Differentiated outbound call configuration generation method and system based on scoring model

Abstract

The present invention discloses a method, system, electronic device, and computer-readable medium for generating differentiated outbound call configurations based on a scoring model. The method comprises: utilizing screened outbound call services to be processed, collecting device data of the outbound call devices corresponding to the outbound call services, and establishing a basic database for storing the outbound call devices and corresponding device data; using a scoring model to score the outbound call devices to determine the different categories to which the outbound call devices belong, dividing them into different queues according to the categories and assigning them to corresponding different configuration agencies, and adding one or more corresponding tags to the outbound call devices; and automatically generating an outbound call configuration result for the outbound call device based on the score, the configuration agency skills, and the tag. This method solves the problem of flexibly distinguishing individual differences between the outbound call devices and the groups to which they belong to provide a more optimal outbound call configuration, thereby improving the efficiency and accuracy of outbound call configuration.

Description

Technical Field

[0001] The present invention relates to the field of computer information processing technology, and in particular to a method, system, electronic device and computer-readable medium for generating differentiated outbound call configurations based on a scoring model. Background Art

[0002] With the advancement of artificial intelligence (AI) technology, customer service systems based on intelligent voice robots have already captured a significant portion of application scenarios. However, manual outbound call systems still possess advantages that intelligent voice robots lack, such as more efficient, comprehensive, and accurate communication, timely identification and resolution of issues, adjustments to user behavior and transaction processing patterns, timely dissemination of policies, reminders of monitored user information status and recommended actions, and even proactive notifications to prevent missed transactions, unnecessary cost increases, and loss of profits. However, due to manual intervention, objective flaws in management logic, or subjective human error, outbound call configuration in these systems is crude and rudimentary, leading to significant discrepancies and confusion in outbound call processing, resulting in a poor user experience. In particular, outbound call processing can be inflexible, with some devices receiving calls when they shouldn't, while others failing to receive calls when they should. This inefficient outbound call processing fails to align with actual user needs, leading to frequent errors and user complaints. Furthermore, the outbound call configuration of manual outbound call systems cannot be effectively integrated with the customer service systems powered by intelligent voice robots.

[0003] Therefore, there is a need for an improved technology for automatic outbound call configuration. Summary of the Invention

[0004] In view of this, the main purpose of the present invention is to propose a method, system, electronic device, and computer-readable medium for generating differentiated outbound call configurations based on a scoring model. The methods address the following technical problems: how to flexibly distinguish individual differences between outbound call devices and the groups they belong to in order to provide corresponding, highly targeted, or even more optimal, outbound call configurations, thereby improving the efficiency and accuracy of outbound call configurations. This method allows both the customer and the service provider to jointly determine the outbound call strategy, thereby achieving flexible, customized outbound call configurations and outbound call service processing. Furthermore, the methods address how to achieve flexible, customized outbound call configurations by dividing and matching outbound call device types and outbound call configuration organization groups, and implementing targeted, multi-dimensional control within the organization. This method, based on a big data platform, automatically generates outbound call configurations and constructs outbound call service strategies. This method is applicable to both the logical management and control of manual outbound call systems and the logical control of intelligent robot customer service systems. Furthermore, for manual systems, it avoids the drawbacks of using crude strategies, such as the inability to provide effective configuration iteration data and respond to social emergencies, due to the significant differences in the actual attributes of different customers in the same queue and the uneven skills of the service providers.

[0005] In order to solve the above technical problems, the first aspect of the present invention proposes a differentiated outbound call configuration generation method based on a scoring model, comprising: screening outbound call services that meet the outbound call conditions in network interaction services, collecting device data of outbound call devices corresponding to the outbound call services, and establishing a basic database for storing outbound call devices and corresponding device data; analyzing the characteristics of the outbound call devices based on the outbound call devices and the corresponding device data, and scoring the corresponding outbound call devices using a scoring model according to the characteristics to determine the category to which the outbound call devices belong; assigning the outbound call devices to corresponding configuration agencies according to the categories to which they belong; adding one or more corresponding tags to the outbound call devices according to the scores of the outbound call devices according to the scoring model from one or more dimensions; generating an outbound call configuration result corresponding to the outbound call devices according to the scores of the outbound call devices according to the scoring model, the configuration agencies to which each outbound call device is assigned, and the one or more tags of the outbound call devices.

[0006] According to a preferred embodiment of the present invention, device data of devices to be called corresponding to outbound call services are collected, and a basic database is established to store the devices to be called and the corresponding device data. Specifically, the method includes: using a server to analyze and organize the basic attributes of the devices to be called according to the basic attributes of the devices to be called, and writing the corresponding device data sets into the basic database; and / or, synchronizing the source data from an external data source or by file parsing into a data table of the basic database through a shell script.

[0007] According to a preferred embodiment of the present invention, analysis and organization are performed based on the basic attributes of the device to be called out, specifically including: obtaining external data and network behavior data of the customer using the corresponding device from a third-party platform based on the personal information in the application information entered according to the prompt in the device data of the device to be called out.

[0008] According to a preferred embodiment of the present invention, the devices to be called outbound are assigned to corresponding configuration agencies according to the categories to which they belong, specifically including: the server pre-sets different groups of outbound call service configuration agencies according to the different attributes of the outbound call interaction scenarios and outbound call skills; the authority of the configuration agency is determined according to the layer settings of the configuration agency; different queues are divided according to the categories to which the devices to be called outbound belong; different queues have different authority configurations; and the queues are assigned to configuration agencies with matching authority according to the authority configurations.

[0009] According to a preferred embodiment of the present invention, it also includes: generating a multi-digit random number for the device to be called out in the configuration mechanism that matches the weight, specifically including: based on the devices to be called out in each queue divided into different queues, dividing the devices to be called out in multiple queues into different outbound call groups; generating a different random number for each outbound call group through a random number algorithm, and determining the number of outbound calls for the device to be called in the corresponding outbound call group based on the random number.

[0010] According to a preferred embodiment of the present invention, the characteristics of the devices to be called outbound are analyzed based on the devices to be called outbound and the corresponding device data, and the corresponding devices to be called outbound are scored using a scoring model according to the characteristics to determine the category to which the devices to be called outbound belong. Specifically, the following steps are performed: the server analyzes the device data in the basic database from multiple dimensions based on a big data platform, and the device data in the multiple dimensions include historical network behavior data, historical contract fulfillment data, dynamic and / or static basic data, and external service scoring data; the trained scoring model is used to input the device data in the multiple dimensions analyzed in the current database to predict and score the corresponding devices to be called outbound; the different categories to which the corresponding devices to be called outbound are determined based on the classification processing of the scores; wherein the scoring model is constructed using a machine learning algorithm model and trained using the historical devices to be called outbound and the corresponding device data that have been classified as such as the training data.

[0011] According to a preferred embodiment of the present invention, an outbound call configuration result corresponding to the device to be called is generated based on the score of the outbound call device by the scoring model, the configuration agency to which each device to be called is assigned, and one or more tags of the device to be called. Specifically, the outbound call configuration result corresponding to the device to be called is generated by the corresponding configuration agency to which the device to be called is assigned, based on the authority and outbound call skills of the configuration agency, combined with the score of the device to be called, the multi-digit random number generated for the device to be called, and the one or more tags added to the device to be called; and the configuration agency executes the outbound call service for the device to be called according to the outbound call configuration result using the outbound call skills under the authority restriction of the configuration agency.

[0012] In order to solve the above technical problems, the second aspect of the present invention provides a differentiated outbound call configuration generation system based on a scoring model, including: a database construction unit, which is used to collect device data of outbound call devices corresponding to the outbound call services by screening outbound call services that meet outbound call conditions in network interaction services, and establish a basic database for storing the outbound call devices and corresponding device data; a model calculation unit, which is used to analyze the characteristics of the outbound call devices based on the outbound call devices and the corresponding device data, and score the corresponding outbound call devices using a scoring model according to the characteristics to determine the category to which the outbound call devices belong; a queue division unit, which is used to assign the outbound call devices to corresponding configuration mechanisms according to their categories; a label generation unit, which is used to add one or more corresponding labels to the outbound call devices according to the scores of the outbound call devices by the scoring model from one or more dimensions; and a configuration generation unit, which is used to generate an outbound call configuration result corresponding to the outbound call devices based on the scores of the outbound call devices by the scoring model, the configuration mechanisms to which each outbound call device is assigned, and the one or more labels of the outbound call devices.

[0013] In order to solve the above technical problem, the third aspect of the present invention proposes an electronic device, comprising: a processor, and a memory storing computer-executable instructions, wherein the computer-executable instructions, when executed, enable the processor to execute the method of the first aspect.

[0014] In order to solve the above technical problems, the fourth aspect of the present invention proposes a computer-readable medium, wherein the computer-readable medium stores one or more programs, and when the one or more programs are executed by a processor, the method of the first aspect is implemented.

[0015] According to the technical solution of the present invention, a database is constructed for customers who are to make outbound calls and their data, and model scoring is performed based on their characteristics to determine the categories and divide them into corresponding queues, so as to match them with the corresponding outbound call strategy configuration agency. In the configuration agency determined based on the processing technical capabilities and permissions, a unique and targeted outbound call configuration is automatically configured for each device to be called out based on the score, the corresponding random number generated by the big data reference iteration, the labels added corresponding to multiple focuses, etc., combined with the attributes and capabilities of the configuration agency, so as to obtain and execute the corresponding outbound call service strategy.

[0016] In this way, in various business processing scenarios, whether the logical control and management of outbound call equipment and the execution of outbound call strategies are performed manually or by intelligent robots, they are all flexibly carried out according to the specific circumstances of the actual business environment of the outbound call equipment. According to the specific circumstances, the customer groups to which the outbound call equipment belongs are divided and assigned to the corresponding configuration agencies. In combination with the skill level of the agency's business personnel, personalized outbound call strategies for special outbound call equipment are customized. This also provides sales personnel or business robots with accurate, efficient, low-repetition and low-error outbound call control, realizing the flexible strategy configuration and efficient outbound call control management that have been previously achieved, and avoiding defects such as erroneous outbound calls caused by huge individual differences due to various differences in the actual attributes of the outbound call equipment in the same queue.

[0017] Furthermore, for human sales representatives, this can effectively improve work efficiency and communication effectiveness. Furthermore, based on the profile of the device to be called, differentiated outbound call policies can be configured for users of sensitive devices. This can prevent a negative user experience during outbound call processing, reduce the consumption of platform resources such as lines, and reduce the risk of complaints caused by inefficient platform operations, achieving a win-win situation.

[0018] Furthermore, by introducing random numbers to conduct comparative tests on devices with the same attributes waiting for outbound calls, the strategy refinement and accuracy can be continuously increased, and the outbound call configuration can be iteratively optimized through continuous comparative tests. This also provides data support for the subsequent continuous iterative optimization of the outbound call configuration and outbound call business processing strategy, avoiding the defect of being unable to provide effective configuration iteration data under a rough strategy.

[0019] Furthermore, when applied to responding to social emergencies, such as special outbound call service processing strategies for outbound call equipment in high-risk areas in response to epidemics in public health and epidemic prevention, it provides high-efficiency and precise configuration and processing, and has strong scalability. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to make the technical problems solved by the present invention, the technical means adopted, and the technical effects achieved more clearly, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. However, it should be noted that the drawings described below are only drawings of exemplary embodiments of the present invention. Those skilled in the art can derive drawings of other embodiments based on these drawings without inventive effort.

[0021] Figure 1 This is a main flow chart of an embodiment of a dialing strategy configuration method according to the present invention;

[0022] Figure 2 This is a main structural block diagram of an embodiment of a dialing strategy configuration system according to the present invention;

[0023] Figure 3 is a structural block diagram of an embodiment of an electronic device according to the present invention;

[0024] Figure 4 is a structural block diagram of an embodiment of a computer-readable medium according to the present invention. DETAILED DESCRIPTION

[0025] The exemplary embodiments of the present invention will now be described more fully with reference to the accompanying drawings. Although each exemplary embodiment can be implemented in a variety of specific ways, it should not be understood that the present invention is limited to the embodiments set forth herein. Rather, these exemplary embodiments are provided to make the content of the present invention more complete and to more fully convey the inventive concept to those skilled in the art.

[0026] Under the premise of being consistent with the technical concept of the present invention, the structure, performance, effect or other characteristics described in a specific embodiment may be combined with one or more other embodiments in any appropriate manner.

[0027] In the introduction of specific embodiments, the detailed description of the structure, performance, effect or other features is intended to enable those skilled in the art to fully understand the embodiments. However, this does not preclude those skilled in the art from implementing the present invention with a technical solution that does not include the aforementioned structure, performance, effect or other features under specific circumstances.

[0028] The flowcharts in the accompanying drawings are merely illustrative of the process flow and do not necessarily include all of the content, operations, and steps in the flowcharts, nor do they necessarily imply that all of the steps in the flowcharts must be executed in the order shown. For example, some of the steps in the flowcharts may be separated, some may be combined or partially combined, and so on. The execution order shown in the flowcharts may be changed according to actual circumstances without departing from the spirit of the present invention.

[0029] Frames in the accompanying drawings Figure 1 The term "functional entity" generally refers to a functional entity and does not necessarily correspond to a physically independent entity. That is, these functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different networks and / or processor devices and / or microcontroller devices.

[0030] The same reference numerals in the accompanying drawings represent the same or similar elements, components or parts, and thus repeated descriptions of the same or similar elements, components or parts may be omitted below. It should also be understood that although the first, second, third and other numbered adjectives may be used herein to describe various devices, elements, components or parts, these devices, elements, components or parts should not be limited by these adjectives. In other words, these adjectives are only used to distinguish one from another. For example, the first device may also be called the second device, but this does not deviate from the essential technical solution of the present invention. In addition, the terms "and / or" and "and / or" refer to all combinations including any one or more of the listed items.

[0031] [Example 1]

[0032] The following combination Figure 2 The main flow chart of an embodiment of the method according to the present invention is shown to illustrate an implementation example of the present invention. In this embodiment, the outbound call configuration method includes:

[0033] Step S1 : Screening outbound call services that meet outbound call conditions in network interaction services, collecting device data of outbound call devices corresponding to the outbound call services, and establishing a basic database to store the outbound call devices and corresponding device data.

[0034] In one embodiment, device data of the devices to be called corresponding to the outbound call services are collected, and a basic database is established to store the devices to be called and the corresponding device data. For example, the server is used to analyze and organize the basic attributes of the devices to be called and write the corresponding device data sets into the basic database, and / or, the source data is synchronized to the data table of the basic database from an external data source or by file parsing through a shell script.

[0035] Data can be obtained from multiple data sources. For example, a core system can organize basic customer attributes and write data sets, and then synchronize source data from other data sources or by parsing files into the basic database tables for outbound calls using shell scripts. Specifically, analysis and organization based on the basic attributes of the outbound device involves: The device data for the outbound device includes personal information entered in the application information according to the prompts, and external data and network behavior data of the corresponding customers / users using the device is obtained from third-party platforms. This external data can include credit information, and network behavior data can include various current and historical behaviors, statuses, and results of the user / customer using the device.

[0036] Example 1:

[0037] The present invention is applied to an example of a specific business scenario, such as application to outbound call configuration for post-loan monitoring, control of dialing / calling equipment for manual monitoring, etc.

[0038] Step 1: Collect device data and establish a basic database of devices to be called outbound.

[0039] In a specific embodiment, by reading the core system database, the data exchange business cases are screened to select the outbound call business cases that meet the outbound call conditions, and the device data of the outbound call device corresponding to the outbound call business cases is obtained. The outbound call can be a reminder notification or an early warning notification.

[0040] Among them, data interaction can be that the data provider provides data content to the user, and the data user uploads relevant information in real time when using the data. For example, in an enterprise with certain data confidentiality, when an employee queries confidential data, it is determined in real time whether the employee has illegally obtained the confidential data, such as taking screenshots, inserting a USB flash drive, etc.

[0041] For another example, the information of each pending outbound call business case includes: whether the pending outbound call business case is online / offline, the account age of the pending outbound call business case, and if there is a historical outbound call operation for the pending outbound call business case, it may also include outbound call follow-up information. The outbound call follow-up information includes: the number of call notifications for the case in the previous month, the number of dialed numbers, the number of dials made, the number of dials made by the user, whether the connection was lost, etc.

[0042] The device data for the device to be called includes the user of the device to be called or the customer requesting the data, relevant information entered for the data application, the customer's own account information, including the terminal device to be called, and outbound call information. Application information can include basic information authorized by the user, such as service account information, user terminal device identification information, and user location information. Application information can also include behavioral information, such as user behavior in videos or images, user page operation data, user service access duration, and user service access frequency. The specific content of the application information can be determined based on the actual application scenario and is not limited here.

[0043] Furthermore, the device data of the devices to be called out is sorted and the data set is written. For example, the source data can be synchronized to the basic database table of the devices to be called out in the outbound call system from other data sources through shell scripts or by parsing files.

[0044] In one specific embodiment, after a customer enters the application information using their terminal device according to the corresponding prompts, the customer's basic information and network behavior data can be obtained from the third-party platform based on the personal information in the application information. It should be noted that basic information may include business account information, user terminal device identification information, user location information, etc.

[0045] In one embodiment, the data dimensions of online behavior data can be very broad, including, but not limited to, basic attribute information provided by customers when using terminal devices, customer behavior information, customer online purchases, customer app behavior, social circle information, etc. In the context of post-loan monitoring, for example, online behavior data may include user page operation data, user service access duration, user service access frequency, social behavior data, etc. This online behavior data can be obtained from the corresponding partner organization with the authorization of the customer when applying for the service. For example, an online shopping platform can obtain a customer's payment behavior data and online shopping data.

[0046] Furthermore, the collected device data is pre-processed. For example: Data cleaning: removing default and constant variables, checking the missing value ratio, removing variables with high missing value ratios, checking the data definition range, and identifying incorrect variables. For example: Variable derivation and clustering: handling variable outliers, performing variable derivation and clustering analysis based on existing data, and clustering variables by relevance. For example: Data analysis: customer group analysis to determine whether customers need to be modeled separately, and determining the target variable definition and observation period based on data analysis. For example: Descriptive analysis: in-depth analysis of variable distribution, linear relationships between variables, understanding variables based on data analysis, and empirical analysis of variable availability.

[0047] Step S2: Analyze the characteristics of the device to be called based on the device to be called and the corresponding device data, and score the corresponding device to be called using a scoring model according to the characteristics to determine the category to which the device to be called belongs.

[0048] In one embodiment, within the big data platform, by analyzing various dimensional data such as historical behavior data of users / customers who use callable devices, historical fulfillment data of users / customers, static / dynamic basic data of users / customers, external data, etc., a model score is given to the corresponding devices to be called used by these customers / users.

[0049] Specifically, the server analyzes the device data in the basic database from multiple dimensions based on the big data platform. The device data in the multiple dimensions include historical behavior data of the corresponding network, historical performance data, dynamic and / or static basic data, external service rating data, etc.; using the trained rating model, the device data of the multiple dimensions analyzed in the current database is input to predict and score the corresponding devices to be called outbound; according to the classification processing of the rating, different categories are determined for the corresponding devices to be called outbound; wherein, the rating model is constructed using a machine learning algorithm model and is trained using the historical devices to be called outbound that have been classified and the corresponding device data as training data.

[0050] For example, in the case where the embodiment of the present invention is applied to a post-loan monitoring business scenario:

[0051] Step 2: Based on the acquired device data of the devices to be used for outbound calls, analyze the features of the devices to be used for outbound calls and score the terminal devices to be used for outbound calls. That is, score the customers who use the devices and classify the devices to be used for outbound calls into different categories, thereby classifying the customers who use the devices into different categories.

[0052] Specifically, machine learning algorithms can be used to establish scoring models, and outbound call strategies can be formulated based on the model scores. This can be used in post-loan processes such as financial institutions. This model can be used to quantify customer delinquencies using specific model scores. Common examples include outbound call model scores, loss of contact model scores, and aging rollover rate model scores. Each model type can be divided into models with different delinquency status based on specific scenarios. The machine learning algorithm can include at least one of a classifier, logistic regression, or neural network.

[0053] Thus, a scoring model can be used to select data from the various information obtained in the first step, such as historical customer behavior data (such as historical borrowing or repayment behavior data of overdue customers), static / dynamic basic customer data, external data, and other dimensional data, to perform feature analysis on the customer. Based on the analysis and processing results, risk assessment information of the customer (such as an overdue customer) can be predicted or determined. The risk assessment information can be used to determine the risk level of the customer. Risk levels can include high risk, medium risk, and low risk.

[0054] Taking post-loan monitoring as an example, a loan is classified as high risk if any of the following conditions are met: 1) The customer has been overdue for more than five days in the past six months; 2) The loan amount is over 250,000 yuan in the past three months and there has been a past due payment; 3) The application score is below 580 points in the past three months and there has been a past due payment; 4) There have been two or more past due payments in the past three months; 5) If the loan is repaid in installments, there has been a current overdue payment. A loan is classified as medium risk if any of the following conditions are met: a) The first installment payment is overdue; b) There has been a single overdue payment in the past six months; c) The loan balance is over 50,000 yuan and the billing date is after the 25th of the month; d) The loan amount is over 250,000 yuan in the past three months; e) The application score is below 580 points in the past three months. Low risk refers to any of the following conditions except for high and medium risk.

[0055] Step S3: Allocate the devices to be called outbound to corresponding configuration agencies according to their categories.

[0056] In one embodiment, the server pre-configures different groups of outbound call service configuration organizations based on the different attributes of outbound call interaction scenarios and outbound call skills; determines the authority of the configuration organization based on the layer settings of the configuration organization; divides different queues according to the type of devices to be called out; different queues have different authority configurations; and assigns the queues to configuration organizations with matching authority based on the authority configurations. Furthermore, a multi-digit random number is generated for the devices to be called out in the configuration organizations with matching authority items, specifically including: dividing the devices to be called out in multiple queues into different outbound call groups based on the devices to be called out in each queue; generating a different random number for each outbound call group using a random number algorithm; and determining the number of outbound calls for the devices to be called in the corresponding outbound call group based on the random number. Based on all the devices waiting for outbound calls that are assigned or changed to the same queue on the same day, the devices waiting for outbound calls in each queue are divided into different outbound call groups, a corresponding random number is generated for each outbound call group, and the number of calls for the devices waiting for outbound calls in the outbound call group is set according to the random number. After determining the corresponding configuration mechanism for the devices waiting for outbound calls, this solution determines different numbers of calls for devices waiting for outbound calls of the same category to achieve test comparison; after the devices waiting for outbound calls in the same queue complete the calls according to the number of calls corresponding to the random number, the complaint rate and purpose completion rate of each device waiting for outbound calls can be determined, and the number of calls can be iteratively updated according to the complaint rate and purpose completion rate, which can continuously increase the precision and accuracy of the dialing times.

[0057] For example, the devices waiting for outbound calls in the same queue are divided into two groups, A and B, and random numbers are generated for groups A and B respectively. For example, the random number of group A is 5, and the random number of group B is 8. Group A is set to make 5 calls a day, and group B is set to make 8 calls a day. Then the complaint rate and recovery rate of the two groups are observed to iterate the calling strategy. That is, if the devices waiting for outbound calls in the queue do not change in the future, the number of outbound calls for the devices waiting for outbound calls in the queue is adjusted according to the outbound call results of the test control scheme based on the random number scheme. Moreover, further iteration can be performed to group the devices waiting for outbound calls in the queue again. At this time, the number of outbound calls can be set, and slight adjustments can be made based on the determined number of outbound calls. For example, different numbers of outbound calls can be obtained by adding or subtracting one from the number of outbound calls, and different numbers of outbound calls are used for outbound calls to devices waiting for outbound calls in different groups.

[0058] Furthermore, outbound call services can be provided through manual service or voice robots. For manual service, the different attributes of the outbound call skills of manual service can be obtained by following up the service processes of different employees, obtaining the employees' voice service records, and then determining the employee's outbound call skill level through voice service records, user post-service ratings, and service purpose achievement rates. Furthermore, the employee's outbound call skill level for different service purposes can also be determined, for example, the outbound call skill level for services at different stages such as customer acquisition, return visits, and after-sales. In this embodiment, each employee is set to a different outbound call business configuration organization group according to the outbound call interaction scenario and the level of the outbound call skills; for voice robots, the different attributes of the voice robot's outbound call skills can be the data used when constructing the voice robot. Different voice robots can perform different outbound call services due to different data sources. In the process of enterprise services, in order to make the voice robot more humanized, a more accurate data source will be used when building the voice robot to avoid mixing different data to train the voice robot and improve the response efficiency of the voice robot. Although this will train voice robots for different services, the voice robots obtained in this way can provide customers with more professional services and avoid answering questions that are not asked. For example, a voice robot built by training outbound voice calls made by customer acquisition employees may not be able to complete after-sales service or other services well. In this embodiment, each voice robot is set to a different outbound call service configuration organization group according to the outbound call interaction scenario and the business category to which the outbound call skills belong.

[0059] In this embodiment, the authority of the configuration agency can be determined according to the attributes of the outbound call skills. The outbound call skill level of each subject can be determined through the voice records of each subject making outbound calls, the user's post-service rating, and the service purpose achievement rate. The higher the outbound call skill level, the higher the authority of the corresponding configuration agency, so as to realize the division of subjects with different outbound call skills into different configuration agencies, so as to provide services for different devices to be called out.

[0060] Furthermore, the random number generator uses an algorithm to uniformly generate multi-digit random numbers for all customers who change their queues that day. This ensures uniformity and randomness by generating random numbers in large quantities. This ensures that different functional tests do not interfere with each other.

[0061] Furthermore, organizations, here referring to configuration organizations, can include salespeople or business robots (e.g., salespeople in post-loan monitoring scenarios), divided into different organization groups based on various organizational attributes such as workplace and skills. Furthermore, through the configuration of the organization layer, various permissions for each configuration organization, such as data, menu, and function permissions, can be determined.

[0062] Furthermore, based on the results of steps S1 and S2, the pending customers are divided into different groups, also known as queues, according to their respective categories. Within these queues, cases are assigned to different organizations based on their assigned permissions, each with different skill attributes. A multi-digit random number is also generated for these customers for use in testing and comparison of various functions.

[0063] Take the post-loan monitoring business scenario as an example:

[0064] Step 3: Divide the devices waiting for outbound calls into different queues and assign them to different organizations for processing.

[0065] Specifically, during outbound call processing, clients (e.g., overdue customers) are divided into different groups, called queues, based on their objective attributes (e.g., remaining principal, overdue stage, credit limit utilization rate, etc.) and subjective evaluations (e.g., model scores, cumulative customer behavior scores, and past post-loan performance). Each queue has different handling skills and available information, enabling refined outbound call processing strategies (e.g., reminders, alerts, etc.) tailored to the specific situation.

[0066] In one specific implementation, a salesperson database (e.g., post-loan salesperson or sales robot) can also be included. This database contains all assignable salespersons (e.g., assignable post-loan salespersons), and each salesperson's basic information is also stored as a salesperson tag in the database. For example, the basic information of a post-loan salesperson may include: work experience, the amount of repayments received last month, the number of completed cases, the number of outbound calls, and other performance information.

[0067] Based on the results of steps 1 and 2, the devices waiting for outbound call cases are divided into different queues. Within each queue, cases are assigned to different organizations based on their assigned permissions, and each organization has different business skill attributes.

[0068] Specifically, dividing the devices to be called outbound into different queues means dividing the devices to be called outbound into major categories according to different decision variables in the basic information of the services to be called outbound.

[0069] In post-loan monitoring scenarios, for example, devices waiting for outbound calls can be categorized by the customers / users using the devices, such as business card customers, I-debit customers, M-debit customers, B-debit customers, customers with fraud disputes, ultra-small loan customers, special fund loan customers, and ordinary customers. I-debit customers, M-debit customers, and B-debit customers are customers who are overdue for more than 30 days, more than 60 days, and more than 90 days, respectively.

[0070] For example, customers can also be divided into: customers who have actively repaid their debts, customers who have made outbound calls and repaid their debts, and customers who have made outbound calls but have not repaid their debts.

[0071] For example, customers can be categorized as those with successful outbound calls and those with failed outbound calls. A successful outbound call can mean that the customer answered the call and made the repayment on time or according to the time promised on the call, or that the customer answered the call and made the repayment on the same day as the outbound call. Correspondingly, a failed outbound call can mean that the customer did not answer the call, or that the customer answered the call but hung up midway, or that the customer answered the call but did not make the repayment on time or did not make the repayment on the same day as the outbound call, or did not make the repayment on the same day as the outbound call, etc.

[0072] For example, customers can be divided into six queues: A, B, C, D, E, and F, which are:

[0073] A, corresponding to the first overdue payment (repayment within the interest-free period);

[0074] B, corresponding to the first three periods;

[0075] C, corresponding to historical overdue days of more than 30 / 60 days;

[0076] D, corresponding to the determination of loss of contact;

[0077] E, corresponding to extension, renewal, and overdue after restructuring;

[0078] F, corresponding to difficult issues (GPS offline / vehicle preservation / accident / litigation).

[0079] The decision on which agency to assign the payment to is primarily based on cost and operational capacity. If a customer can be assigned to a low-cost agency for timely reminders, there's no need to assign the customer to a high-cost agency. Initially, SMS and IVR can be used as the primary means of handling overdue payments. If the customer still fails to repay, the process can be gradually upgraded to junior, mid-level, and senior-level skilled staff.

[0080] On the other hand, outbound calls can be handled by robots or humans, depending on the difficulty level. Depending on the situation, outbound calls can be assigned to different outbound sales representatives. For example, for customers facing financial difficulties, highly skilled human operators can be assigned to handle the calls.

[0081] At the same time, multi-digit random numbers are generated for these customers and used for testing and comparison of several functions.

[0082] Random numbers are generated by applying an algorithm to uniformly generate multi-digit random numbers for all devices waiting for outbound calls that have changed queues that day. This large-scale random number generation ensures uniformity and randomness. This multi-digit random number generation can be used to randomly distribute cases to one group of organizations or another, ensuring that different functional tests do not interfere with each other.

[0083] In addition, by introducing random numbers to distinguish customers with the same attributes, the outbound calling strategy can be continuously tested and iterated to optimize.

[0084] For example, outbound call cases or outbound call equipment can be reallocated between configuration organizations.

[0085] For example, the outbound business cases or devices to be called out will be evenly distributed among the organizations. In other embodiments, an uneven distribution of outbound business cases or devices to be called out can be used.

[0086] Step S4: adding one or more corresponding tags to the device to be called based on the rating of the device to be called by the rating model from one or more dimensions.

[0087] This method of adding tags is to specifically identify the devices to be used for outbound calls. Through big data analysis, especially data related to the use of devices by customers and users, different dimensions of tagging, display, and screening are achieved for the devices used by customers and users.

[0088] In one implementation, the big data platform assigns multiple tags to devices waiting for outbound calls, based on different dimensional emphases. A single device can have an unlimited number of tags. By focusing on different dimensional emphases, the characteristics of devices waiting for outbound calls (such as overdue customers in post-loan monitoring) are characterized from multiple dimensions, creating a device profile or customer profile. Multiple tags are assigned to devices, or to customers using them.

[0089] Furthermore, device profiling or device profiling mainly assists in the formulation and implementation of outbound call strategies from the distribution of characteristics of different customer groups using the device. Customer characteristics can be mainly basic information, including age, gender, marital status, years of work, educational level, housing type, etc. In the application scenario of post-loan monitoring, for example, for the same overdue amount, the risk level of different ages is quite different. The risk level of overdue customers aged 40+ is significantly higher than that of overdue customers aged 30-35. The main reason is that the overall qualifications and capabilities of the two age groups are quite different. Therefore, in the context of comprehensively considering various different key feature dimensions (such as the overdue account period, overdue balance, model score and other dimensions of overdue customers in the post-loan monitoring scenario), the feature reference of customer profiling is also an important factor in the formulation of specific outbound call strategies.

[0090] Post-loan monitoring application scenario:

[0091] Step 4: Characterize overdue customers from multiple dimensions by focusing on different dimensions and assign them multiple labels. The same overdue customer can have an unlimited number of labels.

[0092] For example, based on repayment ability and willingness, you can set the following tags:

[0093] High repayment ability and willingness: These customers are often those who forget to repay or are habitually overdue. They have the highest reminder rate and are higher than the overall market.

[0094] High repayment ability, low repayment willingness: This type of customer has the repayment ability and is likely to repay early on when reminders are given. However, the repayment probability gradually decreases over time.

[0095] Low repayment ability, high repayment willingness: This type of customer often defaults because they lack the ability to repay temporarily, but have a high repayment willingness. Some customers will only repay after being reminded for a period of time.

[0096] Low repayment ability and low repayment willingness: This type of customer is more difficult to call out.

[0097] For example, you can classify them according to the sensitivity to outbound calls:

[0098] The first category: customers who refuse to repay regardless of reminders or not;

[0099] The second category: customers who repay when not reminded, but do not repay even after being reminded;

[0100] The third category: customers who will repay regardless of reminders or not;

[0101] The fourth category: customers who will not return the money unless reminded, and will return it only after being reminded.

[0102] Comprehensively considering multiple dimensions or all dimensions can more accurately, precisely and efficiently automatically customize the corresponding outbound call configuration based on big data.

[0103] Step S5: generating an outbound call configuration result corresponding to the device to be called based on the score of the device to be called by the scoring model, the configuration mechanism to which each device to be called is assigned, and one or more tags of the device to be called.

[0104] In one embodiment, the devices to be called outbound in the queue are assigned to corresponding configuration agencies, which generate outbound call configuration results corresponding to the devices to be called outbound based on the authority and outbound call skills of the configuration agency, combined with the scores of the devices to be called outbound, the multi-digit random numbers generated for the devices to be called outbound, and one or more added tags. The configuration agency then uses the outbound call skills to execute the outbound call service for the devices to be called outbound based on the outbound call configuration results under the authority restrictions of the configuration agency.

[0105] By combining multiple dimensions such as the model score, outbound call mechanism, random numbers, and multiple tags, a targeted outbound call configuration for the device to be called is automatically customized or generated. When outbound call processing needs to be performed, the outbound call configuration strategy formed by the configuration is used, and accurate outbound call control is achieved by executing the strategy.

[0106] In actual applications, outbound calling strategies can be specifically customized through automatically customized outbound calling configurations to determine the number of times a customer can be called outbound on the same day, the number of times a call can be connected, three-party outbound calling strategies, and other complete outbound calling strategies based on customer attributes.

[0107] In the application scenario of post-loan monitoring:

[0108] Step 5: Based on the four parameters determined in the previous steps, the outbound call configuration of the device to be used for outbound calls is automatically customized, and the policy control of the customer's outbound call service is executed in real time.

[0109] In this real-time control strategy, specific outbound call strategies may include: the upper limit of outbound calls for customers using corresponding terminal devices within one day, the upper limit of connections / notifications, the upper limit of three-party outbound calls, the upper limit of three-party connections / notifications, whether three parties can be contacted, whether three parties can be viewed, the time when outbound calls are available, etc.

[0110] By combining the four parameters for outbound call configuration, different customer-specific parameters can be divided to further automatically and efficiently customize refined and differentiated outbound call strategies based on big data information.

[0111] In post-loan monitoring scenarios, for example, for rebellious customers, the frequency of outbound calls using the wrong terminal devices may need to be reduced, while for reformed customers, the frequency may need to be increased. For example, different outbound call times, ring times, and call frequency can be set for different customers based on their industry's work patterns, thereby improving loan outbound call efficiency. Another example is the ability to tailor different call scripts to different industries and job profiles, such as businesspeople, employees, and the unemployed, further enhancing outbound call control during loan monitoring.

[0112] In this way, by allowing the most suitable salesperson or business robot to make the most appropriate outbound call at the most appropriate time, through the most appropriate outbound call processing, that is, outbound call strategy, to achieve intelligent outbound calling, the goal of optimizing resource allocation and improving resource allocation efficiency and accuracy can be achieved.

[0113] Furthermore, the records of outbound call service processing that have been executed can be fed back to various databases, such as the database where the device data is located, to update or improve the device data.

[0114] The existing monitoring outbound call system can only roughly define the outbound call strategy of each outbound call device in different queues on the same day. But in fact, even if the customers are in the same queue, the differences between customers are very large. If the unified strategy is obviously too rigid, it will cause unnecessary efficiency reduction, too many errors, poor accurate monitoring capabilities, and even an increase in complaints. By using the big data platform introduced by the present invention, customers can be finely profiled, and different customers can be labeled differently based on various aspects of customer history, static and dynamic qualification status, external data, etc., to achieve a personalized dialing strategy for each customer. In combination with the skills of salesmen or business robots, they are divided into different organizations, and different strategies are set for different organizations, which can also achieve a personalized dialing strategy for each salesman or business robot. In addition, by introducing random numbers to distinguish customers with the same attributes, the dialing strategy can be continuously optimized by comparison, testing, and iteration.

[0115] [Example 2]

[0116] Figure 2 This is a functional module architecture block diagram of an embodiment of an outbound call configuration system according to the present invention. The system includes at least:

[0117] Database construction unit 210 is a unit for constructing a basic database for devices waiting for outbound calls. It is configured to screen network interaction services for outbound call services that meet outbound call requirements, collect device data for the devices waiting for outbound calls corresponding to the outbound call services, and establish a basic database storing the devices waiting for outbound calls and their corresponding device data. The specific functions implemented by unit 210 are described in step S1 of Example 1 and are not further detailed here.

[0118] Model calculation unit 220 is configured to analyze the characteristics of the devices to be called based on the devices and their corresponding device data. Based on these characteristics, the scoring model is used to score the devices to be called, thereby determining the category to which the devices belong. The specific functions implemented by unit 220 are described in step S2 of Example 1 and are not further detailed here.

[0119] The queue division unit 230 is used to assign the devices to be called out to the corresponding configuration mechanism according to their categories. The specific functions implemented by the unit 230 refer to step S3 in embodiment 1, and the specific contents are not repeated here.

[0120] Label generation unit 240 is a unit for generating labels for devices to be called outbound. It is configured to assign one or more corresponding labels to the devices to be called outbound based on the scoring model's evaluation of the devices from one or more dimensions. The specific functions implemented by unit 240 are described in step S4 of Example 1 and are not further detailed here.

[0121] Configuration generation unit 250, which generates the configuration for the outbound call policy, is configured to generate an outbound call configuration result for each device to be called based on the scoring model's scores for the devices to be called, the configuration mechanisms assigned to each device to be called, and one or more tags for the device to be called. The specific functions implemented by unit 250 are described in step S5 of Example 1 and are not further detailed here.

[0122] In the process of outbound call configuration for outbound call devices, in addition to changing or assigning them to corresponding queues and allocating them to corresponding configuration agencies, the system also automatically generates or customizes different proprietary outbound call configurations for the same agency under the same queue for the outbound call devices based on the different attributes and skills of the configuration agencies themselves and the labels of random numbers and multi-dimensional reference data of other outbound call devices. This accurately and efficiently determines the outbound call strategy when implementing real-time outbound call business processing on them, and executes the outbound call business processing to make effective outbound calls to the outbound call devices.

[0123] Those skilled in the art will appreciate that the various units in the above device embodiments may be distributed in the system as described, or may be modified accordingly and distributed in one or more systems different from the above embodiments. The units in the above embodiments may be combined into one unit or further divided into multiple sub-units.

[0124] [Example 3]

[0125] The following describes an electronic device embodiment of the present invention. This electronic device can be considered a physical implementation of the method and apparatus embodiments of the present invention described above. Details described in the electronic device embodiment of the present invention should be considered supplementary to the above-described method or apparatus embodiments; details not disclosed in the electronic device embodiment of the present invention can be implemented with reference to the above-described method or apparatus embodiments.

[0126] Figure 3 is a structural block diagram of an exemplary embodiment of an electronic device according to the present invention. Figure 3 The electronic device shown is only an example and should not limit the functions and scope of use of the embodiments of the present invention.

[0127] like Figure 3 As shown, the electronic device 400 of this exemplary embodiment is in the form of a general-purpose data processing device. Components of the electronic device 400 may include, but are not limited to, at least one processing unit 410, at least one storage unit 420, a bus 430 connecting different electronic device components (including the storage unit 420 and the processing unit 410), a display unit 440, and the like.

[0128] The storage unit 420 stores a computer-readable program, which may be a source program or a code of a read-only program. The program may be executed by the processing unit 410, so that the processing unit 410 performs the steps of various embodiments of the present invention. For example, the processing unit 410 may perform the following steps: Figure 1 Steps shown.

[0129] The storage unit 420 may include a readable medium in the form of a volatile storage unit, such as a random access memory unit (RAM) 4201 and / or a cache memory unit 4202, and may further include a read-only memory unit (ROM) 4203. The storage unit 420 may also include a program / utility 4204 having a set (at least one) of program modules 4205. Such program modules 4205 include, but are not limited to, operating the electronic device, one or more application programs, other program modules, and program data. Each of these examples or some combination thereof may include the implementation of a network environment.

[0130] Bus 430 may represent one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, a processing unit, or a local bus using any of a variety of bus architectures.

[0131] The electronic device 400 can also communicate with one or more external devices 100 (e.g., keyboard, display, network device, Bluetooth device, etc.), so that the client can interact with the electronic device 400 via these external devices 100, and / or enable the electronic device 400 to communicate with one or more other data processing devices (e.g., router, modem, etc.). Such communication can be carried out through the input / output (I / O) interface 450, and can also be carried out with one or more networks (e.g., local area network (LAN), wide area network (WAN) and / or public network, such as the Internet) through the network adapter 460. The network adapter 460 can communicate with other modules of the electronic device 400 through the bus 430. It should be understood that although Figure 4 Not shown, other hardware and / or software modules may be used in the electronic device 400, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID electronic devices, tape drives, and data backup storage electronic devices.

[0132] [Example 4]

[0133] Figure 4 Schematic diagram of a computer readable medium embodiment of the present invention. Figure 4As shown, the computer program can be stored on one or more computer-readable media. The computer-readable medium can be a readable signal medium or a readable storage medium. The readable storage medium can be, for example, but not limited to, an electronic device, device or component of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination thereof. More specific examples (non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. When the computer program is executed by one or more data processing devices, the computer-readable medium is enabled to implement the above-mentioned method of the present invention, namely: obtaining Chinese keywords of each corporate entity; semantically splitting the keywords according to pictographic elements, and outputting the word vector of the corporate entity according to the semantic splitting result; identifying the word vector through a semantic relevance interpretation model to obtain the public opinion recognition result of the corporate entity, the semantic relevance interpretation model is used to identify the semantic relevance between the keywords and each word or Chinese character in the text training data, and using the words or Chinese characters whose semantic relevance meets the threshold as the public opinion recognition result to perform semantic level interpretation on the keywords; constructing an enterprise knowledge graph based on the public opinion recognition results of each corporate entity and the economic relationship between the corporate entities; and determining blacklisted enterprises based on the enterprise knowledge graph.

[0134] Through the description of the above embodiments, it is easy for those skilled in the art to understand that the exemplary embodiments described in the present invention can be implemented by software or by combining software with necessary hardware. Therefore, the technical solution according to the embodiment of the present invention can be embodied in the form of a software product, which can be stored in a computer-readable storage medium (which can be a CD-ROM, USB flash drive, mobile hard disk, etc.) or on a network, and includes several instructions to enable a data processing device (which can be a personal computer, server, or network device, etc.) to execute the above method according to the present invention.

[0135] The computer-readable storage medium may include a data signal propagated in baseband or as part of a carrier wave, wherein the readable program code is carried. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. The readable storage medium may also be any readable medium other than a readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction-executing electronic device, apparatus, or device. The program code contained on the readable storage medium may be transmitted using any suitable medium, including but not limited to wireless, wired, optical cable, RF, etc., or any suitable combination thereof.

[0136] The program code for performing the operations of the present invention may be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java, C++, and the like, as well as conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the client computing device, partially on the client device, as a stand-alone software package, partially on the client computing device and partially on a remote computing device, or entirely on a remote computing device or server. In cases involving a remote computing device, the remote computing device may be connected to the client computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0137] In summary, the present invention can be implemented by a method, apparatus, electronic device or computer-readable medium that executes a computer program. In practice, a general-purpose data processing device such as a microprocessor or a digital signal processor (DSP) can be used to implement some or all of the functions of the present invention.

[0138] The specific embodiments described above further illustrate the objectives, technical solutions, and beneficial effects of the present invention. It should be understood that the present invention is not inherently related to any specific computer, virtual device, or electronic device, and various general-purpose devices can also implement the present invention. The above description is only a specific embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention shall be included in the scope of protection of the present invention.

Claims

1. A method for generating differentiated outbound call configurations based on a scoring model, characterized in that: include: By screening the outbound call services that meet the outbound call conditions in the network interaction services, collecting the device data of the outbound call devices corresponding to the outbound call services, and establishing a basic database to store the outbound call devices and corresponding device data; Analyze the characteristics of the devices to be called based on the devices and corresponding device data, and score the corresponding devices to be called using a scoring model based on the characteristics to determine the category to which the devices to be called belong; Allocate the devices to be called outbound to the corresponding configuration organization according to their categories, including: The server pre-sets different outbound call service configuration organization groups based on the different attributes of outbound call interaction scenarios and outbound call skills; The authority of the configuration organization is determined according to the configuration organization's internal settings; Different queues are divided according to the categories of devices to be called out; Different queues have different permission configurations; Assigning queues to configuration organizations with matching permissions based on the permissions configuration; and, Generating a multi-digit random number for a waiting outbound call device in a configuration mechanism with matching authority, including: dividing the waiting outbound call devices in multiple queues into different outbound call groups based on the waiting outbound call devices divided into different queues, generating a different random number for each outbound call group using a random number algorithm, and determining an outbound call count for the waiting outbound call device in the corresponding outbound call group based on the random number; Based on one or more dimensions, the scoring model is used to score the outbound call device, and one or more corresponding tags are added to the outbound call device. An outbound call configuration result corresponding to the device to be called is generated based on the score of the device to be called by the scoring model, the configuration mechanism to which each device to be called is assigned, and one or more tags of the device to be called.

2. The method according to claim 1, wherein Collect the device data of the outbound call devices corresponding to the outbound call services, and establish a basic database to store the outbound call devices and corresponding device data, including: The server is used to analyze and organize the basic attributes of the device to be called out, and the corresponding device data set is written into the basic database. And / or, the source data is synchronized from the external data source or through file parsing to the data table of the basic database through shell scripts.

3. The method according to claim 2, wherein Analyze and organize the basic attributes of the device to be called, including: The personal information in the application information entered according to the prompts in the device data of the outbound call device is used to obtain the external data and network behavior data of the customer using the corresponding device from a third-party platform.

4. The method according to claim 1, wherein After allocating the queue to a configuration organization with matching permissions according to the permission configuration, it also includes: Determine different call times for outbound call devices of the same category for testing and comparison.

5. The method according to claim 4, wherein Also includes: After the outbound call devices in the same queue complete the calls according to the number of calls corresponding to the random number, the complaint rate and purpose completion rate of each outbound call device are determined; The number of calls is iteratively updated based on the complaint rate and goal completion rate.

6. The method according to claim 1, wherein The characteristics of the devices to be called and the corresponding device data are analyzed. The corresponding devices to be called are scored using a scoring model based on the characteristics to determine the category to which the devices to be called belong. Specifically, the following are included: The server analyzes the device data in the basic database from multiple dimensions based on the big data platform. The device data in multiple dimensions includes historical network behavior data, historical contract fulfillment data, dynamic and / or static basic data, and external service rating data. Using the trained scoring model, input the device data of multiple dimensions analyzed in the current database to predict and score the corresponding devices to be called outbound; Determine different categories for the corresponding outbound call device based on the classification processing of the scores; The scoring model is constructed using a machine learning algorithm model and trained using historical outbound call devices and corresponding device data that have been classified as such as the original data.

7. The method according to any one of claims 1 to 6, wherein: Based on the scores of the outbound call devices assigned by the scoring model, the configuration agencies to which the devices are assigned, and one or more tags of the devices, an outbound call configuration result corresponding to the devices is generated, specifically including: The configuration agency assigned to the device to be called outbound generates an outbound call configuration result corresponding to the device to be called outbound based on the authority and outbound calling skills of the configuration agency, combined with the score of the device to be called outbound, the multi-digit random number generated for the device to be called outbound, and one or more tags added to the device to be called outbound; and The configuration agency uses the outbound call skills according to the outbound call configuration results and executes the outbound call service for the outbound call device within the authority restrictions of the configuration agency.

8. A differentiated outbound call configuration generation system based on a scoring model, characterized in that: include: A database construction unit is used to screen outbound call services that meet outbound call conditions in network interactive services, collect device data of outbound call devices corresponding to the outbound call services, and establish a basic database for storing the outbound call devices and corresponding device data; A model calculation unit is used to analyze the characteristics of the device to be called based on the device to be called and the corresponding device data, and to score the corresponding device to be called using a scoring model according to the characteristics to determine the category to which the device to be called belongs; A queue division unit is used to assign devices to be called out to corresponding configuration agencies according to their categories, including: the server pre-setting different groups of outbound call service configuration agencies based on different attributes of outbound call interaction scenarios and outbound call skills; determining the authority of the configuration agency based on the layer settings of the configuration agency; dividing different queues according to the categories to which the devices to be called out belong; different queues have different authority configurations; assigning queues to configuration agencies with matching authority according to the authority configuration; and generating multi-digit random numbers for the devices to be called out in the configuration agencies with matching authority, including: dividing the devices to be called in multiple queues into different outbound call groups based on the devices to be called that are divided into different queues, generating different random numbers for each outbound call group through a random number algorithm, and determining the number of outbound calls for the devices to be called in the corresponding outbound call group based on the random numbers; A label generation unit, configured to add one or more corresponding labels to the device to be called based on the score of the device to be called according to the scoring model from one or more dimensions; The configuration generating unit is configured to generate an outbound call configuration result corresponding to the device to be called based on the score of the device to be called by the scoring model, the configuration mechanism to which each device to be called is assigned, and one or more tags of the device to be called.

9. An electronic device comprising a processor and a memory storing computer-executable instructions; characterized in that: The computer executable instructions, when executed, cause a processor to perform the method according to any one of claims 1 to 7.

10. A computer-readable medium, wherein: The computer-readable medium stores one or more programs, and is characterized in that when the one or more programs are executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images