Specifically, data security is ensured through unified authentication, and the speed of data retrieval and the responsiveness of the system are improved through a multi-level caching mechanism (Redis and HBase). If the required data is not in the cache, the system will be downgraded to the source data table query to ensure the integrity and accuracy of the data.

[0084] In a feasible implementation, when the customer service system needs to query customer information or historical interaction records, the system will first check the legitimacy of the request and then try to obtain data from Redis. If there is no data in Redis, the system will query the HBase cache. If there is no data in the HBase cache either, the system will retrieve the data from the source data table in HBase. This coherent process ensures that the system can respond to data query requests quickly and securely, while optimizing the performance and efficiency of data retrieval.

[0085] For example, in a feasible implementation, the system first uses a preset stream processing engine to collect customer service operation data, which provides the original input for the generation of operation indicator labels. Subsequently, the system uses a stream processing framework to perform quasi-real-time analysis and calculation on these data to generate operation indicator labels that reflect the current operation status. These labels include but are not limited to "high load" and "low satisfaction", which provide decision support for the formulation of resource matching strategies.

[0086] After generating the operational indicator labels, the system will enter the label combination stage. In this stage, the system will combine different labels based on the preset label combination logic to form a resource matching strategy. The formulation of this strategy involves determining the target scenario for customer service resource allocation. For example, when identifying the "high load" and "low satisfaction" labels, the system may formulate a strategy to increase customer service staff or optimize service processes to meet current operational challenges.

[0087] In addition, the system will build a skill queue-customer service resource relationship based on the resource matching strategy to ensure that customer service personnel with appropriate skills can be assigned to the appropriate service queue. Finally, the system will use the deep learning DQN scheduling model to perform real-time customer service resource scheduling based on the skill queue-customer service resource relationship. The DQN model can continuously learn and optimize based on historical data and real-time feedback to further improve the accuracy and efficiency of scheduling. Through this series of steps, the system can achieve real-time intelligent scheduling of telephone customer service resources, effectively respond to various challenges in operations, improve customer satisfaction, and optimize resource utilization efficiency.

[0088] Step S2000: constructing a skill queue-customer service resource relationship according to the resource matching strategy;

[0089] Step S3000: Perform customer service resource scheduling based on the skill queue-customer service resource relationship and the preset reinforcement learning DQN scheduling model.

[0090] It should be noted that the skill queue-customer service resource relationship refers to the process of matching specific customer service skill requirements with customer service resources. The purpose of this process is to ensure that customer service personnel with appropriate skills can be assigned to the appropriate service queue to meet the customer's service needs. The basis for building this relationship is the resource matching strategy, which provides guidance on how to allocate resources. The system will match the customer service personnel's skills with customer needs according to the rules in the strategy and build the corresponding skill queue.

[0091] It should be noted that the reinforcement learning DQN scheduling model is an intelligent scheduling model based on deep learning, which can optimize scheduling decisions based on historical data and real-time feedback. The DQN model adjusts and optimizes the allocation of customer service resources by learning the effects of different scheduling strategies.

[0092] In addition, it should be noted that the purpose of customer service resource scheduling is to achieve the optimal allocation of customer service resources to improve service efficiency and customer satisfaction. The system will dynamically adjust the allocation of customer service personnel based on the skill queue-customer service resource relationship and the guidance of the DQN model.

[0093] For example, in a specific implementation, the system learns through the DQN model that increasing the number of customer service personnel in a certain period of time can significantly improve the connection rate and customer satisfaction. Therefore, based on this strategy, the system adjusts the allocation of customer service resources in real time to cope with the peak period of customer calls.

[0094] In a feasible implementation manner, step S2000 may include steps S2100 to S2400:

[0095] Determine the target scenarios for customer service resource allocation;

[0096] Step S2100: determining a number of target operation indicator tags according to the customer service resource allocation target scenario;

[0097] Step S2200: determining a target tag combination logic based on the scenario characteristics of the customer service resource allocation target scenario;

[0098] Step S2300: combining the operation indicator tags based on the target tag combination logic to obtain a resource matching strategy.

[0099] It should be noted that the process of building a resource matching strategy is a key step in achieving effective customer service resource scheduling. This process begins with determining the target scenarios for customer service resource allocation, such as peak hours, low satisfaction periods, or specific promotional events. The determination of the target scenario provides a specific direction and basis for resource allocation. Figure 5 , Figure 5 This is a flow chart of the target tag combination process of the customer service resource scheduling method of this application. In a feasible implementation, the system first defines a series of scheduling rules in the rule configuration stage, such as "If the customer waiting time exceeds 5 minutes, it is considered a high load situation." These rules are parsed and distributed in different levels. At the first level, the system may check the current waiting time and the number of customer service staff. If the waiting time exceeds the threshold, the corresponding tag rule is triggered. Each layer processes these rules concurrently and generates intermediate results based on the "or" and "and" relationships.

[0100] The system then combines these intermediate results according to the preset logical relationship. For example, if the results of the first and second layers both indicate that customer service resources need to be increased, the final decision result will clearly indicate that resource adjustments are needed. This decision result will be used as a direct basis for the scheduling system to adjust customer service resources, ensuring that the system can respond quickly and accurately based on real-time data and preset rules.

[0101] Based on these target scenarios, the system will determine a set of relevant operational indicator tags, such as "high load", "low satisfaction" or "promotional activities", which are directly related to current operational challenges and customer needs. The system will then determine the combination logic of these tags based on the characteristics of the scenario, such as historical data, customer feedback and business trends.

[0102] In addition, it should be noted that the label combination logic is a set of rules that defines how to combine different labels to form a specific resource matching strategy. For example, if the "high load" and "low satisfaction" labels appear at the same time, the system may increase the priority and trigger a specific resource matching strategy, such as increasing customer service manpower or adjusting customer service shifts.

[0103] For example, in one feasible implementation, the system first identifies that the current target scenario is the peak period, and then determines the relevant operational indicator tags based on this scenario, such as "high load" and "long waiting time". The system then determines the tag combination logic based on these tags and historical scenario characteristics, such as giving priority to solving the "high load" problem. Based on this logic, the system combines the tags into a resource matching strategy, such as increasing the number of customer service staff during peak hours and adjusting their shifts to ensure that there is enough manpower to handle the high volume of calls.

[0104] Through this process, the system can ensure that the resource matching strategy not only meets the current operational needs, but also can flexibly respond to changing customer needs and market conditions. The implementation of this strategy makes the scheduling of customer service resources more accurate and efficient, thereby improving the service quality and customer satisfaction of the entire customer service system.

[0105] The present embodiment provides a customer service resource scheduling method, which generates operation indicator labels based on pre-acquired customer service operation data; combines the operation indicator labels based on a preset label combination logic to obtain a resource matching strategy; constructs a skill queue-customer service resource relationship according to the resource matching strategy; and performs customer service resource scheduling based on the skill queue-customer service resource relationship and a preset reinforcement learning DQN scheduling model. Through the above scheme, the operation indicator labels are combined based on a specific combination logic to form a corresponding resource matching strategy, and the corresponding relationship between the skill queue and the customer service resources is further constructed. Finally, the customer service resources are scheduled based on the deep learning model and the corresponding relationship between each skill queue and the customer service resources, thereby realizing accurate and efficient scheduling of telephone customer service resources, thereby improving the accuracy and efficiency of customer service resource scheduling and optimizing resource allocation.

[0106] Based on the first embodiment of the present application, in the second embodiment of the present application, the same or similar contents as those in the above-mentioned embodiment 1 can be referred to the above introduction, and will not be repeated in the following. Figure 6 , step S3000 of the customer service resource scheduling method further includes steps S3100 to S3400:

[0107] Step S3100: generating a customer service resource allocation queue based on the skill queue-customer service resource relationship;

[0108] It should be noted that the skill queue-customer resource relationship refers to the process of matching customer service skills and needs with available customer service resources. The purpose of generating customer service resource allocation queues is to create a priority queue based on the current skill queue and customer service resource status to ensure correct and efficient allocation of customer service resources.

[0109] For example, in a specific implementation, the system will analyze the current skill queue and customer service resource status, and generate a customer service resource allocation queue taking into account factors such as the customer service skill level, past performance, and customer needs. This queue may give priority to customer service staff with high skill matching and short response time to improve service efficiency and customer satisfaction.

[0110] Step S3200: performing customer service resource scheduling according to the customer service resource allocation queue;

[0111] It should be noted that customer service resource scheduling refers to assigning customer service personnel to appropriate service locations or skill queues based on the generated allocation queues. The purpose of this step is to ensure that the right customer service personnel are available to provide services to customers at the right time and place.

[0112] In addition, it should be noted that in this embodiment, the system will adjust the allocation of customer service personnel in real time according to the allocated queue. For example, when it is identified that the customer waiting time in a certain skill queue is long, the system may allocate more customer service personnel to the queue to reduce customer waiting time.

[0113] Step S3300: obtaining system feedback data in the customer service resource scheduling process through a preset DQN scheduling model;

[0114] It should be noted that the DQN scheduling model is a scheduling model based on reinforcement learning, which can continuously learn and optimize based on historical data and real-time feedback. The purpose of obtaining system feedback data is to evaluate the effect of the current scheduling strategy and provide a basis for further optimization.

[0115] For example, in a specific implementation, the DQN scheduling model analyzes feedback data during the customer service resource scheduling process, such as customer satisfaction, customer service processing time, and service success rate, to evaluate the performance of the current scheduling strategy and find room for improvement.

[0116] Step S3400: adjusting the customer service resource scheduling based on the system feedback data.

[0117] It should be noted that adjusting customer service resource scheduling based on system feedback data is to continuously optimize scheduling strategies to adapt to changing customer needs and operating conditions. The purpose of this step is to achieve dynamic adjustment and improve resource utilization efficiency and service quality.

[0118] In addition, it should be noted that the system may adjust the allocation ratio of customer service personnel, optimize the skill matching algorithm, or adjust the scheduling strategy based on feedback data to ensure that the scheduling strategy can better meet actual needs.

[0119] For example, in a specific implementation, if the system feedback data shows that customer satisfaction has decreased over a certain period of time, the system may analyze the reasons and adjust the scheduling strategy accordingly, such as increasing the number of customer service staff during peak hours or adjusting the customer service staff's shift plan to improve service quality. Through this real-time feedback and adjustment mechanism, the system can continuously improve the effect of customer service resource scheduling.

[0120] In a feasible implementation manner, the steps of constructing the DQN scheduling model include steps A1000 to A5000:

[0121] Step A1000: cleaning the customer service operation data and the pre-acquired historical customer service operation data for outliers, and extracting standardized feature data;

[0122] Step A2000: constructing a model training set, a model validation set and a model test set based on the standardized feature data;

[0123] Step A3000: Performing model training on the pre-built original scheduling model using the model training set to obtain a DQN scheduling training model;

[0124] Step A4000: adjusting the target hyperparameters of the DQN scheduling training model through the model verification set to obtain a DQN scheduling adjustment model;

[0125] Step A5000: After the performance of the DQN scheduling adjustment model is evaluated through the model test set, a DQN scheduling model is obtained.

[0126] It should be noted that the construction of the DQN scheduling model is a systematic process involving multiple links such as data preprocessing, model training, hyperparameter adjustment and performance evaluation. Outlier cleaning is a key step in data preprocessing, and its purpose is to ensure the quality of data used in model training, thereby improving the accuracy and robustness of the model. The extraction of standardized feature data is to convert the data into a unified format, eliminate the influence of different magnitudes and dimensions, and enable the model to learn the relationship between data more accurately.

[0127] In addition, it should be noted that building a model training set, a model validation set, and a model test set is a standard practice in the machine learning modeling process. The model training set is used to train the original scheduling model, the model validation set is used to adjust the model's hyperparameters, and the model test set is used to evaluate the model's performance. The reasonable division of these three data sets ensures that the model can be fully trained and can also be effectively verified and tested.

[0128] For example, in a feasible implementation, the system first conducts a comprehensive analysis of customer service operation data and historical customer service operation data, cleans up outliers through preset rules and algorithms, and extracts standardized feature data. Next, the system divides these data into training sets, validation sets, and test sets. Using the training set, the system trains an initial DQN scheduling model, and then adjusts the model's hyperparameters through the validation set to optimize the model's performance. Finally, the system uses the test set to evaluate the adjusted model to ensure that it meets the preset performance standards. Through this series of steps, the system can build an efficient and accurate DQN scheduling model to provide decision support for customer service resource scheduling.

[0129] In addition, it should be noted that the training and adjustment process of the DQN scheduling model is a dynamic and iterative process. The system will continuously collect new operational data to update and optimize the model. In this way, the DQN scheduling model can gradually adapt to new operating environments and demand changes, thereby achieving continuous performance improvement.

[0130] In this embodiment, the construction and application of the DQN scheduling model makes customer service resource scheduling more intelligent and automated. The model can not only respond quickly based on real-time data, but also learn from past scheduling experience and continuously optimize scheduling strategies. Through this data-driven intelligent scheduling method, the system can significantly improve the utilization efficiency of customer service resources and ensure that high-quality services can be provided to customers in different operating scenarios.

[0131] In a feasible implementation manner, step S3400 may include steps S3410 to S3420:

[0132] Step S3410: generating target condition adjustment parameters through the DQN scheduling model;

[0133] Step S3420: adjusting parameters according to the target conditions, updating the customer service resource allocation queue, and executing the step of obtaining system feedback data in the customer service resource scheduling process through the DQN scheduling model.

[0134] It should be noted that the DQN model generates target condition adjustment parameters and updates the customer service resource allocation queue in real time to ensure that scheduling decisions can adapt to current operating conditions and customer needs.

[0135] In this embodiment, the system first analyzes real-time operation data and historical data through the DQN scheduling model to identify key factors that affect customer service efficiency and customer satisfaction. Then, the model generates target condition adjustment parameters, which include but are not limited to the allocation ratio of customer service personnel, skill matching threshold, and customer waiting time. Based on these adjustment parameters, the system updates the customer service resource allocation queue in order to achieve a better resource configuration.

[0136] In addition, it should be noted that after updating the customer service resource allocation queue, the system will continue to monitor the system feedback data during the scheduling process. These data include key indicators such as customer satisfaction, customer service processing time, and service connection rate. The system analyzes these feedback data to evaluate the effectiveness of the current scheduling strategy, and further adjusts the DQN model accordingly to achieve closed-loop optimization.

[0137] For example, in one feasible implementation, the system identifies through the DQN model that during peak hours, customer satisfaction is highly correlated with the response speed of customer service personnel. Therefore, the target condition adjustment parameter generated by the model may indicate an increase in the allocation ratio of highly skilled customer service personnel during peak hours. The system updates the customer service resource allocation queue based on this parameter so that customer requests can be responded to quickly during peak hours. Subsequently, the system collects updated scheduling feedback data, such as the improvement in customer satisfaction, to evaluate the effect of the adjustment. If the data shows that customer satisfaction has improved, the system will continue to use the current strategy; if the effect is not good, the system will further adjust the parameters through the DQN model to seek a better scheduling strategy.

[0138] Through this dynamic adjustment mechanism based on the DQN scheduling model, the system can continuously learn and optimize, continuously improve the efficiency and effectiveness of customer service resource scheduling, and ultimately achieve real-time intelligent scheduling of customer service resources.

[0139] It should be noted that the above examples are only used to understand the present application and do not constitute a limitation on the customer service resource scheduling method of the present application. More simple transformations based on this technical concept are all within the protection scope of the present application.

[0140] Please refer to Figure 7 The present application also provides a customer service resource scheduling device, the customer service resource scheduling device comprising:

[0141] A combination module 71, used to combine pre-generated operation indicator labels based on a preset label combination logic to obtain a resource matching strategy;

[0142] A construction module 72, configured to construct a skill queue-customer service resource relationship according to the resource matching strategy;

[0143] The scheduling module 73 is used to schedule customer service resources based on the skill queue-customer service resource relationship and the preset reinforcement learning DQN scheduling model.

[0144] The customer service resource scheduling device provided by the present application adopts the customer service resource scheduling method in the above embodiment, which can solve the technical problem of customer service resource scheduling. Compared with the prior art, the beneficial effects of the customer service resource scheduling device provided by the present application are the same as the beneficial effects of the customer service resource scheduling method provided by the above embodiment, and other technical features in the customer service resource scheduling device are the same as the features disclosed in the above embodiment method, which will not be repeated here.

[0145] The present application provides a customer service resource scheduling device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the customer service resource scheduling method in the above-mentioned embodiment one.

[0146] Reference below Figure 8 , which shows a schematic diagram of the structure of a customer service resource scheduling device suitable for implementing the embodiment of the present application. The customer service resource scheduling device in the embodiment of the present application may include but is not limited to mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 8 The customer service resource scheduling device shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0147] like Figure 8As shown, the customer service resource scheduling device may include a processing device 1001 (e.g., a central processing unit, a graphics processor, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM: Read Only Memory) 1002 or a program loaded from a storage device 1003 to a random access memory (RAM: Random Access Memory) 1004. In RAM1004, various programs and data required for the operation of the customer service resource scheduling device are also stored. The processing device 1001, ROM1002, and RAM1004 are connected to each other through a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Generally, the following systems can be connected to the I / O interface 1006: an input device 1007 including, for example, a touch screen, a touch pad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output device 1008 including, for example, a liquid crystal display (LCD: Liquid Crystal Display), a speaker, a vibrator, etc.; a storage device 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 can allow the customer service resource scheduling device to communicate with other devices wirelessly or by wire to exchange data. Although the figure shows a customer service resource scheduling device with various systems, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems can be implemented or have alternatively.

[0148] In particular, according to the embodiments disclosed in the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiment disclosed in the present application are executed.

[0149] The customer service resource scheduling device provided in the present application adopts the customer service resource scheduling method in the above-mentioned embodiment. Compared with the prior art, the beneficial effects of the customer service resource scheduling device provided in the present application are the same as the beneficial effects of the customer service resource scheduling method provided in the above-mentioned embodiment, and other technical features in the customer service resource scheduling device are the same as the features disclosed in the method of the previous embodiment, which will not be repeated here.

[0150] It should be understood that the various parts disclosed in this application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0151] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0152] The present application provides a computer-readable storage medium having computer-readable program instructions (ie, computer programs) stored thereon, and the computer-readable program instructions are used to execute the customer service resource scheduling method in the above-mentioned embodiment.

[0153] The computer-readable storage medium provided in the present application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, systems or devices, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, system or device. The program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination of the above.

[0154] The computer-readable storage medium may be included in the customer service resource scheduling device; or may exist independently without being assembled into the customer service resource scheduling device.

[0155] Computer program code for performing the operations of the present application may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0156] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present application. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0157] The modules involved in the embodiments described in this application may be implemented by software or hardware, wherein the name of the module does not constitute a limitation on the unit itself in some cases.

[0158] The readable storage medium provided in this application is a computer-readable storage medium, which stores computer-readable program instructions (i.e., computer programs) for executing the above-mentioned customer service resource scheduling method, and can solve the technical problem of customer service resource scheduling. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as the beneficial effects of the customer service resource scheduling method provided in the above-mentioned embodiment, and will not be repeated here.

[0159] The present application also provides a computer program product, including a computer program, which implements the steps of the customer service resource scheduling method as described above when executed by a processor.

[0160] The computer program product provided in this application can solve the technical problem of customer service resource scheduling. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as the beneficial effects of the customer service resource scheduling method provided in the above embodiment, which will not be repeated here.

[0161] The above descriptions are only some embodiments of the present application, and are not intended to limit the patent scope of the present application. All equivalent structural changes made using the contents of the present application specification and drawings under the technical concept of the present application, or direct / indirect applications in other related technical fields are included in the patent protection scope of the present application.

Claims

1. A customer service resource scheduling method, characterized in that: The method comprises: Combine the pre-generated operation indicator labels based on the preset label combination logic to obtain the resource matching strategy; Building a skill queue-customer service resource relationship according to the resource matching strategy; Customer service resource scheduling is performed based on the skill queue-customer service resource relationship and the preset reinforcement learning DQN scheduling model.

2. The method according to claim 1, characterized in that Before the step of combining the pre-generated operation indicator labels based on the preset label combination logic to obtain the resource matching strategy, the step further includes: Collect customer service operation data generated by the telephone customer service system through the preset stream processing engine; The customer service operation data is analyzed and calculated through a preset streaming processing framework to generate operation indicator labels.

3. The method according to claim 1, characterized in that The step of combining the operation indicator labels based on the preset label combination logic to obtain the resource matching strategy includes: Determine the target scenarios for customer service resource allocation; Determine several target operation indicator labels according to the customer service resource allocation target scenario; Determine the target tag combination logic based on the scenario characteristics of the customer service resource allocation target scenario; The operation indicator tags are combined based on the target tag combination logic to obtain a resource matching strategy.

4. The method according to claim 1, characterized in that The step of scheduling customer service resources based on the skill queue-customer service resource relationship and the preset reinforcement learning DQN scheduling model includes: Generate a customer service resource allocation queue based on the skill queue-customer service resource relationship; Performing customer service resource scheduling according to the customer service resource allocation queue; Obtaining system feedback data in the customer service resource scheduling process through a preset DQN scheduling model; The customer service resource scheduling is adjusted based on the system feedback data.

5. The method according to claim 4, characterized in that The steps of constructing the DQN scheduling model include: Cleaning the customer service operation data and the previously acquired historical customer service operation data for outliers, and extracting standardized feature data; Constructing a model training set, a model validation set, and a model test set based on the standardized feature data; Performing model training on the pre-built original scheduling model through the model training set to obtain a DQN scheduling training model; Adjust the target hyperparameters of the DQN scheduling training model through the model verification set to obtain a DQN scheduling adjustment model; After the performance of the DQN scheduling adjustment model is evaluated through the model test set, a DQN scheduling model is obtained.

6. The method according to claim 5, characterized in that After the step of adjusting the customer service resource scheduling based on the system feedback data, the following step further includes: Generate target condition adjustment parameters through the DQN scheduling model; The parameters are adjusted according to the target conditions, the customer service resource allocation queue is updated, and the steps are performed: the system feedback data in the customer service resource scheduling process is obtained through the DQN scheduling model.

7. A customer service resource scheduling device, characterized in that: The device comprises: A combination module is used to combine pre-generated operation indicator labels based on a preset label combination logic to obtain a resource matching strategy; A construction module, used to construct a skill queue-customer service resource relationship according to the resource matching strategy; The scheduling module is used to schedule customer service resources based on the skill queue-customer service resource relationship and a preset reinforcement learning DQN scheduling model.

8. A customer service resource scheduling device, characterized in that: The device comprises: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the customer service resource scheduling method according to any one of claims 1 to 6.

9. A storage medium, characterized in that: The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, the steps of the customer service resource scheduling method according to any one of claims 1 to 6 are implemented.

10. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the steps of the customer service resource scheduling method according to any one of claims 1 to 6 are implemented.

Images