Employee job number behavior early warning method and device and readable storage medium

Abstract

The invention provides an employee job number behavior early warning method and device and a readable storage medium, and the method comprises the steps: obtaining the query and acceptance order conditions of a target employee job number in a plurality of account periods, and the plurality of account periods comprise a current account period and at least one historical account period; obtaining an early warning index of the job number of the target employee in each account period according to the query and acceptance order condition of the job number of the target employee in each account period in the plurality of account periods; calculating a target early warning score of the target employee job number by using a preset early warning scoring rule according to the early warning index of the target employee job number in each account period; and judging whether the behavior of the target employee job number is abnormal or not according to the target early warning score of the target employee job number. According to the method, the device and the readable storage medium, the problem that an existing employee job number behavior early warning method is single in consideration factor, so that an early warning result is not accurate enough can be solved.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to an employee work number behavior early warning method, device and readable storage medium. Background Art

[0002] As a key driving force for the development of the digital economy, communication technology has been continuously innovating and iterating in recent decades, profoundly affecting the transformation of industrial value and business models. In this process, telecom operators are increasingly closely connected with the production and operation of enterprises and the daily lives of the people, and their production and operation behaviors are directly related to the vital interests of the general public. Today, telecommunications services have become a necessity for the normal operation of enterprises and institutions and the daily lives of urban and rural residents, and have almost become a basic public requirement of society.

[0003] Given the popularity and importance of telecommunications services, telecommunications operators must implement a people-centered development philosophy and service concept. This means that operators need to enhance service awareness, deeply safeguard consumer rights, and continuously improve service levels to meet people's pursuit of higher-quality telecommunications services. In order to provide users with better services, it is particularly important to carefully monitor and warn the behavior of operators' internal employees. By establishing a sound early warning mechanism, it is possible to quickly identify abnormal behaviors that may exist in employee IDs within a specific time period, thereby ensuring the professionalism and reliability of services and safeguarding user rights. This will not only promote the overall improvement of operators' service quality, but will also further enhance the public's trust and satisfaction with telecommunications services.

[0004] However, the existing employee ID behavior early warning method considers only a single factor, resulting in inaccurate early warning results. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide an employee work number behavior warning method, device and readable storage medium in response to the above-mentioned shortcomings of the prior art, so as to solve the problem that the existing employee work number behavior warning method considers a single factor, resulting in inaccurate warning results.

[0006] In a first aspect, the present invention provides an employee work number behavior early warning method, comprising:

[0007] Obtaining the query and order acceptance status of the target employee's employee number in multiple billing periods, wherein the multiple billing periods include the current billing period and at least one historical billing period;

[0008] According to the query and acceptance order status of the target employee ID in each of the multiple account periods, an early warning indicator of the target employee ID in each account period is obtained, wherein the early warning indicator includes at least one of the following: the number of query orders, the query acceptance ratio, the number of outbound users in the query user group, the proportion of query contracting groups, the proportion of query orders within the same channel, the average number of full orders per day, and the average number of query orders per day;

[0009] According to the warning indicators of the target employee ID in each account period, the target warning score of the target employee ID is calculated using the preset warning scoring rules;

[0010] According to the target warning score of the target employee ID, it is determined whether the behavior of the target employee ID is abnormal.

[0011] Further, the early warning indicator of the target employee number in each account period obtained according to the query and order acceptance status of the target employee number in each account period in the multiple account periods includes at least one of the following:

[0012] According to the query and acceptance of orders by the target employee ID in each account period, the number of queried orders and the number of accepted orders by the target employee ID in each account period are counted;

[0013] According to the ratio of the number of query orders and the number of accepted orders of the target employee ID in each account period, the query acceptance ratio of the target employee ID in each account period is obtained;

[0014] According to all the query orders in the query order situation of the target employee number in each account period, the query user group of the target employee number in each account period is obtained, and according to the query user group of the target employee number in each account period, the number of outbound users and the proportion of query contracting groups of the target employee number in the query user group in each account period are obtained;

[0015] According to the number of query orders of the target employee ID in each account period, the proportion of query orders of the target employee ID in the same channel in each account period is obtained;

[0016] According to the query and acceptance order status, the number of queried orders and the number of accepted orders of the target employee ID in each account period, the daily average number of full orders and the daily average number of queried orders of the target employee ID in each account period are obtained.

[0017] Further, obtaining the query user group of the target employee ID in each account period according to all query orders in the query order situation of the target employee ID in each account period specifically includes:

[0018] De-duplicate the corresponding query users of the target employee number in all query orders in each account period, obtain all the users who query the target employee number in each account period, and use all the users who query the target employee number in each account period as the query user group of the target employee number in each account period.

[0019] Further, obtaining the number of outbound users of the target employee's work number in the query user group of each bill period according to the target employee's work number in the query user group of each bill period specifically includes:

[0020] For the query user group of the target employee's work number in each account period, the number of outbound users in the query user group of the account period is counted according to the online identification of each user in the query user group of the account period.

[0021] Further, according to the query user group of the target employee ID in each account period, the query contract group ratio of the target employee ID in each account period is obtained, which specifically includes:

[0022] In response to the target employee being a customer service employee with a contracting group relationship, associating the query user group of the target employee's work number in each account period with the contracting user group in the contracting group of the target employee, and counting the number of query users from the contracting group in the query user group of the target employee's work number in each account period;

[0023] According to the ratio of the number of query users from the contracting group to the total number of query users in the query user group of the target employee ID in each account period, the proportion of the query contracting group of the target employee ID in each account period is obtained.

[0024] Further, obtaining the proportion of inquiries of orders by the target employee ID in the same channel in each billing period according to the number of inquiries of the target employee ID in each billing period specifically includes:

[0025] According to the ratio of the number of query orders of the target employee ID in each account period to the number of query orders of all employee IDs in the same channel of the target employee ID in each account period, the query order ratio of the target employee ID in the same channel in each account period is obtained.

[0026] Further, according to the query and acceptance order situation, the number of queried orders and the number of accepted orders of the target employee ID in each account period, the daily average number of full orders and the daily average number of queried orders of the target employee ID in each account period are obtained, which specifically includes:

[0027] According to the query and order acceptance status of the target employee ID in each billing period, the number of dates on which the target employee ID has orders in each billing period is obtained;

[0028] Calculate the daily average total number of orders for the target employee ID in each account period based on the number of queried orders, the number of accepted orders, and the number of dates on which orders exist for the target employee ID in each account period;

[0029] According to the number of query orders of the target employee ID in each billing period and the number of dates on which orders exist, the daily average number of query orders of the target employee ID in each billing period is calculated.

[0030] Furthermore, the warning indicators include: the number of query orders, the query acceptance ratio, the number of outbound users in the query user group, the proportion of query contracting groups, the proportion of query orders within the same channel, the daily average number of full orders and the daily average number of query orders. The target warning score of the target employee ID is calculated based on the warning indicators of the target employee ID in each account period using the preset warning scoring rules, specifically including:

[0031] Calculate the query order score of the target employee ID in each account period according to the number of query orders and the proportion of query contracting groups of the target employee ID in each account period;

[0032] Calculate the query acceptance ratio score of the target employee ID in each account period according to the query acceptance ratio of the target employee ID in each account period;

[0033] Calculate the score of the number of outbound users of the target employee ID in each billing period according to the number of outbound users of the target employee ID in the query user group in each billing period;

[0034] Calculate the score of the order proportion of the target employee ID in the same channel in each account period according to the order proportion of the target employee ID in the same channel in each account period;

[0035] Calculate the warning score of the target employee ID based on the change scenario in two adjacent account periods according to the query order score, query acceptance ratio score, outbound user number score, query order ratio score within the same channel, daily average number of full orders and daily average number of query orders of the target employee ID in each account period;

[0036] Calculate the warning score of the target employee ID in each billing period according to the query order score, query acceptance ratio score, outbound user number score, and query order ratio score within the same channel of the target employee ID in each billing period;

[0037] The target warning score of the target employee number in the current account period is calculated according to the warning scores of the target employee number based on the change scenarios in two adjacent account periods and the warning score within each account period.

[0038] Further, the target warning score of the target employee number in the current account period is calculated according to the warning scores of the target employee number based on the change scenarios in two adjacent account periods and the warning score in each account period, specifically including:

[0039] The target warning score of the target employee ID in this account period is calculated using the following formula:

[0040]

[0041] in, is the target warning score of the target employee ID in the t-account period, The warning score of the target employee's ID number during the account period of account period t. The warning score of the target employee's ID number during the t-1 billing period. The warning score of the target employee's ID number during the t-2 account period. is the warning score based on the change scenario of the target employee's ID number in the t account period and the t-1 account period. It is the early warning score of the target employee's work number in the t-1 account period and the t-2 account period based on the change scenario.

[0042] Further, judging whether the behavior of the target employee ID is abnormal based on the target warning score of the target employee ID specifically includes:

[0043] If the target warning score of the target employee ID is greater than a preset threshold, it is determined that the behavior of the target employee ID is abnormal.

[0044] In a second aspect, the present invention provides an employee work number behavior early warning device, comprising:

[0045] An acquisition module, used to acquire the query and order acceptance status of the target employee's employee number in multiple account periods, wherein the multiple account periods include the current account period and at least one historical account period;

[0046] A obtaining module, connected to the acquisition module, is used to obtain the early warning index of the target employee number in each account period according to the query and acceptance order status of the target employee number in each account period in the multiple account periods, wherein the early warning index includes at least one of the following: the number of query orders, the query acceptance ratio, the number of outbound users in the query user group, the proportion of query contracting groups, the proportion of query orders within the same channel, the average number of full orders per day, and the average number of query orders per day;

[0047] A calculation module, connected to the obtaining module, for calculating the target warning score of the target employee number according to the warning index of the target employee number in each account period using a preset warning scoring rule;

[0048] The early warning module is connected to the calculation module and is used to determine whether the behavior of the target employee ID is abnormal according to the target early warning score of the target employee ID.

[0049] In a third aspect, the present invention provides an employee work number behavior warning device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to implement the employee work number behavior warning method described in the first aspect above.

[0050] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the employee work number behavior warning method described in the first aspect is implemented.

[0051] The employee work number behavior warning method, device and readable storage medium provided by the present invention first obtain the query and order acceptance status of the target employee work number in multiple account periods, wherein the multiple account periods include the current account period and at least one historical account period; then, based on the query and order acceptance status of the target employee work number in each account period in the multiple account periods, the warning index of the target employee work number in each account period is obtained, wherein the warning index includes at least one of the following: the number of query orders, the query acceptance ratio, the number of outbound users in the query user group, the proportion of query contracting groups, the proportion of query orders within the same channel, the average daily number of full orders and the average daily number of query orders; then, based on the warning index of the target employee work number in each account period, the target warning score of the target employee work number is calculated using a preset warning scoring rule; finally, based on the target warning score of the target employee work number, it is determined whether the behavior of the target employee work number is abnormal. The present invention obtains early warning indicators of the target employee's work number in each account period that are related to the specific business scenario and work number attributes by counting the query and acceptance of orders of the target employee's work number in different account periods, and uses the designed early warning scoring rules to finally obtain the target early warning score of the target employee's work number, thereby providing the audit department with accurate and comprehensive early warning results, solving the problem that the existing employee work number behavior early warning method considers a single factor, resulting in inaccurate early warning results. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 This is a flow chart of an employee work number behavior early warning method according to Embodiment 1 of the present invention;

[0053] Figure 2 A flowchart of obtaining a query acceptance ratio according to an embodiment of the present invention;

[0054] Figure 3 A flowchart of obtaining a query user group according to an embodiment of the present invention;

[0055] Figure 4A flowchart of obtaining a query carry-out ratio according to an embodiment of the present invention;

[0056] Figure 5 A flow chart of obtaining the proportion of query contract groups according to an embodiment of the present invention;

[0057] Figure 6 A flow chart of obtaining the ratio of orders queried within the same channel according to an embodiment of the present invention;

[0058] Figure 7 This is a schematic diagram of an image of a scoring function for querying the number of orders according to an embodiment of the present invention;

[0059] Figure 8 A box plot showing the proportion of query contract groups in three account periods according to an embodiment of the present invention;

[0060] Fig. 9 It is a schematic diagram of an image of a parameter function that can reflect the proportion of query contract groups according to an embodiment of the present invention;

[0061] Fig.10 A schematic diagram of a query acceptance ratio score function according to an embodiment of the present invention;

[0062] Fig.11 This is a schematic diagram of a scoring function for the proportion of orders queried within the same channel according to an embodiment of the present invention;

[0063] Fig.12 A box plot of the final warning score distribution of an embodiment of the present invention;

[0064] Fig.13 A histogram of the final warning score distribution of the embodiment of the present invention;

[0065] Fig.14 A violin plot of the final warning score distribution of an embodiment of the present invention;

[0066] Fig.15 This is a structural schematic diagram of an employee work number behavior early warning device according to Embodiment 2 of the present invention;

[0067] Fig.16 This is a structural diagram of an employee work number behavior early warning device according to Example 3 of the present invention. DETAILED DESCRIPTION

[0068] In order to enable those skilled in the art to better understand the technical solution of the present invention, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings.

[0069] It should be understood that the specific embodiments and drawings described herein are only used to explain the present invention rather than to limit the present invention.

[0070] It can be understood that, in the absence of conflict, the various embodiments of the present invention and the various features in the embodiments can be combined with each other.

[0071] It can be understood that, for the convenience of description, the drawings of the present invention only show the parts related to the present invention, while the parts irrelevant to the present invention are not shown in the drawings.

[0072] It can be understood that each unit and module involved in the embodiments of the present invention may correspond to only one physical structure, or may be composed of multiple physical structures, or multiple units and modules may be integrated into one physical structure.

[0073] It can be understood that, without conflict, the functions and steps marked in the flowcharts and block diagrams of the present invention may occur in an order different from that marked in the drawings.

[0074] It is understood that the flowcharts and block diagrams of the present invention illustrate the possible architectures, functions, and operations of the systems, devices, equipment, and methods according to the various embodiments of the present invention. Each box in the flowchart or block diagram may represent a unit, module, program segment, or code, which contains executable instructions for implementing the specified functions. Moreover, each box or combination of boxes in the block diagram and flowchart may be implemented by a hardware-based system that implements the specified functions, or may be implemented by a combination of hardware and computer instructions.

[0075] It can be understood that the units and modules involved in the embodiments of the present invention can be implemented by software or hardware. For example, the units and modules can be located in a processor.

[0076] Application Overview

[0077] A common method for "in-process monitoring and early warning" of operator employee ID behavior is to count the number of orders queried through the system and the number of orders accepted within a certain period of time for all IDs, and find the ID with the larger ratio by calculating the ratio between the two, so as to achieve the early warning effect. However, the current audit and early warning of operator ID behavior has the following shortcomings:

[0078] 1. The existing operator employee number behavior audit and early warning model has a relatively simple "in-process monitoring and evaluation mechanism", which is limited to counting the ratio between the number of inquiries and the number of acceptances for each employee number in a specific time period (such as the current month). If the ratio exceeds the threshold, the corresponding employee number is considered to have a problem. However, this method ignores some other potential factors that affect employee number query behavior.

[0079] 2. The original operator employee number behavior audit and early warning model focuses on the employee number's behavior data for the current month and lacks the extraction and integration of historical month information.

[0080] 3. The existing operator employee number behavior audit warning model often judges whether the employee number is "problematic (i.e. abnormal)" based on a single fixed condition. If the employee number meets one of the corresponding conditions, it is considered problematic. This method lacks a comprehensive and comprehensive "scoring rules" to comprehensively evaluate various factors and reflect them in the final results in the form of quantified scores.

[0081] To sum up, the work numbers of operator employees come from multiple different departments, and the number of inquiries accepted by each department varies greatly. Among them, customer service employees (and corresponding work numbers) need special attention. In addition, the changes in employee work number behavior between a period of time also need to be clearly considered, because if there are possible problems with employee behavior, it must have started at a certain point in time. By comparing the behaviors before and after, it can be determined at what point in time the employee is suspected of having problems. At the same time, the above-mentioned different factors need to use a set of overall scoring calculation rules to obtain the final warning score, so that the risk warning is more scientific in caliber and business logic interpretation. That is, the existing employee work number behavior warning method considers a single factor, resulting in inaccurate warning results.

[0082] In response to the above technical problems, the present application provides an employee work number behavior warning method, device and readable storage medium. By counting the query and acceptance of orders of the target employee work number in different account periods, the warning indicators of the target employee work number and the specific business scenarios and work number attributes in each account period are obtained, and the designed warning scoring rules are used to finally obtain the target warning score of the target employee work number, thereby providing the audit department with accurate and comprehensive warning results, so as to at least solve the problem that the existing employee work number behavior warning method considers a single factor, resulting in inaccurate warning results.

[0083] After introducing the basic principles of the present application, various non-limiting embodiments of the present application will be described in detail with reference to the accompanying drawings.

[0084] Embodiment 1:

[0085] This embodiment provides an employee work number behavior early warning method, such as Figure 1 As shown, the method includes:

[0086] Step S101: obtaining the query and order acceptance status of the target employee's employee number in multiple account periods, wherein the multiple account periods include the current account period and at least one historical account period.

[0087] It should be noted that query orders are orders with the order business type of "Comprehensive Information Query", and acceptance orders are all other orders that are not query orders.

[0088] Step S102: According to the query and acceptance order status of the target employee ID in each of the multiple account periods, obtain the warning indicator of the target employee ID in each account period, wherein the warning indicator includes at least one of the following: the number of query orders, the query acceptance ratio, the number of outbound users in the query user group, the proportion of query contracting groups, the proportion of query orders within the same channel, the average daily number of full orders and the average daily number of query orders.

[0089] In this embodiment, in order to comprehensively analyze the behavior of the target employee's work number, based on the query and order acceptance status of the target employee's work number in different account periods, the early warning indicators of the target employee's work number in each account period are calculated.

[0090] In an optional embodiment, obtaining the early warning indicator of the target employee number in each account period according to the query and order acceptance status of the target employee number in each account period in the multiple account periods includes at least one of the following:

[0091] According to the query and acceptance of orders by the target employee ID in each account period, the number of queried orders and the number of accepted orders by the target employee ID in each account period are counted;

[0092] According to the ratio of the number of query orders and the number of accepted orders of the target employee ID in each account period, the query acceptance ratio of the target employee ID in each account period is obtained;

[0093] According to all the query orders in the query order situation of the target employee number in each account period, the query user group of the target employee number in each account period is obtained, and according to the query user group of the target employee number in each account period, the number of outbound users and the proportion of query contracting groups of the target employee number in the query user group in each account period are obtained;

[0094] According to the number of query orders of the target employee ID in each account period, the proportion of query orders of the target employee ID in the same channel in each account period is obtained;

[0095] According to the query and acceptance order status, the number of queried orders and the number of accepted orders of the target employee ID in each account period, the daily average number of full orders and the daily average number of queried orders of the target employee ID in each account period are obtained.

[0096] Specifically, the number of query orders and accepted orders of the target employee ID in each account period is counted, and the query acceptance ratio of the target employee ID in each account period is calculated. The specific calculation formula is as follows:

[0097] Inquiry acceptance ratio = number of inquiry orders / number of accepted orders

[0098] In an optional embodiment, obtaining the query user group of the target employee ID in each account period according to all query orders in the query order situation of the target employee ID in each account period specifically includes:

[0099] De-duplicate the corresponding query users of the target employee number in all query orders in each account period, obtain all the users who query the target employee number in each account period, and use all the users who query the target employee number in each account period as the query user group of the target employee number in each account period.

[0100] Specifically, for all query orders of the target employee's work number in each account period, the corresponding query users in all query orders of the target employee's work number in each account period are deduplicated, and summarized by users to obtain all users who query the target employee's work number in each account period, which is called the query user group of the work number in each account period. The number of users involved in the query user group of the work number in each account period is counted as the number of query users of the work number in each account period.

[0101] In an optional embodiment, obtaining the number of outbound users of the target employee's work number in the query user group in each bill period according to the target employee's work number in the query user group in each bill period specifically includes:

[0102] For the query user group of the target employee's work number in each account period, the number of outbound users in the query user group of the account period is counted according to the online identification of each user in the query user group of the account period.

[0103] Specifically, for the query user group of the target employee's work number in each account period, the number of users carried out in the query user group in the account period is counted according to the online identification of each user in the query user group in the account period.

[0104] In an optional embodiment, obtaining the query contract group ratio of the target employee ID in each bill period according to the query user group of the target employee ID in each bill period specifically includes:

[0105] In response to the target employee being a customer service employee with a contracting group relationship, associating the query user group of the target employee's work number in each account period with the contracting user group in the contracting group of the target employee, and counting the number of query users from the contracting group in the query user group of the target employee's work number in each account period;

[0106] According to the ratio of the number of query users from the contracting group to the total number of query users in the query user group of the target employee ID in each account period, the proportion of the query contracting group of the target employee ID in each account period is obtained.

[0107] Specifically, if the target employee is an employee of the customer service department and has a contracting group relationship, the query user group of the target employee's work number in each account period is associated with the contracting user group in the contracting group of the target employee, and the query users from the contracting group in the query user group of the target employee's work number in each account period are obtained. The number of query users from the contracting group for the target employee's work number in each account period is counted, and the proportion of the number of query users from the contracting group for the target employee's work number in each account period in the corresponding total number of query users is calculated. The specific calculation formula is as follows:

[0108] Query contracting group ratio = (number of query users from the contracting group) / (total number of query users)

[0109] In an optional embodiment, obtaining the proportion of inquiries of orders by the target employee ID in the same channel in each billing period according to the number of inquiries of orders by the target employee ID in each billing period specifically includes:

[0110] According to the ratio of the number of query orders of the target employee ID in each account period to the number of query orders of all employee IDs in the same channel of the target employee ID in each account period, the query order ratio of the target employee ID in the same channel in each account period is obtained.

[0111] Specifically, the ratio of the number of query orders of the target employee ID in each account period to the number of query orders of all employee IDs in the same channel in each account period is calculated to obtain the query order ratio of the target employee ID in the same channel in each account period.

[0112] In an optional embodiment, obtaining the daily average number of full orders and the daily average number of queried orders of the target employee ID in each account period according to the query and acceptance order status, the number of queried orders, and the number of accepted orders of the target employee ID in each account period specifically includes:

[0113] According to the query and order acceptance status of the target employee ID in each billing period, the number of dates on which the target employee ID has orders in each billing period is obtained;

[0114] Calculate the daily average total number of orders for the target employee ID in each account period based on the number of queried orders, the number of accepted orders, and the number of dates on which orders exist for the target employee ID in each account period;

[0115] According to the number of query orders of the target employee ID in each billing period and the number of dates on which orders exist, the daily average number of query orders of the target employee ID in each billing period is calculated.

[0116] Specifically, the number of dates on which orders exist for the target employee ID in each billing period is obtained, and based on the number of dates on which orders exist for the target employee ID in each billing period, the daily average number of full orders and the daily average number of queried orders for the target employee ID in each billing period are calculated. The specific calculation formula is as follows:

[0117] Average total number of orders per day = (number of queried orders + number of accepted orders) / number of dates with orders

[0118] Average number of queried orders per day = number of queried orders / number of dates with existing orders

[0119] Step S103: Calculate the target warning score of the target employee ID according to the warning index of the target employee ID in each account period using a preset warning scoring rule.

[0120] In this embodiment, in order to make the risk warning more scientific in terms of caliber and business logic interpretation, the warning indicators of the target employee's work number in each account period are comprehensively evaluated by using the preset warning scoring rules to obtain the target warning score of the target employee's work number.

[0121] In an optional embodiment, the early warning indicators include: the number of query orders, the query acceptance ratio, the number of outbound users in the query user group, the proportion of query contracting groups, the proportion of query orders in the same channel, the daily average number of full orders and the daily average number of query orders. The target early warning score of the target employee number is calculated according to the early warning indicators of the target employee number in each account period using the preset early warning scoring rules, which specifically includes:

[0122] Calculate the query order score of the target employee ID in each account period according to the number of query orders and the proportion of query contracting groups of the target employee ID in each account period;

[0123] Calculate the query acceptance ratio score of the target employee ID in each account period according to the query acceptance ratio of the target employee ID in each account period;

[0124] Calculate the score of the number of outbound users of the target employee ID in each billing period according to the number of outbound users of the target employee ID in the query user group in each billing period;

[0125] Calculate the score of the order proportion of the target employee ID in the same channel in each account period according to the order proportion of the target employee ID in the same channel in each account period;

[0126] Calculate the warning score of the target employee ID based on the change scenario in two adjacent account periods according to the query order score, query acceptance ratio score, outbound user number score, query order ratio score within the same channel, daily average number of full orders and daily average number of query orders of the target employee ID in each account period;

[0127] Calculate the warning score of the target employee ID in each billing period according to the query order score, query acceptance ratio score, outbound user number score, and query order ratio score within the same channel of the target employee ID in each billing period;

[0128] The target warning score of the target employee number in the current account period is calculated according to the warning scores of the target employee number based on the change scenarios in two adjacent account periods and the warning score within each account period.

[0129] Specifically, for each account period of the target employee's work number, the calculation formula for the query order score of the target employee's work number in the account period is:

[0130] y order =h quantity· g contract

[0131]

[0132] Among them, y order is the query order score of the target employee’s ID in the payment period, h quantit y is the score of the number of query orders of the target employee ID in the payment period, g contract is the parameter that can reflect the proportion of the contract group in the query of the target employee's work number in the billing period, x quantity is the number of orders queried by the target employee ID during the payment period, x th_quantity is the threshold of the number of query orders for the target employee ID in the billing period, x contract The proportion of the query contract group for the target employee's work number in the billing period;

[0133] Before calculating the query acceptance ratio score of the target employee ID in the account period, considering the actual data distribution, reconstruct the definition of the query acceptance ratio and transform the query acceptance ratio as follows: 1. If the number of query orders and the number of accepted orders are both 0, then it is 0; 2. If the number of query orders is not 0, but the number of accepted orders is 0, take the number of query orders / "manually set number of accepted orders"; 3. If the number of query orders and the number of accepted orders are both not 0, take "number of query orders" / ("number of accepted orders"*"a certain fixed proportional coefficient"); 4. If the calculation results of parts 1-3 above are greater than a certain artificially set threshold, set it to the threshold, otherwise maintain the original status. After the above steps, the transformed query acceptance ratio is obtained, and the calculation formula for the query acceptance ratio score of the target employee ID in the account period is:

[0134]

[0135] Among them, y queryacceptanceis the query acceptance ratio score of the target employee’s ID number during the billing period, x queryacceptance is the query acceptance ratio of the target employee’s work number after the change of the account period, x th_queryacceptance The threshold value of the query acceptance ratio after the target employee's work number is changed in the account period;

[0136] The calculation formula for the number of outbound users of the target employee ID in the account period is:

[0137]

[0138] Among them, y portout is the score of the number of outbound users of the target employee ID during the billing period, x portout is the number of users who port out the target employee ID during the billing period, x th_portout The threshold value for the number of users who port out of the target employee ID during the billing period;

[0139] The calculation formula for the score of the order proportion of the target employee ID in the same channel during the account period is:

[0140]

[0141] Among them, y channel Query the order ratio score of the target employee ID in the same channel during the payment period, x channel Query the order ratio of the target employee ID in the same channel during the payment period, x th_channel The threshold for querying the order ratio for the target employee ID within the same channel during the billing period;

[0142] The calculation formula for the early warning score based on the change scenario of the target employee's work number in two adjacent account periods is:

[0143]

[0144] in, is the warning score based on the change scenario of the target employee's ID number in the t account period and the t-1 account period. The query order score of the target employee's ID in the t account period and the t-1 account period. is the query acceptance ratio score of the target employee's ID in the t account period and the t-1 account period. is the score of the number of outbound users of the target employee ID in the t account period and the t-1 account period, Query the order proportion score for the target employee ID in the same channel in the t account period and the t-1 account period. The daily average total number of orders for the target employee ID in the t period and the t-1 period, is the average daily query order score of the target employee ID in the t account period and the t-1 account period, is the daily average total number of orders for the target employee ID in the t accounting period, is the average daily total number of orders for the target employee in the t-1 account period, is the average number of daily query orders for the target employee ID in the t accounting period, The average number of orders queried daily by the target employee ID in the t-1 account period;

[0145] The calculation formula for the warning score of the target employee ID during the billing period is:

[0146] y period =y order +y queryacceptance +y portout +y channel

[0147] Among them, y period The warning score for the target employee's work number during the billing period;

[0148] The target warning score of the target employee ID in this account period is calculated using the following formula:

[0149]

[0150] in, is the target warning score of the target employee ID in the t-account period, The warning score of the target employee's ID number during the account period of account period t. The warning score of the target employee's ID number during the t-1 billing period. The warning score of the target employee's ID number during the t-2 account period. is the warning score based on the change scenario of the target employee's ID number in the t account period and the t-1 account period. It is the early warning score of the target employee's work number in the t-1 account period and the t-2 account period based on the change scenario.

[0151] Step S104: judging whether the behavior of the target employee ID is abnormal according to the target warning score of the target employee ID.

[0152] Specifically, if the target warning score of the target employee ID is greater than a preset threshold, it is determined that the behavior of the target employee ID is abnormal.

[0153] In a specific embodiment, a method for early warning of employee work number behavior for a "in-process monitoring and evaluation" scenario is provided, and the specific process is as follows: Step 1: Select the employee's work number and dynamically update it according to a certain time period; Step 2: Obtain the query / acceptance order status of these work numbers, and calculate the corresponding statistical indicators (statistics) of these work numbers for the current month; Step 3: Use the calculation method of Step 2 to calculate the indicators of the work number of the current month in the previous month and the previous month, and their derived statistical fields; Step 4: Summarize the derived statistical fields in Steps 2 and 3, and process the corresponding indicators of the two adjacent account periods according to certain rules to obtain the changes between different months, use the corresponding early warning scoring rules, and finally obtain the early warning score for each work number. If the final early warning score exceeds the set threshold, it is considered that there may be problems with the work number behavior.

[0154] The following is an introduction to the specific contents of each part of the process:

[0155] 1. Employee ID selection and dynamic update

[0156] At the beginning of each month, the employee ID table is updated and adjusted, and the employee ID range is selected and the corresponding information is extracted as follows:

[0157] Job number range:

[0158] The employee IDs in use, excluding system IDs, interface IDs (such as signal control interface IDs), and IDs in the whitelist;

[0159] The corresponding information extracted:

[0160] Extract the employee ID, employee name, channel ID, channel name, channel type, and the mobile phone number used for employee registration;

[0161] It should be noted that if the employee ID is from the customer service department, in addition to the above information, the employee's user contracting group relationship must also be considered (that is, some employees from the customer service department will contract a certain number of online business numbers), and the contracting group ID and the mobile phone number of the contracting group manager must be extracted. Then, "mobile phone number" is used as the association condition (mobile phone number of employee ID = mobile phone number of contracting group manager), and the user contracting group ID of the customer service department employee is introduced into the above employee ID information.

[0162] 2. Employee ID query / order acceptance status

[0163] 2.1 Pick up order

[0164] For each employee ID, extract the total number of orders accepted by the employee ID in the current month (divided into unfinished orders and completed orders according to order status), which is called "full order records". The extraction criteria are as follows:

[0165] Unfinished orders:

[0166] Remove the order type automatically generated by the system:

[0167] Perform regular matching based on order notes to remove work orders that are automatically postponed by backend contracts;

[0168] Remove based on order type: reissue of appointment business instructions, information control orders (such as high-value partial suspension, partial suspension due to arrears, suspension due to arrears, account cancellation due to arrears, and speed limit triggered by information control);

[0169] Completed orders:

[0170] Orders whose order status is completed normally;

[0171] Remove the order type automatically generated by the system:

[0172] Perform regular matching based on order notes to remove work orders that are automatically postponed by backend contracts;

[0173] Remove based on order type: reissue of appointment business instructions, information control orders (such as high-value partial suspension, partial suspension due to arrears, suspension due to arrears, account cancellation due to arrears, and speed limit triggered by information control);

[0174] Extract the following fields from the full order record:

[0175] Employee ID, order ID, order business type ID, user ID, business number, acceptance time (in the format of 'yyyy-mm-ddhh:mm:ss').

[0176] 2.2 Job Number Inquiry Form and Acceptance Form

[0177] Based on the full order records, we extract the orders with the order business type of "Comprehensive Information Query", which are called "Query Orders" (i.e. Query Orders) of the work number; all other orders that are not "Query Orders" are called "Acceptance Orders" (i.e. Acceptance Orders) of the work number; Figure 2 As shown, the number of inquiry orders and acceptance orders for each employee number in the month is counted, and the ratio of inquiry orders and acceptance orders of the employee number to the total number of orders of the employee number, as well as the inquiry-acceptance ratio, are calculated:

[0178] Inquiry order ratio = number of inquiry orders / (number of inquiry orders + number of accepted orders)

[0179] Accepted order ratio = number of accepted orders / (number of queried orders + number of accepted orders)

[0180] Inquiry acceptance ratio = number of inquiry orders / number of accepted orders

[0181] It should be noted that if the denominator is 0, the value is 0.

[0182] 2.3 Query user groups of employee numbers and their derived fields

[0183] 2.3.1 Query user groups

[0184] For each user, each time the user is queried in the foreground, a corresponding query record will be generated in the background; accordingly, if a work number queries the same user N times within the time range, there will be N query records for the work number. Therefore, Figure 3 As shown, for each query form of the work number, the users who are queried are deduplicated to obtain a unique identifier, which is then aggregated by user to obtain all users who have queried the work number in the current month (referred to as the "query user group" of the work number). For each query user group of the work number, the number of users involved in the user group is counted as the number of query users for the work number.

[0185] 2.3.2 Query the portability of user groups

[0186] like Figure 4 As shown in the figure, for the query user group of each employee number, the number of users who have been ported out in the month is counted according to the user's online identification, and the proportion of the corresponding number of users to the number of query users under each employee number is calculated:

[0187] Query outbound ratio = Query user group_Number of outbound users / Total number of query users

[0188] It should be noted that if the denominator is 0, the value is 0.

[0189] 2.3.3 Query the contract status of the user group

[0190] like Figure 5 As shown in the figure, for the employees of the customer service department, if they have a contracting group relationship, the contracting relationship id to which each user in the query user group of the employee number belongs is associated with the contracting relationship id of the employee's employee number itself, and the number of user contracting relationships in the query user group of the employee's employee number that have the same id as the query employee's own employee number id (that is, the number of query users from the contracting group) is obtained, and the proportion of the number of query users of the employee is calculated:

[0191] Query contracting group ratio = (number of query users from the contracting group) / (total number of query users)

[0192] 2.4 Comparison of the number of inquiries between work numbers in the same channel in the same month

[0193] like Figure 6 As shown, calculate the proportion of enquiry orders for the employee ID in the same channel in the current month: calculate the ratio between the number of enquiry orders for the employee ID in the current month and the number of enquiry orders for all employee IDs in the same channel in the current month.

[0194] 3. Statistics for different months, and the average number of full orders and average number of inquiries per day for each monthly account period

[0195] Considering that the data of historical months and their changes will also have an effect and impact on the employee number behavior warning task, the calculation method of the corresponding indicators of the current month mentioned above is used to continue to calculate the corresponding indicators of the employee number in the previous month and the month before that; in addition, the daily average number of full orders and the daily average number of query orders of users in each monthly account period are continued to be calculated:

[0196] Average total number of orders per day = (number of queried orders + number of accepted orders) / number of dates with orders

[0197] Average number of queried orders per day = number of queried orders / number of dates with existing orders

[0198] It should be noted that in the process of calculating the "average number of query orders per day", the denominator uses the "number of dates with orders" instead of {"total number of dates experienced in the month", "number of dates with query orders"}. The specific reasons are as follows: although the "total number of dates experienced in the month" can take all dates into account, employees do not use their work numbers every day to query or accept business in the system, and the dates on which each employee uses their work numbers to generate orders are different. They should be averaged according to the employee's own situation (for example, the shifts of different work numbers are staggered). Intuitively speaking, it is more intuitive to use "number of dates with query orders" instead of "number of dates with orders". However, from the perspective of employees' use of work numbers, the probability of generating query orders and generating acceptance orders should be equal. Just because a work number only generates acceptance orders on a certain date, it cannot be assumed that there is "no motivation to generate query orders" on that day.

[0199] It should be noted that the use of the monthly "time range" is, on the one hand, that the monthly time range has a relatively large span, which can better eliminate the behavioral surge that may be caused by a few days.

[0200] 4. Summarize the results in 2 and 3, and use the corresponding warning scoring rules to calculate the final warning score for each employee number (that is, the target warning score of the employee number)

[0201] After obtaining the previous derivative field processing results, use the corresponding warning scoring rules to calculate the final warning score for each employee number. The final warning score includes the following parts:

[0202] 4.1 Query the number of orders:

[0203] If there are too many orders for employee ID query, it may mean that there is a problem. If a linear operation is used to directly subtract the number of employee query orders from the threshold, the magnitude of the final value will be relatively large for the case of "large number of queries". Considering the actual scenario, the span of data magnitude is also relatively large, which is not conducive to analysis. It is hoped that the change of this difference can be "slowed down" when the number of queries is "large" and the change can be "increased" when the number of queries is "small". In this way, the maximum data magnitude and the overall data magnitude span can be reduced, which is conducive to data analysis. The use of logarithmic functions can well reflect this. Therefore, for the part of querying the number of orders, the calculation method of "decibel number" using logarithmic functions is borrowed. The specific formula is as follows:

[0204]

[0205] Among them, h quantity is the score of the number of query orders, x quantity is the number of queried orders, x th_quantity Represents the threshold for querying the number of orders. Here, x th_quantity Set to 400.

[0206] like Figure 7 As shown, Figure 7 Comparing this function with the linear function y2, we can see that the logarithmic function has a higher growth rate when the value is small, and when the value is large, its growth rate is smaller than the linear operation. Subsequently, based on the decibel number, it is used as part of the "Query Order Score" item.

[0207] 4.2 Query the contract status of the user group

[0208] For the work numbers of the customer service department with a contracting relationship, it is hoped that under normal circumstances, its query user group should overlap with its contracting users as much as possible. If the overlap between the query user group and the contracting user group is small, it is considered that there may be a problem with the work number in the query of the contracting user.

[0209] Based on the above ideas, Figure 8 As shown in the figure, the overlap ratio of the query contracting rate of the query user group and the contracting user group of the customer service department's employee number in three account periods (that is, the proportion of query contracting groups) is counted. It can be seen that in most scenarios, the overlap ratio of the query contracting rate is in the range of [0.4, 0.8].

[0210] Considering that the query contract information is processed based on the query information, part of the "query order score" item above quantity On this basis, continue to multiply by the item that can reflect the query contract rate:

[0211] y order =h quantity ·g contract

[0212] This item contract The following characteristics need to be considered: after multiplying this item, the new score is greater than or equal to the original score, that is, the lower limit of the value range corresponding to this item is 1; the domain of the query contracting group ratio is [0, 1], and when it is above a positive threshold of the domain, the value of this item should be relatively small and decrease slowly in the direction of (positive threshold, 1) (the value should slowly approach 1); when it is below the positive threshold, it can quickly grow to a constant value in the direction of (positive threshold, 0).

[0213] In view of these two points, a possible function construction form comes to mind: y = P(x)Q(x)+1, where Q(x) is a function that "tries to characterize the trend" using an exponential / logarithmic function, and is a polynomial function used to adjust the speed of increase or decrease of Q(x). The addition of 1 at the end is for "multiplication". At the same time, in order to make the function graph pass through the points (0, 2) and (1, 1), the corresponding proportional coefficient is adjusted. The final formula is as follows:

[0214] The work number has a contract relationship

[0215] Among them, g contract Represents the parameter that reflects the proportion of query contract groups, x contract ∈[0,1] represents the proportion of query contract groups (x contract Only for the work number with "contracting relationship", if it is not a work number with contracting relationship, g contract =1).

[0216] At the same time, comparisons with other possible functions are plotted, such as Fig. 9 As shown in the diagram, we can see that y1 (linear function) changes evenly, but cannot express well the "when it is higher than a positive threshold in the domain, the value is relatively small and decreases slowly in the direction of (positive threshold, 1)". The remaining three lines are not represented by This item can better reflect the changes. The difference between the three lines lies in the polynomial P(x): y2 can be regarded as P(x)=2, while y3 and y4 use different polynomials of P(x). It can be seen that the decrease of y4 is better than that of y3, but considering the threshold of this task itself, the function corresponding to y3 is finally adopted.

[0217] 4.3 Inquiry Acceptance Ratio

[0218] For each employee ID, if the employee ID has more order inquiries than accepted orders within a certain period of time, that is, the ratio of the number of inquiries to the number of accepted orders is large, it is considered that there may be a problem (it is not desirable to use a dedicated employee ID for inquiries). Considering the actual data distribution, the definition of "inquiry acceptance ratio" is reconstructed and the "inquiry acceptance ratio" is transformed as follows:

[0219] 1. If the number of queried orders and the number of accepted orders are both 0, then the value is 0; 2. If the number of queried orders is not 0, but the number of accepted orders is 0, take the number of queried orders / "artificially set number of accepted orders". Here, after experiments, the "artificially set number of accepted orders" is set to 50; 3. If the number of queried orders and the number of accepted orders are both not 0, take "number of queried orders" / ("number of accepted orders"*"a certain fixed proportional coefficient"). Here, the "fixed proportional coefficient" is set to 5; 4. If the calculation results of parts 1-3 above are greater than a certain artificially set threshold, set it to the threshold, otherwise maintain the original status and set the threshold of this step to 15.

[0220] After the above steps, we get the "transformed query acceptance ratio". Then, we use the following formula to get the "query acceptance ratio" score for each job number:

[0221]

[0222] Among them, y queryacceptance is the query acceptance ratio score, x queryacceptance is the transformed query acceptance ratio, x th_queryacceptance is the threshold of the query acceptance ratio after transformation.

[0223] The image of the corresponding function is Fig.10 As shown, here, the result obtained using the above formula has a clear upper limit value of 5, while x / 10 constructed according to the linear relationship has no upper limit value. The score "based on the query part" needs to have an upper limit value, so a formula with a clear upper limit value is used.

[0224] 4.4 Query the user's carry-out status

[0225] For example, among the user groups querying employee IDs, there are many users who have carried out the service, so it is considered that there may be a problem. Based on the idea of ​​"Querying the number of orders" above, the idea of ​​using "decibel number" in "Querying the number of orders" above is still used to complete the calculation of the corresponding warning score:

[0226]

[0227] where x portout is the number of outbound users in the query user group, x th_portoutis the threshold of the number of users ported out. The 0.95 quantiles of the number of users ported out for each employee ID in the query user groups in the three account periods are (7, 13, 7) respectively. The threshold in the formula is set to 30.

[0228] 4.5 Comparison of query numbers between work numbers within a channel

[0229] For multiple work numbers from the same channel, we do not want a certain work number to be responsible for inquiries in the same channel: if the inquiry forms of a certain work number account for a higher proportion of the inquiry forms in the channel where the work number is located, it is considered that there is a problem with the work number.

[0230] If the employee number is below a certain ratio threshold, it is considered to be ok and the corresponding score is 0. Otherwise, the score is the decibel number between the value and the threshold. The calculation method of this part of the warning score is as follows:

[0231]

[0232] Among them, y channel is the score of the order proportion in the same channel, x channel is the ratio of the number of orders queried by a certain employee number to the number of orders queried by the channel in which it is located (that is, the ratio of orders queried within the same channel), x th_channel is the threshold of the proportion of orders queried within the same channel (taking into account x channel >0,x th_channel ≥0, in the following logarithmic operation, in order to ensure that this piecewise function is channel =x th_channel is continuous, no additional "1+") is added. The function graph is as follows Fig.11 shown.

[0233] It should be noted that the threshold setting principle is as follows: the 0.9 quantiles of the same-channel query ratio of each employee number in the query user group of the three account periods are (0.29, 0.3, 0.28) respectively. Therefore, the threshold x of this part is set to th_channel Set to 0.3.

[0234] 4.6 Proportional coefficients for values ​​from different monthly accounting periods

[0235] After obtaining the corresponding score for the current period, continue to calculate the same indicators and scores for the previous two months. For the previous month's account period, the corresponding scores are multiplied by the "time proportion coefficient" of 0.5. The "time proportion coefficient" of the previous two months' account periods is 0.25.

[0236] 4.7 Changes between different accounting periods

[0237] After obtaining the scores of the employee number in each monthly account period, the corresponding "change" is calculated as the "warning score based on the change scenario" to reflect the "change" (mainly increase) of the employee number behavior. The warning score corresponding to the change during the monthly account period is composed of the following score items:

[0238]

[0239] The introduction of each item is as follows: For the previous warning scores, the difference between the values ​​of two adjacent account periods is directly used to reflect the change relationship. Considering the actual scenario, the "sudden increase" scenario is more concerned. For the various score items mentioned above "within each monthly account period", the changes during each monthly account period are taken as an example. The corresponding formal formula is given by taking the query acceptance ratio of the work number as an example:

[0240]

[0241] in, Represents the query acceptance ratio score corresponding to the employee number in the t account period, Represents the query acceptance ratio score corresponding to the employee number in the t-1 account period, and the same type of values ​​in other different monthly account periods are taken in the same way. Calculate the difference between the calculation results of these monthly account periods. If the difference is <0, it is 0, otherwise it is the corresponding difference.

[0242] It should be noted that before calculating the score between the previous month and the month before last, it is multiplied by the corresponding time proportion coefficient of 0.5.

[0243] For the two indicators of "average total orders per day" and "average query orders per day", we use another calculation method based on the previous method of calculating decibels:

[0244] For the indicators "Daily average total number of orders" and "Daily average number of inquiries", the indicator changes during each monthly account period, and the corresponding formula is as follows (between the t-month account period and the t-1-month account period):

[0245]

[0246] They refer to the daily average total number of orders for the work number in the t period and the t-1 period, respectively. They refer to the average daily number of query orders for the employee number in the t account period and the t-1 account period respectively.

[0247] 4.8 Final warning score formula

[0248]

[0249] The early warning score for each monthly account period is composed of the following scoring items:

[0250] y period =y order +y queryacceptance +y portout +y channel

[0251] 5. Experiment and Verification

[0252] The corresponding employee number behavior data of a certain account period and the two previous account periods were selected to verify the availability and effectiveness of the model itself. According to statistics, a total of 6288 employee numbers participated in the behavior warning analysis. The box plot of the final warning score distribution is as follows: Fig.12 As shown, it can be seen that the final warning scores of 8 work numbers exceed 60 points, and they are "independent of the score range of other work numbers". They are work numbers with "risky usage behavior". After subsequent communication with the employees corresponding to these work numbers, it is learned that these work numbers are used normally and are specifically used for porting query / maintenance within the branch. Therefore, there will be a large number of query users and a high porting ratio. In addition to the scenario of the 8 work numbers mentioned above, the "upper limit value" of the box plot is calculated to be 34.192, which is slightly larger than the 98% quantile of the score distribution (34.068). According to statistics, there are 121 work numbers whose final scores are greater than the "upper limit value" of the box plot, indicating that the statistical results of this method can more accurately reflect the usage of employee work numbers. The histogram of the final warning score distribution is shown as follows: Fig.13 As shown in the figure, the violin plot of the final warning score distribution is as follows Fig.14 shown.

[0253] 6. Work number whitelist mechanism

[0254] After the statistics are completed and the operation is confirmed to be compliant with the corresponding work number user, the work number will be manually added to the whitelist. The work numbers counted in the whitelist will no longer participate in the work number early warning audit within a certain period of time when the whitelist is effective (for example, the work number mentioned above will no longer be counted after the situation is clarified).

[0255] It is worth mentioning that the present invention aims to construct an operator employee number behavior audit and early warning model with a "comprehensive scoring rule", which is used for the "in-process monitoring and evaluation" link of the employee number behavior audit process. Through the employee's employee number and the user's contracting relationship, the complete information of the employee is obtained, and the order query and acceptance status of the employee number in different monthly account periods are counted, and the corresponding derived fields within each period are calculated. By comparing the statistical indicators between different channels within the current period and the statistical indicators of the same employee number between different periods, the statistical indicators between the current channels and the changes in statistical indicators between different periods are obtained respectively. These two types of derivative indicators are combined with the current statistical indicators, the current channel statistical indicators, and the changes in statistical indicators between different periods. According to a certain scoring mechanism, the target early warning score of the employee number is calculated. For the employee IDs participating in the evaluation, the corresponding statistical indicators in different monthly account periods are counted to obtain the derived fields related to the specific business scenarios and ID attributes. For each derived field, a designed scoring mechanism is used to finally obtain the "early warning score" of each ID, which is used for the "in-process monitoring and evaluation" of the IDs. The model of the present invention summarizes the operator ID information of different months for a more comprehensive analysis. The model first selects the employee IDs participating in the early warning according to certain rules, and obtains their order query information in the current month and historical months to generate the corresponding derived fields. For the IDs of the employees in the customer service department, the model compares their query user group with their contracted user group to obtain the contract rate overlap ratio of the customer service query users. Subsequently, based on the consideration of changes in user behavior between different months, by using the designed scoring rules, the model finally quantifies and outputs the early warning score of each ID for the current month, providing the audit department with accurate and comprehensive "in-process monitoring and evaluation" early warning results.

[0256] The employee work number behavior warning method provided by the embodiment of the present invention first obtains the query and acceptance order status of the target employee work number in multiple account periods, wherein the multiple account periods include the current account period and at least one historical account period; then, according to the query and acceptance order status of the target employee work number in each account period in the multiple account periods, the warning index of the target employee work number in each account period is obtained, wherein the warning index includes at least one of the following: the number of query orders, the query acceptance ratio, the number of outbound users in the query user group, the proportion of query contracting groups, the proportion of query orders within the same channel, the average daily number of full orders and the average daily number of query orders; then, according to the warning index of the target employee work number in each account period, the target warning score of the target employee work number is calculated using a preset warning scoring rule; finally, according to the target warning score of the target employee work number, it is judged whether the behavior of the target employee work number is abnormal. The present invention obtains early warning indicators of the target employee's work number in each account period that are related to the specific business scenario and work number attributes by counting the query and acceptance of orders of the target employee's work number in different account periods, and uses the designed early warning scoring rules to finally obtain the target early warning score of the target employee's work number, thereby providing the audit department with accurate and comprehensive early warning results, solving the problem that the existing employee work number behavior early warning method considers a single factor, resulting in inaccurate early warning results.

[0257] Embodiment 2:

[0258] like Fig.15 As shown, this embodiment provides an employee work number behavior early warning device, which is used to execute the above-mentioned employee work number behavior early warning method, including:

[0259] The acquisition module 11 is used to obtain the query and order acceptance status of the target employee's employee number in multiple account periods, wherein the multiple account periods include the current account period and at least one historical account period;

[0260] The obtaining module 12 is connected to the obtaining module 11, and is used to obtain the early warning index of the target employee number in each account period according to the query and acceptance order status of the target employee number in each account period in the multiple account periods, wherein the early warning index includes at least one of the following: the number of query orders, the query acceptance ratio, the number of outbound users in the query user group, the proportion of query contracting groups, the proportion of query orders within the same channel, the average number of full orders per day, and the average number of query orders per day;

[0261] A calculation module 13, connected to the obtaining module 12, is used to calculate the target warning score of the target employee number according to the warning index of the target employee number in each account period using a preset warning scoring rule;

[0262] The early warning module 14 is connected to the calculation module 13 and is used to determine whether the behavior of the target employee ID is abnormal according to the target early warning score of the target employee ID.

[0263] Further, the obtaining module 12 includes at least one of the following:

[0264] A first statistical unit is used to count the number of inquiries and orders accepted by the target employee ID in each billing period according to the inquiries and orders accepted by the target employee ID in each billing period;

[0265] A first obtaining unit is used to obtain the query acceptance ratio of the target employee number in each account period according to the ratio of the number of query orders and the number of accepted orders of the target employee number in each account period;

[0266] The second obtaining unit is used to obtain the query user group of the target employee number in each account period according to all the query orders in the query order situation of the target employee number in each account period, and obtain the number of outbound users and the proportion of query contracting groups of the target employee number in the query user group of each account period according to the query user group of the target employee number in each account period;

[0267] The third obtaining unit is used to obtain the proportion of the inquiries of the target employee number in the same channel in each billing period according to the number of inquiries of the target employee number in each billing period;

[0268] The fourth obtaining unit is used to obtain the daily average total number of orders and the daily average number of queried orders of the target employee number in each billing period according to the query and acceptance order status, the number of queried orders and the number of accepted orders of the target employee number in each billing period.

[0269] Further, the second obtaining unit includes:

[0270] The deduplication unit is used to deduplicate the users corresponding to the query of the target employee number in all query orders in each account period, obtain all users queried by the target employee number in each account period, and use all users queried by the target employee number in each account period as the query user group of the target employee number in each account period.

[0271] Furthermore, the second obtaining unit further includes:

[0272] The second statistical unit is used for counting the number of outbound users in the query user group of each account period according to the online identification of each user in the query user group of the account period for the target employee's work number.

[0273] Furthermore, the second obtaining unit further includes:

[0274] an associating unit, configured to, in response to the target employee being a customer service employee with a contracting group relationship, associate the query user group of the target employee's work number in each account period with the contracting user group in the contracting group of the target employee, and count the number of query users from the contracting group in the query user group of the target employee's work number in each account period;

[0275] The fifth obtaining unit is used to obtain the proportion of the query contracting group of the target employee number in each account period according to the ratio of the number of query users from the contracting group in the query user group of the target employee number in each account period to the total number of query users.

[0276] Further, the third obtaining unit specifically includes:

[0277] The sixth obtaining unit is used to obtain the query order ratio of the target employee number in the same channel in each billing period according to the ratio of the query order number of the target employee number in each billing period to the query order number of all employee numbers in the same channel of the target employee number in each billing period.

[0278] Further, the fourth obtaining unit specifically includes:

[0279] An acquisition unit, configured to acquire the number of dates on which orders for the target employee number exist in each billing period according to the query and order acceptance status of the target employee number in each billing period;

[0280] The first calculation unit is used to calculate the daily average total number of orders of the target employee ID in each billing period according to the number of queried orders, the number of accepted orders, and the number of dates on which orders exist for the target employee ID in each billing period;

[0281] The second calculation unit is used to calculate the daily average number of query orders of the target employee number in each billing period according to the number of query orders of the target employee number in each billing period and the number of dates on which orders exist.

[0282] Further, the early warning indicators include: the number of query orders, the query acceptance ratio, the number of outbound users in the query user group, the proportion of query contracting groups, the proportion of query orders in the same channel, the daily average number of full orders and the daily average number of query orders. The calculation module 13 specifically includes:

[0283] A third calculation unit is used to calculate the query order score of the target employee number in each account period according to the number of query orders and the query contract group ratio of the target employee number in each account period;

[0284] A fourth calculation unit, configured to calculate a query acceptance ratio score of the target employee's work number in each billing period according to the query acceptance ratio of the target employee's work number in each billing period;

[0285] A fifth calculation unit, configured to calculate a score of the number of outbound users of the target employee ID in each billing period according to the number of outbound users of the target employee ID in the query user group in each billing period;

[0286] A sixth calculation unit, configured to calculate a score for the proportion of orders queried by the target employee ID in the same channel in each billing period according to the proportion of orders queried by the target employee ID in the same channel in each billing period;

[0287] A seventh calculation unit is used to calculate the warning score based on the change scenario of the target employee ID in two adjacent account periods according to the query order score, query acceptance ratio score, outbound user number score, query order ratio score within the same channel, daily average full order number and daily average query order number of the target employee ID in each account period;

[0288] An eighth calculation unit is used to calculate the warning score of the target employee ID in each account period according to the query order score, query acceptance ratio score, outbound user number score, and query order ratio score within the same channel of the target employee ID in each account period;

[0289] The ninth calculation unit is used to calculate the target warning score of the target employee number in the current account period according to the warning scores of the target employee number based on the change scenarios in two adjacent account periods and the warning score within each account period.

[0290] Furthermore, the ninth computing unit is specifically configured to:

[0291] The target warning score of the target employee ID in this account period is calculated using the following formula:

[0292]

[0293] in, is the target warning score of the target employee ID in the t-account period, The warning score of the target employee's ID number during the account period of account period t. The warning score of the target employee's ID number during the t-1 billing period. The warning score of the target employee's ID number during the t-2 account period. is the warning score based on the change scenario of the target employee's ID number in the t account period and the t-1 account period. It is the early warning score of the target employee's work number in the t-1 account period and the t-2 account period based on the change scenario.

[0294] Furthermore, the early warning module 14 specifically includes:

[0295] The judgment unit is used to judge that the behavior of the target employee ID is abnormal if the target warning score of the target employee ID is greater than a preset threshold.

[0296] Embodiment 3:

[0297] refer to Fig.16 This embodiment provides an employee work number behavior warning device, including a memory 21 and a processor 22. The memory 21 stores a computer program, and the processor 22 is configured to run the computer program to execute the employee work number behavior warning method in Example 1.

[0298] The memory 21 is connected to the processor 22. The memory 21 may be a flash memory, a read-only memory or other memory. The processor 22 may be a central processing unit or a single-chip microcomputer.

[0299] Embodiment 4:

[0300] This embodiment provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the employee work number behavior early warning method in the above-mentioned embodiment 1 is implemented.

[0301] The computer-readable storage medium includes volatile or non-volatile, removable or non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, computer program modules or other data). Computer-readable storage media include, but are not limited to, RAM (Random Access Memory), ROM (Read-Only Memory), EEPROM (Electrically Erasable Programmable read only memory), flash memory or other memory technology, CD-ROM (Compact Disc Read-Only Memory), digital versatile disk (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by a computer.

[0302] In summary, the employee work number behavior warning method, device and readable storage medium provided by the embodiments of the present invention first obtain the query and order acceptance status of the target employee work number in multiple account periods, wherein the multiple account periods include the current account period and at least one historical account period; then, based on the query and order acceptance status of the target employee work number in each account period in the multiple account periods, the warning indicator of the target employee work number in each account period is obtained, wherein the warning indicator includes at least one of the following: the number of query orders, the query acceptance ratio, the number of outbound users in the query user group, the proportion of query contracting groups, the proportion of query orders within the same channel, the average daily number of full orders and the average daily number of query orders; then, based on the warning indicator of the target employee work number in each account period, the target warning score of the target employee work number is calculated using a preset warning scoring rule; finally, based on the target warning score of the target employee work number, it is determined whether the behavior of the target employee work number is abnormal. The present invention obtains early warning indicators of the target employee's work number in each account period that are related to the specific business scenario and work number attributes by counting the query and acceptance of orders of the target employee's work number in different account periods, and uses the designed early warning scoring rules to finally obtain the target early warning score of the target employee's work number, thereby providing the audit department with accurate and comprehensive early warning results, solving the problem that the existing employee work number behavior early warning method considers a single factor, resulting in inaccurate early warning results.

[0303] It is to be understood that the above embodiments are merely exemplary embodiments used to illustrate the principles of the present invention, but the present invention is not limited thereto. For those of ordinary skill in the art, various modifications and improvements can be made without departing from the spirit and essence of the present invention, and these modifications and improvements are also considered to be within the scope of protection of the present invention.

Claims

1. A method for early warning of employee work number behavior, characterized in that: The method comprises: Obtaining the query and order acceptance status of the target employee's employee number in multiple billing periods, wherein the multiple billing periods include the current billing period and at least one historical billing period; According to the query and acceptance order status of the target employee ID in each of the multiple account periods, an early warning indicator of the target employee ID in each account period is obtained, wherein the early warning indicator includes at least one of the following: the number of query orders, the query acceptance ratio, the number of outbound users in the query user group, the proportion of query contracting groups, the proportion of query orders within the same channel, the average number of full orders per day, and the average number of query orders per day; According to the warning indicators of the target employee ID in each account period, the target warning score of the target employee ID is calculated using the preset warning scoring rules; According to the target warning score of the target employee ID, it is determined whether the behavior of the target employee ID is abnormal.

2. The method according to claim 1, characterized in that The obtaining of the early warning indicator of the target employee number in each account period according to the query and order acceptance status of the target employee number in each account period in the multiple account periods includes at least one of the following: According to the query and acceptance of orders by the target employee ID in each account period, the number of queried orders and the number of accepted orders by the target employee ID in each account period are counted; According to the ratio of the number of query orders and the number of accepted orders of the target employee ID in each account period, the query acceptance ratio of the target employee ID in each account period is obtained; According to all the query orders in the query order situation of the target employee number in each account period, the query user group of the target employee number in each account period is obtained, and according to the query user group of the target employee number in each account period, the number of outbound users and the proportion of query contracting groups of the target employee number in the query user group in each account period are obtained; According to the number of query orders of the target employee ID in each account period, the proportion of query orders of the target employee ID in the same channel in each account period is obtained; According to the query and acceptance order status, the number of queried orders and the number of accepted orders of the target employee ID in each account period, the daily average number of full orders and the daily average number of queried orders of the target employee ID in each account period are obtained.

3. The method according to claim 2, characterized in that The obtaining the query user group of the target employee ID in each account period according to all the query orders of the target employee ID in the query order situation in each account period specifically includes: De-duplicate the corresponding query users of the target employee number in all query orders in each account period, obtain all the users who query the target employee number in each account period, and use all the users who query the target employee number in each account period as the query user group of the target employee number in each account period.

4. The method according to claim 2, characterized in that: The obtaining, according to the query user group of the target employee ID in each account period, the number of outbound users in the query user group of the target employee ID in each account period specifically includes: For the query user group of the target employee's work number in each account period, the number of outbound users in the query user group of the account period is counted according to the online identification of each user in the query user group of the account period.

5. The method according to claim 2, characterized in that: According to the query user group of the target employee ID in each account period, the query contract group ratio of the target employee ID in each account period is obtained, specifically including: In response to the target employee being a customer service employee with a contracting group relationship, associating the query user group of the target employee's work number in each account period with the contracting user group in the contracting group of the target employee, and counting the number of query users from the contracting group in the query user group of the target employee's work number in each account period; According to the ratio of the number of query users from the contracting group to the total number of query users in the query user group of the target employee ID in each account period, the proportion of the query contracting group of the target employee ID in each account period is obtained.

6. The method according to claim 2, characterized in that The step of obtaining the proportion of orders queried by the target employee ID in the same channel in each account period according to the number of queried orders by the target employee ID in each account period specifically includes: According to the ratio of the number of query orders of the target employee ID in each account period to the number of query orders of all employee IDs in the same channel of the target employee ID in each account period, the query order ratio of the target employee ID in the same channel in each account period is obtained.

7. The method according to claim 2, characterized in that The step of obtaining the daily average number of full orders and the daily average number of queried orders of the target employee ID in each account period according to the query and accepted order status, the number of queried orders, and the number of accepted orders of the target employee ID in each account period specifically includes: According to the query and order acceptance status of the target employee ID in each billing period, the number of dates on which the target employee ID has orders in each billing period is obtained; Calculate the daily average total number of orders for the target employee ID in each account period based on the number of queried orders, the number of accepted orders, and the number of dates on which orders exist for the target employee ID in each account period; According to the number of query orders of the target employee ID in each billing period and the number of dates on which orders exist, the daily average number of query orders of the target employee ID in each billing period is calculated.

8. The method according to claim 1, characterized in that The early warning indicators include: the number of query orders, the query acceptance ratio, the number of outbound users in the query user group, the proportion of query contracting groups, the proportion of query orders within the same channel, the daily average number of full orders and the daily average number of query orders. The target early warning score of the target employee ID is calculated based on the early warning indicators of the target employee ID in each account period using the preset early warning scoring rules, which specifically includes: Calculate the query order score of the target employee ID in each account period according to the number of query orders and the proportion of query contracting groups of the target employee ID in each account period; Calculate the query acceptance ratio score of the target employee ID in each account period according to the query acceptance ratio of the target employee ID in each account period; Calculate the score of the number of outbound users of the target employee ID in each billing period according to the number of outbound users of the target employee ID in the query user group in each billing period; Calculate the score of the order proportion of the target employee ID in the same channel in each account period according to the order proportion of the target employee ID in the same channel in each account period; Calculate the warning score of the target employee ID based on the change scenario in two adjacent account periods according to the query order score, query acceptance ratio score, outbound user number score, query order ratio score within the same channel, daily average number of full orders and daily average number of query orders of the target employee ID in each account period; Calculate the warning score of the target employee ID in each billing period according to the query order score, query acceptance ratio score, outbound user number score, and query order ratio score within the same channel of the target employee ID in each billing period; The target warning score of the target employee number in the current account period is calculated according to the warning scores of the target employee number based on the change scenarios in two adjacent account periods and the warning score within each account period.

9. The method according to claim 8, characterized in that Calculating the target warning score of the target employee number in the current account period according to the warning scores of the target employee number based on the change scenarios in two adjacent account periods and the warning score in each account period specifically includes: The target warning score of the target employee ID in this account period is calculated using the following formula: in, is the target warning score of the target employee ID in the t-account period, The warning score of the target employee's ID number during the account period of account period t. The warning score of the target employee's ID number during the t-1 billing period. The warning score of the target employee's ID number during the t-2 account period. is the warning score based on the change scenario of the target employee's ID number in the t account period and the t-1 account period. It is the early warning score of the target employee's work number in the t-1 account period and the t-2 account period based on the change scenario.

10. The method according to claim 1, characterized in that The determining, based on the target warning score of the target employee ID, whether the behavior of the target employee ID is abnormal specifically includes: If the target warning score of the target employee ID is greater than a preset threshold, it is determined that the behavior of the target employee ID is abnormal.

11. An employee work number behavior early warning device, characterized in that: include: An acquisition module, used to acquire the query and order acceptance status of the target employee's employee number in multiple account periods, wherein the multiple account periods include the current account period and at least one historical account period; A obtaining module, connected to the acquisition module, is used to obtain the early warning index of the target employee number in each account period according to the query and acceptance order status of the target employee number in each account period in the multiple account periods, wherein the early warning index includes at least one of the following: the number of query orders, the query acceptance ratio, the number of outbound users in the query user group, the proportion of query contracting groups, the proportion of query orders within the same channel, the average number of full orders per day, and the average number of query orders per day; A calculation module, connected to the obtaining module, for calculating the target warning score of the target employee number according to the warning index of the target employee number in each account period using a preset warning scoring rule; The early warning module is connected to the calculation module and is used to determine whether the behavior of the target employee ID is abnormal according to the target early warning score of the target employee ID.

12. An employee work number behavior early warning device, characterized in that: It comprises a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to implement the employee work number behavior early warning method according to any one of claims 1 to 10.

13. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the employee work number behavior early warning method according to any one of claims 1 to 10 is implemented.

Images