WO2018086428A1 - 一种目标计费规则确定方法、相关设备及系统 - Google Patents

一种目标计费规则确定方法、相关设备及系统 Download PDF

Info

Publication number
WO2018086428A1
WO2018086428A1 PCT/CN2017/104086 CN2017104086W WO2018086428A1 WO 2018086428 A1 WO2018086428 A1 WO 2018086428A1 CN 2017104086 W CN2017104086 W CN 2017104086W WO 2018086428 A1 WO2018086428 A1 WO 2018086428A1
Authority
WO
WIPO (PCT)
Prior art keywords
user
bmp
target
users
service
Prior art date
Application number
PCT/CN2017/104086
Other languages
English (en)
French (fr)
Inventor
谭卫国
汪芳山
Original Assignee
华为技术有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 华为技术有限公司 filed Critical 华为技术有限公司
Publication of WO2018086428A1 publication Critical patent/WO2018086428A1/zh

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04M—TELEPHONIC COMMUNICATION
    • H04M15/00—Arrangements for metering, time-control or time indication ; Metering, charging or billing arrangements for voice wireline or wireless communications, e.g. VoIP
    • H04M15/82—Criteria or parameters used for performing billing operations
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L12/00—Data switching networks
    • H04L12/02—Details
    • H04L12/14—Charging, metering or billing arrangements for data wireline or wireless communications
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L12/00—Data switching networks
    • H04L12/02—Details
    • H04L12/14—Charging, metering or billing arrangements for data wireline or wireless communications
    • H04L12/1403—Architecture for metering, charging or billing
    • H04L12/1407—Policy-and-charging control [PCC] architecture
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04M—TELEPHONIC COMMUNICATION
    • H04M15/00—Arrangements for metering, time-control or time indication ; Metering, charging or billing arrangements for voice wireline or wireless communications, e.g. VoIP
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
    • H04W4/24—Accounting or billing

Definitions

  • the present invention relates to the field of data processing technologies, and in particular, to a method for determining a target charging rule, and related devices and systems.
  • the process of the operator launching the new package is generally: determining the initial weight of each package indicator, obtaining the function of the new package through the initial weight, solving the function, that is, correspondingly obtaining a charging rule, and then calculating corresponding corresponding by the charging rule
  • the package indicator wherein the package indicator may include revenue, matching degree, number of subscription users, etc. If the package indicator does not meet the expected target, the initial weight of each package indicator is adjusted until the package indicator corresponding to the charging rule reaches the expected target. In this way, the operator will launch a new package that includes billing rules that enable the package metrics to achieve the desired goals.
  • the embodiment of the invention discloses a method for determining a target charging rule, a related device and a system, which can solve the problem that the determining efficiency of the charging rule in the new package is low.
  • the embodiment of the present invention discloses a first aspect, which discloses a method for determining a target charging rule, and the method may include:
  • the Business Management Point can determine the multi-objective optimization function group according to the target service and the plurality of to-be-optimized indicators, thereby determining an effective solution of the multi-objective optimization function group, wherein the effective solution includes more target services. Billing rules.
  • the BMP also predicts the potential users of the package containing the target service, and calculates the value of each of the plurality of to-be-optimized indicators of the potential users for each of the plurality of charging rules, thereby satisfying the optimization.
  • the charging rule corresponding to the value of the target to be optimized indicator is determined as the target charging rule.
  • the target service may refer to one of local calls, long distance calls, roaming calls, traffic, short messages, and value-added services, such as ring tones, call reminders, video calls, and missed call reminders.
  • the indicators to be optimized may refer to income, package matching degree, number of package subscribers, and the like.
  • a potential user is a user who might order a new package that includes the target business.
  • the specific manner in which the BMP determines the multi-objective optimization function group according to the target service and the plurality of to-be-optimized indicators may be: the BMP first determines a function of each of the plurality of to-be-optimized indicators for the target service, and thus the determined Multiple functions are used as a multi-objective optimization function group.
  • the efficiency of determining the charging rules of the standard service is determining the charging rules of the standard service.
  • the method may further include:
  • BMP selects the target user and determines the target business from multiple services.
  • the specific way for the BMP to determine the target service from multiple services may be:
  • the BMP For each of the multiple services, the BMP collects the average service usage of the target user for the service within a preset time period, and counts the average service usage of the service for all users in the preset time period, and then For each of the multiple services, the BMP recalculates the ratio of the average service usage of the service to the average service usage of the service for all users, and determines the service corresponding to the maximum ratio as the target service.
  • the preset time period may be one month, three months, or half a year, which is not limited in the embodiment of the present invention.
  • the service corresponding to the maximum ratio indicates that the average usage of the target user is large, which indicates that the service has the greatest impact on the target user to a certain extent, so that the service corresponding to the maximum ratio can be determined as the target service.
  • the specific way for the BMP to select the target user may be:
  • the specific manner in which the BMP selects the target user may also be:
  • clustering algorithms such as k-means or hierarchical clustering
  • the preset rule may be a manually defined rule, for example, "age 18 years old or older, hometown is off-site, consumption in January- May and May-August is 1.5 times or more of other months", “age 18-24 years old” Download the Super Course App,” “The consumption of long-distance calls accounts for more than 60% of the total consumption”, and so on.
  • the target user is a user who has expanded the user selected according to the rules, and the designed package can more accurately reflect the needs of the user group to which the user selected according to the rule belongs.
  • the specific way for the BMP to predict the potential users of the package containing the target service may be:
  • the user data may include a target service, and historical feedback information of the user on the package including the target service.
  • the classification algorithm may include, but is not limited to, a decision tree algorithm, a logistic regression algorithm, and the like.
  • the method may further include:
  • the BMP sends potential users and multiple charging rules to the Convergent Billing Point (CBP).
  • CBP Convergent Billing Point
  • the simulated billing result is obtained, and the BMP can receive the simulated billing result from the CBP and calculate more based on the simulated billing result.
  • the rehearsal of the charging rules is for the potential users, that is, the simulated charging is performed on the potential users by using the obtained charging rules, so that the value of the optimization index calculated by simulating the accounting result can more reflect the new charging rules for most users. The impact of this improves the accuracy of the preview.
  • the second aspect of the embodiment of the present invention discloses a BMP, which may include a determining module, a predicting module, a calculating module, a selecting module, and a communication module, for performing the target charging rule determining method disclosed in the above first aspect.
  • a third aspect of the embodiments of the present invention discloses another BMP, which may include a processor, a communication device, a memory, and a communication bus, where: the processor, the communication device, and the memory are connected by using a communication bus; and the communication device is controlled by the processor.
  • the control is for transmitting and receiving messages; the memory is for storing a set of program codes, and the processor is configured to call the program code stored in the memory to execute the target charging rule determining method disclosed in the first aspect above.
  • the fourth aspect of the embodiment of the present invention discloses another method for determining a target charging rule, and the method may include:
  • the CBP After the CBP receives the plurality of charging rules for the target service from the BMP and the potential users of the package including the target service, the CBP can perform the simulated accounting for the potential users based on each of the plurality of charging rules.
  • the simulated billing result is obtained, so that the obtained simulated billing result is sent to the BMP, so that the BMP calculates the value of at least one of the plurality of to-be-optimized indicators of the potential users based on the simulated billing rules.
  • the CBP performs a simulated account for the potential users based on each of the plurality of charging rules, and the specific manner of obtaining the simulated accounting result may be:
  • the historical CDR data includes historical usage information of the target service, and for each charging rule of the multiple charging rules, calculating each potential based on the historical usage information
  • the user uses the tariff information of the target business to obtain a simulated billing result.
  • the historical bill data may include historical usage information including the target service, historical usage of other services, historical fees of the user, and the like.
  • the historical usage information may refer to the number of used services by the user in the past period of time, such as the usage of traffic within one month, the number of minutes of calls within the city within one month, and the like.
  • CBP simulates the potential users based on each of the plurality of charging rules, so that the value of the optimized indicator calculated by the simulated accounting result can better reflect the impact of the new charging rule on most users, thereby Improve the accuracy of the preview.
  • the fifth aspect of the embodiment of the present invention discloses a CBP, which may include a communication module and a billing module, and is configured to execute the target billing rule determining method disclosed in the fourth aspect.
  • a sixth aspect of the embodiments of the present invention discloses another CBP, which may include a processor, a communication device, a memory, and a communication bus, wherein: the processor, the communication device, and the memory are connected by using a communication bus; and the communication device is controlled by the processor.
  • the control is used to send and receive messages; the memory is used to store a set of program codes, and the processor is configured to call the program code stored in the memory to execute the target charging rule determining method disclosed in the fourth aspect above.
  • a seventh aspect of the embodiments of the present invention discloses a target charging rule determining system, which may include the BMP disclosed in the second aspect and the CBP disclosed in the fifth aspect, for performing the first aspect and the fifth aspect described above.
  • a method for determining a target charging rule may include the BMP disclosed in the second aspect and the CBP disclosed in the fifth aspect, for performing the first aspect and the fifth aspect described above.
  • the BMP may determine the multi-objective optimization function group according to the target service and the plurality of to-be-optimized indicators, and then solve the effective solution of the multi-objective optimization function group in one time, that is, multiple sets of optimal charging rules, and then Make sure that more The to-be-optimized indicator satisfies the charging rule of the optimization target, so that the weight of the optimization indicator does not need to be adjusted repeatedly, so that the determination efficiency of the charging rule for the target service in the new package can be improved.
  • the target service is the service that has the greatest impact on the target user, and the target user is the user who has expanded the user selected according to the rule.
  • the designed package can truly reflect the demand of the user group to which the user selected according to the rule belongs.
  • the preview of the charging rule is performed for the potential user, that is, the potential user is simulated by using the obtained charging rule, so that the value of the optimization index calculated by simulating the accounting result can better reflect the new charging rule pair.
  • FIG. 1 is a schematic structural diagram of a convergent charging system according to an embodiment of the present invention
  • FIG. 2 is a schematic flowchart of a method for determining a target charging rule according to an embodiment of the present invention
  • FIG. 3a is a schematic diagram of a distribution function of total usage of a target service according to an embodiment of the present invention.
  • FIG. 3b is a schematic diagram showing a distribution of effective solutions of a multi-objective optimization function group according to an embodiment of the present invention
  • FIG. 4 is a schematic structural diagram of a BMP according to an embodiment of the present invention.
  • FIG. 5 is a schematic structural diagram of another BMP according to an embodiment of the present disclosure.
  • FIG. 6 is a schematic structural diagram of a CBP according to an embodiment of the present invention.
  • FIG. 7 is a schematic structural diagram of another CBP according to an embodiment of the present disclosure.
  • FIG. 8 is a schematic structural diagram of a target charging rule determining system according to an embodiment of the present invention.
  • the embodiment of the invention discloses a method for determining a target charging rule, a related device and a system. It can improve the efficiency of determining the charging rules in the new package. The details are described below separately.
  • CBS Convergent Billing System
  • the CBS applied in the embodiments of the present invention mainly has the following functions: customer management, product and tariff management, system management, pricing and billing engine, billing, billing, payment, arrears, online adaptation, offline adaptation, Recharge management, electronic recharge, reporting, operation and maintenance (O&M), and interactive voice response (IVR). among them:
  • Customer Management including customer information management, user management, account management, and sales order management.
  • customer information management includes modifying customer basic information, managing customer billing arrangements (Billing Arrangement, BA) information, etc.
  • user management includes user account opening, activation, modification of user basic information, downtime, re-opening, modification of user password, loss reporting, solution Loss of account, etc.
  • account management includes modifying account basic information, managing account credit, querying account payment relationship, etc.
  • sales order management includes ordering, canceling, ordering sales, and switching sales for users.
  • Product and Tariffs Management Provides definitions, testing, release processes, and tariff configurations for products and sales.
  • the product management model is a standard "sales-product-service" three-tier model.
  • System Management Provides system-level data management for system runtime, including rights management, organization and user management, resource management, log management, scheduled task management, network node management, and nationalized settings management.
  • Unified Rating & Charging Engine Provides unified user rating and billing functions, including fee reservation, accumulation, pricing, billing, balance management, rewards, credit control, offline reminders, online reminders , monthly knots, etc.
  • Billing Provides functions such as bill customization, wrong order management, formal billing, real-time billing, test billing, and bill re-granting.
  • Invoicing Provides bill formatting to generate bills for various media for customers, such as paper bills, email bills, and more.
  • Payment Provides functions such as recharge payment, payment back, refund, transfer, transfer, bad account write-off, general ledger, etc.
  • Debt Collection Reduces the operator's revenue loss by checking and monitoring the payment status of the account and performing reminders and reminders on accounts that are not due in arrears according to pre-defined reminders.
  • Online Mediation Provides unified access and control for voice and SMS services, and sends billing requests to the Unified Rating & Charging Engine for billing.
  • Offline Mediation Collect offline CDRs and provide them to the Unified Rating & Charging Engine for offline billing.
  • Voucher Management Provides recharge card recharge and recharge card management.
  • E-Topup Provides electronic recharge.
  • Reports Provides the function of report presentation. It mainly extracts the raw data needed by the business report from the other functional modules of CBS, and extracts and transforms the data.
  • the various business data are displayed in the form of graphs and tables. Business management personnel keep abreast of business operation and maintenance.
  • O&M Provides unified network management functions, including topology management, fault alarms, performance statistics, and performance monitoring.
  • IVR Provides the ability to play voice to the user. Users can learn about current service usage, account balance, and lifecycle status based on the voice played by the system. Users can also recharge, check balances, change passwords, etc. through voice self-service.
  • FIG. 1 is a schematic diagram of a CBS architecture applied according to an embodiment of the present invention.
  • the following network elements are included: CBP, BMP, Service Control Point (SCP), Accounts Receivable (AR) system, and arrears collection (Debt Collection, DC) System, Billing, Invoicing, DCCProxy, Mediation, Synchronization Status Message front end processor (SSMFEP), Network Management System (I2000) ), Uniform Voucher Center (UVC), General Front End Processor (GFEP), Reporting System (Reporting System), Signaling Access Unit (USAU), and Billing Interface Processor (Record) Bill Interface, RBI) and Bill Query. among them,
  • CBP It is the core network element of CBS. It provides functions including billing and pricing, and supports real-time charging and offline charging.
  • CBP is a network element with convergent charging capability, which can handle services under multiple network types and services of multiple dimensions.
  • the charging process of CBP to users includes pre-processing, authentication, rating, accounting, generating bills and credit control. Management account letter Interest, account balance and user life cycle.
  • CBP and BMP work together to handle various services, such as querying service information, subscribing to products, and modifying sub-brands.
  • CBP cooperates with SCP and UVC to process the recharge card recharge business and co-process the cash recharge service with BMP.
  • BMP The functions provided by BMP include defining unique product tariffs (unique product tariffs are defined by BMP and then synchronized to CBP), interfaces between CBP and SCP and external third-party systems, and Web service interfaces to external systems to connect to CBS systems. Communication with CBP, SCP, resource management, log management, and customer management.
  • SCP The functions provided include controlling and handling intelligent calls and business logic management.
  • AR system A subsystem of CBS that provides functions such as payment, transfer, transfer, refund, bulk transaction, and automatic payment.
  • DC system It mainly undertakes the arrears collection function that the post-paid users do not pay before the payment deadline, and reminds the users through various methods (sms, single stop, double stop, manual call, etc.) pay.
  • Billing processing account-related business, providing calculation account-level discounts, discounts, gifts, realizing test-out, formal payment, immediate payment, and can generate bills in xml format for billing.
  • Billing It can send billing information to customers according to the bills ordered by customers. Currently, it supports three types of billing: SMS, Email, and paper bill.
  • DCCProxy Used for network element access to CBP, routing of messages, and sorting of files.
  • Network management system Provides management functions for network elements in the CBS, such as system management, topology management, configuration management, performance management, and fault management.
  • UVC Provides a unified system for recharging and payment services for operators and users.
  • GFEP Implements data exchange services between CBS network elements and external platforms, such as bank interfaces and telephone payment interfaces.
  • Reporting System Provides a set of flexible and convenient reporting application services for report generation, management and display.
  • USAU Provides narrowband and wideband signaling protocols.
  • RBI It is a channel for transferring CDR files between CBS network elements and other systems. It can also be used as a channel for transferring files between any two external entities. During the transmission process, RBI has functions such as file collection, file transfer, file filtering, file merging, and file compression.
  • Billing Query Provides the query function for each bill.
  • each function of the CBS requires several network elements to complete together.
  • the accounting function needs to be completed by multiple network elements such as SCP and CBP.
  • the above plurality of network elements communicate by using an Internet interface, and are connected to each other through a local area network (LAN).
  • LAN local area network
  • CBP and BMP can jointly determine the optimal charging rule in the new package.
  • FIG. 2 is a schematic flowchart diagram of a method for determining a target charging rule according to an embodiment of the present invention. As shown in FIG. 2, the method may include the following steps:
  • the BMP selects a target user and determines a target service from multiple services.
  • the user database stores relevant information of each user of the operator, such as user data, service subscription data, payment data, and business behavior data.
  • user data such as user data, service subscription data, payment data, and business behavior data.
  • service subscription data such as service subscription data
  • payment data such as payment data
  • business behavior data such as business behavior data.
  • the service of the operator may include, but is not limited to, local calls, long distance calls, roaming calls, traffic, short messages, and value-added services, such as ring tones, call reminders, video calls, missed call reminders, and the like. Therefore, the BMP can acquire a plurality of services of the operator, thereby determining the target service required for the new package.
  • the specific manner in which the BMP selects the target user may be:
  • the BMP selects a user set including multiple users from all users based on a preset rule, and acquires at least one feature of each user in the user set, and calculates, for each feature of the at least one feature, a user who has the feature in the user set.
  • the preset rule may be a manually defined rule, for example, “age 18 years old or older, hometown is off-site, consumption in January- May and May-August is 1.5 times or more of other months”, “age 18- At the age of 24, I downloaded the Super Course App, "The consumption of long-distance calls accounted for more than 60% of the total consumption,” and so on. Therefore, the BMP can input related information of each user of the specified operator stored in the user database, and mine all users satisfying the preset rule and related information from the related information of the users according to a preset rule, and obtain more information.
  • the user collection of the user, the user in the user collection can be regarded as a seed user, and then the user collection is continuously expanded by the lookalike method according to the relevant characteristics of the seed user. That is to say, after determining the seed user, the BMP also determines at least one feature of each user from the related information of the seed user, wherein the feature may refer to an age, a place of residence, a frequently active area, a student, Office workers, often staying up late, working overtime, game players, frequent online shopping, large traffic consumption, etc., are not limited in the embodiment of the present invention.
  • the BMP determines a user who has the feature in the user set, and determines a user who has the feature among all users, and then calculates the number of users having the feature in the user set and the feature of all users.
  • Other users of the corresponding feature are added to the user set (that is, all users having the feature corresponding to the minimum information entropy are used as seed users), thereby achieving the purpose of expanding the user set, that is, the seed user.
  • the BMP compares the minimum information entropy with the preset information entropy threshold. If the minimum information entropy is greater than the preset information entropy threshold, it indicates that the user having the feature is insufficient to distinguish the seed user, so that other users with the minimum information entropy corresponding feature are not added to the user set, that is, the seed user ends. If the minimum information entropy is less than or equal to the preset information entropy threshold, then other users having the feature corresponding to the minimum information entropy are added to the user set, and then repeatedly executed, except for the features corresponding to the minimum information in the at least one feature.
  • Each of the other features using the added user set, calculating the number of users having the feature in the user set and the users having the feature among all users, except for the user having the feature in the user set.
  • the information entropy of the number of other users the operation of adding other users with the minimum information entropy corresponding feature to the user set until the minimum information entropy is greater than the preset information entropy threshold, thereby determining the user in the added user set For the target user.
  • BMP first manually defines a rule for college students, such as “age 18-24 years old, download and use super curriculum app”, through this rule to the user database. All users in the filter can get a collection of users of some college students, but this does not represent the entire college students. Among them, the users in the user collection are called intra-group users, and the users in the non-user collection are called extra-group users.
  • entropy -p 1 *logp 1 -p 2 *logp 2 .
  • the feature with the smallest information entropy is taken as a key feature, and the out-of-group user of the key feature is added to the user set, and the above steps are repeated. Until the minimum information entropy is greater than the preset information entropy threshold, each user in the added user set is finally determined as the target user.
  • the specific manner in which the BMP selects the target user may also be:
  • the BMP selects a user set including multiple users from all users based on a preset rule, and then uses a clustering algorithm, such as k-means or hierarchical clustering, to divide all users into clusters, which may be based on each user.
  • a clustering algorithm such as k-means or hierarchical clustering
  • Feature similarity aggregates all users into several clusters, and then for each cluster, counts the proportion of users belonging to the user set in the cluster to all users of the cluster, so that all users in the cluster with the largest proportion, and the user collection The user is identified as the target user.
  • the target user is a user who expands the user selected according to the rule, and the designed package can more accurately reflect the demand of the user group to which the user selected according to the rule belongs.
  • the specific manner in which the BMP determines the target service from multiple services may be:
  • the BMP For each of the multiple services, the BMP counts the average service usage of the target user for the service within a preset time period, and the average service usage of the service for all users in the preset time period, and then for each service.
  • the service calculates the ratio of the average service usage of the target user to the average service usage of the service for all users, and determines the service corresponding to the maximum ratio as the target service.
  • the preset time period may be one month, three months, or half a year, which is not limited by the embodiment of the present invention.
  • BMP can first obtain multiple services of the operator, local calls, long distance calls, roaming calls, traffic, SMS, and value-added services, such as ring tones, call reminders, video calls, missed call reminders, etc., for each service, determine all users.
  • Each user uses the usage of the service within a preset time period, and then calculates the average service usage of the target user and the average service usage of all users, thereby obtaining the ratio of the two average service usages of the service.
  • the service corresponding to the maximum ratio indicates that the average usage of the target user is large, which indicates that the service has the greatest impact on the target user to a certain extent, so that the service corresponding to the maximum ratio can be determined as the target service.
  • BMP counts the average usage of college students per business in one month and the average usage of all users, and calculates the average usage of college students in each business and the average usage of all users.
  • the ratio is obtained as shown in Table 1 below:
  • the ratio of the average usage of college students whose traffic is traffic to the average usage of all users is the largest, which is 1.5, indicating that the business that has the greatest impact on college students is traffic, so that the traffic is determined as the target service, then the new package is
  • the design can be adjusted and optimized for traffic.
  • the BMP may first determine the target service from multiple services, and then determine the target user according to the target service. For example, the operator needs to adjust the local charging rule and introduce a new package, then User data of all users, such as hometown, ID number, age, address, etc., can be obtained, thereby selecting local users and users (such as college students) who are often resident in a short period of time as target users.
  • the BMP determines a multi-objective optimization function group according to the target service and the plurality of to-be-optimized indicators.
  • a new package is generally designed with multiple package indicators to be optimized.
  • the BMP After determining the target user and the target service, the BMP also determines the package indicator to be optimized, which is referred to as the to-be-optimized indicator, and the BMP is based on the target service. And determining a plurality of indicators to be optimized to determine a multi-objective optimization function group.
  • the to-be-optimized indicator may refer to the revenue, the package matching degree, the number of the subscription users, and the like, which are not limited in the embodiment of the present invention.
  • the specific manner in which the BMP determines the multi-objective optimization function group according to the target service and the plurality of to-be-optimized indicators may be:
  • the BMP first determines a function of each of the plurality of to-be-optimized indicators with respect to the target service, thereby using the determined plurality of functions as a multi-objective optimization function group.
  • the BMP needs to adjust the traffic tariff.
  • the determination of the multi-objective optimization function group is exemplified by the income and package matching degree as an example.
  • the adjusted unit price of the traffic tariff ie, the charging rule of the target service
  • the variable x yuan/MB
  • d(x, y) the distribution function of the usage of traffic, which is represented by d(x, y), where x is the unit price, y It is the usage amount, and d(x, y) is the proportion of the user whose usage amount is y in the case where the unit price is x.
  • FIG. 3a which is a schematic diagram of a distribution function of the total usage of the target service disclosed in the embodiment of the present invention, that is, the distribution function of d(y) is as shown in FIG. 3a, and the solution of d(y) can be used as a solution.
  • the target user uses the historical data of the traffic to complete, specifically:
  • Polling a plurality of distribution functions which may be preset typical distribution functions, such as a Gaussian distribution function, a Laplacian distribution function, and the like.
  • the fitting degree of the distribution function with the historical data of the target user is calculated, thereby obtaining a distribution function with the best fitting degree.
  • the fitness can be compared by the actual distribution probability and the theoretical distribution probability of the distribution function to calculate the mean square error.
  • the actual distribution and theoretical distribution of traffic are shown in Table 2 below:
  • the above function represents the revenue of a new package with a unit price of x in traffic.
  • f2(x) is used to indicate the change function of the package matching degree as the price x changes.
  • the matching degree f2(x) m-n*x, where m and n are the same as a and b in the function of income, which is constant and can be set according to business experience.
  • the function of the package matching degree may be more complicated.
  • the charging rule of the traffic includes a traffic packet, the amount of the traffic packet, the unit price beyond the traffic packet, and the unit price may also be the ladder pricing. ,and many more.
  • the number of variables that need to be solved is more; the usage function and the attenuation function of the package matching degree and the flow price are not necessarily linear decay functions, but may also be exponential decay functions, but the solution ideas are solved with the single variables described above. As with the idea, the embodiment of the present invention does not describe this case in detail herein.
  • the BMP can determine a function of each of the plurality of to-be-optimized indicators with respect to the target service, and combine the determined plurality of functions to obtain a plurality of multi-objective optimization function groups of the indicators to be optimized.
  • the BMP determines an effective solution of the multi-objective optimization function group.
  • the BMP after determining the multi-objective optimization function group, the BMP can solve the effective solution of the function group. It can be understood that it is assumed that the functions of the plurality of indicators to be optimized regarding the target service are respectively represented by f1(x), f2(x)...fn(x), where n is the number of indicators to be optimized.
  • the effective solution for solving multi-objective optimization functions can be multi-objective optimization algorithms, such as multi-objective genetic algorithms, multi-objective evolutionary algorithms, and so on.
  • solving the multi-objective optimization function can be described as: determining the feasible domain of x, denoted by S, given a feasible point x * ⁇ S, if any one of x is such that f(x * ) ⁇ f(x), then x * can be called the absolute optimal solution of the multi-objective optimization function. If x ⁇ S is not present such that f(x) > f(x * ), then x * is called the effective solution of the multi-objective optimization function.
  • the multi-objective optimization function group is determined according to the target service and the plurality of to-be-optimized indicators, and then the effective solution of the multi-objective optimization function group is solved at one time, that is, multiple sets of optimal charging rules are determined, and The plurality of to-be-optimized indicators satisfy the charging rule of the optimization target, so that the weight of the optimization indicator does not need to be adjusted repeatedly, which can improve the determining efficiency of the charging rule in the new package.
  • FIG. 3b is a schematic diagram of a distribution of effective solutions of the multi-objective optimization function group disclosed in the embodiment of the present invention.
  • the multi-objective optimization function group includes a function f1 and a function f2.
  • a, b, c, d, e, g, and h are respectively solutions of the multi-objective optimization function group, that is, the functions f1 and f2. Common solution.
  • the BMP can thus determine e, g, and h as the effective solution of the multi-objective optimization function group. Assuming that the effective solution of the multi-objective optimization function group of the income and matching degree is 0.34, 0.28, and 0.19, then the three effective solutions respectively correspond to three charging rules, that is, the unit price of the traffic is 0.34 (yuan/MB, respectively). ), 0.28 (yuan/MB) and 0.19 (yuan/MB).
  • the BMP predicts potential users of the package including the target service.
  • the BMP may also determine a potential user of the package including the target service, where the potential user refers to a user who may subscribe to a new package including the target service.
  • step 204 may be performed at the same time as the steps 202 to 203, which are not limited in the embodiment of the present invention.
  • the specific manner in which the BMP predicts potential users of the package including the target service may be:
  • the BMP obtains the user data of each user of all the users, and then learns the user data through the classification algorithm to obtain a classification model, and calculates the probability of the user subscription package corresponding to the user data based on the user data and the classification model, thereby probabilistically greater than the pre-
  • the user with the probability threshold is determined to be a potential user.
  • the user data may include a target service, and historical feedback information of the user on the package including the target service.
  • the user data includes information related to the target service.
  • the information related to the target service may refer to the unit price (ie, charging rule) of the target service of the currently used package of the user, and the previously used package.
  • the unit price of the target service, etc. and the historical feedback information of the user's package containing the target service.
  • the historical feedback information refers to whether the user has changed the package, or whether the user has subscribed to the recommended package after the user recommends the package in the history record.
  • the user data may further include the duration of the user's network, the frequency of the package replacement, the price of the currently used package, the unit price of other services of the currently used package, the price of the previously used package, the unit price of other services of the previously used package, etc. Wait.
  • Table 3 shows the user data of each user obtained by the BMP:
  • the historical feedback information of the package is 1, indicating that the user has changed the package, and is 0, indicating that the user has not subscribed to the recommended package after the user recommends the package.
  • the BMP can obtain the user data of each user of all the users, and use the classification algorithm to learn the obtained user data to obtain a classification model.
  • the BMP inputs the user data of each user into the classification model, and can predict whether the user is a potential user who subscribes to the package containing the target service according to the classification model, thereby outputting the potential user.
  • the classification model may refer to multiple classification rules, such as whether the user's network duration exceeds 12 months, whether the package has been replaced, and the charging rule of the target service of the currently used package, whether the user subscribes to the recommended package after the user recommends the package.
  • classification rules such as whether the user's network duration exceeds 12 months, whether the package has been replaced, and the charging rule of the target service of the currently used package, whether the user subscribes to the recommended package after the user recommends the package.
  • Classification algorithms may include, but are not limited to, decision tree algorithms, logistic regression algorithms, and the like.
  • the BMP inputs the user data of each user into the classification model, and the specific manner for predicting whether the user is a potential user who subscribes to the package containing the target service according to the classification model may be: BMP according to these
  • the classification rule is used to judge the probability that the user corresponding to the user data will subscribe to the package containing the target service, and if the probability is greater than the preset probability threshold, it is determined to be a potential user.
  • the preset probability threshold may be a preset probability threshold, such as 80% or 90%, which is not limited in the embodiment of the present invention.
  • the classification rule is: after the user recommends the package, the user subscribes to the recommended package, and the current unit price of the package used exceeds 0.35 yuan/MB, and the ratio of the two is 0.4, and 0.6, respectively.
  • the probability threshold is 0.85.
  • the probability that the user subscribes to the new package containing the traffic is 0.6; if there is a recommended package described in the user data of another user, the user subscribes to the recommended package, and the current unit price of the package currently used by the user is 0.3.
  • the target service is used as the traffic, and the preset probability threshold is 0.85.
  • the classification rule is: after the user recommends the package, the user subscribes to the recommended package, and the current unit price of the package used exceeds 0.35 yuan/MB, and In the network for more than 12 months, the proportion of the three is 0.2, 0.5 and 0.3. After the BMP obtains the user data of each user, if the user does not subscribe to the recommended package after the recommended package is recorded in the user data of the user, the current unit price of the package currently used by the user is 0.4 yuan/MB.
  • the classification model obtained by the decision tree algorithm is “the package replacement frequency is greater than or equal to 3, the order probability is 0.8; the package replacement frequency is less than 3, and the current unit price of the package is less than 0.2 yuan/MB, and the order probability is 0.25; the replacement frequency of the package is less than 3, the unit price of the current use package is greater than or equal to 0.2 yuan / MB, the unit price of the previous use package is greater than or equal to 0.4 yuan / MB, the order probability is 0.7; the package replacement frequency is less than 3, the current use of the package The unit price of the traffic is greater than or equal to 0.2 yuan/MB.
  • the unit price of the previously used package is less than 0.4 yuan/MB, and the ordering probability is 0.65”. If the frequency of package replacement for a user is 1, the unit price of the current usage package is 0.3 yuan/MB, and the unit price of the previous usage package is 0.35, then the above classification model shows that the user will order the probability of a new package containing traffic. It is 0.65.
  • the BMP sends multiple charging rules and potential users to the CBP.
  • the BMP calculates the value of the to-be-optimized indicator.
  • a plurality of determined charging rules, and predicted potential users are also sent to the CBP, so that the CBP simulates the posting of the potential users based on each of the plurality of charging rules.
  • the CBP receives multiple charging rules and potential users from the BMP, and performs simulated accounting on the potential users based on each of the plurality of charging rules to obtain a simulated accounting result.
  • the CBP after receiving the plurality of charging rules and potential user information sent by the BMP, the CBP simulates and accounts the potential users based on each of the plurality of charging rules, and obtains a simulation.
  • the result of the account can be as follows:
  • the CBP obtains historical CDR data of the potential user, wherein the historical CDR data includes historical usage information of the target service, and for each charging rule of the plurality of charging rules, the CBP calculates each based on the historical usage information.
  • a potential user uses the tariff information of the target business to obtain a simulated billing result.
  • the CBP also sends the obtained simulation result to the BMP, so that the BMP calculates at least one indicator to be optimized based on the simulated accounting result.
  • the historical bill data may include historical usage information including the target service, historical usage of other services, historical fees of the user, and the like.
  • the historical usage information may refer to the number of used services by the user in the past period of time, such as the usage of traffic within one month, the number of minutes of calls within the city within one month, and the like.
  • the target service is traffic and there are revenues to be optimized in the plurality of indicators to be optimized
  • the CBP needs to obtain potential users for three months.
  • the amount of traffic used within the account thereby calculating the product of each billing rule (ie, the unit price of the new package) and the traffic usage of each potential user, respectively, to obtain the tariff for each potential user.
  • the CBP sends a simulated accounting result to the BMP.
  • the BMP receives the simulated accounting result from the CBP, and calculates, for each of the plurality of charging rules, a value of each of the plurality of to-be-optimized indicators of the potential user, including the simulation based on the simulation.
  • the result of the accounting calculates the value of at least one of the plurality of indicators to be optimized.
  • the BMP calculates the value of each of the plurality of to-be-optimized indicators of the potential users, which may be: for each charging rule, Calculate the value of each target to be optimized for each potential user separately, and then calculate the value of the target to be optimized for all potential users of the same to-be-optimized indicator under the same charging rule, thereby obtaining each to-be-matched corresponding to the charging rule. Optimize the value of the indicator.
  • the BMP can obtain the charging rule of the target service of the current user's current package. If the matching degree of the new package needs to be previewed, the BMP can calculate for each charging rule in the multiple charging rules of the new package. The degree of matching between the two charging rules, thereby obtaining the average matching degree of each charging rule and the potential users among the plurality of charging rules of the new package. If it is necessary to rehearse the revenue of the new package, the BMP can calculate the value of the potential user's income based on the received simulated accounting result for each of the plurality of charging rules of the new package.
  • the specific manner in which the BMP calculates the value of at least one of the plurality of to-be-optimized indicators based on the simulated accounting result may be: the BMP receives the usage target simulated by each potential user for each charging rule.
  • the tariff information of the service where the tariff information can be regarded as a part of the value of the indicator to be optimized. If the BMP needs to obtain the value of the to-be-optimized indicator, it is required to calculate the tariff of all potential users using the target service under the same charging rule. And obtaining the value of the indicator to be optimized corresponding to the charging rule.
  • the potential user performs a package preview on the charging rule of the target service of the new package, that is, the simulated charging is performed on the potential user by using the obtained charging rule, so that the optimization index calculated by the simulation result is simulated.
  • the value better reflects the impact of the new billing rules on most users, thus improving the accuracy of the preview.
  • the steps 205-207 are optional steps, and only if there are multiple indicators to be optimized that need to be simulated to obtain a specific value to be optimized, if not, the BMP will not be executed based on the BMP.
  • the simulation result is an operation of calculating a value of at least one of the plurality of indicators to be optimized.
  • the multiple billing rules determined by the BMP are 0.34, 0.28, and 0.19, respectively.
  • the CBP simulates the billing for each billing rule
  • the simulated billing result is sent to the billing result.
  • BMP BMP calculates the matching degree of each charging rule in multiple charging rules, which may be an average matching degree, and calculates the income of each charging rule based on the simulated accounting result, as shown in Table 4.
  • the BMP determines, according to the value of the to-be-optimized indicator that satisfies the optimization target, a charging rule that is a target charging rule.
  • the BMP after calculating the value of the to-be-optimized indicator corresponding to each charging rule, the BMP further determines whether the value of the to-be-optimized indicator meets the optimization target. Therefore, the BMP needs to obtain each indicator to be optimized in advance. Optimize your goals. For example, your revenue will increase by 5% based on a certain package, and the matching will increase by 8% based on the package.
  • the charging rule corresponding to the value of the to-be-optimized indicator that meets the optimization goal is determined according to the obtained optimization target, and the charging rule is used as the target charging rule to apply the charging rule of the target service in the new package.
  • the revenue corresponding to the specified package is increased by 9%, while the matching degree is only increased by 3%, and the traffic unit price is 0.28.
  • the revenue increased by 6%, the matching degree increased by 5%, and the traffic unit price of 0.19 corresponds to the income of the specified package increased by 4%, and the matching degree increased by 7%. If the revenue is increased by 5% and the matching degree is increased by 5%, then based on the above data, the charging rule with the unit price of 0.28 will be determined as the charging rule for the new package.
  • the BMP can determine the multi-objective optimization function group according to the target service and the plurality of to-be-optimized indicators, and then solve the effective solution of the multi-objective optimization function group in one time, that is, multiple sets of better calculations.
  • the fee rule further determines the charging rule that makes the plurality of indicators to be optimized satisfy the optimization goal, so that the weight of the optimization indicator does not need to be adjusted repeatedly, so that the determining efficiency of the charging rule for the target service in the new package can be improved.
  • the target service is the service that has the greatest impact on the target user, and the target user is the user who has expanded the user selected according to the rule.
  • the designed package can truly reflect the demand of the user group to which the user selected according to the rule belongs.
  • the preview of the charging rule is performed for the potential user, that is, the potential user is simulated by using the obtained charging rule, so that the value of the optimization index calculated by simulating the accounting result can better reflect the new charging rule pair.
  • FIG. 4 is a schematic structural diagram of a BMP according to an embodiment of the present invention.
  • the BMP described in FIG. 4 can be applied to the above method embodiments.
  • the BMP can include:
  • the determining module 401 is configured to determine a multi-objective optimization function group according to the target service and the plurality of to-be-optimized indicators, and determine an effective solution of the multi-objective optimization function group, wherein the effective solution may include multiple charging rules for the target service.
  • the prediction module 402 is configured to predict potential users of the package including the target service.
  • the calculating module 403 is configured to calculate, for each of the plurality of charging rules, a value of each of the plurality of to-be-optimized indicators of the potential users.
  • the determining module 401 is further configured to determine that the charging rule corresponding to the value of the to-be-optimized indicator that satisfies the optimization target is the target charging rule.
  • the target service may refer to one of local calls, long distance calls, roaming calls, traffic, short messages, and value-added services, such as ring tones, call reminders, video calls, and missed call reminders.
  • the metric to be optimised may be the income, the degree of the package, the number of the users of the package, and the like, which are not limited in the embodiment of the present invention.
  • a potential user is a user who might order a new package that includes the target business.
  • the specific manner in which the determining module 401 determines the multi-objective optimization function group according to the target service and the plurality of to-be-optimized indicators may be:
  • the BMP can also include:
  • a module 404 is selected for selecting a target user.
  • the determining module 401 is further configured to determine the target service from the plurality of services.
  • the specific manner in which the determining module 401 determines the target service from the multiple services may be:
  • the average service usage of the target user for the service in a preset time period and the average service usage of the user for the service within a preset time period, and then for each The service calculates the ratio of the average service usage of the service to the average service usage of the service for all users, and determines the service corresponding to the maximum ratio as the target service.
  • the preset time period may be one month, three months, or half a year, which is not limited in the embodiment of the present invention.
  • the service corresponding to the maximum ratio indicates that the average usage of the target user is large, which indicates that the service has the greatest impact on the target user to a certain extent, so that the service corresponding to the maximum ratio can be determined as the target service.
  • the specific manner in which the selection module 404 selects the target user may be:
  • the specific manner in which the selection module 404 selects the target user may also be:
  • clustering algorithms such as k-means or hierarchical clustering
  • the preset rule may be a manually defined rule, for example, "age 18 years old or older, hometown is off-site, consumption in January- May and May-August is 1.5 times or more of other months", “age 18-24 years old” Download the Super Course App,” “The consumption of long-distance calls accounts for more than 60% of the total consumption”, and so on.
  • the specific manner in which the prediction module 402 predicts potential users of the package including the target service may be:
  • the user data may include a target service, and historical feedback information of the user on the package including the target service.
  • the classification algorithm may include, but is not limited to, a decision tree algorithm, a logistic regression algorithm, and the like.
  • the BMP can also include:
  • a communication module 405 configured to send a potential user and a plurality of charging rules to the CBP, and receive a simulated accounting result from the CBP, wherein the simulated accounting result is calculated by the CBP based on each of the plurality of charging rules Simulated billing for potential users.
  • the calculation module 403 is further configured to calculate, according to the simulated billing result, a value of at least one of the plurality of to-be-optimized indicators to be optimized.
  • the embodiment of the present invention discloses another BMP.
  • FIG. 5 is a schematic structural diagram of another BMP according to an embodiment of the present invention.
  • the BMP described in FIG. 5 can be applied to the above method embodiments.
  • the BMP may include at least one processor 501 such as a CPU, a communication device 502, a memory 503, and at least one communication bus 504.
  • the processor 501, the communication device 502, and the memory 503 are connected by a bus 504.
  • the communication device 502 can be a receiver, a transmitter, and controlled by the processor 501 for exchanging messages with an external device.
  • the above memory 503 may be a high speed RAM memory or a non-volatile memory such as a disk memory. Optionally, it may also be at least one storage device located in the processor.
  • the foregoing memory 503 is configured to store a set of program codes, and the processor 501 is configured to call the program code stored in the memory 503 to perform the following operations:
  • the processor 501 is configured to determine a multi-target optimization function group according to the target service and the plurality of to-be-optimized indicators, determine an effective solution of the multi-objective optimization function group, and predict a potential user of the package including the target service.
  • the effective solution may include multiple charging rules for the target service.
  • the processor 501 is further configured to calculate, for each of the plurality of charging rules, a value of each of the plurality of to-be-optimized indicators of the potential users, and determine a value of the to-be-optimized indicator that satisfies the optimization target.
  • the corresponding charging rule is the target charging rule.
  • the communication device 502 may further send multiple charging rules and potential users to the CBP, so that the CBP simulates and accounts the potential users based on each of the multiple charging rules, and obtains a simulation.
  • the result is billed and then the simulated billing result sent by CBP is received.
  • the processor 501 can thereby calculate a value of at least one of the plurality of to-be-optimized indicators to be optimized based on the simulated billing result.
  • the BMP can determine the multi-objective optimization function group according to the target service and the plurality of to-be-optimized indicators, and then solve the effective solution of the multi-objective optimization function group in one time, that is, multiple groups.
  • the optimal charging rule determines the charging rules that make the plurality of indicators to be optimized meet the optimization goal, so that the weight of the optimization indicator does not need to be adjusted repeatedly, so that the determining efficiency of the charging rule for the target service in the new package can be improved.
  • the target service is the service that has the greatest impact on the target user, and the target user is the user who has expanded the user selected according to the rule.
  • the designed package can truly reflect the demand of the user group to which the user selected according to the rule belongs. Further, the preview of the charging rule is performed for the potential user, that is, the potential user is simulated by using the obtained charging rule, so that the value of the optimization index calculated by simulating the accounting result can better reflect the new charging rule pair. The impact of most users, which improves the accuracy of the preview.
  • FIG. 6 is a schematic structural diagram of a CBP according to an embodiment of the present invention.
  • the CBP described in FIG. 6 can be applied to the above method embodiments.
  • the CBP can include:
  • the communication module 601 is configured to receive a plurality of charging rules for the target service from the BMP and potential users of the package including the target service.
  • the billing module 602 is configured to perform simulated billing on the potential users based on each of the plurality of billing rules, and obtain a simulated billing result.
  • the communication module 601 is further configured to send a simulated billing result to the BMP, so that the BMP calculates the potential based on the simulated billing rule.
  • the value of at least one of the plurality of to-be-optimized indicators of the user to be optimized.
  • the billing module 602 simulates the billing of the potential users based on each of the plurality of billing rules, and the specific manner of obtaining the simulated billing result may be:
  • the historical CDR data includes historical usage information of the target service, and for each charging rule of the multiple charging rules, calculating each potential based on the historical usage information
  • the user uses the tariff information of the target business to obtain a simulated billing result.
  • the historical bill data may include historical usage information including the target service, historical usage of other services, historical fees of the user, and the like.
  • the historical usage information may refer to the number of used services by the user in the past period of time, such as the usage of traffic within one month, the number of minutes of calls within the city within one month, and the like.
  • the embodiment of the present invention discloses another CBP.
  • FIG. 7 is a schematic structural diagram of another CBP according to an embodiment of the present invention.
  • the CBP described in FIG. 7 can be applied to the above method embodiments.
  • the BMP may include at least one processor 701 such as a CPU, a communication device 702, a memory 703, and at least one communication bus 704.
  • the processor 701, the communication device 702, and the memory 703 are connected by a bus 704.
  • the communication device 702 may be a receiver, a transmitter, and controlled by the processor 501 for exchanging messages with an external device.
  • the above memory 703 may be a high speed RAM memory or a non-volatile memory such as a disk memory. Optionally, it may also be at least one storage device located in the processor.
  • the above-mentioned memory 703 is used for storing a set of program codes, and the processor 701 and the communication device 702 are used to call the program code stored in the memory 703, and perform the following operations:
  • the communication device 702 is configured to receive a plurality of charging rules for the target service from the BMP and potential users of the package including the target service.
  • the processor 701 is configured to perform simulated accounting on the potential users based on each of the plurality of charging rules, to obtain a simulated accounting result.
  • the communication device 702 is further configured to send a simulated billing result to the BMP, so that the BMP calculates a value of at least one of the plurality of to-be-optimized indicators of the potential user based on the simulated billing rule.
  • CBP simulates the potential users based on each of the plurality of charging rules, so that the value of the optimization index calculated by simulating the accounting result is more capable. Reflects the impact of the new billing rules on most users, which improves the accuracy of the preview.
  • FIG. 8 is a schematic structural diagram of a target charging rule determining system according to an embodiment of the present invention. As shown in Figure 8, the system can include BMP 801 and CBP 802, where:
  • the BMP 801 may determine a multi-objective optimization function group according to the target service and the plurality of to-be-optimized indicators, and determine an effective solution of the multi-objective optimization function group, where the effective solution includes multiple charging rules for the target service.
  • the BMP also predicts the potential users of the package containing the target service, that is, the user who may subscribe to the new package including the target service, and then sends multiple charging rules and predicted potential users to the CBP802.
  • the CBP 802 After receiving a plurality of charging rules and potential users from the BMP 801, the CBP 802 simulates the billing of the potential users based on each of the plurality of charging rules, thereby obtaining a simulated billing result, and then simulating the simulation.
  • the result of the payment is sent to BMP801.
  • the BMP 801 may calculate the value of each of the plurality of to-be-optimized indicators of the potential users for each of the plurality of charging rules. Among them, including based on simulation Calculate the value of at least one indicator to be optimized. Further, after calculating the value of each to-be-optimized indicator corresponding to each charging rule, the BMP 801 determines the charging rule corresponding to the value of the to-be-optimized indicator that satisfies the optimization target as the target charging rule, thereby It acts as a billing rule for the target business in the new package and releases a new package.
  • the BMP 801 can also pre-select the target user and determine the target service from multiple services.
  • the target user is selected, that is, the seed user is selected according to a preset rule, and then the seed user is extended by a lookalike method or a clustering algorithm, so that the user who finally expands is used as the target user of the new package.
  • the determination of the target service is to select the service that has the greatest impact on the target user from among multiple services.
  • the BMP can determine the multi-objective optimization function group according to the target service and the plurality of indicators to be optimized, and then solve the effective solution of the multi-objective optimization function group at one time, that is, multiple sets of better calculations.
  • the fee rule further determines the charging rule that makes the plurality of indicators to be optimized satisfy the optimization goal, so that the weight of the optimization indicator does not need to be adjusted repeatedly, so that the determining efficiency of the charging rule for the target service in the new package can be improved.
  • the target service is the service that has the greatest impact on the target user, and the target user is the user who has expanded the user selected according to the rule.
  • the designed package can truly reflect the demand of the user group to which the user selected according to the rule belongs.
  • the preview of the charging rule is performed for the potential user, that is, the potential user is simulated by using the obtained charging rule, so that the value of the optimization index calculated by simulating the accounting result can better reflect the new charging rule pair.
  • modules in the BMP and the CBP in the embodiments of the present invention may be combined, divided, and deleted according to actual needs.
  • the BMP and the CBP in the embodiment of the present invention may be implemented by a general-purpose integrated circuit, such as a CPU (Central Processing Unit), or an ASIC (Application Specific Integrated Circuit).
  • a CPU Central Processing Unit
  • ASIC Application Specific Integrated Circuit
  • the storage medium may be a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Business, Economics & Management (AREA)
  • Accounting & Taxation (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

本发明实施例公开了一种目标计费规则确定方法、相关设备及系统。其中，该方法包括：BMP可以根据目标业务以及多个待优化指标确定多目标优化函数组，从而确定出多目标优化函数组的有效解，其中，该有效解包括对目标业务的多个计费规则。BMP还会预测包含目标业务的套餐的潜在用户，并对于多个计费规则中的每个计费规则，计算潜在用户的多个待优化指标中每个待优化指标的值，从而将满足优化目标的待优化指标的值所对应的计费规则确定为目标计费规则。通过本发明实施例，BMP可以一次性得到多个计费规则，从而不需要反复调整优化指标的权重，这样可以提高新套餐中关于目标业务的计费规则的确定效率。

Description

一种目标计费规则确定方法、相关设备及系统
本申请要求于2016年11月11日提交中国专利局、申请号为201610998206.7，发明名称为“一种目标计费规则确定方法、相关设备及系统”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本发明涉及数据处理技术领域，具体涉及一种目标计费规则确定方法及、相关设备及系统。
背景技术
由于不同用户对运营商的通信业务的需求所有不同，例如，有的用户本地通话偏多，有的用户漫游通话偏多，而有的用户数据流量偏多，为了吸引客户，运营商通常会针对不同用户群推出不同的套餐，以满足用户的不同需求，降低用户的离网风险。
运营商推出新套餐的流程一般是：确定每个套餐指标的初始权重，通过初始权重得到新套餐的函数，求解得到该函数，即对应得到一个计费规则，然后通过该计费规则计算对应的套餐指标，其中，套餐指标可以包括收入、匹配度、订购用户数等，如果套餐指标未达到预期目标，则会调整每个套餐指标的初始权重，直到计费规则对应的套餐指标达到预期目标为止，这样运营商就会推出包括使得套餐指标达到预期目标的计费规则的新套餐。
在实践中发现，由于每个套餐指标的维度所不同，为了使新套餐的套餐指标达到预期目标，如果采用上述方式确定计费规则，则需要不断调整套餐指标的权重，这样会降低新套餐中计费规则的确定效率。
发明内容
本发明实施例公开了一种目标计费规则确定方法、相关设备及系统，可以解决新套餐中计费规则的确定效率较低的问题。
本发明实施例公开了第一方面公开了一种目标计费规则确定方法，该方法可以包括：
业务管理点(Business Management Point，BMP)可以根据目标业务以及多个待优化指标确定多目标优化函数组，从而确定出多目标优化函数组的有效解，其中，该有效解包括对目标业务的多个计费规则。BMP还会预测包含目标业务的套餐的潜在用户，并对于多个计费规则中的每个计费规则，计算潜在用户的多个待优化指标中每个待优化指标的值，从而将满足优化目标的待优化指标的值所对应的计费规则确定为目标计费规则。
其中，目标业务可以是指市话、长途通话、漫游通话、流量、短信以及增值服务，如彩铃、来电提醒、可视通话、漏接来电提醒等中的一种。待优化指标可以是指收入、套餐匹配度、套餐订购用户数等。潜在用户是可能会订购包含目标业务的新套餐的用户。
具体的，BMP根据目标业务以及多个待优化指标确定多目标优化函数组的具体方式可以为：BMP首先确定多个待优化指标中每个待优化指标关于目标业务的函数，从而将确定出的多个函数作为多目标优化函数组。
将多个待优化指标关于目标业务的问题确定为多目标优化函数组，然后一次性求解出多目标优化函数组的有效解，即多组较优的计费规则，再从中确定出使得多个待优化指标满足优化目标的计费规则，从而不需要反复调整优化指标的权重，这样可以提高新套餐中关于目 标业务的计费规则的确定效率。
可选的，在根据目标业务以及多个待优化指标确定多目标优化函数组之前，该方法还可以包括：
BMP选取目标用户，以及从多个业务中确定目标业务。
其中，BMP从多个业务中确定目标业务的具体方式可以为：
对于多个业务中的每个业务，BMP统计在预设时间段内目标用户对于该业务的平均业务使用量，以及统计在该预设时间段内所有用户对于该业务的平均业务使用量，然后对于多个业务中的每个业务，BMP再计算目标用户对于该业务的平均业务使用量与所有用户对于该业务的平均业务使用量的比值，从而将最大比值所对应的业务确定为目标业务。
其中，预设时间段可以是一个月，三个月，也可以是半年，本发明实施例不做限定。最大比值所对应的业务说明目标用户的平均使用量较大，在一定程度上说明该业务对目标用户的影响最大，从而可以将最大比值所对应的业务确定为目标业务。
可选的，BMP选取目标用户的具体方式可以为：
基于预设规则从所有用户中选取包括多个用户的用户集合，并获取用户集合中每个用户的至少一个特征，对于至少一个特征中的每个特征，计算用户集合中具备该特征的用户数量与所有用户中具备该特征的用户中除该用户集合中具备该特征的用户之外的其他用户数量的信息熵，从而将具备最小信息熵对应特征的其他用户添加至用户集合中，得到添加后的用户集合，然后对于至少一个特征中除最小信息熵对应特征之外的其他特征中的每个特征，重复执行计算用户集合中具备该特征的用户数量与所有用户中具备该特征的用户中除该用户集合中具备该特征的用户之外的其他用户数量的信息熵，以及将具备最小信息熵对应特征的其他用户添加至用户集合的操作，直至最小信息熵大于预设信息熵阈值为止，最后将添加后的用户集合中的用户确定为目标用户。
可选的，BMP选取目标用户的具体方式还可以为：
基于预设规则从所有用户中选取包括多个用户的用户集合，然后使用聚类算法，如k-means或者层次聚类等，将所有用户分为若干个簇，可以是基于每个用户的特征相似度将所有用户聚为若干个簇，然后针对每个簇，统计簇中属于用户集合的用户占该簇的所有用户的比例，从而将比例最大的簇中的所有用户，以及用户集合中的用户确定为目标用户。
其中，预设规则可以是人工定义的规则，例如“年龄18岁以上，籍贯为异地，1-3月和5-8月的消费量是其它月份的1.5倍以上”，“年龄18-24岁，下载使用过超级课程表App”，“长途通话的消费量占总消费量的60％以上”，等等。
目标用户为对按照规则选取的用户进行扩展后的用户，这样设计的套餐更能真实反映出按照规则选取的用户所属的用户群体的需求。
可选的，BMP预测包含目标业务的套餐的潜在用户的具体方式可以为：
获取所有用户中每个用户的用户数据，然后通过分类算法对用户数据进行学习，得到分类模型，并基于用户数据和分类模型计算该用户数据对应的用户订购套餐的概率，从而将概率大于预设概率阈值的用户确定为潜在用户。其中，该用户数据可以包括目标业务，以及用户对包含目标业务的套餐的历史反馈信息。
其中，分类算法可以包括但不限于决策树算法、逻辑回归算法等。
可选的，该方法还可以包括：
BMP向融合计费点(Convergent Billing Point，CBP)发送潜在用户和多个计费规则， 以便于CBP基于多个计费规则中的每个计费规则对潜在用户进行模拟出账，得到模拟出账结果，BMP从而可以接收来自CBP的模拟出账结果，并基于模拟出账结果计算多个待优化指标中的至少一个待优化指标的值。
对计费规则的预演针对潜在用户，即，采用得到的计费规则对潜在用户进行模拟出账，这样通过模拟出账结果计算的优化指标的值更能反映出新计费规则对大部分用户的影响，从而提高了预演的准确性。
相应的，本发明实施例第二方面公开了一种BMP，可以包括确定模块、预测模块、计算模块、选取模块以及通信模块，用于执行上述第一方面公开的目标计费规则确定方法。
相应的，本发明实施例第三方面公开了另一种BMP，可以包括处理器、通信装置、存储器以及通信总线，其中：处理器、通信装置、存储器通过通信总线连接；通信装置受处理器的控制用于收发消息；存储器用于存储一组程序代码，处理器用于调用存储器中存储的程序代码执行上述第一方面公开的目标计费规则确定方法。
相应的，本发明实施例第四方面公开了另一种目标计费规则确定方法，该方法可以包括：
当CBP接收到来自BMP的对目标业务的多个计费规则和包含目标业务的套餐的潜在用户后，CBP可以基于多个计费规则中的每个计费规则对潜在用户进行模拟出账，得到模拟出账结果，从而将得到的模拟出账结果发送给BMP，以便于BMP基于模拟出账规则计算潜在用户的多个待优化指标中的至少一个待优化指标的值。
具体的，CBP基于多个计费规则中的每个计费规则对潜在用户进行模拟出账，得到模拟出账结果的具体方式可以为：
获取潜在用户的历史话单数据，其中，该历史话单数据中包含有目标业务的历史使用量信息，对于多个计费规则中的每个计费规则，基于历史使用量信息计算每个潜在用户使用目标业务的资费信息，从而得到模拟出账结果。
其中，历史话单数据可以包括含有目标业务的历史使用量信息，含有其他业务的历史使用量，用户的历史资费，等等。其中，历史使用量信息可以是指用户过去的一段时间内使用目标业务的数量，如一个月内的流量使用量，一个月内市内通话的分钟数，等等。
CBP基于多个计费规则中的每个计费规则对潜在用户进行模拟出账，这样通过模拟出账结果计算的优化指标的值更能反映出新计费规则对大部分用户的影响，从而提高了预演的准确性。
相应的，本发明实施例第五方面公开了一种CBP，可以包括通信模块和出账模块，用于执行上述第四方面公开的目标计费规则确定方法。
相应的，本发明实施例第六方面公开了另一种CBP，可以包括处理器、通信装置、存储器以及通信总线，其中：处理器、通信装置、存储器通过通信总线连接；通信装置受处理器的控制用于收发消息；存储器用于存储一组程序代码，处理器用于调用存储器中存储的程序代码执行上述第四方面公开的目标计费规则确定方法。
相应的，本发明实施例第七方面公开了一种目标计费规则确定系统，该系统可以包括第二方面公开的BMP以及第五方面公开的CBP，用于执行上述第一方面以及第五方面公开的目标计费规则确定方法。
实施本发明实施例，具有如下有益效果：
本发明实施例中，BMP可以根据目标业务和多个待优化指标确定多目标优化函数组，然后一次性求解出多目标优化函数组的有效解，即多组较优的计费规则，再从中确定出使得多 个待优化指标满足优化目标的计费规则，从而不需要反复调整优化指标的权重，这样可以提高新套餐中关于目标业务的计费规则的确定效率。其中，目标业务为对目标用户影响最大的业务，而目标用户为对按照规则选取的用户进行扩展后的用户，这样设计的套餐更能真实反映出按照规则选取的用户所属的用户群体的需求。进一步的，对计费规则的预演针对潜在用户，即，采用得到的计费规则对潜在用户进行模拟出账，这样通过模拟出账结果计算的优化指标的值更能反映出新计费规则对大部分用户的影响，从而提高了预演的准确性。
附图说明
为了更清楚地说明本发明实施例中的技术方案，下面将对实施例描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
图1为本发明实施例公开的融合计费系统的架构示意图；
图2为本发明实施例公开的一种目标计费规则确定方法的流程示意图；
图3a为本发明实施例公开的目标业务的总使用量的分布函数的示意图；
图3b为本发明实施例公开的多目标优化函数组的有效解的分布示意图；
图4为本发明实施例公开的一种BMP的结构示意图；
图5为本发明实施例公开的另一种BMP的结构示意图；
图6为本发明实施例公开的一种CBP的结构示意图；
图7为本发明实施例公开的另一种CBP的结构示意图；
图8为本发明实施例公开的一种目标计费规则确定系统的结构示意图。
具体实施方式
下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅仅是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。
本发明实施例公开了一种目标计费规则确定方法、相关设备及系统。可以提高新套餐中计费规则的确定效率。以下分别进行详细说明。
为了便于理解本发明实施例公开的技术方案，下面首先对本发明所应用的融合计费系统(Convergent Billing System，CBS)架构进行简要地介绍。
本发明实施例所应用的CBS主要具备以下功能：客户管理、产品和资费管理、系统管理、批价计费引擎、出账、账单、缴费、欠费催款、在线适配、离线适配、充值管理、电子充值、报表、操作维护(Operation and Maintenance，O&M)、交互式语音应答(Interactive Voice Response，IVR)。其中：
客户管理(Customer Management)：包括客户信息管理、用户管理、账户管理、销售品订购管理。其中，客户信息管理包括修改客户基本信息、管理客户计费安排(Billing Arrangement，BA)信息等；用户管理包括用户开户、激活、修改用户基本信息、停机、复机、修改用户密码、挂失、解挂失等；账户管理包括修改账户基本信息、管理账户信用度、查询账户支付关系等；销售品订购管理包括为用户订购、取消订购销售品，为用户切换销售品等。
产品和资费管理(Product and Tariffs Management)：提供产品和销售品的定义、测试、发布流程，以及产品的资费配置。其中，产品管理模型为标准的“销售品-产品-服务”三层模型。
系统管理(System Management)：提供系统运行时系统级数据的管理，包括权限管理、机构与用户管理、资源管理、日志管理、定时任务管理、网络节点管理、国家化设置管理等。
批价计费引擎(Unified Rating & Charging Engine)：提供统一的用户批价计费功能，包括费用预留、累计、批价、计费、余额管理、奖励赠送、信用控制、离线提醒、在线提醒、月结等。
出账(Billing)：提供账单定制、错单管理、正式出账、实时出账、测试出账、话单重批等功能。
账单(Invoicing)：提供账单格式化，为客户生成各种介质的账单，例如纸质账单、Email账单等。
缴费(Payment)：提供充值缴费、缴费回退、退款、转账、调账、呆坏账核销、总分类账等功能。
欠费催缴(Debt Collection)：通过检查和监控账户的缴费状态，并且根据预先定义的催缴规则对到期未缴欠费的账户执行提醒和催缴动作，以减少运营商的收入损失。
在线适配(Online Mediation)：为话音、短信业务提供统一的接入和控制，并向Unified Rating & Charging Engine发送计费请求进行计费。
离线适配(Offline Mediation)：采集离线话单，提供给Unified Rating & Charging Engine进行离线计费。
充值管理(Voucher Management)：提供充值卡充值和充值卡管理。
电子充值(E-Topup)：提供电子充值功能。
报表(Reports)：提供报表展现功能，主要通过从CBS的其他功能模块中提取业务报表需要的原始数据经过数据抽取、转换和加载，将各类业务数据以图、表的形式展现出来，以方便业务管理人员及时了解业务运维情况。
O&M：提供统一的网管功能，包括拓扑管理、故障告警、性能统计、性能监视等。
IVR：提供向用户播放语音的功能。用户可以根据系统播放的语音，了解当前业务使用情况、账户余额、生命周期状态等。用户还可以通过语音自助业务进行充值、查询余额、修改密码等。
请参阅图1，为本发明实施例应用的CBS架构示意图。在图1所描述的CBS架构中，包括以下网元：CBP、BMP、业务控制点(Service Control Point，SCP)、账务应收(Accounts Receivable，AR)系统、欠费催缴(Debt Collection，DC)系统、出账(Billing)、账单(Invoicing)、网元接入(DCCProxy)、适配(Mediation)、同步状态信息前置机(Synchronization Status Message front end processor，SSMFEP)、网管系统(I2000)、充值中心(Uniform Voucher Center，UVC)、通用前置机(General Front End Processor，GFEP)、报表系统(Report)、信令单元(Universal Signaling Access Unit，USAU)、话单接口处理机(Record Bill Interface，RBI)以及账单查询(Bill Query)等。其中，
CBP：为CBS的核心网元，提供功能包括计费和批价，支持实时计费和离线计费两种模式。CBP是一个具有融合计费能力的网元，能处理多种网络类型和多种维度业务下的业务。CBP对用户的计费过程包括预处理、鉴权、批价、入账、生成话单和信用控制。管理账户信 息，账户余额和用户的生命周期。CBP与BMP协同处理各种服务，例如查询服务信息、订阅产品、修改子品牌等业务。CBP和SCP、UVC协同处理充值卡充值业务，并与BMP协同处理现金充值业务。
BMP提供的功能包括定义特有的产品资费(特有的产品资费由BMP定义后同步到CBP)，CBP和SCP与外部第三方系统连接的接口，Web服务接口给外部系统连接到CBS系统，使之能与CBP、SCP通讯，资源管理，日志管理以及客户管理等。
SCP：提供的功能包括控制和处理智能呼叫以及业务逻辑管理等。
AR系统：是CBS的一个子系统，提供了付款、调账、转账、退款、批量交易、自动付款等功能。
DC系统：主要承担后付费用户在出账完成后未在缴费截止日之前缴费的欠费催缴功能，并通过各种方式(短信、单停、双停、人工催缴等)对用户进行催缴。
出账：处理账户相关业务，提供计算账户级的优惠、折扣、赠送，实现测试出账、正式出账、立即出账，并可以生成xml格式的账单提供给账单。
账单：可根据客户订购的账单情况，发送账单信息给客户，目前支持短信、Email、纸质账单三种账单类型。
DCCProxy：用于网元接入CBP，处理消息的路由分发，以及文件的分拣。
适配：用于采集话单。
SSMFEP：负责各种协议的转换。
网管系统：提供对CBS中各网元的管理功能，例如：系统管理、拓扑管理、配置管理、性能管理和故障管理等。
UVC：为运营商和用户提供了统一的充值、缴费服务的系统。
GFEP：实现CBS各网元与外部平台，如银行接口、电话付费接口等之间的数据交换业务。
报表系统：提供了报表生成、管理和展示的一整套灵活方便的报表应用服务。
USAU：提供窄带和宽带信令协议。
RBI：是CBS网元之间及其他系统之间传输话单文件的通道，也可以作为任何两个外部实体之间传输文件的通道。在传输过程中，RBI具有文件采集、文件传送、文件过滤、文件合并、文件压缩等功能。
账单查询：提供各账单的查询功能。
需要说明的是，CBS每个功能的实现需要若干个网元共同完成，例如出账功能需要SCP和CBP等多个网元共同完成。上述若干个网元采用互联网接口通信，通过局域网(Local Area Network，LAN)相互连接。
在图1所示的CBS架构中，如果运营商需要推出新的套餐，CBP和BMP可以共同实现新套餐中最优计费规则的确定。
基于图1所示的CBS架构，本发明实施例公开了一种目标计费规则确定方法。请参阅图2，为本发明实施例公开的一种目标计费规则确定方法的流程示意图。如图2所示，该方法可以包括以下步骤：
201、BMP选取目标用户，以及从多个业务中确定目标业务。
本发明实施例中，用户数据库中存储有运营商各个用户的相关信息，如用户资料、业务订购数据、缴费数据、业务行为数据等。当运营商需要推出新套餐时，可以选取新套餐的目 标用户，以及从多个业务中确定目标业务。
其中，运营商的业务可以包括但不限于市话、长途通话、漫游通话、流量、短信以及增值服务，如彩铃、来电提醒、可视通话、漏接来电提醒等。因此，BMP可以获取运营商的多个业务，从而从中确定出新套餐需要的目标业务。
作为一种可行的实施方式，BMP选取目标用户的具体方式可以为：
BMP基于预设规则从所有用户中选取包括多个用户的用户集合，并获取用户集合中每个用户的至少一个特征，对于至少一个特征中的每个特征，计算用户集合中具备该特征的用户数量与所有用户中具备该特征的用户中除该用户集合中具备该特征的用户之外的其他用户数量的信息熵，从而将具备最小信息熵对应特征的其他用户添加至用户集合中，得到添加后的用户集合，然后对于至少一个特征中除最小信息熵对应特征之外的其他特征中的每个特征，重复执行计算用户集合中具备该特征的用户数量与所有用户中具备该特征的用户中除该用户集合中具备该特征的用户之外的其他用户数量的信息熵，以及将具备最小信息熵对应特征的其他用户添加至用户集合的操作，直至最小信息熵大于预设信息熵阈值为止，最后将添加后的用户集合中的用户确定为目标用户。
具体实现中，预设规则可以是人工定义的规则，例如“年龄18岁以上，籍贯为异地，1-3月和5-8月的消费量是其它月份的1.5倍以上”，“年龄18-24岁，下载使用过超级课程表App”，“长途通话的消费量占总消费量的60％以上”，等等。因此，BMP可以将用户数据库中存储的指定运营商的各个用户的相关信息输入，根据预设规则，从这些用户的相关信息中挖掘出满足预设规则的所有用户以及其相关信息，得到包括多个用户的用户集合，用户集合中的用户可以看作是种子用户，然后根据种子用户的相关特征采用lookalike方法不断扩展用户集合。也即是说，BMP在确定出种子用户后，还会从种子用户的相关信息中确定每个用户的至少一个特征，其中，该特征可以是指年龄、居住地、经常活动的区域、学生、上班族、经常熬夜、加班族、游戏玩家、经常网购、流量消耗大等，本发明实施例不做限定。
进一步的，BMP会针对每个特征，确定用户集合中具备该特征的用户，以及确定所有用户中具备该特征的用户，然后计算用户集合中具备该特征的用户数量与所有用户中具备该特征的用户中除了用户集合中具备该特征的用户之外的其他用户数量的信息熵，每个特征得到对应的一个信息熵，BMP从而从这些信息熵中确定出最小信息熵，再将具备最小信息熵所对应特征的其他用户添加至该用户集合(也即是说，将具备最小信息熵所对应特征的所有用户都作为种子用户)，从而达到对用户集合即种子用户进行扩充的目的。
需要说明的是，BMP在确定出最小信息熵后，会将该最小信息熵与预设信息熵阈值进行比较。如果最小信息熵大于预设信息熵阈值，则说明具备该特征的用户已不足以区分种子用户，从而不会将具备最小信息熵所对应特征的其他用户添加至该用户集合，即结束对种子用户的扩充；如果最小信息熵小于或等于预设信息熵阈值，则将具备最小信息熵所对应特征的其他用户添加至该用户集合，然后重复执行，针对至少一个特征中除了最小信息上所对应特征之外的其他特征中的每个特征，利用添加后的用户集合，计算该用户集合中具备该特征的用户数量与所有用户中具备该特征的用户中除该用户集合中具备该特征的用户之外的其他用户数量的信息熵，将具备最小信息熵对应特征的其他用户添加至该用户集合的操作，直到最小信息熵大于预设信息熵阈值为止，从而将添加后的用户集合中的用户确定为目标用户。
举例来说，假设运营商现在需要针对“大学生”群体制定套餐，BMP首先人工定义一个大学生用户的规则，例如“年龄18-24岁，下载使用过超级课程表App”，通过这个规则对用 户数据库中的所有用户进行过滤，可以得到一部分大学生的用户集合，但这并不能代表整个大学生。其中，用户集合中的用户称为群内用户，非用户集合中的用户称为群外用户。BMP然后获取当前用户集合中每个用户的特征，针对每个特征，计算群内用户数量和与群外用户数量的信息熵，其中，信息熵公式为：entropy＝-p1*logp1-p2*logp2。假设某一特征为“居住在大学城”，BMP针对该特征，统计出群内用户有10000人，群外用户有2000人，从而得到p1＝10000/(10000+2000)＝0.83，p2＝2000/(10000+2000)＝0.17，从而得到该特征的信息熵为-0.83*log0.83-0.17*log0.17＝0.20。通过这种方式计算出每个特征的群内用户数量与群外用户数量的信息熵后，取信息熵最小的特征作为关键特征，将该关键特征的群外用户添加至用户集合，重复上述步骤，直到最小的信息熵大于预设信息熵阈值为止，最终将添加后的用户集合中的每个用户确定为目标用户。
作为另一种可行的实施方式，BMP选取目标用户的具体方式还可以为：
BMP基于预设规则从所有用户中选取包括多个用户的用户集合，然后使用聚类算法，如k-means或者层次聚类等，将所有用户分为若干个簇，可以是基于每个用户的特征相似度将所有用户聚为若干个簇，然后针对每个簇，统计簇中属于用户集合的用户占该簇的所有用户的比例，从而将比例最大的簇中的所有用户，以及用户集合中的用户确定为目标用户。
本发明实施例中，目标用户为对按照规则选取的用户进行扩展后的用户，这样设计的套餐更能真实反映出按照规则选取的用户所属的用户群体的需求。
具体的，BMP从多个业务中确定目标业务的具体方式可以为：
对于多个业务中的每个业务，BMP统计在预设时间段内目标用户对于该业务的平均业务使用量，以及在预设时间段内所有用户对于该业务的平均业务使用量，然后针对每个业务，计算目标用户对于该业务的平均业务使用量与所有用户对于该业务的平均业务使用量的比值，从而将最大比值对应的业务确定为目标业务。
具体实现中，预设时间段可以是一个月，三个月，也可以是半年，本发明实施例不做限定。BMP首先可以获取运营商的多个业务，市话、长途通话、漫游通话、流量、短信以及增值服务，如彩铃、来电提醒、可视通话、漏接来电提醒等，针对每个业务，确定所有用户中每个用户在预设时间段内使用该业务的使用量，然后分别计算目标用户的平均业务使用量以及所有用户的平均业务使用量，从而得到该业务的两平均业务使用量的比值。最大比值所对应的业务说明目标用户的平均使用量较大，在一定程度上说明该业务对目标用户的影响最大，从而可以将最大比值所对应的业务确定为目标业务。
举例来说，假设目标用户为大学生，BMP统计一个月内每个业务的大学生的平均使用量以及所有用户的平均使用量，以及计算每个业务中大学生的平均使用量与所有用户的平均使用量的比值，得到如下表1所示的数据：
表1
业务 大学生的平均使用量 所有用户的平均使用量 比值
市话(分钟) 120 200 0.6
长途通话(分钟) 80 60 1.3
漫游通话(分钟) 15 30 0.5
流量(MB) 450 300 1.5
由上表可知，业务为流量的大学生的平均使用量与所有用户的平均使用量的比值最大，为1.5，表示对大学生影响最大的业务为流量，从而将流量确定为目标业务，那么新套餐的 设计可以主要针对流量进行资费调整和优化。
作为又一种可行的实施方式，BMP还可以先从多个业务中确定出目标业务，然后根据目标业务来确定目标用户，例如，运营商需要针对市话计费规则进行调整，推出新套餐，那么可以获取所有用户的用户资料，如籍贯，身份证号，年龄，住址，等等，从而选出本地的用户以及短时间内常居住于本地的用户(如大学生)作为目标用户。
202、BMP根据目标业务以及多个待优化指标确定多目标优化函数组。
本发明实施例中，设计新的套餐一般有多个套餐指标需要优化，BMP在确定出目标用户以及目标业务之后，还会确定需要优化的套餐指标，以下简称待优化指标，BMP从而根据目标业务以及确定的多个待优化指标来确定多目标优化函数组。其中，待优化指标可以指收入、套餐匹配度、套餐订购用户数，等等，本发明实施例不做限定。
具体的，BMP根据目标业务以及多个待优化指标确定多目标优化函数组的具体方式可以为：
BMP首先确定多个待优化指标中每个待优化指标关于目标业务的函数，从而将确定出的多个函数作为多目标优化函数组。
结合上述两个例子，针对大学生客户群，BMP需要对流量资费进行调整，这里以收入和套餐匹配度为例来对多目标优化函数组的确定进行举例说明。假设调整后的流量资费单价(即目标业务的计费规则)为变量x(元/MB)，这里针对流量相关的收入和套餐匹配度两个目标进行同时优化。
1、对收入来说，收入是价格与总使用量的乘积，为了得到总使用量，BMP需要先得到流量的使用量的分布函数，用d(x，y)表示，其中x是单价，y是使用量，d(x，y)是在单价为x的情况下，使用量为y的用户比例。首先假设d(x，y)＝d(y)*(a-b*x)，其中a和b是常量，表示随着价格的升高，使用量的衰减速度，可以依据经验设定，例如a＝1，b＝0.01。请一并参阅图3a，为本发明实施例公开的目标业务的总使用量的分布函数的示意图，即，d(y)的分布函数如图3a所示，d(y)的求解可以用拟合目标用户使用流量的历史数据的方法来完成，具体为：
(1)轮询多个分布函数，可以是预先设定的典型的分布函数，如高斯分布函数、拉普拉斯分布函数等。
高斯分布
Figure PCTCN2017104086-appb-000001
拉普拉斯分布
Figure PCTCN2017104086-appb-000002
(2)获取目标用户使用流量的历史数据，然后对每个分布函数，用最大似然估计方法拟合历史数据，估计得到分布的参数。例如，对于高斯分布，估计参数μ和σ2的拟合公式为：
Figure PCTCN2017104086-appb-000003
(3)对于使用目标用户的历史数据拟合出的每个分布函数，计算该分布函数与目标用户的历史数据的拟合度，从而得到拟合度最好的分布函数。其中，拟合度可以用实际的分布概率和分布函数的理论分布概率进行比较，计算均方差。例如，流量的实际分布和理论分布如下表2：
表2
流量区间(M) 实际的分布概率 分布函数的分布概率
0-50 0.1 0.1
50-100 0.2 0.25
100-300 0.25 0.3
300-500 0.3 0.25
500以上 0.15 0.1
则拟合度为：
Figure PCTCN2017104086-appb-000004
通过上述方式确定出d(y)的分布函数后，从而得到d(x，y)的函数，这样BMP就可以确定出收入(用f1(x)表示)的函数为：
f1(x)＝∫d(x,y)*x*ydy
上述函数表示在流量的单价为x的新套餐的收入。
2、对套餐匹配度来说，用f2(x)表示随着价格x的变化，套餐匹配度的变化函数。和收入的函数f1(x)一样，可以假设匹配度f2(x)＝m-n*x，其中，m和n和收入的函数中a和b一样，为常量，可以根据业务经验设定。在实际的计费规则中，套餐匹配度的函数可能更为复杂，例如，流量的计费规则包括一个流量包、流量包包含的量、超出流量包之外的单价，单价还可能是阶梯定价，等等。也就是需要求解的变量数目更多，；使用量和套餐匹配度与流量价格的衰减函数，也不一定是线性衰减函数，也可能是指数衰减函数，但是求解思想与上面描述的单个变量的求解思想一样，本发明实施例在此不对这种情况进行详细描述。
BMP通过上述方式从而可以确定出多个待优化指标中每个待优化指标关于目标业务的函数，将确定出的多个函数组合，从而得到多个待优化指标的多目标优化函数组。
203、BMP确定多目标优化函数组的有效解。
本发明实施例中，BMP在确定出多目标优化函数组后，可以求解该函数组的有效解。可以理解的是，假设多个待优化指标的关于目标业务的函数分别用f1(x)、f2(x)…fn(x)表示，其中，n为待优化指标的数量。求解多目标优化函数的有效解可以用多目标优化算法，如多目标遗传算法、多目标进化算法，等等。
其中，求解多目标优化函数可以描述为：确定x的可行域，用S表示，给定一个可行点x*∈S，如果存在任意一个x都使得f(x*)≥f(x)，那么x*就可以称为多目标优化函数的绝对最优解，如果不存在x∈S，使得f(x)＞f(x*)，那么x*就称为多目标优化函数的有效解。
本发明实施例中，根据目标业务和多个待优化指标确定多目标优化函数组，然后一次性求解出多目标优化函数组的有效解，即多组较优的计费规则，从中确定出使得多个待优化指标满足优化目标的计费规则，从而不需要反复调整优化指标的权重，这样可以提高新套餐中计费规则的确定效率。
举例来说，请一并参阅图3b，为本发明实施例公开的多目标优化函数组的有效解的分布示意图。如图3b所示，多目标优化函数组包括函数f1以及函数f2，图中a、b、c、d、e、g、h分别为多目标优化函数组的解，即函数f1与函数f2的共同解。从图中可以看出，在各 个解之间，显然d比a好，e比d好，g比b好，h比c好，等等，但是对于e、g以及h来说，无法确定解的优劣，但又不存在比其更好的解，BMP从而可以将e、g以及h确定为该多目标优化函数组的有效解。假设得到的收入和匹配度的多目标优化函数组关于流量的有效解为0.34、0.28以及0.19，那么这三个有效解分别对应三个计费规则，即流量的单价分别为0.34(元/MB)、0.28(元/MB)以及0.19(元/MB)。
204、BMP预测包含目标业务的套餐的潜在用户。
本发明实施例中，BMP在确定出目标业务后，还可以确定包含目标业务的套餐的潜在用户，其中，潜在用户是指可能会订购包含目标业务的新套餐的用户。
需要说明的是，步骤204可以与步骤202～203同时执行，本发明实施例不做限定。
具体的，BMP预测包含目标业务的套餐的潜在用户的具体方式可以为：
BMP获取所有用户中每个用户的用户数据，然后通过分类算法对用户数据进行学习，得到分类模型，并基于用户数据和分类模型计算该用户数据对应的用户订购套餐的概率，从而将概率大于预设概率阈值的用户确定为潜在用户。其中，该用户数据可以包括目标业务，以及用户对包含目标业务的套餐的历史反馈信息。
具体实现中，用户数据包括与目标业务相关的信息，假设目标业务为流量，那么与目标业务相关的信息可以是指用户当前所用套餐的目标业务的单价(即计费规则)，以前所用套餐的目标业务的单价等；以及用户对包含目标业务的套餐的历史反馈信息，历史反馈信息是指用户是否更换过套餐，或者历史记录中为用户推荐套餐后用户是否订购了推荐套餐等。可选的，用户数据还可以包括用户的在网时长、套餐更换频率、当前所用套餐的价格、当前所用套餐的其他业务的单价，以前所用套餐的价格，以前所用套餐的其他业务的单价，等等。如表3所示，表3为BMP获取的每个用户的用户数据：
表3
Figure PCTCN2017104086-appb-000005
需要说明的是，对套餐的历史反馈信息为1，表示该用户更换过套餐，为0，表示为该用户推荐套餐后该用户未订购推荐套餐。
因此，BMP在确定出目标业务后，可以获取所有用户中每个用户的用户数据，采用分类算法对获得的这些用户数据进行学习，得到分类模型。BMP将每个用户的用户数据输入到该分类模型，就可以依据该分类模型预测出该用户是否为订购包含目标业务的套餐的潜在用户，从而输出潜在用户。
其中，分类模型可以是指多个分类规则，如用户在网时长是否超过12个月，是否更换过套餐，当前所用套餐的目标业务的计费规则，为用户推荐套餐后用户是否订购了推荐套餐，等等，本发明实施例不做限定。分类算法可以包括但不限于决策树算法、逻辑回归算法等。
可以理解的是，BMP将每个用户的用户数据输入到该分类模型，就可以依据该分类模型预测出该用户是否为订购包含目标业务的套餐的潜在用户的具体方式可以为：BMP根据这些 分类规则来评判用户数据所对应的用户会订购包含目标业务的套餐的概率，如果概率大于预设概率阈值，则会确定其为潜在用户。其中，预设概率阈值可以是预先设定的概率阈值，比如为80％或者90％，本发明实施例不做限定。
举例来说，假设目标业务为流量，分类规则为：为用户推荐套餐后用户订购了推荐套餐，当前所用套餐的流量单价超过0.35元/MB，两者占比分别为0.4，和0.6，预设概率阈值为0.85。那么BMP在获取到每个用户的用户数据后，如果存在某一用户的用户数据中记载的为其推荐套餐后该用户未订购推荐套餐，该用户当前所用套餐的流量单价为0.4元/MB，可以计算出该用户订购包含流量的新套餐的概率为0.6；如果存在另一用户的用户数据中记载的为其推荐套餐后该用户订购了推荐套餐，且该用户当前所用套餐的流量单价为0.3元/MB，可以计算出该用户订购包含流量的新套餐的概率为0.4+0.3/0.35*0.6＝0.91，从而可以得出后者可以确定为订购包含流量的新套餐的潜在用户。
又举例来说，还是以目标业务为流量，预设概率阈值为0.85为例，假设分类规则为：为用户推荐套餐后用户订购了推荐套餐，当前所用套餐的流量单价超过0.35元/MB，且在网时长超过12个月，三者的占比分别为0.2、0.5以及0.3。那么BMP在获取到每个用户的用户数据后，如果存在某一用户的用户数据中记载的为其推荐套餐后该用户未订购推荐套餐，该用户当前所用套餐的流量单价为0.4元/MB，且该用户的在网时长为10个月，可以计算出该用户订购包含流量的新套餐的概率为0.5+10/12*0.3＝0.75，从而可知，经过上述预测方式，该用户不为订购包含流量的新套餐的潜在用户。
又举例来说，假设用决策树算法得到的分类模型是“套餐更换频率大于或等于3，订购概率为0.8；套餐更换频率小于3，当前使用套餐的流量单价小于0.2元/MB，订购概率为0.25；套餐更换频率小于3，当前使用套餐的单价大于或等于0.2元/MB，以前使用套餐的流量单价大于或等于0.4元/MB，订购概率为0.7；套餐更换频率小于3，当前使用套餐的流量单价大于或等于0.2元/MB，以前使用套餐的流量单价小于0.4元/MB，订购概率为0.65”。如果某一用户的套餐更换频率为1，当前使用套餐的流量单价为0.3元/MB，以前使用套餐的流量单价为0.35，那么由上述分类模型可知，该用户会订购包含流量的新套餐的概率为0.65。
205、BMP向CBP发送多个计费规则和潜在用户。
本发明实施例中，在多个待优化指标中存在需要进行模拟出账才能计算出具体值的待优化指标的情况下，如待优化指标为收入，那么BMP在计算待优化指标的值之前，还会将确定出的多个计费规则，以及预测的潜在用户发送给CBP，以便CBP基于多个计费规则中的每个计费规则对潜在用户进行模拟出账。
206、CBP接收来自BMP的多个计费规则和潜在用户，并基于多个计费规则中的每个计费规则对潜在用户进行模拟出账，得到模拟出账结果。
本发明实施例中，CBP在接收到BMP发送的多个计费规则和潜在用户的信息后，会基于多个计费规则中的每个计费规则对潜在用户进行模拟出账，得到模拟出账结果，其具体方式可以为：
CBP获取潜在用户的历史话单数据，其中，该历史话单数据中包含有目标业务的历史使用量信息，对于多个计费规则中的每个计费规则，CBP基于历史使用量信息计算每个潜在用户使用目标业务的资费信息，从而得到模拟出账结果。
进一步的，CBP还会将得到的模拟出账结果发送给BMP，以便BMP基于模拟出账结果计算至少一个待优化指标。
其中，历史话单数据可以包括含有目标业务的历史使用量信息，含有其他业务的历史使用量，用户的历史资费，等等。其中，历史使用量信息可以是指用户过去的一段时间内使用目标业务的数量，如一个月内的流量使用量，一个月内市内通话的分钟数，等等。
举例来说，假设目标业务为流量，多个待优化指标中存在收入这一项待优化指标，那么BMP在向CBP发送多个计费规则以及潜在用户后，CBP需要获取潜在用户在三个月内的流量使用量，从而计算每个计费规则(即新套餐的流量单价)分别与每个潜在用户的流量使用量的乘积，得到每个潜在用户使用流量的资费。
207、CBP向BMP发送模拟出账结果。
208、BMP接收来自CBP的模拟出账结果，并对于多个计费规则中的每个计费规则，计算潜在用户的多个待优化指标中每个待优化指标的值，其中包括基于该模拟出账结果计算多个待优化指标中至少一个待优化指标的值。
本发明实施例中，对于多个计费规则中的每个计费规则，BMP计算潜在用户的多个待优化指标中每个待优化指标的值，具体可以是：针对每个计费规则，分别计算每个潜在用户的每个待优化指标的值，然后统计同一计费规则下，同一待优化指标的所有潜在用户的待优化指标的值，从而得到该计费规则所对应的每个待优化指标的值。
举例来说，BMP可以获取潜在用户当前所用套餐的目标业务的计费规则，如果需要预演新套餐的匹配度，BMP则可以针对新套餐的多个计费规则中的每个计费规则，计算两计费规则之间的匹配度，从而得到新套餐的多个计费规则中的每个计费规则与潜在用户的平均匹配度。如果需要预演新套餐的收入，BMP则可以针对新套餐的多个计费规则中的每个计费规则，基于接收的模拟出账结果计算潜在用户的收入的值。
具体实现中，BMP基于该模拟出账结果计算多个待优化指标中至少一个待优化指标的值的具体方式可以为：BMP接收的是每个潜在用户针对每个计费规则模拟出的使用目标业务的资费信息，此处的资费信息可以看作一个待优化指标的值的一部分，如果BMP需要得到该待优化指标的值，则需要统计同一计费规则下，所有潜在用户使用目标业务的资费，从而得到该计费规则所对应的待优化指标的值。
本发明实施例中，采用潜在用户对新套餐的目标业务的计费规则进行套餐预演，即，采用得到的计费规则对潜在用户进行模拟出账，这样通过模拟出账结果计算的优化指标的值更能反映出新计费规则对大部分用户的影响，从而提高了预演的准确性。
需要说明的是，步骤205～207为可选步骤，只有多个待优化指标中存在需要模拟出账才能得到具体值的待优化指标的情况下才执行，如果不存在，BMP则不会执行基于该模拟出账结果计算多个待优化指标中至少一个待优化指标的值的操作。
举例来说，假设目标业务为流量，BMP确定出的多个计费规则分别为0.34、0.28以及0.19，CBP针对每个计费规则对潜在用户进行模拟出账后，将模拟出账结果发送给BMP，BMP计算多个计费规则中每个计费规则的匹配度，可以是平均匹配度，以及基于模拟出账结果计算每个计费规则的收入，如表4所示。
表4
价格 收入 匹配度
0.34 200000 0.6
0.28 180000 0.8
0.19 170000 0.9
由上表可知，当流量的单价为0.19时，套餐的匹配度更高、当流量的单价为0.34时，套餐的收入更高。
209、BMP确定满足优化目标的待优化指标的值所对应的计费规则为目标计费规则。
本发明实施例中，BMP在计算出每个计费规则所对应的待优化指标的值之后，会进一步判断待优化指标的值是否满足优化目标，因此，BMP需要预先获取每个待优化指标的优化目标，例如，收入在某一套餐的基础上提升5％，匹配度在该套餐的基础上提升8％。按照获取的优化目标确定出满足优化目标的待优化指标的值所对应的计费规则，从而将该计费规则作为目标计费规则，用以应用在新套餐的目标业务的计费规则。
举例来说，如表4所示，假设流量单价为0.34所对应的收入与指定套餐的收入相比，提升了9％，而匹配度只提升了3％，流量单价为0.28所对应的收入与指定套餐的收入相比，提升了6％，匹配度提升了5％，流量单价为0.19所对应的收入与指定套餐的收入相比，提升了4％，匹配度提升了7％，如果优化目标是收入提升5％，匹配度提升5％，那么基于上述数据，则会将流量单价为0.28的计费规则确定为新套餐关于流量的计费规则。
可见，在图2所描述的方法中，BMP可以根据目标业务和多个待优化指标确定多目标优化函数组，然后一次性求解出多目标优化函数组的有效解，即多组较优的计费规则，再从中确定出使得多个待优化指标满足优化目标的计费规则，从而不需要反复调整优化指标的权重，这样可以提高新套餐中关于目标业务的计费规则的确定效率。其中，目标业务为对目标用户影响最大的业务，而目标用户为对按照规则选取的用户进行扩展后的用户，这样设计的套餐更能真实反映出按照规则选取的用户所属的用户群体的需求。进一步的，对计费规则的预演针对潜在用户，即，采用得到的计费规则对潜在用户进行模拟出账，这样通过模拟出账结果计算的优化指标的值更能反映出新计费规则对大部分用户的影响，从而提高了预演的准确性。
基于图1所示的CBS架构，本发明实施例公开了一种BMP。请参阅图4，为本发明实施例公开的一种BMP的结构示意图。其中，图4所描述的BMP可以应用于上述方法实施例。如图4所示，BMP可以包括：
确定模块401，用于根据目标业务以及多个待优化指标确定多目标优化函数组，并确定多目标优化函数组的有效解，其中，有效解可以包括对目标业务的多个计费规则。
预测模块402，用于预测包含目标业务的套餐的潜在用户。
计算模块403，用于对于多个计费规则中的每个计费规则，计算潜在用户的多个待优化指标中每个待优化指标的值。
确定模块401，还用于确定满足优化目标的待优化指标的值所对应的计费规则为目标计费规则。
其中，目标业务可以是指市话、长途通话、漫游通话、流量、短信以及增值服务，如彩铃、来电提醒、可视通话、漏接来电提醒等中的一种。待优化指标可以是指收入、套餐匹配度、套餐订购用户数等，本发明实施例不做限定。潜在用户是可能会订购包含目标业务的新套餐的用户。
可选的，确定模块401根据目标业务以及多个待优化指标确定多目标优化函数组的具体方式可以为：
确定多个待优化指标中每个待优化指标关于目标业务的函数，从而将确定出的多个函数 作为多目标优化函数组。
可选的，BMP还可以包括：
选取模块404，用于选取目标用户。
那么确定模块401，还用于从多个业务中确定目标业务。
具体的，确定模块401从多个业务中确定目标业务的具体方式可以为：
对于多个业务中的每个业务，统计在预设时间段内目标用户对于该业务的平均业务使用量，以及在预设时间段内所有用户对于该业务的平均业务使用量，然后针对每个业务，计算目标用户对于该业务的平均业务使用量与所有用户对于该业务的平均业务使用量的比值，从而将最大比值对应的业务确定为目标业务。
其中，预设时间段可以是一个月，三个月，也可以是半年，本发明实施例不做限定。最大比值所对应的业务说明目标用户的平均使用量较大，在一定程度上说明该业务对目标用户的影响最大，从而可以将最大比值所对应的业务确定为目标业务。
可选的，选取模块404选取目标用户的具体方式可以为：
基于预设规则从所有用户中选取包括多个用户的用户集合，并获取用户集合中每个用户的至少一个特征，对于至少一个特征中的每个特征，计算用户集合中具备该特征的用户数量与所有用户中具备该特征的用户中除该用户集合中具备该特征的用户之外的其他用户数量的信息熵，从而将具备最小信息熵对应特征的其他用户添加至用户集合中，得到添加后的用户集合，然后对于至少一个特征中除最小信息熵对应特征之外的其他特征中的每个特征，重复执行计算用户集合中具备该特征的用户数量与所有用户中具备该特征的用户中除该用户集合中具备该特征的用户之外的其他用户数量的信息熵，以及将具备最小信息熵对应特征的其他用户添加至用户集合的操作，直至最小信息熵大于预设信息熵阈值为止，最后将添加后的用户集合中的用户确定为目标用户。
可选的，选取模块404选取目标用户的具体方式还可以为：
基于预设规则从所有用户中选取包括多个用户的用户集合，然后使用聚类算法，如k-means或者层次聚类等，将所有用户分为若干个簇，可以是基于每个用户的特征相似度将所有用户聚为若干个簇，然后针对每个簇，统计簇中属于用户集合的用户占该簇的所有用户的比例，从而将比例最大的簇中的所有用户，以及用户集合中的用户确定为目标用户。
其中，预设规则可以是人工定义的规则，例如“年龄18岁以上，籍贯为异地，1-3月和5-8月的消费量是其它月份的1.5倍以上”，“年龄18-24岁，下载使用过超级课程表App”，“长途通话的消费量占总消费量的60％以上”，等等。
可选的，预测模块402预测包含目标业务的套餐的潜在用户的具体方式可以为：
获取所有用户中每个用户的用户数据，然后通过分类算法对用户数据进行学习，得到分类模型，并基于用户数据和分类模型计算该用户数据对应的用户订购套餐的概率，从而将概率大于预设概率阈值的用户确定为潜在用户。其中，该用户数据可以包括目标业务，以及用户对包含目标业务的套餐的历史反馈信息。
其中，分类算法可以包括但不限于决策树算法、逻辑回归算法等。
可选的，BMP还可以包括：
通信模块405，用于向CBP发送潜在用户和多个计费规则，以及接收来自CBP的模拟出账结果，其中，该模拟出账结果由CBP基于多个计费规则中的每个计费规则对潜在用户进行模拟出账得到。
其中，计算模块403，还用于基于模拟出账结果计算多个待优化指标中的至少一个待优化指标的值。
基于图1所示的CBS架构，本发明实施例公开了另一种BMP。请参阅图5，为本发明实施例公开的另一种BMP的结构示意图。其中，图5所描述的BMP可以应用于上述方法实施例。如图5所示，BMP可以包括：至少一个处理器501，如CPU，通信装置502、存储器503以及至少一个通信总线504，上述处理器501、通信装置502和存储器503通过总线504连接。
其中，上述通信装置502可以为接收器、发送器、受处理器501的控制用于与外部设备进行消息的交换。上述存储器503可以是高速RAM存储器，也可为非不稳定的存储器(non-volatile memory)，例如磁盘存储器。可选的，还可以是至少一个位于处理器的存储装置。上述存储器503用于存储一组程序代码，上述处理器501用于调用存储器503中存储的程序代码，执行如下操作：
处理器501，用于根据目标业务以及多个待优化指标确定多目标优化函数组，确定多目标优化函数组的有效解，并预测包含目标业务的套餐的潜在用户。其中，有效解可以包括多个对目标业务的多个计费规则。
处理器501，还用于对于多个计费规则中的每个计费规则，计算潜在用户的多个待优化指标中每个待优化指标的值，并确定满足优化目标的待优化指标的值所对应的计费规则为目标计费规则。
可选的，上述通信装置502，还可以将多个计费规则和潜在用户发送给CBP，以便于CBP基于多个计费规则中的每个计费规则对潜在用户进行模拟出账，得到模拟出账结果，然后接收CBP发送的模拟出账结果。
处理器501从而可以基于模拟出账结果计算多个待优化指标中的至少一个待优化指标的值。
可见，在图4和图5所描述的BMP中，BMP可以根据目标业务和多个待优化指标确定多目标优化函数组，然后一次性求解出多目标优化函数组的有效解，即多组较优的计费规则，再从中确定出使得多个待优化指标满足优化目标的计费规则，从而不需要反复调整优化指标的权重，这样可以提高新套餐中关于目标业务的计费规则的确定效率。其中，目标业务为对目标用户影响最大的业务，而目标用户为对按照规则选取的用户进行扩展后的用户，这样设计的套餐更能真实反映出按照规则选取的用户所属的用户群体的需求。进一步的，对计费规则的预演针对潜在用户，即，采用得到的计费规则对潜在用户进行模拟出账，这样通过模拟出账结果计算的优化指标的值更能反映出新计费规则对大部分用户的影响，从而提高了预演的准确性。
基于图1所示的CBS架构，本发明实施例公开了一种CBP。请参阅图6，为本发明实施例公开的一种CBP的结构示意图。其中，图6所描述的CBP可以应用于上述方法实施例。如图6所示，CBP可以包括：
通信模块601，用于接收来自BMP的对目标业务的多个计费规则和包含目标业务的套餐的潜在用户。
出账模块602，用于基于多个计费规则中的每个计费规则对潜在用户进行模拟出账，得到模拟出账结果。
通信模块601，还用于向BMP发送模拟出账结果，以便BMP基于模拟出账规则计算潜在 用户的多个待优化指标中的至少一个待优化指标的值。
具体的，出账模块602基于多个计费规则中的每个计费规则对潜在用户进行模拟出账，得到模拟出账结果的具体方式可以为：
获取潜在用户的历史话单数据，其中，该历史话单数据中包含有目标业务的历史使用量信息，对于多个计费规则中的每个计费规则，基于历史使用量信息计算每个潜在用户使用目标业务的资费信息，从而得到模拟出账结果。
其中，历史话单数据可以包括含有目标业务的历史使用量信息，含有其他业务的历史使用量，用户的历史资费，等等。其中，历史使用量信息可以是指用户过去的一段时间内使用目标业务的数量，如一个月内的流量使用量，一个月内市内通话的分钟数，等等。
基于图1所示的CBS架构，本发明实施例公开了另一种CBP。请参阅图7，为本发明实施例公开的另一种CBP的结构示意图。其中，图7所描述的CBP可以应用于上述方法实施例。如图7所示，BMP可以包括：至少一个处理器701，如CPU，通信装置702、存储器703以及至少一个通信总线704，上述处理器701、通信装置702和存储器703通过总线704连接。
其中，上述通信装置702可以为接收器、发送器、受处理器501的控制用于与外部设备进行消息的交换。上述存储器703可以是高速RAM存储器，也可为非不稳定的存储器(non-volatile memory)，例如磁盘存储器。可选的，还可以是至少一个位于处理器的存储装置。上述存储器703用于存储一组程序代码，上述处理器701和通信装置702用于调用存储器703中存储的程序代码，执行如下操作：
通信装置702，用于接收来自BMP的对目标业务的多个计费规则和包含目标业务的套餐的潜在用户。
处理器701，用于基于多个计费规则中的每个计费规则对潜在用户进行模拟出账，得到模拟出账结果。
通信装置702，还用于向BMP发送模拟出账结果，以便BMP基于模拟出账规则计算潜在用户的多个待优化指标中的至少一个待优化指标的值。
可见，在图6和图7所描述的CBP中，CBP基于多个计费规则中的每个计费规则对潜在用户进行模拟出账，这样通过模拟出账结果计算的优化指标的值更能反映出新计费规则对大部分用户的影响，从而提高了预演的准确性。
基于图1所示的CBS架构，本发明实施例公开了一种目标计费规则确定系统。请参阅图8，为本发明实施例公开的一种目标计费规则确定系统的结构示意图。如图8所示，该系统可以包括BMP801以及CBP802，其中：
BMP801可以根据目标业务以及多个待优化指标确定多目标优化函数组，确定多目标优化函数组的有效解，其中，该有效解包括对目标业务的多个计费规则。BMP还会预测包含目标业务的套餐的潜在用户，即预设可能会订购包括目标业务的新套餐的用户，然后向CBP802发送多个计费规则以及预测出的潜在用户。
CBP802在接收到来自BMP801的多个计费规则和潜在用户后，会基于多个计费规则中的每个计费规则，对潜在用户进行模拟出账，从而得到模拟出账结果，然后将模拟出账结果发送给BMP801。
BMP801在接收到来自CBP802的模拟出账结果后，可以对于多个计费规则中的每个计费规则计算潜在用户的多个待优化指标中的每个待优化指标的值。其中，包括基于模拟出账结 果计算至少一个待优化指标的值。进一步的，BMP801在计算出每个计费规则所对应的每个待优化指标的值后，会将满足优化目标的待优化指标的值所对应的计费规则确定为目标计费规则，从而将其作为新套餐中对目标业务的计费规则，并发布新套餐。
可选的，BMP801还可以预先选取目标用户，以及从多个业务中确定目标业务。具体的，目标用户的选取，即为，按照预设规则选取种子用户，然后采用lookalike方法或者聚类算法扩展种子用户，从而将最终扩展得到的用户作为新套餐的目标用户。目标业务的确定，即为，从多个业务中选取对目标用户影响最大的业务。
可见，在图8所描述的系统中，BMP可以根据目标业务和多个待优化指标确定多目标优化函数组，然后一次性求解出多目标优化函数组的有效解，即多组较优的计费规则，再从中确定出使得多个待优化指标满足优化目标的计费规则，从而不需要反复调整优化指标的权重，这样可以提高新套餐中关于目标业务的计费规则的确定效率。其中，目标业务为对目标用户影响最大的业务，而目标用户为对按照规则选取的用户进行扩展后的用户，这样设计的套餐更能真实反映出按照规则选取的用户所属的用户群体的需求。进一步的，对计费规则的预演针对潜在用户，即，采用得到的计费规则对潜在用户进行模拟出账，这样通过模拟出账结果计算的优化指标的值更能反映出新计费规则对大部分用户的影响，从而提高了预演的准确性。
需要说明的是，在上述实施例中，对各个实施例的描述都各有侧重，某个实施例中没有详细描述的部分，可以参见其他实施例的相关描述。其次，本领域技术人员也应该知悉，说明书中所描述的实施例均属于优选实施例，所涉及的动作和模块并不一定是本发明所必须的。
本发明实施例方法中的步骤可以根据实际需要进行顺序调整、合并和删减。
本发明实施例BMP和CBP中的模块可以根据实际需要进行合并、划分和删减。
本发明实施例中所述BMP和CBP，可以通过通用集成电路，例如CPU(Central Processing Unit，中央处理器)，或通过ASIC(Application Specific Integrated Circuit，专用集成电路)来实现。
本领域普通技术人员可以理解实现上述实施例方法中的全部或部分流程，是可以通过计算机程序来指令相关的硬件来完成，所述的程序可存储于计算机可读取存储介质中，该程序在执行时，可包括如上述各方法的实施例的流程。其中，所述的存储介质可为磁碟、光盘、只读存储记忆体(Read-Only Memory，ROM)或随机存储记忆体(Random Access Memory，RAM)等。
以上对本发明实施例公开的一种目标计费规则确定方法、相关设备及系统进行了详细介绍，本文中应用了具体实例对本发明的原理及实施方式进行了阐述，以上实施例的说明只是用于帮助理解本发明及其核心思想；同时，对于本领域的一般技术人员，依据本发明的思想，在具体实施方式及应用范围上均会有改变之处，综上所述，本说明书内容不应理解为对本发明的限制。

Claims (19)

  1. 一种目标计费规则确定方法，其特征在于，所述方法包括：
    业务管理点BMP根据目标业务以及多个待优化指标确定多目标优化函数组；
    所述BMP确定所述多目标优化函数组的有效解，所述有效解包括对所述目标业务的多个计费规则；
    所述BMP预测包含所述目标业务的套餐的潜在用户；
    对于所述多个计费规则中的每个计费规则，所述BMP计算所述潜在用户的所述多个待优化指标中每个待优化指标的值；
    所述BMP确定满足优化目标的待优化指标的值所对应的计费规则为目标计费规则。
  2. 根据权利要求1所述的方法，其特征在于，所述BMP根据目标业务以及多个待优化指标确定多目标优化函数组，包括：
    所述BMP确定所述多个待优化指标中每个待优化指标关于所述目标业务的函数；
    所述BMP将确定出的多个函数作为所述多目标优化函数组。
  3. 根据权利要求1所述的方法，其特征在于，所述BMP根据目标业务以及多个待优化指标确定多目标优化函数组之前，所述方法还包括：
    所述BMP选取目标用户；
    所述BMP从多个业务中确定目标业务；
    其中，所述BMP从多个业务中确定目标业务，包括：
    对于所述多个业务中的每个业务，所述BMP统计在预设时间段内所述目标用户对于该业务的平均业务使用量；
    对于所述多个业务中的每个业务，所述BMP统计在所述预设时间段内所有用户对于该业务的平均业务使用量；
    对于所述多个业务中的每个业务，所述BMP计算所述目标用户对于该业务的平均业务使用量与所述所有用户对于该业务的平均业务使用量的比值，并将最大比值对应的业务确定为目标业务。
  4. 根据权利要求3所述的方法，其特征在于，所述BMP选取目标用户，包括：
    所述BMP基于预设规则从所述所有用户中选取包括多个用户的用户集合，并获取所述用户集合中每个用户的至少一个特征；
    对于所述至少一个特征中的每个特征，所述BMP计算所述用户集合中具备该特征的用户数量与所述所有用户中具备该特征的用户中除所述用户集合中具备该特征的用户之外的其他用户数量的信息熵；
    所述BMP将具备最小信息熵对应特征的所述其他用户添加至所述用户集合，并对于所述至少一个特征中除所述最小信息熵对应特征之外的其他特征中的每个特征，重复执行所述计算所述用户集合中具备该特征的用户数量与所述所有用户中具备该特征的用户中除所述用户集合中具备该特征的用户之外的其他用户数量的信息熵，以及所述将具备最小信息熵对应特征的所述其他用户添加至所述用户集合的操作，直至最小信息熵大于预设信息熵阈值为止；
    所述BMP将添加后的用户集合中的用户确定为目标用户。
  5. 根据权利要求1所述的方法，其特征在于，所述BMP预测包含所述目标业务的套餐的潜在用户，包括：
    所述BMP获取所述所有用户中每个用户的用户数据，所述用户数据包括所述目标业务，以及用户对包含所述目标业务的套餐的历史反馈信息；
    所述BMP通过分类算法对所述用户数据进行学习，得到分类模型；
    所述BMP基于所述用户数据和所述分类模型计算所述用户数据对应的用户订购所述套餐的概率；
    所述BMP将概率大于预设概率阈值的用户确定为潜在用户。
  6. 根据权利要求1～5任一项所述的方法，其特征在于，所述方法还包括：
    所述BMP向融合计费点CBP发送所述潜在用户和所述多个计费规则；
    所述BMP接收来自所述CBP的模拟出账结果，所述模拟出账结果由所述CBP基于所述多个计费规则中的每个计费规则对所述潜在用户进行模拟出账得到；
    所述BMP基于所述模拟出账结果计算所述多个待优化指标中的至少一个待优化指标的值。
  7. 一种目标计费规则确定方法，其特征在于，所述方法包括：
    融合计费点CBP接收来自业务管理点BMP的对目标业务的多个计费规则和包含所述目标业务的套餐的潜在用户；
    所述CBP基于所述多个计费规则中的每个计费规则对所述潜在用户进行模拟出账，得到模拟出账结果；
    所述CBP向所述BMP发送所述模拟出账结果，以便所述BMP基于所述模拟出账结果计算所述潜在用户的多个待优化指标中的至少一个待优化指标的值。
  8. 根据权利要求7所述的方法，其特征在于，所述CBP基于所述多个计费规则中的每个计费规则对所述潜在用户进行模拟出账，得到模拟出账结果，包括：
    所述CBP获取所述潜在用户的历史话单数据，所述历史话单数据包含所述目标业务的历史使用量信息；
    对于所述多个计费规则中的每个计费规则，所述CBP基于所述历史使用量信息计算每个潜在用户使用所述目标业务的资费信息。
  9. 一种业务管理点BMP，其特征在于，所述BMP包括：
    确定模块，用于根据目标业务以及多个待优化指标确定多目标优化函数组，并确定所述多目标优化函数组的有效解，所述有效解包括对所述目标业务的多个计费规则；
    预测模块，用于预测包含所述目标业务的套餐的潜在用户；
    计算模块，用于对于所述多个计费规则中的每个计费规则，计算所述潜在用户的所述多个待优化指标中每个待优化指标的值；
    所述确定模块，还用于确定满足优化目标的待优化指标的值所对应的计费规则为目标计费规则。
  10. 根据权利要求9所述的BMP，其特征在于，所述确定模块根据目标业务以及多个待优化指标确定多目标优化函数组的具体方式为：
    确定所述多个待优化指标中每个待优化指标关于所述目标业务的函数；
    将确定出的多个函数作为所述多目标优化函数组。
  11. 根据权利要求9所述的BMP，其特征在于，所述BMP还包括：
    选取模块，用于选取目标用户；
    所述确定模块，还用于从多个业务中确定目标业务；
    其中，所述确定模块从多个业务中确定目标业务的具体方式为：
    对于所述多个业务中的每个业务，统计在预设时间段内所述目标用户对于该业务的平均业务使用量；
    对于所述多个业务中的每个业务，统计在所述预设时间段内所有用户对于该业务的平均业务使用量；
    对于所述多个业务中的每个业务，计算所述目标用户对于该业务的平均业务使用量与所述所有用户对于该业务的平均业务使用量的比值，并将最大比值对应的业务确定为目标业务。
  12. 根据权利要求11所述的BMP，其特征在于，所述选取模块选取目标用户的具体方式为：
    基于预设规则从所述所有用户中选取包括多个用户的用户集合，并获取所述用户集合中每个用户的至少一个特征；
    对于所述至少一个特征中的每个特征，计算所述用户集合中具备该特征的用户数量与所述所有用户中具备该特征的用户中除所述用户集合中具备该特征的用户之外的其他用户数量的信息熵；
    将具备最小信息熵对应特征的所述其他用户添加至所述用户集合，并对于所述至少一个特征中除所述最小信息熵对应特征之外的其他特征中的每个特征，重复执行所述计算所述用户集合中具备该特征的用户数量与所述所有用户中具备该特征的用户中除所述用户集合中具备该特征的用户之外的其他用户数量的信息熵，以及所述将具备最小信息熵对应特征的所述其他用户添加至所述用户集合的操作，直至最小信息熵大于预设信息熵阈值为止；
    将添加后的用户集合确定为目标用户。
  13. 根据权利要求9所述的BMP，其特征在于，所述预测模块预测包含所述目标业务的套餐的潜在用户的具体方式为：
    获取所述所有用户中每个用户的用户数据，所述用户数据包括所述目标业务，以及用户对包含所述目标业务的套餐的历史反馈信息；
    通过分类算法对所述用户数据进行学习，得到分类模型；
    基于所述用户数据和所述分类模型计算所述用户数据对应的用户订购所述套餐的概率；
    将概率大于预设概率阈值的用户确定为潜在用户。
  14. 根据权利要求9～13任一项所述的BMP，其特征在于，所述BMP还包括：
    通信模块，用于向融合计费点CBP发送所述潜在用户和所述多个计费规则，以及接收来自所述CBP的模拟出账结果，所述模拟出账结果由所述CBP基于所述多个计费规则中的每个计费规则对所述潜在用户进行模拟出账得到；
    其中，所述计算模块，还用于基于所述模拟出账结果计算所述多个待优化指标中的至少一个待优化指标的值。
  15. 一种业务管理点BMP，其特征在于，所述BMP包括处理器、通信装置、存储器以及通信总线，其中：所述处理器、所述通信装置、所述存储器通过所述通信总线连接；所述通信装置受所述处理器的控制用于收发消息；所述存储器用于存储一组程序代码，所述处理器 用于调用所述存储器中存储的程序代码执行如权利要求1～6任一项所述的方法。
  16. 一种融合计费点CBP，其特征在于，所述CBP包括：
    通信模块，用于接收来自业务管理点BMP的对目标业务的多个计费规则和包含所述目标业务的套餐的潜在用户；
    出账模块，用于基于所述多个计费规则中的每个计费规则对所述潜在用户进行模拟出账，得到模拟出账结果；
    所述通信模块，还用于向所述BMP发送所述模拟出账结果，以便所述BMP基于所述模拟出账结果计算所述潜在用户的多个待优化指标中的至少一个待优化指标的值。
  17. 根据权利要去16所述的CBP，其特征在于，所述出账模块基于所述多个计费规则中的每个计费规则对所述潜在用户进行模拟出账，得到模拟出账结果的具体方式为：
    获取所述潜在用户的历史话单数据，，所述历史话单数据包含所述目标业务的历史使用量信息；
    对于所述多个计费规则中的每个计费规则，基于所述历史使用量信息计算每个潜在用户使用所述目标业务的资费信息。
  18. 一种融合计费点CBP，其特征在于，所述CBP包括处理器、通信装置、存储器以及通信总线，其中：所述处理器、所述通信装置、所述存储器通过所述通信总线连接；所述通信装置受所述处理器的控制用于收发消息；所述存储器用于存储一组程序代码，所述处理器用于调用所述存储器中存储的程序代码执行如权利要求7或8所述的方法。
  19. 一种目标计费规则确定系统，其特征在于，所述系统包括业务管理点BMP和融合计费点CBP，其中：
    所述BMP，用于根据目标业务以及多个待优化指标确定多目标优化函数组，并确定所述多目标优化函数组的有效解，所述有效解包括对所述目标业务的多个计费规则；
    所述BMP，还用于预测包含所述目标业务的套餐的潜在用户；
    所述BMP，还用于向所述CBP发送所述多个计费规则和所述潜在用户；
    所述CBP，用于接收来自所述BMP的所述多个计费规则和所述潜在用户；
    所述CBP，还用于基于所述多个计费规则中的每个计费规则对所述潜在用户进行模拟出账，得到模拟出账结果，并向所述BMP发送所述模拟出账结果；
    所述BMP，还用于接收来自所述CBP的所述模拟出账结果；
    所述BMP，还用于对于所述多个计费规则中的每个计费规则，基于所述模拟出账结果计算所述多个待优化指标中至少一个待优化指标的值；
    所述BMP，还用于确定满足优化目标的待优化指标的值所对应的计费规则为目标计费规则。
PCT/CN2017/104086 2016-11-11 2017-09-28 一种目标计费规则确定方法、相关设备及系统 WO2018086428A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201610998206.7A CN108076242B (zh) 2016-11-11 2016-11-11 一种目标计费规则确定方法、相关设备及系统
CN201610998206.7 2016-11-11

Publications (1)

Publication Number Publication Date
WO2018086428A1 true WO2018086428A1 (zh) 2018-05-17

Family

ID=62109392

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2017/104086 WO2018086428A1 (zh) 2016-11-11 2017-09-28 一种目标计费规则确定方法、相关设备及系统

Country Status (2)

Country Link
CN (1) CN108076242B (zh)
WO (1) WO2018086428A1 (zh)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110233741A (zh) * 2019-05-21 2019-09-13 深圳壹账通智能科技有限公司 服务计费方法、装置、设备及存储介质
CN111061773A (zh) * 2019-11-25 2020-04-24 深圳壹账通智能科技有限公司 数据统计方法及服务器
CN111381768A (zh) * 2018-12-29 2020-07-07 北京亿阳信通科技有限公司 一种数据监控的方法和装置
CN112418425A (zh) * 2020-11-18 2021-02-26 联通(浙江)产业互联网有限公司 用户服务匹配方法及装置

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110557339B (zh) * 2018-05-30 2023-06-30 阿里巴巴集团控股有限公司 一种流量规划方法、装置、计算机设备、存储介质

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7184977B1 (en) * 1997-12-31 2007-02-27 Gte Communications Corporation CLEC convergent billing system
CN102291704A (zh) * 2011-07-28 2011-12-21 华为软件技术有限公司 计费方法、计费设备、交换设备和计费系统
CN103535059A (zh) * 2012-10-16 2014-01-22 华为终端有限公司 计费处理方法、装置和系统
CN104301874A (zh) * 2014-09-15 2015-01-21 华为软件技术有限公司 系统通知发送方法、设备及系统
CN105681614A (zh) * 2015-12-29 2016-06-15 华为软件技术有限公司 一种业务处理方法及装置

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7184977B1 (en) * 1997-12-31 2007-02-27 Gte Communications Corporation CLEC convergent billing system
CN102291704A (zh) * 2011-07-28 2011-12-21 华为软件技术有限公司 计费方法、计费设备、交换设备和计费系统
CN103535059A (zh) * 2012-10-16 2014-01-22 华为终端有限公司 计费处理方法、装置和系统
CN104301874A (zh) * 2014-09-15 2015-01-21 华为软件技术有限公司 系统通知发送方法、设备及系统
CN105681614A (zh) * 2015-12-29 2016-06-15 华为软件技术有限公司 一种业务处理方法及装置

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111381768A (zh) * 2018-12-29 2020-07-07 北京亿阳信通科技有限公司 一种数据监控的方法和装置
CN110233741A (zh) * 2019-05-21 2019-09-13 深圳壹账通智能科技有限公司 服务计费方法、装置、设备及存储介质
CN110233741B (zh) * 2019-05-21 2022-06-17 深圳壹账通智能科技有限公司 服务计费方法、装置、设备及存储介质
CN111061773A (zh) * 2019-11-25 2020-04-24 深圳壹账通智能科技有限公司 数据统计方法及服务器
CN112418425A (zh) * 2020-11-18 2021-02-26 联通(浙江)产业互联网有限公司 用户服务匹配方法及装置

Also Published As

Publication number Publication date
CN108076242A (zh) 2018-05-25
CN108076242B (zh) 2020-04-28

Similar Documents

Publication Publication Date Title
WO2018086428A1 (zh) 一种目标计费规则确定方法、相关设备及系统
JP5289588B2 (ja) ユーザーのサービス量に基づいて、オンラインで料金を計算する方法及びシステム
CN111162925B (zh) 网络流量预测方法、装置、电子设备及存储介质
CN110418022A (zh) 为多个用户标识调整流量套餐的方法及装置
Dullaghan et al. Integration of machine learning techniques to evaluate dynamic customer segmentation analysis for mobile customers
CN103843374A (zh) 基于sy的集成的策略和计费控制
WO2005111878A1 (en) Information verification in a telecommunications network
US20170032432A1 (en) Integrated System
CN110866698A (zh) 用于评定服务提供方的服务分值的装置
CN111182479B (zh) 信息发送的控制方法及装置
US10944874B2 (en) Telecommunication system for monitoring and controlling of a network providing resource to a user
EP3104550B1 (en) Mobile digital cellular telecommunication system with advanced functionality for rating correction
CN105848127B (zh) 一种精确补单方法和装置
CN110930139A (zh) 基于客户消费量的差异化燃气缴费服务系统
CN108076443A (zh) 一种在线计费系统中流量的分配方法及在线计费系统
CN101751611A (zh) 渠道酬金结算方法、装置及系统
CN106503977B (zh) 数据的处理方法、系统及装置
EP3331196A1 (en) Telecommunication system for monitoring and controlling of a network providing resource to a user
CN110288366B (zh) 一种资源发放模型的评估方法及装置
CN112149398A (zh) 一种快速生成电商企业经营利润报表的系统
CN101511075A (zh) 一种通信服务计费方法和通信服务计费系统
US20220147954A1 (en) Prediction of billing disputes
CN106888187B (zh) 业务处理方法和装置
CN102883294A (zh) 一种关联用户行为的分时段计费方法
CN103024207B (zh) 后付费套餐费用的查询处理方法和系统、分解业务服务器

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 17870139

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 17870139

Country of ref document: EP

Kind code of ref document: A1