CN115018556A - A method for determining account set based on federated learning and its device and electronic device - Google Patents
A method for determining account set based on federated learning and its device and electronic deviceDownload PDFInfo
- Publication number
- CN115018556A CN115018556ACN202210764418.4ACN202210764418ACN115018556ACN 115018556 ACN115018556 ACN 115018556ACN 202210764418 ACN202210764418 ACN 202210764418ACN 115018556 ACN115018556 ACN 115018556A
- Authority
- CN
- China
- Prior art keywords
- target
- account
- data
- consumption
- prediction model
- Prior art date
- Legal status (The legal status 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 status listed.)
- Granted
Links
Images
Classifications
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0207—Discounts or incentives, e.g. coupons or rebates
- G06Q30/0212—Chance discounts or incentives
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0207—Discounts or incentives, e.g. coupons or rebates
- G06Q30/0215—Including financial accounts
- G06Q30/0216—Investment accounts
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/06—Asset management; Financial planning or analysis
Landscapes
- Business, Economics & Management (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Engineering & Computer Science (AREA)
- Development Economics (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Operations Research (AREA)
- Technology Law (AREA)
- Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)
Abstract
Description
Translated fromChinese技术领域technical field
本发明涉及金融科技领域，具体而言，涉及一种基于联邦学习的账户集合确定方法及其装置、电子设备。The present invention relates to the field of financial technology, and in particular, to a method for determining an account set based on federated learning, an apparatus and an electronic device thereof.
背景技术Background technique
在万物互联的时代，用户在消费支付时越来越方便，大多用户已经习惯使用二维码等电子方式进行消费支付，这对传统金融机构的业务造成了不小的影响，使得大量用户大量流失。In the era of the Internet of Everything, it is becoming more and more convenient for users to pay for consumption. Most users are accustomed to using electronic methods such as QR codes to pay for consumption. This has caused a considerable impact on the business of traditional financial institutions, causing a large number of users to lose a lot of money. .
传统金融机构为了防止用户流失和沉睡用户的促活，采取了多种促销方式，例如，发放优惠券(例如，满减券、商户券等)等虚拟物品，以促使用户消费。然而，相关技术中，由于发放的用户群体固定，并且仅仅采用金融机构本身留存数据进行机器学习建模，难以预测合适的发放群体，会造成优惠券等虚拟物品发放后，投入产出比不高，难以提升用户活跃度以及难以长期对用户进行留存的问题。In order to prevent the loss of users and the promotion of sleeping users, traditional financial institutions have adopted a variety of promotion methods, such as issuing coupons (for example, full discount coupons, merchant coupons, etc.) and other virtual items to encourage users to consume. However, in the related technologies, due to the fixed user groups, and only the data retained by the financial institution itself is used for machine learning modeling, it is difficult to predict the appropriate distribution group, which will result in a low input-output ratio after virtual items such as coupons are issued. , it is difficult to increase user activity and it is difficult to retain users for a long time.
针对上述的问题，目前尚未提出有效的解决方案。For the above problems, no effective solution has been proposed yet.
发明内容SUMMARY OF THE INVENTION
本发明实施例提供了一种基于联邦学习的账户集合确定方法及其装置、电子设备，以至少解决相关技术中无法确定合适的用户群体，导致虚拟物品发放后，投入产出比不高，难以提升用户活跃度以及难以长期对用户进行留存的技术问题。The embodiments of the present invention provide a method for determining an account set based on federated learning, a device, and an electronic device, so as to at least solve the problem that a suitable user group cannot be determined in the related art, resulting in a low input-output ratio after virtual items are issued, and it is difficult to Improve user activity and technical problems that are difficult to retain users for a long time.
根据本发明实施例的一个方面，提供了一种基于联邦学习的账户集合确定方法，包括：获取目标金融机构的特征数据，并基于所述特征数据，构建初始分类模型和初始预测模型；基于预设联邦策略，扩展所述特征数据，得到目标特征数据；基于所述目标特征数据，调整所述初始分类模型和所述初始预测模型，得到目标分类模型和目标预测模型；基于所述目标分类模型，计算多个账户集合的绑卡率，并基于所述目标预测模型，计算所述多个账户集合的消费提升率；确定所述绑卡率大于第一预设阈值且所述消费提升率大于第二预设阈值的账户集合为目标账户集合，其中，所述目标金融机构对所述目标账户集合中的每个金融账户发放虚拟物品。According to an aspect of the embodiments of the present invention, a method for determining an account set based on federated learning is provided, including: acquiring characteristic data of a target financial institution, and building an initial classification model and an initial prediction model based on the characteristic data; Set up a federation strategy, expand the feature data, and obtain target feature data; based on the target feature data, adjust the initial classification model and the initial prediction model to obtain a target classification model and a target prediction model; based on the target classification model , calculate the card binding rate of multiple account sets, and based on the target prediction model, calculate the consumption increase rate of the multiple account sets; determine that the card binding rate is greater than the first preset threshold and the consumption increase rate is greater than The account set with the second preset threshold is a target account set, wherein the target financial institution issues virtual items to each financial account in the target account set.
可选地，在获取目标金融机构的特征数据之前，还包括：获取预设历史时间段内各个金融账户的账户数据和虚拟物品数据；基于所述账户数据和所述虚拟物品数据，采用预设分类策略，确定所述金融账户所属的所述账户集合。Optionally, before acquiring the characteristic data of the target financial institution, the method further includes: acquiring account data and virtual item data of each financial account within a preset historical time period; based on the account data and the virtual item data, using a preset A classification strategy to determine the account set to which the financial account belongs.
可选地，在基于所述目标特征数据，调整所述初始分类模型和所述初始预测模型，得到目标分类模型和目标预测模型之后，还包括：基于所述目标分类模型和所述目标预测模型，调整所述账户集合，其中，调整后的所述账户集合包括下述至少之一：使用其他金融机构卡但未用目标金融机构卡的账户集合、具有社交属性的账户集合、发放虚拟物品后绑定所述目标金融机构卡的账户集合、未使用所述目标金融机构卡进行消费的账户集合。Optionally, after adjusting the initial classification model and the initial prediction model based on the target feature data to obtain the target classification model and the target prediction model, the method further includes: based on the target classification model and the target prediction model. , adjust the account set, wherein the adjusted account set includes at least one of the following: an account set that uses other financial institution cards but does not use the target financial institution card, an account set with social attributes, and a A set of accounts bound to the target financial institution card, and a set of accounts that do not use the target financial institution card for consumption.
可选地，所述特征数据至少包括：用户属性数据、虚拟物品属性数据、用户消费数据，其中，所述用户属性数据包括下述至少之一：性别、年龄、地区号、金融机构客户类型、金融机构员工标志，所述虚拟物品属性数据包括下述至少之一：虚拟物品类型、各平台已发放次数、各平台核销率，所述用户消费数据包括下述至少之一：预设时间内各平台的总消费次数、预设时间内各平台的总消费金额、领取前的最后消费时间。Optionally, the feature data includes at least: user attribute data, virtual item attribute data, and user consumption data, wherein the user attribute data includes at least one of the following: gender, age, area code, financial institution customer type, An employee logo of a financial institution, the virtual item attribute data includes at least one of the following: the type of virtual item, the number of times each platform has issued, and the write-off rate of each platform, and the user consumption data includes at least one of the following: within a preset time The total consumption times of each platform, the total consumption amount of each platform within the preset time, and the last consumption time before claiming.
可选地，基于预设联邦策略，扩展所述特征数据，得到目标特征数据的步骤，包括：基于所述预设联邦策略，获取多个外部机构的外部特征数据，其中，所述外部特征数据包括下述至少之一：用户行为数据、用户兴趣数据、标识映射数据、设备信息、网络特征数据；基于预设特征维度，从所述外部特征数据中筛选目标数据；基于所述目标数据，扩展所述特征数据，得到所述目标特征数据。Optionally, the step of expanding the feature data based on a preset federal strategy to obtain target feature data includes: acquiring external feature data of multiple external institutions based on the preset federal strategy, wherein the external feature data It includes at least one of the following: user behavior data, user interest data, identification mapping data, device information, and network feature data; based on preset feature dimensions, screening target data from the external feature data; based on the target data, expanding and obtaining the target feature data from the feature data.
可选地，基于所述目标分类模型，计算多个账户集合的绑卡率的步骤，包括：采用所述目标分类模型，确定所述多个账户集合中各金融账户在领完虚拟物品的第一预设时间段内是否进行绑卡，得到确定结果；基于所述确定结果，统计所述多个账户集合在所述第一预设时间段内的绑卡量；基于所述绑卡量，计算所述多个账户集合的所述绑卡率。Optionally, based on the target classification model, the step of calculating the card binding rates of multiple account sets includes: using the target classification model to determine that each financial account in the multiple account sets is on the first day after receiving the virtual item. A determination result is obtained whether card binding is performed within a preset time period; based on the determination result, the amount of card binding of the multiple account sets within the first preset time period is counted; based on the card binding amount, The card binding rate of the plurality of account sets is calculated.
可选地，基于所述目标预测模型，计算所述多个账户集合的消费提升率的步骤，包括：采用所述目标预测模型，确定所述多个账户集合中各金融账户在领完虚拟物品的第二预设时间段内的第一消费次数，并确定各所述金融账户在领取虚拟物品前的第三预设时间段内的第二消费次数；基于所述第一消费次数和所述第二消费次数，计算所述多个账户集合的所述消费提升率。Optionally, based on the target prediction model, the step of calculating the consumption improvement rates of the multiple account sets includes: using the target prediction model to determine that each financial account in the multiple account sets is finishing receiving virtual items. the first consumption times within the second preset time period of the 2019-01 For the second consumption times, the consumption promotion rate of the multiple account sets is calculated.
根据本发明实施例的另一方面，还提供了一种基于联邦学习的账户集合确定装置，包括：获取单元，用于获取目标金融机构的特征数据，并基于所述特征数据，构建初始分类模型和初始预测模型；扩展单元，用于基于预设联邦策略，扩展所述特征数据，得到目标特征数据；调整单元，用于基于所述目标特征数据，调整所述初始分类模型和所述初始预测模型，得到目标分类模型和目标预测模型；计算单元，用于基于所述目标分类模型，计算多个账户集合的绑卡率，并基于所述目标预测模型，计算所述多个账户集合的消费提升率；确定单元，用于确定所述绑卡率大于第一预设阈值且所述消费提升率大于第二预设阈值的账户集合为目标账户集合，其中，所述目标金融机构对所述目标账户集合中的每个金融账户发放虚拟物品。According to another aspect of the embodiments of the present invention, an apparatus for determining an account set based on federated learning is also provided, including: an acquisition unit configured to acquire characteristic data of a target financial institution, and based on the characteristic data, construct an initial classification model and an initial prediction model; an expansion unit for expanding the feature data based on a preset federation strategy to obtain target feature data; an adjustment unit for adjusting the initial classification model and the initial prediction based on the target feature data a model to obtain a target classification model and a target prediction model; a computing unit, used to calculate the card binding rate of multiple account sets based on the target classification model, and calculate the consumption of the multiple account sets based on the target prediction model an increase rate; a determination unit configured to determine the account set whose card binding rate is greater than a first preset threshold and the consumption increase rate is greater than a second preset threshold as a target account set, wherein the target financial institution Each financial account in the set of target accounts issues virtual items.
可选地，所述确定装置还包括：第一获取模块，用于在获取目标金融机构的特征数据之前，获取预设历史时间段内各个金融账户的账户数据和虚拟物品数据；第一确定模块，用于基于所述账户数据和所述虚拟物品数据，采用预设分类策略，确定所述金融账户所属的所述账户集合。Optionally, the determining device further includes: a first acquiring module, configured to acquire account data and virtual item data of each financial account within a preset historical time period before acquiring the characteristic data of the target financial institution; the first determining module , for determining the account set to which the financial account belongs by adopting a preset classification strategy based on the account data and the virtual item data.
可选地，所述确定装置还包括：第一调整模块，用于在基于所述目标特征数据，调整所述初始分类模型和所述初始预测模型，得到目标分类模型和目标预测模型之后，基于所述目标分类模型和所述目标预测模型，调整所述账户集合，其中，调整后的所述账户集合包括下述至少之一：使用其他金融机构卡但未用目标金融机构卡的账户集合、具有社交属性的账户集合、发放虚拟物品后绑定所述目标金融机构卡的账户集合、未使用所述目标金融机构卡进行消费的账户集合。Optionally, the determining device further includes: a first adjustment module, configured to adjust the initial classification model and the initial prediction model based on the target feature data to obtain the target classification model and the target prediction model, based on the target classification model and the target prediction model. The target classification model and the target prediction model adjust the account set, wherein the adjusted account set includes at least one of the following: an account set that uses other financial institution cards but does not use the target financial institution card, A set of accounts with social attributes, a set of accounts bound to the target financial institution card after issuing virtual items, and a set of accounts that do not use the target financial institution card for consumption.
可选地，所述特征数据至少包括：用户属性数据、虚拟物品属性数据、用户消费数据，其中，所述用户属性数据包括下述至少之一：性别、年龄、地区号、金融机构客户类型、金融机构员工标志，所述虚拟物品属性数据包括下述至少之一：虚拟物品类型、各平台已发放次数、各平台核销率，所述用户消费数据包括下述至少之一：预设时间内各平台的总消费次数、预设时间内各平台的总消费金额、领取前的最后消费时间。Optionally, the feature data includes at least: user attribute data, virtual item attribute data, and user consumption data, wherein the user attribute data includes at least one of the following: gender, age, area code, financial institution customer type, An employee logo of a financial institution, the virtual item attribute data includes at least one of the following: the type of virtual item, the number of times each platform has issued, and the write-off rate of each platform, and the user consumption data includes at least one of the following: within a preset time The total consumption times of each platform, the total consumption amount of each platform within the preset time, and the last consumption time before claiming.
可选地，所述扩展单元包括：第二获取模块，用于基于所述预设联邦策略，获取多个外部机构的外部特征数据，其中，所述外部特征数据包括下述至少之一：用户行为数据、用户兴趣数据、标识映射数据、设备信息、网络特征数据；第一筛选模块，用于基于预设特征维度，从所述外部特征数据中筛选目标数据；第一扩展模块，用于基于所述目标数据，扩展所述特征数据，得到所述目标特征数据。Optionally, the expansion unit includes: a second acquisition module, configured to acquire external feature data of multiple external institutions based on the preset federal policy, wherein the external feature data includes at least one of the following: a user Behavior data, user interest data, identity mapping data, device information, and network feature data; a first screening module for screening target data from the external feature data based on preset feature dimensions; a first extension module for screening target data based on For the target data, expand the feature data to obtain the target feature data.
可选地，所述计算单元包括：第二确定模块，用于采用所述目标分类模型，确定所述多个账户集合中各金融账户在领完虚拟物品的第一预设时间段内是否进行绑卡，得到确定结果；第一统计模块，用于基于所述确定结果，统计所述多个账户集合在所述第一预设时间段内的绑卡量；第一计算模块，用于基于所述绑卡量，计算所述多个账户集合的所述绑卡率。Optionally, the computing unit includes: a second determination module, configured to use the target classification model to determine whether each financial account in the multiple account sets has completed receiving the virtual item within the first preset time period. Binding a card to obtain a determination result; a first statistics module for counting the amount of card binding of the multiple account sets within the first preset time period based on the determination result; a first calculation module for based on the For the amount of card binding, the card binding rate of the multiple account sets is calculated.
可选地，所述计算单元还包括：第三确定模块，用于采用所述目标预测模型，确定所述多个账户集合中各金融账户在领完虚拟物品的第二预设时间段内的第一消费次数，并确定各所述金融账户在领取虚拟物品前的第三预设时间段内的第二消费次数；第二计算模块，用于基于所述第一消费次数和所述第二消费次数，计算所述多个账户集合的所述消费提升率。Optionally, the computing unit further includes: a third determination module, configured to use the target prediction model to determine the financial accounts in the multiple account sets within the second preset time period after the virtual item has been collected. the first consumption times, and determine the second consumption times of each of the financial accounts within the third preset time period before receiving the virtual item; a second calculation module is configured to use the first consumption times and the second consumption times The consumption times, and the consumption promotion rate of the multiple account sets is calculated.
根据本发明实施例的另一方面，还提供了一种计算机可读存储介质，所述计算机可读存储介质包括存储的计算机程序，其中，在所述计算机程序运行时控制所述计算机可读存储介质所在设备执行上述所述的基于联邦学习的账户集合确定方法。According to another aspect of the embodiments of the present invention, there is also provided a computer-readable storage medium, the computer-readable storage medium comprising a stored computer program, wherein the computer-readable storage medium is controlled when the computer program is executed The device where the medium is located executes the above-mentioned method for determining an account set based on federated learning.
根据本发明实施例的另一方面，还提供了一种电子设备，包括一个或多个处理器和存储器，所述存储器用于存储一个或多个程序，其中，当所述一个或多个程序被所述一个或多个处理器执行时，使得所述一个或多个处理器实现上述所述的基于联邦学习的账户集合确定方法。According to another aspect of the embodiments of the present invention, there is also provided an electronic device, comprising one or more processors and a memory, the memory being used to store one or more programs, wherein when the one or more programs are When executed by the one or more processors, the one or more processors are caused to implement the above-mentioned method for determining an account set based on federated learning.
在本公开中，获取目标金融机构的特征数据，并基于特征数据，构建初始分类模型和初始预测模型，基于预设联邦策略，扩展特征数据，得到目标特征数据，基于目标特征数据，调整初始分类模型和初始预测模型，得到目标分类模型和目标预测模型，基于目标分类模型，计算多个账户集合的绑卡率，并基于目标预测模型，计算多个账户集合的消费提升率，确定绑卡率大于第一预设阈值且消费提升率大于第二预设阈值的账户集合为目标账户集合。在本申请中，可以在获取到的目标金融机构的特征数据的基础上，采用预设联邦策略，扩展该特征数据，通过扩展的目标特征数据，调整初始分类模型和初始预测模型，并通过调整后的目标分类模型和目标预测模型，分别计算多个账户集合的绑卡率以及消费提升率，将绑卡率大于第一预设阈值且消费提升率大于第二预设阈值的账户集合作为发放虚拟物品的目标账户集合，能够确定出虚拟物品最适合发放的用户群体，从而可以有效提升用户活跃度，提高用户粘性，进而解决了相关技术中无法确定合适的用户群体，导致虚拟物品发放后，投入产出比不高，难以提升用户活跃度以及难以长期对用户进行留存的技术问题。In the present disclosure, the characteristic data of the target financial institution is obtained, and based on the characteristic data, an initial classification model and an initial prediction model are constructed, based on a preset federal strategy, the characteristic data is expanded to obtain the target characteristic data, and the initial classification is adjusted based on the target characteristic data. Model and initial prediction model, obtain the target classification model and target prediction model, calculate the card binding rate of multiple account sets based on the target classification model, and calculate the consumption increase rate of multiple account sets based on the target prediction model, and determine the card binding rate The set of accounts that is greater than the first preset threshold and whose consumption increase rate is greater than the second preset threshold is the set of target accounts. In this application, on the basis of the acquired characteristic data of the target financial institution, a preset federal strategy can be adopted to expand the characteristic data, and the initial classification model and the initial prediction model can be adjusted through the expanded target characteristic data, and by adjusting The following target classification model and target prediction model are used to calculate the card binding rate and consumption increase rate of multiple account sets respectively, and the account set with the card binding rate greater than the first preset threshold and the consumption increase rate greater than the second preset threshold is used as the issuance. The set of target accounts of virtual items can determine the user group that is most suitable for issuing virtual items, which can effectively increase user activity and user stickiness, thus solving the problem of inability to determine the appropriate user group in related technologies, resulting in virtual items after distribution. The input-output ratio is not high, it is difficult to increase user activity and it is difficult to retain users for a long time.
附图说明Description of drawings
此处所说明的附图用来提供对本发明的进一步理解，构成本申请的一部分，本发明的示意性实施例及其说明用于解释本发明，并不构成对本发明的不当限定。在附图中：The accompanying drawings described herein are used to provide a further understanding of the present invention and constitute a part of the present application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the attached image:
图1是根据本发明实施例的一种可选的基于联邦学习的账户集合确定方法的流程图；1 is a flowchart of an optional federated learning-based account set determination method according to an embodiment of the present invention;
图2是根据本发明实施例的一种可选的基于联邦学习的账户集合确定装置的示意图；2 is a schematic diagram of an optional federated learning-based account set determination device according to an embodiment of the present invention;
图3是根据本发明实施例的一种用于基于联邦学习的账户集合确定方法的电子设备(或移动设备)的硬件结构框图。FIG. 3 is a block diagram of the hardware structure of an electronic device (or mobile device) for a method for determining an account set based on federated learning according to an embodiment of the present invention.
具体实施方式Detailed ways
为了使本技术领域的人员更好地理解本发明方案，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的实施例，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做出创造性劳动前提下所获得的所有其他实施例，都应当属于本发明保护的范围。In order to make those skilled in the art better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only Embodiments are part of the present invention, but not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.
需要说明的是，本发明的说明书和权利要求书及上述附图中的术语“第一”、“第二”等是用于区别类似的对象，而不必用于描述特定的顺序或先后次序。应该理解这样使用的数据在适当情况下可以互换，以便这里描述的本发明的实施例能够以除了在这里图示或描述的那些以外的顺序实施。此外，术语“包括”和“具有”以及他们的任何变形，意图在于覆盖不排他的包含，例如，包含了一系列步骤或单元的过程、方法、系统、产品或设备不必限于清楚地列出的那些步骤或单元，而是可包括没有清楚地列出的或对于这些过程、方法、产品或设备固有的其它步骤或单元。It should be noted that the terms "first", "second" and the like in the description and claims of the present invention and the above drawings are used to distinguish similar objects, and are not necessarily used to describe a specific sequence or sequence. It is to be understood that the data so used may be interchanged under appropriate circumstances such that the embodiments of the invention described herein can be practiced in sequences other than those illustrated or described herein. Furthermore, the terms "comprising" and "having" and any variations thereof, are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those expressly listed Rather, those steps or units may include other steps or units not expressly listed or inherent to these processes, methods, products or devices.
需要说明的是，本公开中的基于联邦学习的账户集合确定方法及其装置可用于金融科技领域在基于联邦学习对账户集合进行确定的情况下，也可用于除金融科技领域之外的任意领域在基于联邦学习对账户集合进行确定的情况下，本公开中对基于联邦学习的账户集合确定方法及其装置的应用领域不做限定。It should be noted that the method and device for determining an account set based on federated learning in the present disclosure can be used in the field of financial technology. When the account set is determined based on federated learning, it can also be used in any field other than the field of financial technology. In the case of determining the account set based on federated learning, the present disclosure does not limit the application field of the method and device for determining the account set based on federated learning.
需要说明的是，本公开所涉及的相关信息(包括但不限于用户设备信息、用户个人信息等)和数据(包括但不限于用于展示的数据、分析的数据等)，均为经用户授权或者经过各方充分授权的信息和数据。例如，本系统和相关用户或机构间设置有接口，在获取相关信息之前，需要通过接口向前述的用户或机构发送获取请求，并在接收到前述的用户或机构反馈的同意信息后，获取相关信息。It should be noted that the relevant information (including but not limited to user equipment information, user personal information, etc.) and data (including but not limited to display data, analysis data, etc.) involved in this disclosure are authorized by the user or information and data fully authorized by the parties. For example, there is an interface between the system and relevant users or institutions. Before obtaining relevant information, it is necessary to send an acquisition request to the aforementioned users or institutions through the interface, and after receiving the consent information fed back by the aforementioned users or institutions, obtain relevant information. information.
本发明下述各实施例可应用于各种基于联邦学习对账户集合进行确定的系统/应用/设备中。本发明可以针对虚拟物品(即福利物品，例如，优惠券等)智能创建账户集合，在传统金融机构已有的金融类数据基础上，可以采用联邦学习，纵向补充其它机构的互联网数据(例如，绑了其他卡用户且高消费、用户活跃度高、社交影响力强(喜欢分享)等)，优化模型，从而能够预测出虚拟物品最适合发放的账户集合，以提升用户活跃度以及最大程度的激活用户。The following embodiments of the present invention can be applied to various systems/applications/devices for determining account sets based on federated learning. The present invention can intelligently create account sets for virtual items (that is, welfare items, such as coupons, etc.), and based on the existing financial data of traditional financial institutions, federated learning can be used to supplement the Internet data of other institutions vertically (for example, Binding other card users with high consumption, high user activity, strong social influence (like sharing), etc.), optimize the model, so as to predict the account set that is most suitable for issuing virtual items, so as to increase user activity and maximize user activity. Activate user.
下面结合各个实施例来详细说明本发明。The present invention will be described in detail below with reference to each embodiment.
实施例一Example 1
根据本发明实施例，提供了一种基于联邦学习的账户集合确定方法实施例，需要说明的是，在附图的流程图示出的步骤可以在诸如一组计算机可执行指令的计算机系统中执行，并且，虽然在流程图中示出了逻辑顺序，但是在某些情况下，可以以不同于此处的顺序执行所示出或描述的步骤。According to an embodiment of the present invention, an embodiment of a method for determining an account set based on federated learning is provided. It should be noted that the steps shown in the flowchart of the accompanying drawing can be executed in a computer system such as a set of computer-executable instructions. and, although a logical order is shown in the flowcharts, in some cases the steps shown or described may be performed in an order different from that herein.
图1是根据本发明实施例的一种可选的基于联邦学习的账户集合确定方法的流程图，如图1所示，该方法包括如下步骤：FIG. 1 is a flowchart of an optional method for determining an account set based on federated learning according to an embodiment of the present invention. As shown in FIG. 1 , the method includes the following steps:
步骤S101，获取目标金融机构的特征数据，并基于特征数据，构建初始分类模型和初始预测模型。In step S101, characteristic data of the target financial institution is acquired, and based on the characteristic data, an initial classification model and an initial prediction model are constructed.
步骤S102，基于预设联邦策略，扩展特征数据，得到目标特征数据。Step S102, based on the preset federation strategy, expand the feature data to obtain target feature data.
步骤S103，基于目标特征数据，调整初始分类模型和初始预测模型，得到目标分类模型和目标预测模型。Step S103, based on the target feature data, adjust the initial classification model and the initial prediction model to obtain the target classification model and the target prediction model.
步骤S104，基于目标分类模型，计算多个账户集合的绑卡率，并基于目标预测模型，计算多个账户集合的消费提升率。Step S104: Calculate the card binding rates of the multiple account sets based on the target classification model, and calculate the consumption promotion rates of the multiple account sets based on the target prediction model.
步骤S105，确定绑卡率大于第一预设阈值且消费提升率大于第二预设阈值的账户集合为目标账户集合，其中，目标金融机构对目标账户集合中的每个金融账户发放虚拟物品。Step S105: Determine the account set with the card binding rate greater than the first preset threshold and the consumption increase rate greater than the second preset threshold as the target account set, wherein the target financial institution issues virtual items to each financial account in the target account set.
通过上述步骤，可以获取目标金融机构的特征数据，并基于特征数据，构建初始分类模型和初始预测模型，基于预设联邦策略，扩展特征数据，得到目标特征数据，基于目标特征数据，调整初始分类模型和初始预测模型，得到目标分类模型和目标预测模型，基于目标分类模型，计算多个账户集合的绑卡率，并基于目标预测模型，计算多个账户集合的消费提升率，确定绑卡率大于第一预设阈值且消费提升率大于第二预设阈值的账户集合为目标账户集合。在本发明实施例中，可以在获取到的目标金融机构的特征数据的基础上，采用预设联邦策略，扩展该特征数据，通过扩展的目标特征数据，调整初始分类模型和初始预测模型，并通过调整后的目标分类模型和目标预测模型，分别计算多个账户集合的绑卡率以及消费提升率，将绑卡率大于第一预设阈值且消费提升率大于第二预设阈值的账户集合作为发放虚拟物品的目标账户集合，能够确定出虚拟物品最适合发放的用户群体，从而可以有效提升用户活跃度，提高用户粘性，进而解决了相关技术中无法确定合适的用户群体，导致虚拟物品发放后，投入产出比不高，难以提升用户活跃度以及难以长期对用户进行留存的技术问题。Through the above steps, the characteristic data of the target financial institution can be obtained, and based on the characteristic data, an initial classification model and an initial prediction model can be constructed, and based on the preset federal strategy, the characteristic data can be expanded to obtain the target characteristic data, and the initial classification can be adjusted based on the target characteristic data. Model and initial prediction model, obtain the target classification model and target prediction model, calculate the card binding rate of multiple account sets based on the target classification model, and calculate the consumption increase rate of multiple account sets based on the target prediction model, and determine the card binding rate The set of accounts that is greater than the first preset threshold and whose consumption increase rate is greater than the second preset threshold is the set of target accounts. In the embodiment of the present invention, on the basis of the acquired characteristic data of the target financial institution, a preset federal strategy may be adopted to expand the characteristic data, and the initial classification model and the initial prediction model may be adjusted through the expanded target characteristic data, and the Through the adjusted target classification model and target prediction model, the card binding rate and consumption increase rate of multiple account sets are calculated respectively, and the account set with the card binding rate greater than the first preset threshold and the consumption increase rate greater than the second preset threshold is calculated. As a set of target accounts for issuing virtual items, it can determine the user group that is most suitable for issuing virtual items, which can effectively increase user activity and user stickiness, and solve the problem that the appropriate user group cannot be determined in related technologies, resulting in virtual item distribution. After that, the input-output ratio is not high, it is difficult to increase user activity and it is difficult to retain users for a long time.
下面结合上述各步骤对本发明实施例进行详细说明。The embodiments of the present invention will be described in detail below with reference to the above steps.
在本发明实施例中，一种可选的，在获取目标金融机构的特征数据之前，还包括：获取预设历史时间段内各个金融账户的账户数据和虚拟物品数据；基于账户数据和虚拟物品数据，采用预设分类策略，确定金融账户所属的账户集合。In the embodiment of the present invention, an optional, before acquiring the characteristic data of the target financial institution, further comprising: acquiring account data and virtual item data of each financial account within a preset historical time period; based on the account data and virtual items Data, using a preset classification strategy to determine the account set to which the financial account belongs.
在本发明实施例中，可以先获取预设历史时间段内(例如，过去一年内)各个金融账户的账户数据(例如，用户基本属性数据、用户消费数据等)和虚拟物品数据(例如，虚拟物品属性数据等)，基于这些账户数据和虚拟物品数据，采用预设分类策略(例如，采用机器学习训练的分类模型)，确定金融账户所属的账户集合，该账户集合可以包括：有经常使用其他金融机构卡支付的习惯但未用本金融机构卡的账户集合、有社交属性(即愿意分享给其他人)的账户集合、之前本金融机构卡，发虚拟物品后会绑本金融机构卡的账户集合、之前绑了本金融机构卡，但未使用本金融机构卡消费的账户集合等。In this embodiment of the present invention, account data (for example, user basic attribute data, user consumption data, etc.) and virtual item data (for example, virtual item data (for example, virtual item data) of each financial account within a preset historical time period (for example, within the past year) may be acquired first. Item attribute data, etc.), based on these account data and virtual item data, use a preset classification strategy (for example, a classification model trained by machine learning) to determine the account set to which the financial account belongs. The account set may include: The habit of paying by financial institution card but not using the account collection of the financial institution card, the account collection with social attributes (that is, willing to share with others), the previous financial institution card, and the account of the financial institution card will be bound after virtual items are issued Collection, collection of accounts that have been bound with the card of this financial institution before, but have not used the card of this financial institution for consumption, etc.
步骤S101，获取目标金融机构的特征数据，并基于特征数据，构建初始分类模型和初始预测模型。可选的，特征数据至少包括：用户属性数据、虚拟物品属性数据、用户消费数据，其中，用户属性数据包括下述至少之一：性别、年龄、地区号、金融机构客户类型、金融机构员工标志，虚拟物品属性数据包括下述至少之一：虚拟物品类型、各平台已发放次数、各平台核销率，用户消费数据包括下述至少之一：预设时间内各平台的总消费次数、预设时间内各平台的总消费金额、领取前的最后消费时间。In step S101, characteristic data of the target financial institution is acquired, and based on the characteristic data, an initial classification model and an initial prediction model are constructed. Optionally, the feature data includes at least: user attribute data, virtual item attribute data, and user consumption data, wherein the user attribute data includes at least one of the following: gender, age, area code, financial institution customer type, financial institution employee logo , the virtual item attribute data includes at least one of the following: the type of virtual item, the number of times each platform has been issued, the write-off rate of each platform, and the user consumption data includes at least one of the following: The total consumption amount of each platform within the set time, and the last consumption time before collection.
在本发明实施例中，可以先获取目标金融机构的特征数据，并基于特征数据，构建初始分类模型(即绑卡预测二分类模型，用于预测用户在领完虚拟物品后是否进行绑卡，并可以设定当Y＝0表示领完虚拟物品(如各平台优惠券)后未在相应平台上绑卡；Y＝1表示领完虚拟物品后在相应平台上绑了卡)，以及构建初始预测模型(即消费次数提升预测回归模型，用于预测用户在领完虚拟物品的后预设时间内(如一个月内)相对于前预设时间内(如一个月内)的消费次数提升比值)。In the embodiment of the present invention, the characteristic data of the target financial institution may be obtained first, and based on the characteristic data, an initial classification model (that is, a card binding prediction two-classification model, which is used to predict whether the user will bind a card after receiving the virtual item, And can be set when Y=0 means that the card is not bound on the corresponding platform after receiving the virtual item (such as coupons from various platforms); Y=1 means that the card is bound on the corresponding platform after receiving the virtual item), and build the initial The prediction model (that is, the prediction regression model for the increase in consumption times) is used to predict the ratio of the increase in the consumption times of the user within the preset time period after receiving the virtual item (such as within one month) relative to the previous preset time period (such as within one month). ).
本实施例中，特征数据可以包括：用户属性数据、虚拟物品属性数据、用户消费数据等，其中，用户属性数据可以包括：性别、年龄、地区号、金融机构客户类型(例如，持有的金融机构卡类型等)、金融机构员工标志、职业代码、教育程度代码、单位性质代码、婚姻状况代码、行政级别代码、归属单位种类代码、金融机构客户星级、净资产等用户基本属性数据，表1为本实施例一种可选的用户属性数据，如表1所示：In this embodiment, the feature data may include: user attribute data, virtual item attribute data, user consumption data, etc., wherein the user attribute data may include: gender, age, area code, financial institution customer type (for example, financial type of institution card, etc.), financial institution employee logo, occupation code, education level code, unit nature code, marital status code, administrative level code, type code of belonging unit, financial institution customer star rating, net assets and other basic user attribute data, table 1 is a kind of optional user attribute data in this embodiment, as shown in Table 1:
表1Table 1
虚拟物品属性数据可以包括：虚拟物品类型、各平台已发放次数、各平台核销率等，表2为本实施例一种可选的虚拟物品属性数据，如表2所示：The virtual item attribute data may include: the virtual item type, the number of times each platform has issued, the write-off rate of each platform, etc. Table 2 is an optional virtual item attribute data in this embodiment, as shown in Table 2:
表2Table 2
用户消费数据可以包括：预设时间内各平台的总消费次数、预设时间内各平台的总消费金额、领取前的最后消费时间等，表3为本实施例一种可选的用户消费数据，如表3所示：The user consumption data may include: the total consumption times of each platform within the preset time, the total consumption amount of each platform within the preset time, the last consumption time before receiving, etc. Table 3 is an optional user consumption data in this embodiment. ,as shown in Table 3:
表3table 3
另一种可选的，用于构建初始预测模型的用户消费数据还可以包括：领取前在各平台上的绑卡状态、领取前已在各平台领取的个数、领取前已在所有平台领取的总数，表4为本实施例另一种可选的用户消费数据，如表4所示：Alternatively, the user consumption data used to construct the initial prediction model may also include: the card binding status on each platform before the collection, the number of cards that have been collected on each platform before the collection, and the number of cards that have been collected on all platforms before the collection Table 4 is another optional user consumption data of this embodiment, as shown in Table 4:
表4Table 4
步骤S102，基于预设联邦策略，扩展特征数据，得到目标特征数据。Step S102, based on the preset federation strategy, expand the feature data to obtain target feature data.
可选的，基于预设联邦策略，扩展特征数据，得到目标特征数据的步骤，包括：基于预设联邦策略，获取多个外部机构的外部特征数据，其中，外部特征数据包括下述至少之一：用户行为数据、用户兴趣数据、标识映射数据、设备信息、网络特征数据；基于预设特征维度，从外部特征数据中筛选目标数据；基于目标数据，扩展特征数据，得到目标特征数据。Optionally, based on the preset federal strategy, the step of extending the feature data to obtain the target feature data includes: based on the preset federal strategy, acquiring external feature data of multiple external institutions, wherein the external feature data includes at least one of the following : User behavior data, user interest data, identity mapping data, device information, and network feature data; based on preset feature dimensions, filter target data from external feature data; based on target data, expand feature data to obtain target feature data.
在本发明实施例中，可以通过联邦建模纵向扩展外部互联网机构提供的客户互联网行为标签(即基于预设联邦策略，获取多个外部机构的外部特征数据)，之后，可以基于预设特征维度(如金融类特征维度，包括：证券、借贷、保险、银行、基金、期货等)，从外部特征数据中筛选目标数据，在目标金融机构已有的特征数据基础上，扩展得到目标特征数据(即基于目标数据，扩展特征数据，得到目标特征数据)，然后，可以基于扩展得到的目标特征数据，通过多种建模方式，能够深度洞察用户消费需求，为目标金融机构发放的虚拟物品匹配更精准的用户群体。In this embodiment of the present invention, the customer Internet behavior labels provided by external Internet agencies can be extended vertically through federated modeling (that is, based on a preset federal policy, external feature data of multiple external agencies is acquired), and then, based on preset feature dimensions (For example, the dimension of financial features, including: securities, lending, insurance, banks, funds, futures, etc.), filter the target data from external feature data, and expand the target feature data based on the existing feature data of the target financial institution ( That is, based on the target data, expand the feature data to obtain the target feature data), and then, based on the expanded target feature data, through a variety of modeling methods, it is possible to deeply understand the consumer demand of users, and match the virtual items issued by the target financial institutions. precise user groups.
本实施例中，该外部特征数据可以包括：用户行为数据(例如，用户180天内应用app使用情况统计，主要包括app多样性统计特征、频次统计特征、行为变化统计特征等)、用户兴趣数据(可以根据用户访问和安装行为挖掘用户兴趣标签)、标识映射数据(即关联到的全网标识信息，包括：IMEI(International Mobile Equipment Identity，国际移动设备身份码)、IDFA(Identifier For Advertising，iOS独有的广告标识符)、cookie(储存在用户本地终端上的数据)等)、设备信息(包括：操作系统、机型、品牌、价格、设备评级等)、网络特征数据(即基于深度神经网络DNN，并输出中间层作为特征数据)，此外，外部特征数据还可以包括：办公类数据、效率类数据、阅读类数据、拍照美化类数据、影音娱乐类数据、社交通讯类数据、购物比价类数据、新闻资讯类数据、网赚类数据(如，区块链网赚，兼职网赚，新闻网赚，运动网赚，游戏网赚，视频网赚等)、便捷生活类数据、汽车类数据、出行导航类数据、旅游住宿类数据、金融投资类数据、金融借贷类数据等。In this embodiment, the external feature data may include: user behavior data (for example, the user's application statistics within 180 days, mainly including app diversity statistics, frequency statistics, behavior change statistics, etc.), user interest data ( User interest tags can be mined according to user access and installation behavior), identification mapping data (that is, the associated network-wide identification information, including: IMEI (International Mobile Equipment Identity, International Mobile Equipment Identity), IDFA (Identifier For Advertising, iOS unique) some advertising identifiers), cookies (data stored on the user’s local terminal), etc.), device information (including: operating system, model, brand, price, device rating, etc.), network feature data (that is, based on deep neural network DNN, and output the middle layer as feature data), in addition, the external feature data can also include: office data, efficiency data, reading data, photo beautification data, audio-visual entertainment data, social communication data, shopping comparison data Data, news data, online earning data (such as blockchain online earning, part-time online earning, news online earning, sports online earning, game online earning, video online earning, etc.), convenient life data, automobile data , travel navigation data, travel accommodation data, financial investment data, financial lending data, etc.
步骤S103，基于目标特征数据，调整初始分类模型和初始预测模型，得到目标分类模型和目标预测模型。Step S103, based on the target feature data, adjust the initial classification model and the initial prediction model to obtain the target classification model and the target prediction model.
在本发明实施例中，可以通过扩展后的目标特征数据，调整初始分类模型和初始预测模型，以得到目标分类模型和目标预测模型，其中，通过目标分类模型和目标预测模型确定目标金融机构可以发放虚拟物品的账户集合，能够有效提升金融账户的绑卡率(即增加在第三方平台上绑本金融机构的任意种类卡的数量)以及消费笔数。In the embodiment of the present invention, the initial classification model and the initial prediction model can be adjusted through the expanded target feature data to obtain the target classification model and the target prediction model, wherein the target financial institution can be determined through the target classification model and the target prediction model. The collection of accounts that issue virtual items can effectively increase the card binding rate of financial accounts (that is, increase the number of cards of any type bound to the financial institution on the third-party platform) and the number of consumption transactions.
本实施例中，表5为本实施例一种可选的仅采用本金融机构的特征数据构建的初始分类模型的建模效果，如表5所示：In this embodiment, Table 5 shows the modeling effect of an optional initial classification model constructed by using only the characteristic data of this financial institution in this embodiment, as shown in Table 5:
表5table 5
本实施例中，表6为本实施例一种可选的仅采用本金融机构的特征数据构建的初始预测模型的建模效果，如表6所示：In this embodiment, Table 6 shows the modeling effect of an optional initial prediction model constructed by using only the characteristic data of this financial institution in this embodiment, as shown in Table 6:
表6Table 6
本实施例中，表7为本实施例一种可选的联邦建模后的目标分类模型的提升效果，如表7所示：In this embodiment, Table 7 shows the improvement effect of an optional federated modeling target classification model in this embodiment, as shown in Table 7:
表7Table 7
本实施例中，表8为本实施例一种可选的联邦建模后的目标预测模型的提升效果，如表8所示：In this embodiment, Table 8 shows the improvement effect of an optional federated modeling target prediction model in this embodiment, as shown in Table 8:
表8Table 8
可选的，在基于目标特征数据，调整初始分类模型和初始预测模型，得到目标分类模型和目标预测模型之后，还包括：基于目标分类模型和目标预测模型，调整账户集合，其中，调整后的账户集合包括下述至少之一：使用其他金融机构卡但未用目标金融机构卡的账户集合、具有社交属性的账户集合、发放虚拟物品后绑定目标金融机构卡的账户集合、未使用目标金融机构卡进行消费的账户集合。Optionally, after adjusting the initial classification model and the initial prediction model based on the target feature data to obtain the target classification model and the target prediction model, the method further includes: adjusting the account set based on the target classification model and the target prediction model, wherein the adjusted The account set includes at least one of the following: an account set that uses the card of another financial institution but does not use the target financial institution card, an account set with social attributes, an account set bound to the target financial institution card after issuing the virtual item, and the target financial institution card is not used. A collection of accounts for consumption by institutional cards.
在本发明实施例中，联邦建模能够提升仅使用本金融机构内的特征数据建出的模型效果，还能够调整账户集合(即基于目标分类模型和目标预测模型，调整账户集合)，调整后的账户集合可以包括：有经常使用其他金融机构卡支付的习惯但未用本金融机构卡的账户集合(即使用其他金融机构卡但未用目标金融机构卡的账户集合)；有社交属性(即愿意分享给其他人)的账户集合；之前未绑本金融机构行的卡，发券后会绑本金融机构卡的账户集合(即发放虚拟物品后绑定目标金融机构卡的账户集合)；之前绑了本金融机构卡，但目前未使用本金融机构卡消费的账户集合(未使用目标金融机构卡进行消费的账户集合)。In the embodiment of the present invention, federated modeling can improve the effect of a model built using only the feature data in the financial institution, and can also adjust the account set (that is, adjust the account set based on the target classification model and the target prediction model). The set of accounts can include: the set of accounts that have the habit of using other financial institutions' cards to pay but do not use the card of this financial institution (that is, the set of accounts that use the card of other financial institutions but not the card of the target financial institution); have social attributes (ie The set of accounts that are willing to share with others); the set of accounts that have not been bound to the bank of this financial institution before, will be bound to the set of accounts of the card of this financial institution after the issuance of the coupons (that is, the set of accounts bound to the card of the target financial institution after issuing the virtual item); The set of accounts that are bound to the card of this financial institution, but are not currently used for consumption with the card of this financial institution (the collection of accounts that do not use the card of the target financial institution for consumption).
步骤S104，基于目标分类模型，计算多个账户集合的绑卡率，并基于目标预测模型，计算多个账户集合的消费提升率。Step S104: Calculate the card binding rates of the multiple account sets based on the target classification model, and calculate the consumption promotion rates of the multiple account sets based on the target prediction model.
可选的，基于目标分类模型，计算多个账户集合的绑卡率的步骤，包括：采用目标分类模型，确定多个账户集合中各金融账户在领完虚拟物品的第一预设时间段内是否进行绑卡，得到确定结果；基于确定结果，统计多个账户集合在第一预设时间段内的绑卡量；基于绑卡量，计算多个账户集合的绑卡率。Optionally, based on the target classification model, the step of calculating the card binding rate of the multiple account sets includes: using the target classification model to determine that each financial account in the multiple account sets is within a first preset time period after the virtual item has been collected. Whether to perform card binding, a determination result is obtained; based on the determination result, the number of card bindings of multiple account sets within the first preset time period is counted; based on the number of card bindings, the card binding rate of multiple account sets is calculated.
在本发明实施例中，可以采用目标分类模型，确定多个账户集合中各金融账户在各第三方平台上领完虚拟物品(如，优惠券)的第一预设时间段内(如，一周内)是否进行绑卡，从而得到确定结果，本实施例中的目标分类模型可以是XGboost(eXtreme GradientBoosting，极致梯度提升)二分类模型，之后，可以基于确定结果，统计多个账户集合在第一预设时间段内的绑卡量，通过绑卡量，计算多个账户集合的绑卡率。In this embodiment of the present invention, a target classification model may be used to determine the first preset time period (eg, one week) within which each financial account in the multiple account sets has received virtual items (eg, coupons) on each third-party platform. In order to obtain a determination result, the target classification model in this embodiment may be an XGboost (eXtreme GradientBoosting, extreme gradient boosting) two-classification model. The number of card bindings within a preset time period, through which the binding rate of multiple accounts is calculated.
可选的，基于目标预测模型，计算多个账户集合的消费提升率的步骤，包括：采用目标预测模型，确定多个账户集合中各金融账户在领完虚拟物品的第二预设时间段内的第一消费次数，并确定各金融账户在领取虚拟物品前的第三预设时间段内的第二消费次数；基于第一消费次数第一消费次数和第二消费次数，计算多个账户集合的消费提升率。Optionally, based on the target prediction model, the step of calculating the consumption improvement rate of the multiple account sets includes: using the target prediction model to determine that each financial account in the multiple account sets is within a second preset time period after the virtual item has been collected. and determine the second consumption times of each financial account within the third preset time period before receiving the virtual item; based on the first consumption times, the first consumption times and the second consumption times, calculate a set of multiple accounts consumption growth rate.
在本发明实施例中，采用目标预测模型(如，XGboost回归模型)，在多个账户集合中各金融账户在各第三方平台上领完虚拟物品后，可以确定在第二预设时间段内(如，一月内)第一消费次数，以及各金融账户在领取虚拟物品前的第三预设时间段内(如，一月内)的第二消费次数，通过第一消费次数第一消费次数和第二消费次数，计算领取后在第二预设时间段内在各平台上第一消费次数与领取前在第三预设时间段内在各平台上第二消费次数)的差值，并计算差值与领取前在第三预设时间段内在各平台上第二消费次数加上预设次数(如，1)的比值，得到多个账户集合的消费提升率。In the embodiment of the present invention, a target prediction model (eg, XGboost regression model) is used, and after each financial account in the multiple account sets has received the virtual item on each third-party platform, it can be determined that within the second preset time period (eg, within a month) the first consumption count, and the second consumption count of each financial account within the third preset time period (eg, within a month) before receiving the virtual item, the first consumption through the first consumption count number of times and the number of times of second consumption, calculate the difference between the number of times of first consumption on each platform within the second preset time period after collection and the number of times of second consumption on each platform within the third preset time period before collection), and calculate The ratio of the difference to the second number of consumptions on each platform in the third preset time period before the collection plus the preset number of times (eg, 1), to obtain the consumption promotion rate of the multiple account sets.
步骤S105，确定绑卡率大于第一预设阈值且消费提升率大于第二预设阈值的账户集合为目标账户集合，其中，目标金融机构对目标账户集合中的每个金融账户发放虚拟物品。Step S105: Determine the account set with the card binding rate greater than the first preset threshold and the consumption increase rate greater than the second preset threshold as the target account set, wherein the target financial institution issues virtual items to each financial account in the target account set.
在本发明实施例中，找到绑卡率大于第一预设阈值(根据实际情况进行设置)且消费提升率大于第二预设阈值(根据实际情况进行设置)的账户集合，将该账户集合作为目标账户集合，以使目标金融机构对目标账户集合中的每个金融账户发放虚拟物品。In the embodiment of the present invention, an account set with a card binding rate greater than a first preset threshold (set according to the actual situation) and a consumption increase rate greater than a second preset threshold (set according to the actual situation) is found, and the account set is used as A set of target accounts to enable the target financial institution to issue virtual items to each financial account in the set of target accounts.
本发明实施例中，可以采用联邦建模，纵向补充其它机构的互联网数据进行联邦学习，建立目标分类模型和目标预测模型，以确定更适合发放虚拟物品的用户群体，能够提升用户活跃度以及激活用户，增加用户粘度。In the embodiment of the present invention, federated modeling can be used to vertically supplement the Internet data of other institutions for federated learning, and a target classification model and target prediction model can be established to determine a user group that is more suitable for issuing virtual items, which can improve user activity and activation. users, increase user stickiness.
实施例二Embodiment 2
本实施例中提供的一种基于联邦学习的账户集合确定装置包含了多个实施单元，每个实施单元对应于上述实施例一中的各个实施步骤。An apparatus for determining an account set based on federated learning provided in this embodiment includes multiple implementation units, and each implementation unit corresponds to each implementation step in the above-mentioned first embodiment.
图2是根据本发明实施例的一种可选的基于联邦学习的账户集合确定装置的示意图，如图2所示，该确定装置可以包括：获取单元20，扩展单元21，调整单元22，计算单元23，确定单元24，其中，FIG. 2 is a schematic diagram of an optional federated learning-based account set determination device according to an embodiment of the present invention. As shown in FIG. 2 , the determination device may include: an acquisition unit 20 , an expansion unit 21 , an adjustment unit 22 , and a calculation unit 20 . unit 23, determining unit 24, wherein,
获取单元20，用于获取目标金融机构的特征数据，并基于特征数据，构建初始分类模型和初始预测模型；The obtaining unit 20 is used to obtain the characteristic data of the target financial institution, and based on the characteristic data, construct an initial classification model and an initial prediction model;
扩展单元21，用于基于预设联邦策略，扩展特征数据，得到目标特征数据；The expansion unit 21 is used to expand the feature data based on the preset federation strategy to obtain the target feature data;
调整单元22，用于基于目标特征数据，调整初始分类模型和初始预测模型，得到目标分类模型和目标预测模型；Adjustment unit 22, for adjusting the initial classification model and the initial prediction model based on the target feature data to obtain the target classification model and the target prediction model;
计算单元23，用于基于目标分类模型，计算多个账户集合的绑卡率，并基于目标预测模型，计算多个账户集合的消费提升率；The calculation unit 23 is used to calculate the card binding rate of the multiple account sets based on the target classification model, and calculate the consumption promotion rate of the multiple account sets based on the target prediction model;
确定单元24，用于确定绑卡率大于第一预设阈值且消费提升率大于第二预设阈值的账户集合为目标账户集合，其中，目标金融机构对目标账户集合中的每个金融账户发放虚拟物品。The determining unit 24 is configured to determine the set of accounts with the card binding rate greater than the first preset threshold and the consumption increase rate greater than the second preset threshold as the set of target accounts, wherein the target financial institution issues payment to each financial account in the set of target accounts Virtual item.
上述确定装置，可以通过获取单元20获取目标金融机构的特征数据，并基于特征数据，构建初始分类模型和初始预测模型，通过扩展单元21基于预设联邦策略，扩展特征数据，得到目标特征数据，通过调整单元22基于目标特征数据，调整初始分类模型和初始预测模型，得到目标分类模型和目标预测模型，通过计算单元23基于目标分类模型，计算多个账户集合的绑卡率，并基于目标预测模型，计算多个账户集合的消费提升率，通过确定单元24确定绑卡率大于第一预设阈值且消费提升率大于第二预设阈值的账户集合为目标账户集合。在本发明实施例中，可以在获取到的目标金融机构的特征数据的基础上，采用预设联邦策略，扩展该特征数据，通过扩展的目标特征数据，调整初始分类模型和初始预测模型，并通过调整后的目标分类模型和目标预测模型，分别计算多个账户集合的绑卡率以及消费提升率，将绑卡率大于第一预设阈值且消费提升率大于第二预设阈值的账户集合作为发放虚拟物品的目标账户集合，能够确定出虚拟物品最适合发放的用户群体，从而可以有效提升用户活跃度，提高用户粘性，进而解决了相关技术中无法确定合适的用户群体，导致虚拟物品发放后，投入产出比不高，难以提升用户活跃度以及难以长期对用户进行留存的技术问题。The above determination device can obtain the characteristic data of the target financial institution through the acquisition unit 20, and based on the characteristic data, construct an initial classification model and an initial prediction model, and expand the characteristic data based on the preset federal strategy through the expansion unit 21 to obtain the target characteristic data, The adjustment unit 22 adjusts the initial classification model and the initial prediction model based on the target feature data to obtain the target classification model and the target prediction model, and the calculation unit 23 calculates the binding rate of multiple account sets based on the target classification model, and predicts based on the target The model calculates the consumption increase rate of multiple account sets, and the determination unit 24 determines the account set with the card binding rate greater than the first preset threshold and the consumption increase rate greater than the second preset threshold as the target account set. In the embodiment of the present invention, on the basis of the acquired characteristic data of the target financial institution, a preset federal strategy may be adopted to expand the characteristic data, and the initial classification model and the initial prediction model may be adjusted through the expanded target characteristic data, and the Through the adjusted target classification model and target prediction model, the card binding rate and consumption increase rate of multiple account sets are calculated respectively, and the account set with the card binding rate greater than the first preset threshold and the consumption increase rate greater than the second preset threshold is selected. As a set of target accounts for issuing virtual items, it can determine the user group that is most suitable for issuing virtual items, which can effectively increase user activity and user stickiness, and solve the problem that the appropriate user group cannot be determined in related technologies, resulting in virtual item distribution. After that, the input-output ratio is not high, it is difficult to increase user activity and it is difficult to retain users for a long time.
可选的，确定装置还包括：第一获取模块，用于在获取目标金融机构的特征数据之前，获取预设历史时间段内各个金融账户的账户数据和虚拟物品数据；第一确定模块，用于基于账户数据和虚拟物品数据，采用预设分类策略，确定金融账户所属的账户集合。Optionally, the determining device further includes: a first acquiring module, configured to acquire account data and virtual item data of each financial account within a preset historical time period before acquiring the characteristic data of the target financial institution; the first determining module, using Based on account data and virtual item data, a preset classification strategy is used to determine the account set to which the financial account belongs.
可选的，确定装置还包括：第一调整模块，用于在基于目标特征数据，调整初始分类模型和初始预测模型，得到目标分类模型和目标预测模型之后，基于目标分类模型和目标预测模型，调整账户集合，其中，调整后的账户集合包括下述至少之一：使用其他金融机构卡但未用目标金融机构卡的账户集合、具有社交属性的账户集合、发放虚拟物品后绑定目标金融机构卡的账户集合、未使用目标金融机构卡进行消费的账户集合。Optionally, the determining device further includes: a first adjustment module, configured to adjust the initial classification model and the initial prediction model based on the target feature data, and after obtaining the target classification model and the target prediction model, based on the target classification model and the target prediction model, Adjust the account set, wherein the adjusted account set includes at least one of the following: an account set that uses other financial institution cards but not the target financial institution card, an account set with social attributes, and a target financial institution after issuing virtual items. The account collection of the card, the collection of accounts that do not use the target financial institution card for consumption.
可选的，特征数据至少包括：用户属性数据、虚拟物品属性数据、用户消费数据，其中，用户属性数据包括下述至少之一：性别、年龄、地区号、金融机构客户类型、金融机构员工标志，虚拟物品属性数据包括下述至少之一：虚拟物品类型、各平台已发放次数、各平台核销率，用户消费数据包括下述至少之一：预设时间内各平台的总消费次数、预设时间内各平台的总消费金额、领取前的最后消费时间。Optionally, the feature data includes at least: user attribute data, virtual item attribute data, and user consumption data, wherein the user attribute data includes at least one of the following: gender, age, area code, financial institution customer type, financial institution employee logo , the virtual item attribute data includes at least one of the following: the type of virtual item, the number of times each platform has been issued, the write-off rate of each platform, and the user consumption data includes at least one of the following: The total consumption amount of each platform within the set time, and the last consumption time before collection.
可选的，扩展单元包括：第二获取模块，用于基于预设联邦策略，获取多个外部机构的外部特征数据，其中，外部特征数据包括下述至少之一：用户行为数据、用户兴趣数据、标识映射数据、设备信息、网络特征数据；第一筛选模块，用于基于预设特征维度，从外部特征数据中筛选目标数据；第一扩展模块，用于基于目标数据，扩展特征数据，得到目标特征数据。Optionally, the expansion unit includes: a second acquisition module, configured to acquire external feature data of multiple external institutions based on a preset federal policy, wherein the external feature data includes at least one of the following: user behavior data, user interest data , identify mapping data, device information, and network feature data; the first screening module is used to filter target data from external feature data based on the preset feature dimension; the first expansion module is used to expand the feature data based on the target data, and obtain target feature data.
可选的，计算单元包括：第二确定模块，用于采用目标分类模型，确定多个账户集合中各金融账户在领完虚拟物品的第一预设时间段内是否进行绑卡，得到确定结果；第一统计模块，用于基于确定结果，统计多个账户集合在第一预设时间段内的绑卡量；第一计算模块，用于基于绑卡量，计算多个账户集合的绑卡率。Optionally, the computing unit includes: a second determination module, configured to use a target classification model to determine whether each financial account in the multiple account sets is bound to a card within a first preset time period after the virtual item is received, and obtain a determination result. The first statistical module is used to count the card binding amount of multiple account sets within the first preset time period based on the determination result; the first calculation module is used to calculate the binding card amount of multiple account sets based on the card binding amount Rate.
可选的，计算单元还包括：第三确定模块，用于采用目标预测模型，确定多个账户集合中各金融账户在领完虚拟物品的第二预设时间段内的第一消费次数，并确定各金融账户在领取虚拟物品前的第三预设时间段内的第二消费次数；第二计算模块，用于基于第一消费次数和第二消费次数，计算多个账户集合的消费提升率。Optionally, the computing unit further includes: a third determination module, configured to use the target prediction model to determine the first consumption times of each financial account in the multiple account sets within the second preset time period after the virtual item has been collected, and Determine the second consumption times of each financial account within the third preset time period before receiving the virtual item; the second calculation module is used to calculate the consumption promotion rate of the plurality of account sets based on the first consumption times and the second consumption times .
上述的确定装置还可以包括处理器和存储器，上述获取单元20，扩展单元21，调整单元22，计算单元23，确定单元24等均作为程序单元存储在存储器中，由处理器执行存储在存储器中的上述程序单元来实现相应的功能。The above-mentioned determination device may also include a processor and a memory, and the above-mentioned acquisition unit 20, expansion unit 21, adjustment unit 22, calculation unit 23, determination unit 24, etc. are all stored in the memory as program units, and executed by the processor and stored in the memory. The above program unit to realize the corresponding function.
上述处理器中包含内核，由内核去存储器中调取相应的程序单元。内核可以设置一个或以上，通过调整内核参数来确定绑卡率大于第一预设阈值且消费提升率大于第二预设阈值的账户集合为目标账户集合。The above-mentioned processor includes a kernel, and the corresponding program unit is called from the memory by the kernel. The kernel can be set to one or more accounts, and by adjusting the kernel parameters, it is determined that the account set with the card binding rate greater than the first preset threshold and the consumption increase rate greater than the second preset threshold is the target account set.
上述存储器可能包括计算机可读介质中的非永久性存储器，随机存取存储器(RAM)和/或非易失性内存等形式，如只读存储器(ROM)或闪存(flash RAM)，存储器包括至少一个存储芯片。The above-mentioned memory may include non-persistent memory in computer readable medium, random access memory (RAM) and/or non-volatile memory, such as read only memory (ROM) or flash memory (flash RAM), the memory includes at least a memory chip.
本申请还提供了一种计算机程序产品，当在数据处理设备上执行时，适于执行初始化有如下方法步骤的程序：获取目标金融机构的特征数据，并基于特征数据，构建初始分类模型和初始预测模型，基于预设联邦策略，扩展特征数据，得到目标特征数据，基于目标特征数据，调整初始分类模型和初始预测模型，得到目标分类模型和目标预测模型，基于目标分类模型，计算多个账户集合的绑卡率，并基于目标预测模型，计算多个账户集合的消费提升率，确定绑卡率大于第一预设阈值且消费提升率大于第二预设阈值的账户集合为目标账户集合。The present application also provides a computer program product, which, when executed on a data processing device, is suitable for executing a program initialized with the following method steps: acquiring characteristic data of a target financial institution, and building an initial classification model and an initial classification model based on the characteristic data Prediction model, based on the preset federation strategy, expand the feature data, obtain the target feature data, adjust the initial classification model and the initial prediction model based on the target feature data, obtain the target classification model and the target prediction model, and calculate multiple accounts based on the target classification model The card binding rate of the set is calculated, and based on the target prediction model, the consumption increase rate of multiple account sets is calculated, and the account set with the card binding rate greater than the first preset threshold and the consumption increase rate greater than the second preset threshold is determined as the target account set.
根据本发明实施例的另一方面，还提供了一种计算机可读存储介质，计算机可读存储介质包括存储的计算机程序，其中，在计算机程序运行时控制计算机可读存储介质所在设备执行上述的基于联邦学习的账户集合确定方法。According to another aspect of the embodiments of the present invention, a computer-readable storage medium is also provided, where the computer-readable storage medium includes a stored computer program, wherein when the computer program runs, the device where the computer-readable storage medium is located is controlled to execute the above-mentioned A Federated Learning-Based Account Set Determination Method.
根据本发明实施例的另一方面，还提供了一种电子设备，包括一个或多个处理器和存储器，存储器用于存储一个或多个程序，其中，当一个或多个程序被一个或多个处理器执行时，使得一个或多个处理器实现上述的基于联邦学习的账户集合确定方法。According to another aspect of the embodiments of the present invention, there is also provided an electronic device, comprising one or more processors and a memory, the memory is used for storing one or more programs, wherein when the one or more programs are executed by one or more When executed by each processor, one or more processors are made to implement the above-mentioned method for determining account set based on federated learning.
图3是根据本发明实施例的一种用于基于联邦学习的账户集合确定方法的电子设备(或移动设备)的硬件结构框图。如图3所示，电子设备可以包括一个或多个(图中采用302a、302b，……，302n来示出)处理器302(处理器302可以包括但不限于微处理器MCU或可编程逻辑器件FPGA等的处理装置)、用于存储数据的存储器304。除此以外，还可以包括：显示器、输入/输出接口(I/O接口)、通用串行总线(USB)端口(可以作为I/O接口的端口中的一个端口被包括)、网络接口、键盘、电源和/或相机。本领域普通技术人员可以理解，图3所示的结构仅为示意，其并不对上述电子装置的结构造成限定。例如，电子设备还可包括比图3中所示更多或者更少的组件，或者具有与图3所示不同的配置。Fig. 3 is a block diagram of a hardware structure of an electronic device (or mobile device) for a method for determining an account set based on federated learning according to an embodiment of the present invention. As shown in FIG. 3, the electronic device may include one or more processors 302 (illustrated by 302a, 302b, . A processing device such as a device FPGA), a
上述本发明实施例序号仅仅为了描述，不代表实施例的优劣。The above-mentioned serial numbers of the embodiments of the present invention are only for description, and do not represent the advantages or disadvantages of the embodiments.
在本发明的上述实施例中，对各个实施例的描述都各有侧重，某个实施例中没有详述的部分，可以参见其他实施例的相关描述。In the above-mentioned embodiments of the present invention, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference may be made to related descriptions of other embodiments.
在本申请所提供的几个实施例中，应该理解到，所揭露的技术内容，可通过其它的方式实现。其中，以上所描述的装置实施例仅仅是示意性的，例如所述单元的划分，可以为一种逻辑功能划分，实际实现时可以有另外的划分方式，例如多个单元或组件可以结合或者可以集成到另一个系统，或一些特征可以忽略，或不执行。另一点，所显示或讨论的相互之间的耦合或直接耦合或通信连接可以是通过一些接口，单元或模块的间接耦合或通信连接，可以是电性或其它的形式。In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are only illustrative, for example, the division of the units may be a logical function division, and there may be other division methods in actual implementation, for example, multiple units or components may be combined or Integration into another system, or some features can be ignored, or not implemented. On the other hand, the shown or discussed mutual coupling or direct coupling or communication connection may be through some interfaces, indirect coupling or communication connection of units or modules, and may be in electrical or other forms.
所述作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个单元上。可以根据实际的需要选择其中的部分或者全部单元来实现本实施例方案的目的。The units described as separate components may or may not be physically separated, and components shown as units may or may not be physical units, that is, may be located in one place, or may be distributed to multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution in this embodiment.
另外，在本发明各个实施例中的各功能单元可以集成在一个处理单元中，也可以是各个单元单独物理存在，也可以两个或两个以上单元集成在一个单元中。上述集成的单元既可以采用硬件的形式实现，也可以采用软件功能单元的形式实现。In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically alone, or two or more units may be integrated into one unit. The above-mentioned integrated units may be implemented in the form of hardware, or may be implemented in the form of software functional units.
所述集成的单元如果以软件功能单元的形式实现并作为独立的产品销售或使用时，可以存储在一个计算机可读取存储介质中。基于这样的理解，本发明的技术方案本质上或者说对现有技术做出贡献的部分或者该技术方案的全部或部分可以以软件产品的形式体现出来，该计算机软件产品存储在一个存储介质中，包括若干指令用以使得一台计算机设备(可为个人计算机、服务器或者网络设备等)执行本发明各个实施例所述方法的全部或部分步骤。而前述的存储介质包括：U盘、只读存储器(ROM，Read-Only Memory)、随机存取存储器(RAM，Random Access Memory)、移动硬盘、磁碟或者光盘等各种可以存储程序代码的介质。The integrated unit, if implemented in the form of a software functional unit and sold or used as an independent product, may be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium , including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk and other media that can store program codes .
以上所述仅是本发明的优选实施方式，应当指出，对于本技术领域的普通技术人员来说，在不脱离本发明原理的前提下，还可以做出若干改进和润饰，这些改进和润饰也应视为本发明的保护范围。The above are only the preferred embodiments of the present invention. It should be pointed out that for those skilled in the art, without departing from the principles of the present invention, several improvements and modifications can be made. It should be regarded as the protection scope of the present invention.
Claims (10)
Translated fromChinesePriority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210764418.4ACN115018556B (en) | 2022-06-30 | 2022-06-30 | Account set determining method and device based on federal learning and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210764418.4ACN115018556B (en) | 2022-06-30 | 2022-06-30 | Account set determining method and device based on federal learning and electronic equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115018556Atrue CN115018556A (en) | 2022-09-06 |
| CN115018556B CN115018556B (en) | 2025-09-09 |
Family
ID=83079460
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210764418.4AActiveCN115018556B (en) | 2022-06-30 | 2022-06-30 | Account set determining method and device based on federal learning and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115018556B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070100715A1 (en)* | 2005-07-29 | 2007-05-03 | The Prudential Insurance Company Of America | Financial instrument providing a guaranteed growth rate and a guarantee of lifetime payments |
| US20080109338A1 (en)* | 2005-02-04 | 2008-05-08 | Searete Llc, A Limited Liability Corporation | Virtual credit in simulated environments |
| CN111681041A (en)* | 2020-05-26 | 2020-09-18 | 深圳市元征科技股份有限公司 | Electronic coupon issuing method and device, electronic equipment and storage medium |
| US20210264466A1 (en)* | 2020-02-26 | 2021-08-26 | Visa International Service Association | System, Method, and Computer Program Product for Generating a Synthetic Control Group |
| CN114372871A (en)* | 2022-01-07 | 2022-04-19 | 中国工商银行股份有限公司 | Method and device for determining credit score value, electronic device and storage medium |
- 2022
- 2022-06-30CNCN202210764418.4Apatent/CN115018556B/enactiveActive
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080109338A1 (en)* | 2005-02-04 | 2008-05-08 | Searete Llc, A Limited Liability Corporation | Virtual credit in simulated environments |
| US20070100715A1 (en)* | 2005-07-29 | 2007-05-03 | The Prudential Insurance Company Of America | Financial instrument providing a guaranteed growth rate and a guarantee of lifetime payments |
| US20210264466A1 (en)* | 2020-02-26 | 2021-08-26 | Visa International Service Association | System, Method, and Computer Program Product for Generating a Synthetic Control Group |
| CN111681041A (en)* | 2020-05-26 | 2020-09-18 | 深圳市元征科技股份有限公司 | Electronic coupon issuing method and device, electronic equipment and storage medium |
| CN114372871A (en)* | 2022-01-07 | 2022-04-19 | 中国工商银行股份有限公司 | Method and device for determining credit score value, electronic device and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115018556B (en) | 2025-09-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2019196579A1 (en) | Method and apparatus for issuing smart voucher, and method and apparatus for verification and cancellation by using smart voucher | |
| CN108846705A (en) | A kind of digital integration service system, method and electronic device | |
| CN107944593A (en) | A kind of resource allocation methods and device, electronic equipment | |
| CN109583964B (en) | Advertisement putting method and device | |
| JP2011519093A (en) | A model for early adoption and retention of funding sources to fund prize programs | |
| CN109670896A (en) | Investment product recommended method, device, equipment and storage medium | |
| US20160005072A1 (en) | Method and system for online commerce analysis | |
| WO2023070844A1 (en) | Intelligent design method for personalized discount coupon, and electronic apparatus and storage medium | |
| CN108765144A (en) | Processing method, terminal device and the medium of e-sourcing packet | |
| CN120106902B (en) | Activity voucher issuance method, system, intelligent terminal and storage medium | |
| Ying et al. | Understanding the factors influences users continuous intention towards e-wallet in Malaysia: Identifying the gap | |
| Adebayo et al. | Impact of E-payment system on buying behaviour: evidence from retail outlets in Ilorin metropolis, Nigeria | |
| Ahangar | An investigation into the determinant of customers’ preferences and satisfaction of internet banking (empirical study of Iranian banking industry) | |
| CN106997548A (en) | Electronic commerce business point management system | |
| CN112330373A (en) | User behavior analysis method, device and computer-readable storage medium | |
| Bataev | Financial technology: Efficiency evaluation of challenger banks | |
| US20150073863A1 (en) | Method and System for Linking Browsing History to Proprietary Transaction Data | |
| Tang et al. | Quantitative standard of promotion strategy and analysis on the influence of consumer purchase behavior | |
| CN113554428A (en) | An anchored digital asset transaction method and device for a multi-digital asset exchange platform | |
| CN115018556A (en) | A method for determining account set based on federated learning and its device and electronic device | |
| CN112418953A (en) | Contribution degree-based data pricing method, system, electronic device and storage medium | |
| CN107025545A (en) | A kind of transaction processing method and transaction system | |
| CN116227826A (en) | Resource distribution method, device, storage medium and computer equipment | |
| CN115034806A (en) | Advertisement delivery method, device, storage medium and electronic equipment | |
| CN113554425A (en) | Anchored digital asset computing method and device of multi-digital asset exchange platform |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |