Service agent distribution method and device, computer equipment and storage medium

Abstract

The embodiment of the invention belongs to the field of financial science and technology, and relates to a service agent distribution method and device, computer equipment and a storage medium, and the method comprises the following steps: obtaining agent historical service data; performing feature extraction on the seat historical service data to obtain seat capability multi-dimensional feature information; obtaining customer historical communication data, and obtaining customer demand information according to the customer historical communication data; inputting the customer demand information and the seat capability multi-dimensional feature information into a preset matching network to obtain a matching score; and according to a preset matching rule and the matching score, carrying out customer and service seat distribution. According to the invention, the client and the service seat can be correspondingly distributed, so that the service adaptability of the client and the service seat is effectively improved, and the service experience feeling of the client is improved.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, specifically to the field of financial technology, and in particular to a service seat allocation method, device, computer equipment and storage medium. Background Art

[0002] Modern financial customers tend to interact with banks or financial institutions through multiple channels, such as telephone, online chat, social media, mobile applications, etc. Among them, customer service agents, as manual services, have been providing professional consultation and answers to financial customers' questions, thereby effectively solving customer needs and improving customer satisfaction.

[0003] In the business system of financial institutions, as the business services provided by the business system, such as product transactions, after-sales, insurance consulting, etc., involve more and more professional business fields, it is necessary to overcome the need for seats to have certain professional knowledge and capabilities to meet customer service needs. However, in the actual service process, when many customer service seats face complex problems, due to large differences in knowledge levels and skills, and uneven service quality, some seats are unable to accurately understand customer needs, provide effective solutions or handle customer complaints, thus affecting customer experience and loyalty, and bringing certain impacts to the development of financial business. Summary of the invention

[0004] The purpose of the embodiments of the present application is to provide a service seat allocation method, device, computer equipment and storage medium to solve the problem of being unable to accurately and effectively allocate customers and seats.

[0005] In order to solve the above technical problems, the embodiment of the present application provides a service seat allocation method, which adopts the following technical solution:

[0006] Get historical service data of agents;

[0007] Extracting features from the agent's historical service data to obtain multi-dimensional feature information of the agent's capabilities;

[0008] Acquire customer historical communication data, and acquire customer demand information based on the customer historical communication data;

[0009] Inputting the customer demand information and the multi-dimensional feature information of the agent's ability into a preset matching network to obtain a matching score;

[0010] The customer and the service agent are allocated according to the preset matching rules and the matching score. Further, the step of obtaining the agent's historical service data specifically includes:

[0011] Acquire a service data extraction identifier, and extract a historical service data set from a service record database according to the service data extraction identifier;

[0012] Preprocessing the historical service data set to obtain a standard historical service data set;

[0013] The agent identification information is acquired, and information association is performed based on the agent identification information and the standard historical service data set to obtain the agent historical service data.

[0014] Furthermore, the agent historical service data includes agent service time information, agent service record information, agent problem solving status, and customer satisfaction survey information. The step of extracting features from the agent historical service data to obtain multi-dimensional feature information of agent capabilities specifically includes:

[0015] Calculate the number of service requests processed by each agent within a specified time period according to the agent service time information and the agent service record information, and compare and calculate the number of service requests with a preset work standard to obtain the agent work saturation;

[0016] Calculate the number of services provided by each agent in different business areas based on the agent service record information, and evaluate the agent's professional knowledge level based on the preset professional knowledge scoring rules to obtain the agent's field professional score;

[0017] Calculate the service quality of each agent after solving the customer's problem based on the agent's problem solving status and the customer satisfaction survey information to obtain an agent service quality score;

[0018] The agent work saturation, the agent field professional score, and the agent service quality score are used as the multi-dimensional feature information of the agent ability.

[0019] Furthermore, the customer historical communication data includes customer historical voice information, customer historical voice text, and customer historical behavior information. The step of obtaining the customer historical communication data and obtaining the customer demand information according to the customer historical communication data specifically includes:

[0020] Extracting features from the customer's historical voice information to obtain customer voice features;

[0021] Performing semantic analysis on the customer's historical voice text to obtain customer text semantic information;

[0022] Integrate the customer's voice features, the customer's text semantic information, and the customer's historical behavior information to obtain customer behavior feature information;

[0023] Performing clustering algorithm analysis based on the customer behavior feature information to obtain customer category groups;

[0024] Constructing the customer demand model according to the customer category group and the customer behavior characteristic information;

[0025] Acquire customer real-time information, input the customer real-time information into the customer demand model, and obtain the customer demand information.

[0026] Furthermore, the step of inputting the customer demand information and the multi-dimensional feature information of the agent's ability into a preset matching network to obtain a matching score specifically includes:

[0027] Extracting customer problem type, customer problem description information, and problem urgency information from the customer demand information;

[0028] Integrate the customer problem type, the customer problem description information, and the problem urgency information to obtain a customer demand feature vector;

[0029] Converting the multi-dimensional feature information of the agent's ability into a feature vector of the agent's ability;

[0030] The customer demand feature vector and the agent ability feature vector are input into a preset matching network to perform correlation calculation to obtain the matching score.

[0031] Furthermore, before the step of allocating customers and service seats according to the preset matching rules and the matching scores, the following steps are also included:

[0032] Get entity data and relationship information;

[0033] Performing data cleaning on the entity data to obtain valid entity data;

[0034] Analyze the relationship types between entities according to the valid entity data and the relationship information to obtain entity relationship constraint information;

[0035] Determine whether the entity relationship constraint information complies with preset business logic rules;

[0036] If the entity relationship constraint information complies with the preset business logic rule, the entity relationship constraint information is used as the preset matching rule.

[0037] Furthermore, after the step of allocating customers and service seats according to the preset matching rules and the matching scores, the following steps are also included:

[0038] Obtain agent service feedback data;

[0039] Conducting a comprehensive capability evaluation on the agents based on the agent service feedback data;

[0040] Adjust the skills training plan of the agents according to the comprehensive competency evaluation;

[0041] Execute the skills training plan and regularly collect agent capability information to update agent capability indicators;

[0042] Check whether the agent capability index reaches the expected value corresponding to the skill training plan;

[0043] If the agent capability index reaches the expected value corresponding to the skill training plan, the preset matching network is updated according to the agent capability index.

[0044] In order to solve the above technical problems, the embodiment of the present application further provides a service seat allocation device, which adopts the following technical solution:

[0045] Data acquisition module, used to obtain historical service data of agents;

[0046] A feature extraction module is used to extract features from the agent's historical service data to obtain multi-dimensional feature information of the agent's capabilities;

[0047] A demand acquisition module is used to acquire customer historical communication data and acquire customer demand information based on the customer historical communication data;

[0048] A demand matching module, used to input the customer demand information and the multi-dimensional feature information of the agent's ability into a preset matching network to obtain a matching score;

[0049] The service allocation module is used to allocate customers and service seats according to preset matching rules and the matching scores.

[0050] In order to solve the above technical problems, the embodiment of the present application further provides a computer device, which adopts the following technical solution:

[0051] A computer device comprises a memory and a processor, wherein the memory stores computer-readable instructions, and the processor implements the steps of any one of the above service seat allocation methods when executing the computer-readable instructions.

[0052] In order to solve the above technical problems, the embodiment of the present application further provides a computer-readable storage medium, which adopts the following technical solution:

[0053] A computer-readable storage medium stores computer-readable instructions, and when the computer-readable instructions are executed by a processor, the steps of the service seat allocation method as described in any one of the above items are implemented.

[0054] Compared with the prior art, the embodiment of the present application has the following beneficial effects: the embodiment obtains the historical service data of the agent; extracts the features of the historical service data of the agent to obtain the multi-dimensional feature information of the agent's ability; obtains the historical communication data of the customer, and obtains the customer demand information based on the historical communication data of the customer; inputs the customer demand information and the multi-dimensional feature information of the agent's ability into the preset matching network to obtain the matching score; and allocates the customer and the service agent according to the preset matching rules and the matching score. Thus, the corresponding allocation of customers and service agents is achieved, so as to effectively improve the service adaptability of customers and service agents and improve the customer's service experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the scheme in the present application, a brief introduction is given below to the drawings required for use in the description of the embodiments of the present application. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0056] Figure 1 is an exemplary system architecture diagram to which the present application may be applied;

[0057] Figure 2 A flowchart of an embodiment of a service seat allocation method according to the present application;

[0058] Figure 3 yes Figure 2 A flowchart of a specific implementation of step S10;

[0059] Figure 4 yes Figure 2 A flowchart of a specific implementation of step S20;

[0060] Figure 5 yes Figure 2 A flowchart of a specific implementation of step S30;

[0061] Figure 6 yes Figure 2 A flowchart of a specific implementation of step S40;

[0062] Figure 7 is a structural schematic diagram of an embodiment of a service seat allocation device according to the present application;

[0063] Figure 8 It is a structural diagram of an embodiment of a computer device according to the present application. DETAILED DESCRIPTION

[0064] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by technicians in the technical field of the present application; the terms used in the specification of the application herein are only for the purpose of describing specific embodiments and are not intended to limit the present application; the terms "including" and "having" and any variations thereof in the specification and claims of the present application and the above-mentioned drawings are intended to cover non-exclusive inclusions. The terms "first", "second", etc. in the specification and claims of the present application or the above-mentioned drawings are used to distinguish different objects, not to describe a specific order.

[0065] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor does it refer to non-related or alternative embodiments that are mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0066] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings.

[0067] like Figure 1 As shown, the system architecture 100 may include terminal devices 101, 102, 103, a network 104 and a server 105. The network 104 is used to provide a medium for communication links between the terminal devices 101, 102, 103 and the server 105. The network 104 may include various connection types, such as wired, wireless communication links or optical fiber cables, etc.

[0068] Users can use terminal devices 101, 102, 103 to interact with server 105 through network 104 to receive or send messages, etc. Various communication client applications can be installed on terminal devices 101, 102, 103, such as web browser applications, shopping applications, search applications, instant messaging tools, email clients, social platform software, etc.

[0069] Terminal devices 101, 102, 103 can be various electronic devices with display screens and supporting web browsing, including but not limited to smart phones, tablet computers, e-book readers, MP3 players (Moving Picture Experts Group Audio Layer III), MP4 (Moving Picture Experts Group Audio Layer IV) players, laptop computers, desktop computers, etc.

[0070] The server 105 may be a server that provides various services, such as a background server that provides support for web pages displayed on the terminal devices 101 , 102 , and 103 .

[0071] It should be noted that the service seat allocation method provided in the embodiment of the present application is generally executed by a server, and accordingly, the service seat allocation device is generally arranged in the server.

[0072] It should be understood that Figure 1 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to implementation requirements.

[0073] Continue to refer Figure 2 , shows a flow chart of an embodiment of a method for allocating service seats according to the present application. The method for allocating service seats comprises the following steps:

[0074] Step S10, obtaining seat historical service data;

[0075] In this embodiment, the agent's historical service data includes agent service time information, agent service record information, agent problem solving status, and customer satisfaction survey information, wherein the agent service time information is information about the agent's workload and service efficiency, including data such as the agent's working hours, the time points when the service starts and ends, and busy and idle time periods; the agent service record information is information reflecting the agent's service frequency, service type, and service content, including all service requests handled by the agent, such as calls, chats, and emails; the agent's problem solving status is an important indicator for measuring the agent's professional ability and service effectiveness, including data such as the time it takes for the agent to solve customer problems, the success rate of solving problems, and the number of problems that need to be upgraded or transferred; the customer satisfaction survey information is information reflecting the agent's service quality and customer satisfaction, including information such as customers' evaluation of the agent's service attitude, response speed, and problem-solving ability, which can be collected through customer satisfaction follow-up surveys.

[0076] Step S20, extracting features from the agent's historical service data to obtain multi-dimensional feature information of the agent's capabilities;

[0077] In this embodiment, the multi-dimensional feature information of the agent's ability includes the agent's work saturation, the agent's field professional score, and the agent's service quality score. The agent's work saturation is used to reflect the agent's workload, that is, the agent's working hours, service frequency, and the distribution of busy and idle time. High work saturation means that the agent is experienced and has strong processing ability, while low work saturation means that the agent is inexperienced and has weak processing ability. The agent's field professional score refers to the evaluation score of the agent's professional ability in a specific service field or product knowledge. The agent's field professional score is obtained by comprehensive evaluation based on factors such as the agent's training results, historical problem solving, and customer feedback. A high field professional score means that the agent has stronger service capabilities and professional knowledge in a specific field, while a low field professional score means that the agent has poor service capabilities and professional knowledge in a specific field. The agent's service quality score is an evaluation score of the agent's overall service quality, specifically including evaluations of service attitude, response speed, communication skills, problem-solving skills, etc. The agent's service quality score is usually calculated based on customer satisfaction survey information, service quality monitoring information, etc., and is used to measure the comprehensive service performance of the agent.

[0078] Step S30, obtaining customer historical communication data, and obtaining customer demand information based on the customer historical communication data;

[0079] In this embodiment, the customer's historical communication data is the collected customer's past communication records, including customer's historical voice information, customer's historical voice text, and customer's historical behavior information, wherein the customer's historical voice information is the voice interaction data between the historically recorded customer and the enterprise's agent or self-service system. The customer's historical voice text is a textual representation of the customer's historical voice information obtained by converting the information corresponding to the voice interaction content into text through voice recognition technology, which is used to search, analyze and understand the customer's past communication content, so as to gain insight into the customer's needs, preferences and potential problems. The customer's historical behavior information includes various behaviors of the customer during the agent service process, including the frequency and depth of communication between the customer and the agent, and the type of service selected by the customer during the interaction process, such as consultation, complaint, suggestion, repair, etc.

[0080] Step S40, inputting the customer demand information and the multi-dimensional feature information of the agent's ability into a preset matching network to obtain a matching score;

[0081] In this embodiment, the preset matching network includes a word embedding layer, an attention mechanism layer, a multi-layer nonlinear transformation network layer and a matching layer. The word embedding layer is used to map the word information in the feature vector corresponding to the customer demand information and the feature vector corresponding to the multi-dimensional feature information of the agent's ability into a dense vector to obtain the customer demand word embedding and the multi-dimensional feature word embedding of the agent's ability represented by the word embedding; the attention mechanism layer is used to align the relevant information in the multi-dimensional feature word embedding of the agent's ability according to the key information in the customer demand word embedding, and extract the semantic correlation features; the multi-layer nonlinear transformation network is used to extract the deep-level correlation information in the semantic correlation features of the customer demand information and the multi-dimensional feature information of the agent's ability through the processing of multiple hidden layers to form a deep feature vector; the matching layer is used to perform correlation matching calculation on the deep feature vectors of the customer demand information and the multi-dimensional feature information of the agent's ability to obtain a matching score.

[0082] Step S50: Allocate customers and service seats according to the preset matching rules and the matching scores.

[0083] In this embodiment, the preset matching rules are used to define the following: Type of relationship between entities: defines which entities can be associated with each other, such as customers and service businesses, seats and departments, etc. Directionality of relationships: one-way and two-way, one-way refers to the relationship from one entity to another, such as the relationship that the current service business belongs to the current user, and two-way refers to the relationship between two entities, such as the interactive relationship between users and seats. Constraints of relationships: specifies the conditions that must be met when a relationship is established, such as a service business can only correspond to a certain level of customers, but a certain level of customers can correspond to multiple service businesses; or the professional field of the agent corresponds to a certain type of customer, etc. Business logic rules: These rules are based on the best practices or internal regulations in the business field, and are used to ensure the accuracy and consistency of data. They involve the integrity, consistency, accuracy of data and the logical associations between data, which are mainly manifested in the business processes and the order of business nodes in the current business scenario.

[0084] This embodiment obtains the historical service data of the agent; extracts features from the historical service data of the agent to obtain multi-dimensional feature information of the agent's capabilities; obtains the historical communication data of the customer, and obtains the customer demand information based on the historical communication data of the customer; inputs the customer demand information and the multi-dimensional feature information of the agent's capabilities into a preset matching network to obtain a matching score; and allocates customers and service agents based on the preset matching rules and the matching score. This enables corresponding allocation of customers and service agents to effectively improve the service adaptability of customers and service agents and improve the customer's service experience. This embodiment can be applied to financial business services, such as seat allocation during insurance and product transactions, to improve customer satisfaction in financial business services.

[0085] Continue to refer Figure 3 In some optional implementations of this embodiment, step S10 includes the following steps:

[0086] Step S101, obtaining a service data extraction identifier, and extracting a historical service data set from a service record database according to the service data extraction identifier;

[0087] In this embodiment, the service data extraction identifier is the identification information of the corresponding historical service data set, and the identification information can be set based on business needs, such as a specific service professional field, service type, service time, etc. The identification information will be used as a query condition to accurately filter out the required historical service data set from the huge service record database.

[0088] Step S102, preprocessing the historical service data set to obtain a standard historical service data set;

[0089] In this embodiment, preprocessing includes data cleaning: deleting duplicate records, filling missing values, and correcting erroneous data. Data conversion: converting data into a unified format, such as converting date and time strings into date and time types, converting text fields into lowercase or uppercase, etc. By preprocessing the historical service data set, a complete and standardized standard historical service data set can be effectively obtained.

[0090] Step S103, acquiring agent identification information, and performing information association based on the agent identification information and the standard historical service data set to obtain the agent historical service data.

[0091] In this embodiment, an agent refers to a service provider who directly interacts with a customer, and the agent identification information may include the agent's work number, name, department number, etc. The agent identification information may be extracted from a department list or a corresponding database. The agent identification information is used to search for service records related to the agent in a standard historical service data set, and then all service records related to the agent are aggregated to form agent historical service data. The search for the standard historical service data set may be achieved through a data matching algorithm.

[0092] This embodiment obtains a service data extraction identifier, extracts a historical service data set from a service record database according to the service data extraction identifier; preprocesses the historical service data set to obtain a standard historical service data set; obtains agent identification information, and associates information with the agent identification information and the standard historical service data set, thereby obtaining agent historical service data that associates the agent with the historical service data, so as to facilitate subsequent feature extraction processing.

[0093] Continue to refer Figure 4 In some optional implementations of this embodiment, step S20 includes the following steps:

[0094] Step S201, calculating the number of service requests processed by each agent within a specified time period according to the agent service time information and the agent service record information, and comparing and calculating the number of service requests with a preset work standard amount to obtain the agent work saturation;

[0095] In this embodiment, extracting a specified time period: extracting a specified time period from the agent service time information, the specified time period refers to the time period that needs to be evaluated, which can be one day, one week, one month or any custom time range. Counting the number of service requests: according to the agent service record information and the agent service time information, counting the number of service requests processed by each agent within the specified time period. Comparing the work standard amount: presetting the work standard amount for reference comparison, this standard amount can be determined based on historical data, industry standards or company policies. Compare the actual number of service requests for each agent with the preset work standard amount, so as to determine the work saturation of the agent. Among them, the work saturation can be the percentage value of the actual processing volume to the work standard amount, or it can be a grade classification based on this percentage value (such as low, medium, high), and the grade classification is divided based on the percentage threshold. When the work saturation reaches the percentage threshold corresponding to a certain level, it is judged that the work saturation belongs to the corresponding grade classification.

[0096] Step S202, calculating the service times of each agent in different business fields according to the agent service record information, and evaluating the agent's professional knowledge level according to a preset professional knowledge scoring rule to obtain the agent field professional score;

[0097] In this embodiment, the acquisition of the professional score of the agent field includes the following steps: Classify the business field: classify the service requests in the agent service record information to identify different business fields. The classification step can be completed by analyzing the keywords, service type codes or other identifiers in the service record. Count the number of services: for each agent, count the number of services in different business fields. Apply professional knowledge scoring rules: score the performance of each agent in different business fields according to the preset professional knowledge scoring rules. The preset professional knowledge scoring rules can be formulated based on multiple factors such as service result quality, customer feedback, processing time, etc., by assigning different weights to the service result quality, customer feedback, processing time, etc., and scoring according to the corresponding service situation in the agent service record information. For example, the weight of the service result quality is 50% or 0.5, and the corresponding quality situations include: Accurate: If the agent can accurately answer customer questions or solve customer problems without subsequent follow-up or correction, full marks are obtained on this item. Basically correct: If the agent's answer or solution is basically correct, but there are some minor errors or slight adjustments are required, corresponding deductions are given according to the severity and impact of the errors. Obvious Errors: If the agent's answer or solution is obviously wrong, resulting in the customer's problem not being solved or a new problem being created, a low score or zero score will be given for this item. The above score can be quantitative (such as a score or percentage) or qualitative (such as a grade or label).

[0098] Step S203, calculating the service quality of each agent after solving the customer's problem based on the agent problem solving status and the customer satisfaction survey information, and obtaining an agent service quality score;

[0099] In this embodiment, obtaining the agent service quality score includes the following steps: Analyze the problem-solving situation: Based on the agent's problem-solving situation, analyze the agent's performance in solving customer problems, including factors such as the speed and accuracy of problem solving, and whether superior intervention is required. Integrate customer satisfaction survey information: Combine customer satisfaction survey information to understand the customer's overall satisfaction with the agent's service, including satisfaction scores, customer feedback, etc. Calculate service quality scores: Based on the problem-solving situation and customer satisfaction survey information, calculate the service quality score of each agent, where the service quality score can be a comprehensive indicator that takes into account multiple factors, such as the efficiency of problem solving, the enthusiasm of customer feedback, whether the service standards are met, etc. The service quality score can also be obtained by setting the scoring rules as described above for evaluation, or by manual comprehensive consideration and evaluation.

[0100] Step S204: taking the agent work saturation, the agent field professional score, and the agent service quality score as the multi-dimensional feature information of the agent capability.

[0101] This embodiment calculates the number of service requests handled by each agent within a specified time period according to the agent service time information and the agent service record information, and compares and calculates the number of service requests with the preset work standard quantity to obtain the agent work saturation; calculates the number of services provided by each agent in different business fields according to the agent service record information, and evaluates the agent's professional knowledge level according to the preset professional knowledge scoring rules to obtain the agent field professional score; calculates the service quality fed back by each agent after solving the customer's problem according to the agent problem solving situation and the customer satisfaction survey information to obtain the agent service quality score; uses the agent work saturation, the agent field professional score, and the agent service quality score as the agent capability multidimensional feature information. In this way, the agent capability multidimensional feature information that integrates the agent work saturation, the agent professional level, and the agent service quality is effectively obtained to provide reliable data support for subsequent matching processing.

[0102] Continue to refer Figure 5 In some optional implementations of this embodiment, step S30 includes the following steps:

[0103] Step S301, extracting features from the customer's historical voice information to obtain customer voice features;

[0104] In this embodiment, the acoustic features of the voice of the customer's historical voice information are extracted through audio analysis technology (such as MFCC, Mel-Frequency Cepstral Coefficients, etc.), and the extracted acoustic features are used as customer voice features. Customer voice features include pitch, speech speed, volume changes, and other feature information that can reflect the customer's emotional state or communication style.

[0105] Step S302, performing semantic analysis on the customer's historical voice text to obtain customer text semantic information;

[0106] In this embodiment, natural language processing (NLP) technologies such as lexical analysis, syntactic analysis, named entity recognition (NER) and sentiment analysis are used to deeply understand the text content, thereby obtaining customer text semantic information. These analyses will help identify the topics discussed by customers, the emotions expressed (such as positive, negative, neutral), and possible needs or problems.

[0107] Step S303, integrating the customer voice features, the customer text semantic information, and the customer historical behavior information to obtain customer behavior feature information;

[0108] In this embodiment, the acoustic features extracted from the voice of the customer's historical voice information, the semantic information extracted from the customer's historical voice text, and the customer's historical behavior information (such as purchase records, browsing history, service requests, etc.) are integrated to form a customer behavior feature information that comprehensively represents the customer's behavior characteristics. The customer behavior feature information is used to reflect the customer's preferences, habits and needs.

[0109] Step S304, performing clustering algorithm analysis based on the customer behavior feature information to obtain customer category groups;

[0110] In this embodiment, a clustering algorithm (such as K-means, hierarchical clustering, DBSCAN, etc.) is used to group customer behavior feature information to identify customer category groups with similar behavior patterns and needs, thereby providing reliable and effective data support for the construction of a customer demand model.

[0111] Step S305, constructing the customer demand model according to the customer category group and the customer behavior characteristic information;

[0112] In this embodiment, a customer demand model is constructed based on the customer category groups and customer behavior feature information after cluster analysis to predict customer demand. In this embodiment, the customer demand model is a logistic regression model, and the construction of the customer demand model includes the following steps: Model initialization: Select logistic regression as the model type and set initial parameters. Feature adjustment: According to the clustering results, different feature sets are used for different categories of customers or feature weights are adjusted. Model training: Use the training set data (which has been grouped according to the clustering results) to train the logistic regression model.

[0113] Step S306, acquiring customer real-time information, and inputting the customer real-time information into the customer demand model to obtain the customer demand information.

[0114] In this embodiment, the customer's current behavior data (such as the currently browsed page, clicked link, input query word, etc.) and other related information (such as geographic location, timestamp, etc.) are collected in real time. This real-time information is input into the constructed customer demand model, and the model will predict and output the current customer demand information based on the knowledge and rules it has learned, so as to respond to customer needs immediately, provide personalized recommendations or optimize customer experience.

[0115] This embodiment obtains customer voice features by performing feature extraction on the customer's historical voice information; performs semantic analysis on the customer's historical voice text to obtain customer text semantic information; performs data integration on the customer voice features, the customer text semantic information, and the customer's historical behavior information to obtain customer behavior feature information; performs clustering algorithm analysis based on the customer behavior feature information to obtain customer category groups; constructs the customer demand model based on the customer category groups and the customer behavior feature information; obtains real-time customer information and inputs the real-time customer information into the customer demand model, thereby obtaining customer demand information predicted by the customer demand model constructed according to user behavior and user personality, providing a reliable data basis for subsequent matching operations between agents and customers.

[0116] Continue to refer Figure 5 In some optional implementations of this embodiment, step S40 includes the following steps:

[0117] Step S401, extracting customer problem type, customer problem description information, and problem urgency information from the customer demand information;

[0118] In this embodiment, customer problem type refers to the type of problem raised by the customer during the service process, such as technical problems, service inquiries, complaints, etc.; customer problem description information refers to the customer's specific description of the problem, such as severity, occurrence time, frequency, etc.; problem urgency information is used to assess the urgency of the customer's problem, such as urgent, general, non-urgent, etc.

[0119] Step S402, integrating the customer problem type, the customer problem description information, and the problem urgency information to obtain a customer demand feature vector;

[0120] In this embodiment, the customer problem type, customer problem description information, and problem urgency information are encoded or quantified separately, such as using one-hot encoding to process categorical data and using numerical values ​​to represent the urgency, so that the customer problem type, customer problem description information, and problem urgency information are encoded into corresponding information vectors, and then the encoded information vectors are combined into a customer demand feature vector.

[0121] Step S403, converting the multi-dimensional feature information of the agent's ability into a feature vector of the agent's ability;

[0122] In this embodiment, the conversion step is implemented by normalizing and encoding the agent work saturation, agent field professional score, and agent service quality score in the multi-dimensional feature information of the agent capability, thereby obtaining an agent capability feature vector, wherein the normalization process can be performed by scaling the values ​​of the agent work saturation, agent field professional score, and agent service quality score to between 0 and 1 (by subtracting the minimum value and then dividing by the range of the maximum value minus the minimum value), and then defining the feature vector, for example, work saturation (saturation), field professional score (expertise), and service quality score (quality), which can be defined as [saturation, expertise, quality], and finally the value obtained after the normalization process is matched with the definition of the feature vector, thereby effectively obtaining the agent capability feature vector.

[0123] Step S404: input the customer demand feature vector and the agent capability feature vector into a preset matching network to perform correlation calculation to obtain the matching score.

[0124] In this embodiment, the preset matching network includes a word embedding layer, an attention mechanism layer, a multi-layer nonlinear transformation network layer and a matching layer. The word embedding layer is used to map the word information in the customer demand feature vector and the agent ability feature vector into a dense vector to obtain the customer demand word embedding and the agent ability multidimensional feature word embedding represented by the word embedding; the attention mechanism layer is used to align the relevant information in the agent ability multidimensional feature word embedding according to the key information in the customer demand word embedding, and extract the semantic correlation features; the multi-layer nonlinear transformation network is used to extract the deep correlation information in the semantic correlation features of the customer demand information and the agent ability multidimensional feature information through the processing of multiple hidden layers to form a deep feature vector; the matching layer is used to perform correlation matching calculation on the deep feature vectors of the customer demand information and the agent ability multidimensional feature information to obtain a matching score.

[0125] This embodiment extracts the customer problem type, customer problem description information, and problem urgency information from the customer demand information; integrates the customer problem type, the customer problem description information, and the problem urgency information to obtain a customer demand feature vector; converts the multi-dimensional feature information of the agent capability into an agent capability feature vector; inputs the customer demand feature vector and the agent capability feature vector into a preset matching network for correlation calculation, thereby obtaining a matching score representing the degree of matching between customer demand and agent capability, so as to facilitate subsequent allocation of customers and agents based on the matching score.

[0126] In some optional implementations of this embodiment, before step S50, the following steps are further included:

[0127] Get entity data and relationship information;

[0128] In this embodiment, entity data is data items with unique identifiers, such as customers, products, orders, seats, businesses, etc., and relationship information describes how these entities are related to each other, such as the purchasing relationship between customers and orders, the subscription relationship between customers and businesses, etc.

[0129] Performing data cleaning on the entity data to obtain valid entity data;

[0130] In this embodiment, data cleaning is used to eliminate errors, duplications, incompleteness and other problems in entity data and improve data quality, including removing invalid or incorrectly formatted data items, filling missing values, processing outliers, converting data formats, etc.

[0131] Analyze the relationship types between entities according to the valid entity data and the relationship information to obtain entity relationship constraint information;

[0132] In this embodiment, entity relationship constraint information describes specific conditions or rules that must be met between entities, including the relationship type between entities, the attributes of the relationship, and the constraint conditions. The relationship type can be direct (such as a direct association between a user and an order) or indirect (such as an association established through an intermediate entity). For example, each order must have one and only one corresponding user, and the "creator" field of an order must point to a valid user. A user can create multiple orders, but each order can only be created by one user; or an agent employee can belong to multiple departments, and a department can have multiple agent employees. Agent employees and departments are associated through an association table. Each agent employee can have multiple records in the association table, and each record represents a department described by the agent employee.

[0133] Determine whether the entity relationship constraint information complies with preset business logic rules;

[0134] In this embodiment, the preset business logic rules are a series of predefined rules used to guide business process and data processing decisions, such as defining the order and conditions of business operations (such as placing an order before shipping), setting approval links and authority control in the process, and managing exception handling and rollback mechanisms in the process.

[0135] If the entity relationship constraint information conforms to the preset business logic rule, the entity relationship constraint information is used as the preset matching rule;

[0136] If the entity relationship constraint information does not conform to the preset business logic rules, the valid entity data and the relationship information are re-analyzed to obtain new entity relationship constraint information, and whether the new entity relationship constraint information conforms to the preset business logic rules is re-judged until the new entity relationship constraint information conforms to the preset business logic rules.

[0137] This embodiment obtains entity data and relationship information; performs data cleaning on the entity data to obtain valid entity data; analyzes the relationship type between entities according to the valid entity data and the relationship information to obtain entity relationship constraint information; determines whether the entity relationship constraint information conforms to the preset business logic rules; if the entity relationship constraint information conforms to the preset business logic rules, the entity relationship constraint information is used as the preset matching rule. In this way, matching rules that conform to actual business logic rules are effectively obtained to avoid conflicts between matching rules and business logic rules, and ensure the effectiveness and reliability of matching rules.

[0138] In some optional implementations of this embodiment, after step S50, the following steps are further included:

[0139] Obtain agent service feedback data;

[0140] In this embodiment, the agent service feedback data is feedback data collected after the agent is assigned, including customer satisfaction surveys, service time, problem resolution rate, number of complaints, etc., which can be obtained through various channels such as customer feedback systems, call records, chat records, internal systems, etc.

[0141] Conducting a comprehensive capability evaluation on the agents based on the agent service feedback data;

[0142] In this embodiment, a combination of quantitative and qualitative methods can be adopted, such as setting scoring criteria and using machine learning algorithms to evaluate the communication skills, professional knowledge, problem-solving ability, service attitude and other aspects of the agent service. The setting of the scoring criteria can be achieved by referring to the above-mentioned scoring rule setting method, and the evaluation performed by the machine learning algorithm can be achieved by training the corresponding model.

[0143] Adjust the skills training plan of the agents according to the comprehensive competency evaluation;

[0144] In this embodiment, the adjustment of the skill training plan includes training needs analysis: assessing the current seat's skill level, knowledge structure and ability shortcomings; clarifying the training goals and expected results based on business needs, job requirements and industry trends; determining specific skill areas that need to be improved, such as communication skills, product knowledge, problem-solving skills, customer service etiquette, etc. Training content design: designing targeted training courses based on the needs analysis results, including theoretical explanations, case analysis, practical exercises, etc.; preparing teaching materials, PPTs, videos, simulated dialogue scripts and other auxiliary materials required for training; ensuring that the training content keeps pace with the times and contains the latest industry information, product knowledge and technical trends.

[0145] Execute the skills training plan and regularly collect agent capability information to update agent capability indicators;

[0146] In this embodiment, the ways and methods of executing the skill training plan include: using the network platform for remote learning, such as online courses, video tutorials, live lectures, etc.; organizing face-to-face training activities, including classroom lectures, group discussions, role-playing, practical exercises, etc.; combining online and offline methods to flexibly arrange training time and place; encouraging trainees to participate and enhance learning effects through questions, discussions, feedback, etc. Training schedule: formulate a long-term training plan and implement it in stages to ensure the continuity and systematicness of skill improvement; clarify the specific time, location and duration of each training to ensure the smooth progress of training activities; adjust the training time and progress in a timely manner according to the actual situation and trainee feedback. The agent ability information includes communication skills, product knowledge, problem-solving ability, work efficiency, service attitude, teamwork ability, and learning ability. The agent ability index is a specific measurement standard set based on the agent ability information, which is used to quantitatively evaluate the ability level of the agent. It can be quantitative (such as the average time to answer the phone, the one-time resolution rate, etc.) or qualitative (such as service attitude evaluation, teamwork evaluation, etc.).

[0147] Check whether the agent capability index reaches the expected value corresponding to the skill training plan;

[0148] In this embodiment, the expected value corresponding to the skill training plan refers to the quantitative value of the ability standard or goal that the agent should achieve after training, which is set according to business needs, job requirements and the current ability level of the agent when formulating the skill training plan. By checking whether the agent ability indicator reaches the expected value corresponding to the skill training plan, it is determined whether the preset matching network needs to be updated.

[0149] If the agent capability index reaches the expected value corresponding to the skill training plan, updating the preset matching network according to the agent capability index;

[0150] If the agent capability indicator does not reach the expected value corresponding to the skill training plan, the skill training plan is readjusted and the adjusted skill training plan is executed to obtain the adjusted agent capability indicator, and re-judge whether the adjusted agent capability indicator reaches the expected value corresponding to the skill training plan, until the adjusted agent capability indicator reaches the expected value corresponding to the skill training plan.

[0151] This embodiment obtains agent service feedback data; conducts comprehensive ability evaluation of agents based on the agent service feedback data; adjusts the agent's skill training plan based on the comprehensive ability evaluation; executes the skill training plan and regularly collects agent ability information to update agent ability indicators; checks whether the agent ability indicators reach the expected values ​​corresponding to the skill training plan; if the agent ability indicators reach the expected values ​​corresponding to the skill training plan, the preset matching network is updated based on the agent ability indicators. This enables real-time updating and optimization of the matching network based on feedback data after agent matching to improve the matching accuracy of the matching network.

[0152] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through computer-readable instructions, and the computer-readable instructions can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, the aforementioned storage medium can be a non-volatile storage medium such as a disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

[0153] It should be understood that, although the steps in the flowchart of the accompanying drawings are displayed in sequence as indicated by the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least a part of the steps in the flowchart of the accompanying drawings may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be executed in turn or alternately with other steps or at least a part of the sub-steps or stages of other steps.

[0154] Further references Figure 6 , as a response to the above Figure 1 The present application provides an embodiment of a service seat allocation device, and the device embodiment is similar to Figure 1Corresponding to the method embodiment shown, the device can be specifically applied to various electronic devices.

[0155] like Figure 6 As shown, the service seat allocation device 600 described in this embodiment includes: a data acquisition module 601, a feature extraction module 602, a demand acquisition module 603, a demand matching module 604, and a service allocation module 605. Among them:

[0156] Data acquisition module 601, used to acquire seat historical service data;

[0157] A feature extraction module 602 is used to extract features from the agent historical service data to obtain multi-dimensional feature information of agent capabilities;

[0158] The demand acquisition module 603 is used to acquire the customer's historical communication data and acquire the customer's demand information according to the customer's historical communication data;

[0159] The demand matching module 604 is used to input the customer demand information and the multi-dimensional feature information of the agent's ability into a preset matching network to obtain a matching score;

[0160] The service allocation module 605 is used to allocate customers and service seats according to the preset matching rules and the matching scores.

[0161] This embodiment, by using the above-mentioned service seat allocation device, can obtain the historical service data of the seat; extract the features of the historical service data of the seat to obtain the multi-dimensional feature information of the seat ability; obtain the historical communication data of the customer, and obtain the customer demand information based on the historical communication data of the customer; input the customer demand information and the multi-dimensional feature information of the seat ability into the preset matching network to obtain the matching score; and allocate the customer and the service seat according to the preset matching rules and the matching score. In this way, the corresponding allocation of customers and service seats is realized, so as to effectively improve the service adaptability of customers and service seats and improve the service experience of customers.

[0162] To solve the above technical problems, the present application also provides a computer device. Figure 7 , Figure 7 This is a basic structural block diagram of the computer device in this embodiment.

[0163] The computer device 7 includes a memory 71, a processor 72, and a network interface 73 that are interconnected through a system bus. It should be noted that the figure only shows a computer device 7 with components 71-73, but it should be understood that it is not required to implement all the components shown, and more or fewer components can be implemented instead. Among them, those skilled in the art can understand that the computer device here is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes but is not limited to microprocessors, application specific integrated circuits (Application Specific Integrated Circuit, ASIC), programmable gate arrays (Field-Programmable Gate Array, FPGA), digital processors (Digital Signal Processor, DSP), embedded devices, etc.

[0164] The computer device may be a computing device such as a desktop computer, a notebook, a PDA, a cloud server, etc. The computer device may interact with a user through a keyboard, a mouse, a remote controller, a touch pad, or a voice control device.

[0165] The memory 71 includes at least one type of readable storage medium, and the readable storage medium includes flash memory, hard disk, multimedia card, card-type memory (for example, SD or DX memory, etc.), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, disk, optical disk, etc. In some embodiments, the memory 71 can be an internal storage unit of the computer device 7, such as a hard disk or memory of the computer device 7. In other embodiments, the memory 71 can also be an external storage device of the computer device 7, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (FlashCard), etc. equipped on the computer device 7. Of course, the memory 71 can also include both the internal storage unit of the computer device 7 and its external storage device. In this embodiment, the memory 71 is generally used to store the operating system and various application software installed on the computer device 7, such as computer-readable instructions of the service seat allocation method, etc. In addition, the memory 71 can also be used to temporarily store various types of data that have been output or are to be output.

[0166] The processor 72 may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chips in some embodiments. The processor 72 is generally used to control the overall operation of the computer device 7. In this embodiment, the processor 72 is used to run the computer-readable instructions stored in the memory 71 or process data, such as computer-readable instructions for running the service seat allocation method.

[0167] The network interface 73 may include a wireless network interface or a wired network interface. The network interface 73 is generally used to establish a communication connection between the computer device 7 and other electronic devices.

[0168] By using the above-mentioned computer device, this embodiment can obtain the historical service data of the agent; extract the features of the historical service data of the agent to obtain the multi-dimensional feature information of the agent's ability; obtain the historical communication data of the customer, and obtain the customer demand information based on the historical communication data of the customer; input the customer demand information and the multi-dimensional feature information of the agent's ability into the preset matching network to obtain the matching score; and allocate the customer and the service agent according to the preset matching rules and the matching score. In this way, the corresponding allocation of customers and service agents is achieved, so as to effectively improve the service adaptability of customers and service agents and improve the customer's service experience.

[0169] The present application also provides another implementation, namely, providing a computer-readable storage medium, wherein the computer-readable storage medium stores computer-readable instructions, and the computer-readable instructions can be executed by at least one processor to enable the at least one processor to perform the steps of the service seat allocation method as described above.

[0170] By adopting the above-mentioned computer-readable storage medium, this embodiment can obtain the historical service data of the agent; extract the features of the historical service data of the agent to obtain the multi-dimensional feature information of the agent's ability; obtain the historical communication data of the customer, and obtain the customer demand information based on the historical communication data of the customer; input the customer demand information and the multi-dimensional feature information of the agent's ability into the preset matching network to obtain the matching score; and allocate the customer and the service agent according to the preset matching rules and the matching score. In this way, the corresponding allocation of customers and service agents is realized, so as to effectively improve the service adaptability of customers and service agents and improve the customer's service experience.

[0171] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in each embodiment of the present application.

[0172] Obviously, the embodiments described above are only some embodiments of the present application, rather than all embodiments. The preferred embodiments of the present application are given in the accompanying drawings, but they do not limit the patent scope of the present application. The present application can be implemented in many different forms. On the contrary, the purpose of providing these embodiments is to make the understanding of the disclosure of the present application more thorough and comprehensive. Although the present application is described in detail with reference to the aforementioned embodiments, for those skilled in the art, it is still possible to modify the technical solutions recorded in the aforementioned specific implementation methods, or to perform equivalent replacement of some of the technical features therein. Any equivalent structure made using the contents of the specification and drawings of this application, directly or indirectly used in other related technical fields, is similarly within the scope of patent protection of this application.

Claims

1. A service seat allocation method, characterized in that: The steps include: Get historical service data of agents; Extracting features from the agent's historical service data to obtain multi-dimensional feature information of the agent's capabilities; Acquire customer historical communication data, and acquire customer demand information based on the customer historical communication data; Inputting the customer demand information and the multi-dimensional feature information of the agent's ability into a preset matching network to obtain a matching score; Customers and service seats are allocated according to preset matching rules and the matching scores.

2. The service seat allocation method according to claim 1, characterized in that: The step of obtaining the historical service data of the agent specifically includes: Acquire a service data extraction identifier, and extract a historical service data set from a service record database according to the service data extraction identifier; Preprocessing the historical service data set to obtain a standard historical service data set; The agent identification information is acquired, and information association is performed based on the agent identification information and the standard historical service data set to obtain the agent historical service data.

3. The service seat allocation method according to claim 1, characterized in that: The agent historical service data includes agent service time information, agent service record information, agent problem solving status, and customer satisfaction survey information. The step of extracting features from the agent historical service data to obtain multi-dimensional feature information of agent capabilities specifically includes: Calculate the number of service requests processed by each agent within a specified time period according to the agent service time information and the agent service record information, and compare and calculate the number of service requests with a preset work standard to obtain the agent work saturation; Calculate the number of services provided by each agent in different business areas based on the agent service record information, and evaluate the agent's professional knowledge level based on the preset professional knowledge scoring rules to obtain the agent's field professional score; Calculate the service quality of each agent after solving the customer's problem based on the agent's problem solving status and the customer satisfaction survey information to obtain an agent service quality score; The agent work saturation, the agent field professional score, and the agent service quality score are used as the multi-dimensional feature information of the agent ability.

4. The service seat allocation method according to claim 1, characterized in that: The customer historical communication data includes customer historical voice information, customer historical voice text, and customer historical behavior information. The step of obtaining the customer historical communication data and obtaining customer demand information based on the customer historical communication data specifically includes: Extracting features from the customer's historical voice information to obtain customer voice features; Performing semantic analysis on the customer's historical voice text to obtain customer text semantic information; Integrate the customer's voice features, the customer's text semantic information, and the customer's historical behavior information to obtain customer behavior feature information; Performing clustering algorithm analysis based on the customer behavior feature information to obtain customer category groups; Constructing the customer demand model according to the customer category group and the customer behavior characteristic information; Acquire customer real-time information, input the customer real-time information into the customer demand model, and obtain the customer demand information.

5. The service seat allocation method according to claim 1, characterized in that: The step of inputting the customer demand information and the multi-dimensional feature information of the agent's ability into a preset matching network to obtain a matching score specifically includes: Extracting customer problem type, customer problem description information, and problem urgency information from the customer demand information; Integrate the customer problem type, the customer problem description information, and the problem urgency information to obtain a customer demand feature vector; Converting the multi-dimensional feature information of the agent's ability into a feature vector of the agent's ability; The customer demand feature vector and the agent ability feature vector are input into a preset matching network to perform correlation calculation to obtain the matching score.

6. The service seat allocation method according to claim 1, characterized in that: Before the step of allocating customers and service seats according to the preset matching rules and the matching scores, the following steps are also included: Get entity data and relationship information; Performing data cleaning on the entity data to obtain valid entity data; Analyze the relationship types between entities according to the valid entity data and the relationship information to obtain entity relationship constraint information; Determine whether the entity relationship constraint information complies with preset business logic rules; If the entity relationship constraint information complies with the preset business logic rule, the entity relationship constraint information is used as the preset matching rule.

7. The service seat allocation method according to claim 1, characterized in that: After the step of allocating customers and service seats according to the preset matching rules and the matching scores, the following steps are also included: Obtain agent service feedback data; Conducting a comprehensive capability evaluation on the agents based on the agent service feedback data; Adjust the skills training plan of the agents according to the comprehensive competence evaluation; Execute the skills training plan and regularly collect agent capability information to update agent capability indicators; Check whether the agent capability index reaches the expected value corresponding to the skill training plan; If the agent capability index reaches the expected value corresponding to the skill training plan, the preset matching network is updated according to the agent capability index.

8. A service seat allocation device, characterized in that: include: Data acquisition module, used to obtain historical service data of agents; A feature extraction module, used to extract features from the agent's historical service data to obtain multi-dimensional feature information of the agent's capabilities; A demand acquisition module is used to acquire customer historical communication data and acquire customer demand information based on the customer historical communication data; A demand matching module, used to input the customer demand information and the multi-dimensional feature information of the agent's ability into a preset matching network to obtain a matching score; The service allocation module is used to allocate customers and service seats according to preset matching rules and the matching scores.

9. A computer device, characterized in that: The method comprises a memory and a processor, wherein the memory stores computer-readable instructions, and the processor implements the steps of the service seat allocation method according to any one of claims 1 to 7 when executing the computer-readable instructions.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-readable instructions, and when the computer-readable instructions are executed by a processor, the steps of the service seat allocation method according to any one of claims 1 to 7 are implemented.

Images