Distribution model training method, distribution method and device

Abstract

The invention discloses a training method of a distribution model, and a distribution method and device, and relates to the technical field of computers. A specific embodiment of the training method of the distribution model comprises the following steps: determining a customer service quality category of a customer service and a customer quality category of a customer; determining an association relationship between the customer service and the customer according to the customer service quality category and the customer quality category; constructing a training sample set based on the customer service quality category, the customer quality category and the association relationship, training by using the training sample set to obtain a distribution model, and distributing the customer to the customer service based on the distribution model; customer service staff and customers of different quality levels can be analyzed and distributed. A specific embodiment of the distribution method comprises the steps of determining a target customer service corresponding to each to-be-distributed customer in a to-be-distributed customer list by using a distribution model based on the obtained to-be-distributed customer list; and each to-be-allocated customer is allocated to the corresponding target customer service, so that the marketing efficiency and the marketing effect are improved.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and in particular to a training method, allocation method, and apparatus for allocation models. Background Technology

[0002] Currently, the common marketing method is to conduct telephone marketing to selected target customers through dedicated customer service representatives. Target customers are randomly assigned to customer service representatives from the entire customer base. However, due to the varying marketing skills of different customer service representatives, the marketing results can differ even when facing the same target customer.

[0003] In the process of realizing this invention, the inventors discovered at least the following problems in the related technology:

[0004] Random allocation methods are inefficient and have poor marketing results. Summary of the Invention

[0005] In view of this, embodiments of the present invention provide a training method, allocation method, and apparatus for an allocation model, which can analyze and allocate customer service representatives and customers of different quality levels, thereby improving marketing efficiency and marketing effectiveness.

[0006] To achieve the above objectives, according to one aspect of the present invention, a method for training an allocation model is provided, comprising:

[0007] Determine the customer service quality category and the customer's customer quality category;

[0008] Determine the relationship between customer service representatives and customers based on customer service quality categories and customer quality categories;

[0009] A training sample set is constructed based on customer service quality categories, customer quality categories, and relationships. An assignment model is trained using the training sample set to assign customers to customer service representatives based on the assignment model.

[0010] Optionally, the customer service quality category is determined by a clustering model, which is trained through the following steps:

[0011] Obtain basic customer service profile information and marketing behavior information;

[0012] Extract profile features from basic profile information and marketing features from marketing behavior information;

[0013] Using profile features and marketing features as input data for the clustering model, and customer service quality categories as output data, the clustering model is trained based on the first objective function and / or the second objective function.

[0014] Optionally, the first objective function is to maximize the similarity of customer service quality within each preset category group; the second objective function is to minimize the similarity of customer service quality between each preset category group.

[0015] Optionally, a customer's customer quality category is determined using a customer quality classification model, which is trained through the following steps:

[0016] Obtain the intention score, approval score, and overall score of customers in the customer list;

[0017] The intention score, approval score, and comprehensive score are weighted to obtain the final score;

[0018] The customers in the customer list are sorted based on the final score, and the labeling results of the customer quality category are determined based on the sorting results;

[0019] A customer quality classification model is trained based on the customer profile features and the labeling results of customer quality categories.

[0020] Optionally, the relationship between customer service representatives and customers can be determined based on customer service quality categories and customer quality categories, including:

[0021] Obtain historical task volume data corresponding to each customer service quality category;

[0022] Determine the customer volume data corresponding to each customer quality category;

[0023] Based on historical task volume data and customer volume data, a pre-established relationship model is used to obtain the relationship between customer service representatives to be assigned and customers to be assigned.

[0024] Optionally, the pre-established relationship model is obtained through the following steps:

[0025] Decision variables are set based on the historical task volume data corresponding to each customer service quality category and the customer volume data corresponding to each customer quality category;

[0026] The constraints and objective function are constructed based on the decision variables; the objective function is to maximize the number of customers whose applications are approved.

[0027] Establish a correlation model based on constraints and objective functions.

[0028] According to a second aspect of the present invention, an allocation method is provided, comprising:

[0029] Based on the obtained list of customers to be assigned, the target customer service representative corresponding to each customer in the list is determined using an allocation model; wherein, the allocation model is obtained by any of the methods in the first aspect of the present invention.

[0030] Each customer to be assigned is assigned to the corresponding target customer service representative.

[0031] According to a third aspect of the present invention, a training apparatus for an allocation model is provided, comprising:

[0032] The category determination module is used to determine the customer service quality category of the customer service representative and the customer quality category of the customer.

[0033] The relationship determination module is used to determine the relationship between customer service representatives and customers based on customer service quality categories and customer quality categories.

[0034] The model training module is used to build a training sample set based on customer service quality categories, customer quality categories, and relationships. The training sample set is used to train an allocation model, which is then used to allocate customers to customer service representatives.

[0035] According to a fourth aspect of the present invention, a dispensing apparatus is provided, comprising:

[0036] The customer service identification module is used to determine the target customer service representative corresponding to each customer in the acquired list of customers to be assigned, using an allocation model; wherein the allocation model is obtained by any of the methods in the first aspect of the present invention.

[0037] The customer service assignment module is used to assign each customer to a corresponding target customer service representative.

[0038] According to a fifth aspect of the present invention, an electronic device is provided, comprising:

[0039] One or more processors;

[0040] Storage device for storing one or more programs.

[0041] When one or more programs are executed by one or more processors, the one or more processors implement the methods of any of the above embodiments.

[0042] According to a sixth aspect of the present invention, a computer-readable medium is provided having a computer program stored thereon, which, when executed by a processor, implements the method of any of the above embodiments.

[0043] According to a seventh aspect of the present invention, a computer program product is provided, comprising a computer program that, when executed by a processor, implements the method of any of the above embodiments.

[0044] One embodiment of the above invention has the following advantages or beneficial effects: by determining the customer service quality category of the customer service representative and the customer quality category of the customer; determining the association between the customer service representative and the customer based on the customer service quality category and the customer quality category; constructing a training sample set based on the customer service quality category, the customer quality category, and the association, and using the training sample set to train an allocation model, so as to allocate customers to customer service representatives based on the allocation model; thereby, it is possible to analyze and allocate customer service representatives and customers with different quality levels, thereby improving marketing efficiency and marketing effectiveness. Based on the obtained list of customers to be allocated, the allocation model is used to determine the target customer service representative corresponding to each customer to be allocated in the list; wherein, the allocation model is obtained by any of the methods in the first aspect of the present invention; each customer to be allocated is assigned to the corresponding target customer service representative, thereby improving the matching rate between customer service representatives and customers and marketing effectiveness.

[0045] The further effects of the aforementioned unconventional alternative methods will be explained below in conjunction with specific implementation methods. Attached Figure Description

[0046] The accompanying drawings are provided to better understand the invention and are not intended to unduly limit the scope of the invention. Wherein:

[0047] Figure 1 This is a schematic diagram of the main flow of the training method of the allocation model according to an embodiment of the present invention;

[0048] Figure 2 This is a schematic diagram of the main flow of a training method for an allocation model according to a possible embodiment of the present invention;

[0049] Figure 3 This is a schematic diagram of the main flow of a training method for an allocation model according to another applicable embodiment of the present invention;

[0050] Figure 4 This is a schematic diagram of the main flow of a training method for an allocation model according to another embodiment of the present invention;

[0051] Figure 5 This is a schematic diagram of the main flow of the allocation method according to an embodiment of the present invention;

[0052] Figure 6 This is a schematic diagram of the main modules of the training device for the allocation model according to an embodiment of the present invention;

[0053] Figure 7 This is a schematic diagram of the main modules of the distribution device according to an embodiment of the present invention;

[0054] Figure 8 This is an exemplary system architecture diagram in which embodiments of the present invention can be applied;

[0055] Figure 9 This is a schematic diagram of the structure of a computer system suitable for implementing terminal devices or servers of the present invention. Detailed Implementation

[0056] The following description, in conjunction with the accompanying drawings, illustrates exemplary embodiments of the present invention, including various details to aid understanding. These details should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the invention. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0057] It should be noted that the acquisition, storage, and application of personal information involved in the embodiments of the present invention comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0058] Currently, the common marketing method involves dedicated customer service representatives making phone calls to selected target customers. Target customers are randomly assigned to customer service representatives from the entire customer base. However, due to the varying marketing skills of different customer service representatives, the marketing results can differ even when facing the same target customer, leading to low marketing efficiency and poor marketing results.

[0059] In view of this, according to one aspect of the present invention, a method for training an allocation model is provided.

[0060] Figure 1 This is a schematic diagram illustrating the main flow of the training method for the allocation model according to an embodiment of the present invention. Figure 1 As shown, the training method of the allocation model according to an embodiment of the present invention mainly includes the following steps S101 to S103.

[0061] Step S101: Determine the customer service quality category of the customer service representative and the customer quality category of the customer.

[0062] Customer service quality categories refer to classifying customer service staff into different levels or categories based on factors such as their marketing capabilities, communication skills, and historical performance. For example, there could be categories like "high-quality customer service," "medium-quality customer service," and "low-quality customer service," or categories like "Class A customer service," "Class B customer service," "Class C customer service," and "Class D customer service." Customer quality categories refer to classifying customers into different levels or categories based on factors such as their level of interest, historical consumption records, and interaction frequency. For example, there could be categories like "high-value customers," "medium-value customers," and "low-value customers," or categories like "Class A customers," "Class B customers," "Class C customers," and "Class D customers." The specific form of these categories is not limited in this embodiment of the invention.

[0063] Determining customer service quality categories can be based on assessments of historical sales data, customer satisfaction ratings, and frequency of target achievement. Alternatively, it can be achieved through regular skills tests and training results to measure staff competence, or by collecting and analyzing customer feedback as a key evaluation criterion. Determining customer quality categories can be achieved by analyzing historical purchase records, spending amounts, and purchase frequency. It can also be done by assessing customer activity and interest through behavioral data such as interaction logs, click-through rates, and number of inquiries. Furthermore, a comprehensive evaluation combining market research and customer background information can be used to predict future purchasing potential.

[0064] Step S102: Determine the relationship between customer service representatives and customers based on customer service quality category and customer quality category.

[0065] The relationship between customer service representatives and customers refers to the strategic relationship of assigning customer service representatives of different quality categories to customers of corresponding different quality categories. Through the assignment strategy, each customer can communicate or interact with the most suitable customer service representative to maximize conversion rate and customer satisfaction, thereby achieving the best marketing results.

[0066] The relationship between customer service representatives and customers can be allocated based on the representative's capabilities and the customer's value. For example, high-quality representatives can be assigned to high-value customers, medium-quality representatives to medium-value customers, and low-quality representatives to low-value customers. Alternatively, high-quality, medium-quality, and low-quality representatives can be assigned different proportions of high-value, medium-value, and low-value customers, respectively. Furthermore, adjustments can be made dynamically based on real-time data and feedback. The relationship can be adjusted in real-time based on changes in actual marketing effectiveness and customer service performance. For example, if a representative has recently performed exceptionally well, they can be temporarily assigned more high-value customers. Additionally, to prevent customer service fatigue or customer dependency, customers of different qualities can be rotated among representatives of different qualities. This ensures fair resource allocation, maintains good working conditions for representatives, and prevents over-concentration on one type of customer. Machine learning algorithms can also be used for intelligent allocation based on extensive historical data and the current context. The algorithm continuously learns and optimizes the relationship to find the optimal combination of representative and customer allocation, thereby improving overall marketing effectiveness. It is worth noting that the above methods can be flexibly combined and applied according to specific circumstances, and this embodiment of the invention does not impose specific limitations.

[0067] Specifically, clear classification rules and allocation algorithms can be formulated. For example, a certain score range can be set to divide customer service representatives and customers into different quality categories. The algorithm can then be used to allocate customers based on these scores, such as assigning high-scoring customers to high-scoring customer service representatives and low-scoring customers to low-scoring customer service representatives, and so on.

[0068] Step S103: Construct a training sample set based on customer service quality category, customer quality category and relationship, and use the training sample set to train an allocation model to allocate customers to customer service representatives based on the allocation model.

[0069] Specifically, a large amount of historical data is collected and organized as the foundation for the training sample set, including customer service quality categories, customer quality categories, customer service and customer allocation records, and corresponding marketing performance data. When constructing the training sample set, each sample can contain the following information: customer service quality category, customer quality category, the relationship between the two, and the marketing performance after allocation (such as conversion rate, customer satisfaction, sales, etc.). Analyzing the samples can reveal the optimal allocation strategy between customer service representatives of different quality categories and customers of different quality categories. The allocation model is trained using the constructed training sample set. The allocation model can employ machine learning algorithms such as decision trees, random forests, support vector machines, or neural networks. During model training, the model learns the impact of different allocation strategies on marketing performance and identifies the optimal allocation scheme. After training, the allocation model can automatically allocate customers to the most suitable customer service representatives based on new customer service and customer data. For example, when the system receives a high-value customer, the model will prioritize assigning it to a high-quality customer service representative based on the strategies learned during training, or assign it to a specific person within a high-quality customer service representative group, depending on the specific circumstances. Similarly, for medium- or low-value customers, the model will allocate them based on their characteristics and the optimal allocation strategy.

[0070] This invention, through its embodiments, determines the customer service quality category of customer service representatives and the customer quality category of customers; it then determines the relationship between customer service representatives and customers based on these categories; a training sample set is constructed based on the customer service quality category, customer quality category, and relationship; and an allocation model is trained using this training sample set to assign customers to customer service representatives. This enables the analysis and allocation of customer service representatives and customers with different quality levels, thereby improving marketing efficiency and effectiveness. The trained allocation model allows for intelligent allocation of customer service representatives and customers, improving overall marketing efficiency and customer satisfaction, fully leveraging the potential of customer service representatives, and optimizing resource allocation.

[0071] Optionally, Figure 2 This is a schematic diagram illustrating the main flow of a training method for an allocation model according to a possible embodiment of the present invention. Figure 2 As shown, the customer service quality category is determined by a clustering model, which is trained through steps S201 to S203.

[0072] Step S201: Obtain basic customer service profile information and marketing behavior information.

[0073] Step S202: Extract profile features from basic profile information and marketing features from marketing behavior information.

[0074] Step S203: Using profile features and marketing features as input data for the clustering model, and customer service quality categories as output data for the clustering model, the clustering model is trained based on the first objective function and / or the second objective function to obtain the clustering model.

[0075] Basic profile information refers to data about customer service personnel's personal and professional background, including their age, gender, education, work experience, and educational background. Marketing behavior information refers to data on customer service performance or behavioral patterns generated during their work, including sales performance, customer satisfaction ratings, call duration, and conversion rates. Profile features are feature variables extracted from basic profile information for modeling and analysis. These can be directly used basic profile information or variables obtained through data processing and feature engineering, such as age, gender, education level, and years of work experience. Marketing features are feature variables extracted from marketing behavior information for modeling and analysis. These can be directly used marketing behavior information or obtained through data processing and feature engineering, such as indicators like sales performance, customer feedback ratings, call frequency, and number of successful conversions.

[0076] Specifically, basic profile information can be obtained from internal human resources databases or employee file systems, while marketing behavior information can be obtained from customer service systems or other relevant systems. Profile features are extracted from the basic profile information, such as age as a continuous variable, gender converted to a binary variable, and educational background and work experience encoded as categorical variables. Marketing features are extracted from the marketing behavior information, such as monthly sales, average customer satisfaction ratings, and average call duration as continuous variables, and successful conversion rate as a ratio variable. These profile and marketing features are used as input data to the model, with customer service quality categories as output data. The model is trained by setting a first objective function and a second objective function to obtain a clustering model. The clustering model can automatically classify customer service representatives into different customer service quality categories, thus providing data support for subsequent customer matching and resource optimization.

[0077] Optionally, the first objective function can be to maximize the similarity of customer service quality within each preset category group; the second objective function can be to minimize the similarity of customer service quality between each preset category group. Alternatively, the first objective function can also be to minimize the variance within a category, and the second objective function can also be to maximize the distance between categories.

[0078] Through continuous iteration and optimization, this invention trains a clustering model that can accurately classify customer service quality categories, automatically assigning customer service representatives to different categories. Furthermore, by utilizing clustering models based on profile features and marketing features, customer service representatives can be evaluated and classified more scientifically, thereby improving the accuracy of customer service classification.

[0079] Optionally, customer service quality categories can also be determined through a comprehensive analysis of performance evaluations and customer feedback. For example, this involves regularly assessing key performance indicators (KPIs) such as sales performance, customer satisfaction ratings, and the time and quality of task completion. Furthermore, customer feedback can be collected and analyzed through questionnaires or telephone follow-ups to understand customer evaluations of customer service quality and attitude. Based on this data, customer service can be categorized into high-quality, medium-quality, and low-quality categories.

[0080] Optionally, Figure 3 This is a schematic diagram of the main flow of a training method for an allocation model according to another applicable embodiment of the present invention. Figure 3 As shown, the customer quality category is determined by the customer quality classification model, which is trained through the following steps S301 to S304.

[0081] Step S301: Obtain the intention score, approval score, and overall score of customers in the customer list.

[0082] Step S302: Weight the intention score, batch score, and comprehensive score to obtain the final score.

[0083] Step S303: Sort the customers in the customer list based on the final score, and determine the labeling result of the customer quality category based on the sorting result.

[0084] Step S304: Based on the customer profile features and the labeling results of customer quality categories, a customer quality classification model is trained.

[0085] Step S301 involves obtaining the customer's intention score, approval score, and overall score from the customer list. The intention score assesses the customer's willingness to apply for a credit card and can be calculated based on the customer's level of interest in the product or service, such as through indicators like historical purchase records, browsing behavior, and number of inquiries. The approval score assesses the customer's creditworthiness and can be a score based on the customer's credit assessment or qualification review, such as through information like credit score, income level, and repayment ability. The overall score assesses the combined performance of the customer's intention to apply for a credit card and their creditworthiness, and can be a score based on the customer's level of interest and credit history. In addition, customer intent scores, approval scores, and comprehensive scores can be obtained through intent models, approval models, and comprehensive models, respectively. The intent model takes as input all customers who have been contacted and whose calls have been answered (along with customer profile characteristics), and outputs an intent score for each customer. A higher intent score indicates a stronger interest in applying for a credit card. The approval model takes as input all customers who have applied for a credit card (along with customer profile characteristics), and outputs an approval score for each customer. A higher approval score indicates better creditworthiness and a higher likelihood of approval. The comprehensive model takes as input all customers who have been contacted and whose calls have been answered (along with customer profile characteristics), and outputs a comprehensive score for each customer. A higher score indicates a stronger interest in applying for a credit card and better credit. The intent model filters out customers who are interested in applying for a credit card from those who have been contacted; the approval model filters out those who are interested and whose applications will be approved by the bank; and the comprehensive model directly filters out those who are interested in applying for a credit card and whose applications will be approved by the bank from those who have been contacted.

[0086] Specifically, step S302 weights the obtained intention score, batch approval score, and overall score. Weighting involves assigning different weights to each score based on their importance to customer quality, and then calculating a weighted average. For example, the weights for the intention score, batch approval score, and overall score can be set to 0.4, 0.3, and 0.3, respectively, resulting in a final score for each customer. An online experiment can also be conducted (reaching out marketing to the group of customers with the highest final scores). If the business feedback indicates that few of the reached customers are interested, the weight of the intention model can be increased (equivalent to increasing the influence of the intention score on the final score), and vice versa.

[0087] In step S303, customers in the customer list are sorted based on the final score. Customers in each quality category are divided equally according to the number of people. The labeling results of the customer quality category are determined. Customers with higher final scores are considered high-quality customers, customers with medium scores are medium-quality customers, and customers with lower scores are low-quality customers. Each customer is labeled with its quality category.

[0088] In step S304, customer profile features include basic customer information and behavioral information, such as age, gender, geographic location, occupation, and income level, as well as behavioral information such as purchase history, browsing history, and interaction frequency. Using these profile features and labeled customer quality categories, a model capable of automatically classifying customer quality is built through training using machine learning algorithms (such as decision trees, random forests, or support vector machines).

[0089] The embodiments of the present invention can effectively classify customers into different quality categories, which is conducive to more accurate customer management and marketing strategy formulation.

[0090] Alternatively, customer quality categories can be determined through a combination of human scoring and experience-based judgment, such as market research, customer interviews, and questionnaires to understand customer needs, spending power, and purchasing intentions; or, a comprehensive assessment can be conducted by combining historical transaction data, interaction records, and feedback information. By regularly reviewing and adjusting the scoring criteria, the accuracy and dynamic updating of customer quality categories can be ensured, leveraging human experience to supplement the shortcomings of purely data-driven models.

[0091] Figure 4 This is a schematic diagram illustrating the main flow of a training method for an allocation model according to another embodiment of the present invention. Figure 4 As shown, the relationship between customer service representatives and customers is determined based on customer service quality category and customer quality category, including the following steps S401 to S403.

[0092] Step S401: Obtain the historical task volume data corresponding to each customer service quality category.

[0093] Step S402: Determine the customer volume data corresponding to each customer quality category.

[0094] Step S403: Based on historical task volume data and customer volume data, use a pre-established relationship model to obtain the relationship between the customer service representatives to be assigned and the customers to be assigned.

[0095] Specifically, step S401 extracts relevant data from the customer service management system or data warehouse. This data includes the number of tasks handled by each customer service representative over a past period, their efficiency and quality in completing those tasks, etc. This can be achieved through database queries or data reporting tools, statistically analyzing historical task volume data for high-quality, medium-quality, and low-quality customer service representatives. Step S402 retrieves relevant customer information from the customer relationship management system or marketing database. This information includes each customer's basic information, historical purchase records, number of interactions, etc. Customers are categorized into high-value, medium-value, and low-value categories, and the number of customers in each category is counted. This can be done by writing queries or using data analysis tools to summarize and statistically analyze customers by category, obtaining customer volume data for each quality category. In step S403, the relationship model can be based on the previous allocation strategy and optimization algorithms, such as machine learning models or linear programming models. Historical task volume data and customer volume data are used as input variables and substituted into the association model for calculation. The association model generates the optimal allocation plan based on the customer service's processing capacity and the customer's demand level, and determines the best matching combination between each customer service representative to be assigned and the customer to be assigned. The process is automated through programming, and the allocation results are output as an actionable allocation list to guide the actual customer allocation work, thereby improving overall service efficiency and customer satisfaction.

[0096] Optionally, the pre-established correlation model is obtained through the following steps: setting decision variables based on the historical task volume data corresponding to each customer service quality category and the customer volume data corresponding to each customer quality category; constructing constraints and objective functions based on the decision variables; wherein, the objective function is to maximize the number of customers approved; and establishing a correlation model based on the constraints and objective function.

[0097] Specifically, decision variables are set based on historical task volume data corresponding to each customer service quality category and customer volume data corresponding to each customer quality category. Decision variables can represent the allocation of customer service staff to customers, for example, defined as: x ij Where i represents the customer service quality category, j represents the customer quality category, and x ijThis represents the number of customer service representatives of type i to be assigned to customer type j. Constraints and an objective function are constructed based on the decision variables. Constraints may include: the total workload of each customer service type cannot exceed its historical average workload; all customers in each type should be assigned; and the task allocation for each customer service quality category should match its actual processing capacity. The objective function is to maximize the number of customers whose applications are approved. This number can be estimated using the known success rate of customer service representative and customer assignment. A correlation model is established based on the constraints and objective function. This correlation model can be solved using a linear programming algorithm (such as the simplex method) to obtain the optimal solution, which is then used for customer and customer service assignment.

[0098] The embodiments of the present invention can determine the optimal association scheme between each type of customer service and each type of customer by establishing an association relationship model, thereby maximizing the number of approved customers and improving the overall marketing and service effect.

[0099] According to a second aspect of the present invention, an allocation method is provided.

[0100] Figure 5 This is a schematic diagram of the main flow of the allocation method according to an embodiment of the present invention; as shown Figure 5 As shown, the allocation method according to an embodiment of the present invention mainly includes the following steps S501 to S502.

[0101] Step S501: Based on the obtained list of customers to be assigned, the target customer service representative corresponding to each customer in the list of customers to be assigned is determined using the assignment model; wherein, the assignment model is obtained by any of the methods in the first aspect of the present invention.

[0102] Step S502: Assign each customer to be assigned to the corresponding target customer service representative.

[0103] Specifically, the information of each customer in the list of customers to be assigned is input into the assignment model. The model calculates and outputs the most suitable target customer service category and specific customer service representative for each customer based on their profile characteristics (such as age, gender, purchase history, and interaction records). By analyzing the input features and utilizing a pre-trained assignment strategy, the model determines which customer service representative is best suited for each customer. Based on the model's output, an assignment list is generated, listing each customer and their corresponding target customer service representative. Through a customer relationship management system or internal task assignment system, the customer information is automatically assigned to the corresponding customer service accounts. Notifications can also be sent to relevant customer service representatives to inform them of the newly assigned customer list. Simultaneously, the task status in the system is updated to ensure that all customers are effectively assigned without omissions or duplications.

[0104] The embodiments of the present invention can efficiently and accurately assign customers to the most suitable target customer service representatives, optimize resource allocation, improve service quality and customer satisfaction, and enhance marketing efficiency and effectiveness.

[0105] According to a preferred embodiment of the present invention, in the marketing of bank credit card applications, intelligent list allocation is used to achieve optimal marketing conversion results for agents. This embodiment provides four categories of customer lists (A, B, C, and D), which are distributed to agents of categories A, B, C, and D respectively, with the ultimate goal of maximizing revenue. The customer lists of categories A, B, C, and D can be target customers screened from the entire customer base using a model. Target customers are categorized into four types based on their card application intention and credit rating (category A customers have the highest quality data, i.e., strong card application intention and good credit, followed by category B, and so on). The agent categories A, B, C, and D are categorized based on their marketing ability (category A has the strongest marketing ability, followed by category B, and so on). The marketing ability of agents is evaluated based on their historical call data and conversion rates.

[0106] When clustering and modeling agent quality, the goal is to identify four clusters where agents within each cluster have generally similar quality, while there are significant differences in agent quality between different clusters. The model focuses on agent quality, including features such as patience, education level, and business acumen. Generally, even with the same outbound call list, different agents will achieve varying conversion rates. Therefore, the model can be built based on basic agent profile information (such as age, gender, and education) and business acumen, adding marketing features (such as patience, instant hang-up rate, and approval rate). Using these profile and marketing features, an agent quality clustering model is constructed, ultimately grouping agents into four clusters, labeled A, B, C, and D (where A has the highest quality, followed by B, and so on).

[0107] When classifying and modeling outbound call lists, the lists are categorized according to their final conversion rates. A general credit card application model can be used for multi-model fusion scoring. The intention model models customers' willingness to apply for a credit card; the approval model models customers' creditworthiness; and the comprehensive model models both customers' willingness and creditworthiness. Each customer is assigned a score using these three models, and a final score is obtained by weighted summing of the three scores. This final score represents the overall conversion rate for that customer; a higher final score indicates a higher likelihood of conversion. Finally, the outbound call lists are sorted according to their final scores, divided into four categories based on the number of people. For example, if there are 1 million lists, each with a corresponding final score, the lists are sorted in descending order of their final scores. The top 250,000 lists are designated as A (highest overall score), the middle 250,000 as B, the next middle 250,000 as C, and the bottom 250,000 as D (lowest overall score). The final list was divided into four categories: A, B, C, and D (with A being the highest quality, followed by B, and so on), thus completing the quality classification model for outbound call lists.

[0108] After completing the clustering modeling for agent quality and the classification modeling for outbound call list quality, the next step is to model the relationship between agents and outbound call lists. The outbound call list quality classification model determines the size of the customer list for each category, along with other relevant data. For example, as shown in Table 1 below, the size of the customer list for categories A, B, C, and D is only c1, c2, c3, and c4, respectively.

[0109]

[0110] As shown in Table 2 below, the lead generation rate is the ratio of the number of leads to the number of connected calls. Taking E11 as an example, E11 represents the lead generation rate of a Class A agent on a Class A customer list. That is, if a Class A agent obtains a Class A list during outbound marketing, the lead generation rate of this list will reach E11 (where E11 = potential cardholders / outbound call reach). Assuming E11 is 1%, this means that for every 100 customers reached by a Class A agent, there will be one potential customer.

[0111]

[0112] As shown in Table 3 below, the approval rate is the ratio of the number of approved calls to the number of incoming calls. Taking F11 as an example, F11 represents the approval rate of Class A agents on the Class A customer list. That is, if a Class A agent obtains a Class A list during outbound marketing, the approval rate of this list will reach F11 (where F11 = approved customers / interested customers). Assuming F11 is 25%, this means that for every 100 interested customers reached by a Class A agent, 25 will be approved.

[0113]

[0114] As shown in Table 4 below, the allocation of customer lists of different categories on customer service seats of different categories. The i-th type of list in the table represents the classification of customers, and the j-th type of seat represents the classification of customer service. Specifically, the rows represent the categories of customer lists, which are four categories of customer lists from A(1) to D(4) respectively, and the columns represent the categories of customer service seats, which are four categories of customer service seats from Jia(1) to Ding(4) respectively. Xij represents the magnitude of the j-th type of list assigned to the i-th type of seat. For example, X11 represents the number of the 1st type of customer list assigned to the 1st type of customer service seat, X32 represents the number of the 3rd type of customer list assigned to the 2nd type of customer service seat, and so on.

[0115]

[0116] Based on the above data, a linear programming model is constructed, and the objective function is shown in Equation (1):

[0117] MAX(Z) = (1)

[0118] Among them, Xij represents the magnitude of the j-th type of list assigned to the i-th type of seat, Eij represents the incoming rate of the i-th type of seat on the j-th type of list, and Fij represents the approval rate of the i-th type of seat on the j-th type of list. Xij×Eij×Fij represents the number of people finally approved after allocating the Xij magnitude of the j-th type of list to the i-th type of seat.

[0119] The constraint conditions can include list magnitude constraints, incoming volume limits, approval volume limits, approval rate limits, and so on.

[0120] Among them, the list magnitude constraints are shown in Equations (2) to (5). The consumption of each type of seat for a certain level of list cannot be greater than the magnitude requirement of the list.

[0121] X11+X12+X13+X14<=c1 (2)

[0122] X21+X22+X23+X24<=c2 (3)

[0123] X31+X32+X33+X34<=c3 (4)

[0124] X41+X42+X43+X44<=c4 (5)

[0125] The number of incoming calls is limited as shown in equations (6) to (9). Based on the average approval level of the four types of agents (A, B, C, and D) in historical data, the number of incoming calls for each type of agent is required to ensure that the agents have a certain normal income (the agent's salary is generally proportional to the number of incoming calls, so if an agent's number of incoming calls is too low in a day, it will directly affect the agent's income and may eventually lead to the agent's loss). Therefore, in order to ensure that the agent's income is not affected when intelligent allocation is carried out, it is necessary to constrain the average number of incoming calls for each type of agent. According to historical call records, the average number of incoming calls for the four types of agents (A, B, C, and D) is y1, y2, y3, and y4, respectively.

[0126] X11×E11+X12×E12+X13×E13+X14×E14>=y1 (6)

[0127] X21×E21+X22×E22+X23×E23+X24×E24>=y2 (7)

[0128] X31×E31+X32×E32+X33×E33+X34×E34>=y3 (8)

[0129] X41×E41+X42×E42+X43×E43+X44×E44>=y4 (9)

[0130] The approval quantity limit is shown in equations (10) to (13). Based on the average approval level of the four types of agents (A, B, C, and D) in historical data, the approval level of each type of agent is required to ensure that the final performance of intelligent allocation is greater than that of non-intelligent allocation. According to historical call records, the average number of approvals for the four types of agents (A, B, C, and D) are z1, z2, z3, and z4, respectively.

[0131] X11×F11+X12×F12+X13×F13+X14×F14>=z1 (10)

[0132] X21×F21+X22×F22+X23×F23+X24×F24>=z2 (11)

[0133] X31×F31+X32×F32+X33×F33+X34×F34>=z3 (12)

[0134] X41×F41+X42×F42+X43×F43+X44×F44>=z4 (13)

[0135] The approval rate limit is shown in equation (14). In actual business, most banks not only have requirements for the number of applications and the number of approvals, but also have requirements for the approval rate (approval rate / application rate). Generally, banks will propose a specific percentage P for the approval rate, such as 30%. Therefore, this constraint needs to be considered in the optimization algorithm.

[0136] ∑(Xij×Eij×Fij) / ∑(Xij×Eij)>=P (14)

[0137] In formula (14), Xij represents the magnitude of the j-list assigned to i-type agents, Eij represents the application rate of i-type agents on the j-list, Fij represents the approval rate of i-type agents on the j-list, P represents the required approval rate, approval rate = number of approved applicants / number of card applicants, number of approved applicants = Xij × Eij × Fij, number of card applicants = Xij × Eij.

[0138] Given the objective function and the four corresponding constraints (where only Xij is the variable to be solved, and the rest are actual values ​​that can be calculated in advance and replaced directly with the actual values ​​when solving the objective function), the linear programming algorithm is selected to solve for the optimal solution of intelligent allocation. Finally, Xij is calculated, and the number of seats that each type of agent should be allocated to in the four grade lists of A, B, C, and D can be obtained.

[0139] This invention utilizes an algorithmic model to model the differences between customer list quality and agent quality, enabling the allocation of customer lists of specific quality to agents with specific marketing capabilities. This differs from the conventional method of randomly distributing all lists to all agents (where even assigning high-quality lists to agents with low marketing capabilities results in poor conversion rates and wasted lists). This invention's intelligent agent list distribution method significantly improves marketing conversion rates. Current experimental results show significant improvements in key business indicators such as application processing rate and approval rate, leading to a 20%-25% increase in final performance. This invention has high practicality and business value in scenarios similar to credit card application marketing.

[0140] According to a third aspect of the present invention, a training apparatus for an allocation model.

[0141] Figure 6 This is a schematic diagram of the main modules of the training device for the allocation model according to an embodiment of the present invention, as shown below. Figure 6 As shown, the training device 600 for the allocation model mainly includes:

[0142] Category determination module 601 is used to determine the customer service quality category of the customer service representative and the customer quality category of the customer.

[0143] The relationship determination module 602 is used to determine the relationship between customer service representatives and customers based on customer service quality categories and customer quality categories.

[0144] The model training module 603 is used to construct a training sample set based on customer service quality category, customer quality category and relationship, and to train an allocation model using the training sample set to allocate customers to customer service based on the allocation model.

[0145] Optionally, the training device 600 for the allocation model further includes a clustering model training module, which is used for:

[0146] Obtain basic customer service profile information and marketing behavior information;

[0147] Extract profile features from basic profile information and marketing features from marketing behavior information;

[0148] Using profile features and marketing features as input data for the model, and customer service quality categories as output data, a clustering model is trained based on a first objective function and / or a second objective function.

[0149] Optionally, in the clustering model training module, the first objective function is to maximize the similarity of customer service quality within each preset category group; the second objective function is to minimize the similarity of customer service quality between each preset category group.

[0150] Optionally, the training device 600 for the allocation model also includes a customer quality classification model training module, which is used for:

[0151] Obtain the intention score, approval score, and overall score of customers in the customer list;

[0152] The intention score, approval score, and comprehensive score are weighted to obtain the final score;

[0153] The customers in the customer list are sorted based on the final score, and the labeling results of the customer quality category are determined based on the sorting results;

[0154] A customer quality classification model is trained based on the customer profile features and the labeling results of customer quality categories.

[0155] Optionally, the relationship determination module 602 is also used for:

[0156] Obtain historical task volume data corresponding to each customer service quality category;

[0157] Determine the customer volume data corresponding to each customer quality category;

[0158] Based on historical task volume data and customer volume data, a pre-established relationship model is used to obtain the relationship between customer service representatives to be assigned and customers to be assigned.

[0159] Optionally, the training device 600 for the allocation model further includes an association model training module, which is used for:

[0160] Decision variables are set based on the historical task volume data corresponding to each customer service quality category and the customer volume data corresponding to each customer quality category;

[0161] The constraints and objective function are constructed based on the decision variables; the objective function is to maximize the number of customers whose applications are approved.

[0162] Establish a correlation model based on constraints and objective functions.

[0163] It should be noted that the specific implementation details of the training device for the allocation model in this embodiment of the invention have been described in detail in the training method for the allocation model above, so the details will not be repeated here.

[0164] According to a fourth aspect of the present invention, a dispensing device is provided.

[0165] Figure 7 This is a schematic diagram of the main modules of the distribution device according to an embodiment of the present invention, such as... Figure 7 As shown, a dispensing device 700 includes:

[0166] The customer service determination module 701 is used to determine the target customer service representative corresponding to each customer in the customer list to be assigned based on the obtained customer list to be assigned and using an allocation model; wherein, the allocation model is obtained by any of the methods in the first aspect of the present invention.

[0167] The customer service assignment module 702 is used to assign each customer to be assigned to the corresponding target customer service representative.

[0168] It should be noted that the specific implementation details of the distributing device in the embodiments of the present invention have been described in detail in the above distributing method, so the details will not be repeated here.

[0169] According to a fifth aspect of the present invention, an electronic device is provided, comprising:

[0170] One or more processors;

[0171] Storage device for storing one or more programs.

[0172] When one or more programs are executed by one or more processors, the one or more processors implement the methods provided by the first aspect and / or the second aspect of the embodiments of the present invention.

[0173] According to a sixth aspect of the present invention, a computer-readable medium is provided having a computer program stored thereon, which, when executed by a processor, implements the methods provided in the first aspect and / or the second aspect of the present invention.

[0174] According to a seventh aspect of the present invention, a computer program product is provided, including a computer program that, when executed by a processor, implements the method provided in the first aspect of the present invention.

[0175] Figure 8 An exemplary system architecture 800 is shown, which can be used to train the allocation model according to embodiments of the present invention, or to train the allocation model.

[0176] like Figure 8 As shown, system architecture 800 may include terminal devices 801, 802, and 803, a network 804, and a server 805. Network 804 serves as the medium for providing communication links between terminal devices 801, 802, and 803 and server 805. Network 804 may include various connection types, such as wired or wireless communication links or fiber optic cables, etc.

[0177] Customers can use terminal devices 801, 802, and 803 to interact with server 805 via network 804 to receive or send messages, etc. Various communication client applications can be installed on terminal devices 801, 802, and 803, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, social media platform software, etc. (for example only).

[0178] Terminal devices 801, 802, and 803 can be various electronic devices with displays and web browsing capabilities, including but not limited to smartphones, tablets, laptops, and desktop computers.

[0179] Server 805 can be a server providing various services, such as a backend management server supporting shopping websites browsed by customers using terminal devices 801, 802, and 803 (for example only). The backend management server can analyze and process data such as received requests for prediction of item attribute preferences, and feed back the processing results (for example only) to the terminal devices.

[0180] It should be noted that the training method of the allocation model provided in this embodiment of the invention is generally executed by server 805, and correspondingly, the training device of the allocation model is generally set in server 805. The training method of the allocation model provided in this embodiment of the invention can also be executed by terminal devices 801, 802, and 803, and correspondingly, the training device of the allocation model can be set in terminal devices 801, 802, and 803.

[0181] It should be noted that the allocation method provided in this embodiment of the invention is generally executed by server 805, and correspondingly, the allocation device is generally located in server 805. The allocation method provided in this embodiment of the invention can also be executed by terminal devices 801, 802, and 803, and correspondingly, the allocation device can be located in terminal devices 801, 802, and 803.

[0182] It should be understood that Figure 8 The number of terminal devices, networks, and servers shown is merely illustrative. Depending on implementation needs, any number of terminal devices, networks, and servers can be included.

[0183] The following is for reference. Figure 9 It shows a schematic diagram of the structure of a computer system 900 suitable for implementing a terminal device of the present invention. Figure 9 The terminal device shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of the present invention.

[0184] like Figure 9 As shown, the computer system 900 includes a central processing unit (CPU) 901, which can perform various appropriate actions and processes based on programs stored in read-only memory (ROM) 902 or programs loaded from storage section 908 into random access memory (RAM) 903. The RAM 903 also stores various programs and data required for the operation of the system 900. The CPU 901, ROM 902, and RAM 903 are interconnected via a bus 904. An input / output (I / O) interface 905 is also connected to the bus 904.

[0185] The following components are connected to I / O interface 905: an input section 906 including a keyboard, mouse, etc.; an output section 907 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and speakers, etc.; a storage section 908 including a hard disk, etc.; and a communication section 909 including a network interface card such as a LAN card, modem, etc. The communication section 909 performs communication processing via a network such as the Internet. A drive 910 is also connected to I / O interface 905 as needed. A removable medium 911, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on drive 910 as needed so that computer programs read from it can be installed into storage section 908 as needed.

[0186] In particular, according to the embodiments disclosed in this invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 909, and / or installed from removable medium 911. When the computer program is executed by central processing unit (CPU) 901, it performs the functions defined above in the system of this invention.

[0187] It should be noted that the computer-readable medium shown in this invention can be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this invention, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In this invention, a computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media can also be any computer-readable medium other than computer-readable storage media, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wireless, wire, optical fiber, RF, etc., or any suitable combination thereof.

[0188] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0189] The modules described in the embodiments of the present invention can be implemented in software or hardware. The described modules can also be housed in a processor. For example, a processor may include a category determination module, a model training module, and a relationship determination module. The names of these modules do not necessarily limit the module itself. For example, the category determination module may also be described as "a module for determining the customer service quality category of a customer service representative and the customer quality category of a customer." Alternatively, a processor may include a customer service determination module and a customer service allocation module. The names of these modules do not necessarily limit the module itself. For example, the customer service determination module may also be described as "a module for determining the target customer service representative corresponding to each customer in the customer list to be allocated, based on an acquired list of customers to be allocated, using an allocation model."

[0190] In another aspect, the present invention also provides a computer-readable medium, which may be included in the device described in the above embodiments; or it may exist independently and not assembled into the device. The computer-readable medium carries one or more programs, which, when executed by the device, implement the following method: determining the customer service quality category of a customer service representative and the customer quality category of a customer; determining the association relationship between the customer service representative and the customer based on the customer service quality category and the customer quality category; constructing a training sample set based on the customer service quality category, the customer quality category, and the association relationship; training an allocation model using the training sample set; and allocating customers to customer service representatives based on the allocation model. Alternatively, the device implements the following method: based on an acquired list of customers to be allocated, determining the target customer service representative corresponding to each customer to be allocated in the list using the allocation model; wherein the allocation model is obtained by any of the methods in the first aspect of the present invention; and allocating each customer to be allocated to the corresponding target customer service representative.

[0191] According to the technical solution of the present invention, the customer service quality category of the customer service representative and the customer quality category of the customer are determined; the relationship between the customer service representative and the customer is determined based on the customer service quality category and the customer quality category; a training sample set is constructed based on the customer service quality category, the customer quality category and the relationship, and an allocation model is trained using the training sample set to allocate customers to customer service representatives based on the allocation model; thereby, it is possible to analyze and allocate customer service representatives and customers with different quality levels, thereby improving marketing efficiency and marketing effectiveness; based on the obtained list of customers to be allocated, the target customer service representative corresponding to each customer to be allocated in the list is determined using the allocation model; wherein, the allocation model is obtained by any of the methods in the first aspect of the present invention; each customer to be allocated is assigned to the corresponding target customer service representative, thereby improving the matching rate between customer service representatives and customers and the marketing effectiveness.

[0192] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can occur depending on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

[0193] It should be noted that the acquisition, storage, and application of customer personal information involved in the technical solution disclosed herein comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

Claims

1. A training method for an allocation model, characterized in that, include: Determine the customer service quality category and the customer's customer quality category; The relationship between customer service representatives and customers is determined based on the customer service quality category and the customer quality category. A training sample set is constructed based on the customer service quality category, customer quality category, and relationship. An allocation model is trained using the training sample set to allocate customers to customer service representatives based on the allocation model.

2. The method according to claim 1, characterized in that, The customer service quality category is determined by a clustering model, which is trained through the following steps: Obtain basic customer service profile information and marketing behavior information; Image features are extracted from the basic image information, and marketing features are extracted from the marketing behavior information; The clustering model is obtained by using the profile features and marketing features as input data and customer service quality categories as output data, and training based on the first objective function and / or the second objective function.

3. The method according to claim 2, characterized in that, The first objective function is to maximize the similarity of customer service quality within each preset category group; the second objective function is to minimize the similarity of customer service quality between each preset category group.

4. The method according to claim 1, characterized in that, The customer's customer quality category is determined by a customer quality classification model, which is trained through the following steps: Obtain the intention score, approval score, and overall score of customers in the customer list; The intention score, approval score, and comprehensive score are weighted to obtain the final score; Based on the final score, the customers in the customer list are sorted, and the labeling result of the customer quality category is determined according to the sorting result; The customer quality classification model is trained based on the customer profile features and the labeling results of customer quality categories.

5. The method according to claim 1, characterized in that, The relationship between customer service representatives and customers is determined based on the customer service quality category and customer quality category, including: Obtain historical task volume data corresponding to each customer service quality category; Determine the customer volume data corresponding to each customer quality category; Based on the historical task volume data and customer volume data, the relationship between the customer service representatives to be assigned and the customers to be assigned is obtained using a pre-established relationship model.

6. The method according to claim 5, characterized in that, The pre-established relationship model is obtained through the following steps: Decision variables are set based on the historical task volume data corresponding to each customer service quality category and the customer volume data corresponding to each customer quality category; Based on the decision variables, construct constraints and an objective function; wherein, the objective function is to maximize the number of customers whose applications are approved. A correlation model is established based on the constraints and objective function.

7. An allocation method, characterized in that, include: Based on the obtained list of customers to be assigned, the target customer service representative corresponding to each customer in the list is determined using an allocation model; wherein, the allocation model is obtained by the method described in any one of claims 1 to 6; Each of the customers to be assigned is assigned to the corresponding target customer service representative.

8. A training apparatus for an allocation model, characterized in that, include: The category determination module is used to determine the customer service quality category of the customer service representative and the customer quality category of the customer. The relationship determination module is used to determine the association relationship between customer service representatives and customers based on the customer service quality category and the customer quality category. The model training module is used to construct a training sample set based on the customer service quality category, customer quality category and association relationship, and to train an allocation model using the training sample set to allocate customers to customer service representatives based on the allocation model.

9. A dispensing device, characterized in that, include: The customer service identification module is used to determine the target customer service representative corresponding to each customer in the acquired list of customers to be assigned, using an allocation model; wherein the allocation model is obtained by the method described in any one of claims 1 to 6. The customer service assignment module is used to assign each customer to be assigned to the corresponding target customer service representative.

10. An electronic device, characterized in that, include: One or more processors; Storage device for storing one or more programs. When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any one of claims 1-7.

11. A computer-readable medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the method as described in any one of claims 1-7.

12. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1-7.

Images