WO2019071906A1 - 金融产品推荐装置、方法及计算机可读存储介质 - Google Patents

金融产品推荐装置、方法及计算机可读存储介质 Download PDF

Info

Publication number
WO2019071906A1
WO2019071906A1 PCT/CN2018/077631 CN2018077631W WO2019071906A1 WO 2019071906 A1 WO2019071906 A1 WO 2019071906A1 CN 2018077631 W CN2018077631 W CN 2018077631W WO 2019071906 A1 WO2019071906 A1 WO 2019071906A1
Authority
WO
WIPO (PCT)
Prior art keywords
financial product
target customer
feature
customer
classification model
Prior art date
Application number
PCT/CN2018/077631
Other languages
English (en)
French (fr)
Inventor
刘睿恺
吴振宇
王建明
肖京
Original Assignee
平安科技（深圳）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 平安科技（深圳）有限公司 filed Critical 平安科技（深圳）有限公司
Publication of WO2019071906A1 publication Critical patent/WO2019071906A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR 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
    • G06Q40/02—Banking, e.g. interest calculation or account maintenance
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR 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
    • G06Q30/06—Buying, selling or leasing transactions
    • G06Q30/0601—Electronic shopping [e-shopping]
    • G06Q30/0631—Item recommendations

Definitions

  • the present application relates to the field of information processing technologies, and in particular, to a financial product recommendation apparatus, method, and computer readable storage medium.
  • the marketing plan adopted by most banks is based on the traditional marketing system, which counts the customer's transaction data, selects a certain number of customers as potential customers, and recommends bank-designated finance to these customers by phone or SMS.
  • Products and with the development of the financial industry and the Internet industry, there are more and more types of financial products, including funds, wealth management, precious metals, insurance and other products, and each product contains several products, which As a result, it is difficult for customers to choose a product that suits their needs in the face of a large number of financial products.
  • For banks due to lack of manpower, etc., it is impossible to fully promote the products, and generally choose a small number of products.
  • the promotion of more popular financial products, in addition to this marketing model is not targeted, often combined with the business rules of the product to select some customers in batches, to carry out mass marketing, without in-depth mining of customer trading behaviors, personalized precision marketing .
  • the present application provides a financial product recommendation device, method and computer readable storage medium, the main purpose of which is to improve the recommendation success rate of a financial product.
  • the present application provides a financial product recommendation device including a memory and a processor, wherein the memory stores a financial product recommendation program executable on the processor, the financial product recommendation program being The processor implements the following steps when executed:
  • the financial product is recommended to the target customer via the common contact medium.
  • the present application further provides a financial product recommendation method, the method comprising:
  • the financial product is recommended to the target customer via the common contact medium.
  • the present application further provides a computer readable storage medium, where the financial product recommendation program is stored, and the financial product recommendation program can be executed by one or more processors. To implement the steps of the financial product recommendation method as described above.
  • the financial product recommendation device, method and computer readable storage medium proposed by the present application calculate the satisfaction level of the target customer's financial products according to the characteristics of the target customer according to the preset classification model according to the acquired feature characteristics. Then, according to the product data of the financial product held by the target customer and the satisfaction level of the financial product held, the target customer is selected for the financial product to be recommended.
  • the target customer is selected for the financial product to be recommended.
  • the target customer is selected with the appropriate contact medium. In this way, all the financial products of the bank can be integrated, and targeted recommendations can be made for a certain customer to improve the recommendation success rate of the product.
  • FIG. 1 is a schematic diagram of a preferred embodiment of a financial product recommendation device of the present application.
  • FIG. 2 is a schematic diagram of functional modules of a financial product recommendation program in an embodiment of a financial product recommendation device of the present application
  • FIG. 3 is a flow chart of a preferred embodiment of a financial product recommendation method of the present application.
  • the application provides a financial product recommendation device.
  • FIG. 1 a schematic diagram of a preferred embodiment of a financial product recommendation device of the present application is shown.
  • the financial product recommendation device may be a PC (Personal Computer), or may be a portable terminal device having a display function such as a smart phone, a tablet computer, or a portable computer.
  • PC Personal Computer
  • portable terminal device having a display function such as a smart phone, a tablet computer, or a portable computer.
  • the financial product recommendation device includes a memory 11, a processor 12, a communication bus 13, and a network interface 14.
  • the memory 11 includes at least one type of readable storage medium including a flash memory, a hard disk, a multimedia card, a card type memory (for example, an SD or DX memory, etc.), a magnetic memory, a magnetic disk, an optical disk, and the like.
  • the memory 11 may be an internal storage unit of the financial product recommendation device, such as a hard disk of the financial product recommendation device, in some embodiments.
  • the memory 11 may also be an external storage device of the financial product recommendation device in other embodiments, such as a plug-in hard disk equipped with a financial product recommendation device, a smart memory card (SMC), and a secure digital (Secure Digital, SD) card, flash card, etc.
  • the memory 11 may also include both an internal storage unit of the financial product recommendation device and an external storage device.
  • the memory 11 can be used not only for storing application software installed in the financial product recommendation device and various types of data, such as codes of the financial product recommendation program, but also for temporarily storing data that has been output or will be output.
  • the processor 12 may be a Central Processing Unit (CPU), controller, microcontroller, microprocessor or other data processing chip for running program code or processing stored in the memory 11. Data, such as the implementation of financial product recommendation procedures.
  • CPU Central Processing Unit
  • controller microcontroller
  • microprocessor or other data processing chip for running program code or processing stored in the memory 11.
  • Data such as the implementation of financial product recommendation procedures.
  • Communication bus 13 is used to implement connection communication between these components.
  • the network interface 14 can optionally include a standard wired interface, a wireless interface (such as a WI-FI interface), and is typically used to establish a communication connection between the device and other electronic devices.
  • a standard wired interface such as a WI-FI interface
  • Figure 1 shows only financial product recommendation devices having components 11-14 and financial product recommendation procedures, but it should be understood that not all illustrated components may be implemented and that more or fewer components may be implemented instead.
  • the device may further include a user interface
  • the user interface may include a display
  • an input unit such as a keyboard
  • the optional user interface may further include a standard wired interface and a wireless interface.
  • the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode) touch sensor, or the like.
  • the display may also be suitably referred to as a display screen or display unit for displaying information processed in the financial product recommendation device and a user interface for displaying the visualization.
  • a financial product recommendation program is stored in the memory 11; when the processor 12 executes the financial product recommendation program stored in the memory 11, the following steps are implemented:
  • the target customer's satisfaction level with the financial products held is calculated according to the target customer's transaction characteristics.
  • a matter information table is used in advance for storing each customer transaction item and other various items that occur based on the transaction item, for example, a complaint item, a compensation item, a comment item, and a surrender item. Wait.
  • the above transactions include transactions for fund purchases, insurance purchases, etc. for various financial products.
  • various financial product-related matters will be recorded in the above information table.
  • the contact medium at the time of the event is also recorded.
  • the contact medium is, in this embodiment, mainly includes the following channels: a PC client issued by a bank, an APP client, and a telemarketing channel.
  • the identity information of the target customer may be determined when receiving the request for product recommendation to the target customer, wherein the identity information may be identification information of the unique customer in the database by the ID card number or the mobile phone number. According to its identity information, all matters of the customer or items recorded in the past period of time are extracted from the item information table.
  • the step of extracting the item feature corresponding to the target customer from the item information table includes: extracting all items of the target customer from the item information table according to the identity information of the target customer, and filtering out the pre-predetermined items from the extracted items.
  • a matter of the item category; the corresponding item feature is extracted from the item belonging to the preset item category.
  • the items are subjected to dimensionality reduction processing, and the items belonging to the preset category are retained, and the items are not filtered out.
  • a statement that reflects the customer's satisfaction with financial products where one or more items may correspond to a financial product that a user is holding or has held.
  • the following categories of items are set in advance: an account opening category, a buying category, an insurance class, a surrender item, a complaint category, a compensation category, and the like.
  • You can set up more event categories in advance as needed.
  • a record of the surrender item is generated in the item information table, and it can be inferred that the customer has low satisfaction with the insurance product; or If a customer purchases the fund products and insurance products of the bank after opening an account with the bank, an account opening item and two purchase items will be generated correspondingly in the item information table, thereby inferring the customer.
  • High satisfaction with related financial products It can be seen that different issues reflect the different levels of satisfaction of customers with existing financial products. Therefore, in the comprehensive evaluation of the target customers' satisfaction with the financial products they are currently holding or the financial products they have held. When evaluating all historical matters of the target customer.
  • the item feature extraction process in the embodiment mainly extracts content having relevance from the customer satisfaction content as a matter feature from the recorded content content, for example, for the item for purchasing the insurance product, the item feature may be the insurance item.
  • Each item recorded in the item information table has corresponding information items, such as the item name, the item of the item, and the attribute of the item. Therefore, the information items to be extracted in the items under each item category can be set in advance, and the content of the information items set in advance can be extracted when the item features are extracted.
  • the classification model needs to be trained when calculating the satisfaction level corresponding to the target customer by using the preset classification model.
  • the support vector machine classification model is selected as a preset classification model. Obtaining a feature training set, each item feature in the item feature training set has a corresponding satisfaction level, that is, the feature in the training set needs to be artificially pre-determined to reflect the characteristics of each item; The support vector machine classification model is trained to obtain model parameters of the classification model.
  • the satisfaction degree is set to a plurality of levels in advance.
  • the satisfaction degree is set to five levels. The higher the level, the user is interested in the financial product currently held or has been held. The higher the satisfaction of financial products. Extract all the items belonging to the preset categories in the item information table and extract the item characteristics from them, and manually evaluate the satisfaction level of the financial products held by each user according to the characteristics of the items, and set the characteristics of the items. After the label of the satisfaction level is associated, the item feature database is created, and 80% of the item features are selected as the training set for training the model, and the remaining 20% of the item features are used as the verification set.
  • the training set is input into the support vector machine classification model to train the model, and the model parameters are obtained.
  • the training results are evaluated through the verification set.
  • the more users in the training set the more accurate the training parameters are.
  • the resulting model parameters reflect the correlation between the user's default type of matter and its level of satisfaction with the financial products currently held or financial products that have been held.
  • the training set can be continuously adjusted, and the model optimal parameters are obtained after multiple iterations.
  • the extracted target characteristics of the target customer are input into the trained support vector machine classification model, and the satisfaction level of the target customer to the held financial product is calculated.
  • the mapping relationship between each financial product and other one or more financial products under different satisfaction levels is established in advance, and when the recommended product is selected, the user is selected according to the mapping relationship. Financial products.
  • the financial product is recommended to the target customer via the common contact medium.
  • the contact medium used by the customer After determining the financial product to be recommended, analyzing the existing customer group of the financial product, obtaining the contact medium used by each existing customer corresponding to the financial product to be recommended;
  • the contact medium used by the customer performs statistics to determine the probability of distribution of the product to be recommended on each contact medium, and the distribution probability of the contact medium is large, indicating that the probability of the customer purchasing the financial product through the contact medium is higher, and the distribution will be distributed.
  • the most probable contact medium serves as a common contact medium for the target customer. To improve the recommendation success rate.
  • the financial product recommendation device proposed in this embodiment calculates the satisfaction level of the target customer's financial products according to the acquired classification characteristics according to the characteristics of the target customer according to the predetermined classification model, and then according to the target customer's holding level.
  • the product data of the financial products and the level of satisfaction with the financial products held select the financial products to be recommended for the target customers.
  • the target customer is selected with the appropriate contact medium. In this way, all the financial products of the bank can be integrated, and targeted recommendations can be made for a certain customer to improve the recommendation success rate of the product.
  • the financial product recommendation program may also be divided into one or more modules, one or more modules are stored in the memory 11 and are processed by one or more processors (this embodiment) Illustrated by the processor 12) to complete the application, a module referred to herein refers to a series of computer program instructions that are capable of performing a particular function for describing the execution of a financial product recommendation program in a financial product recommendation device.
  • FIG. 2 it is a schematic diagram of a function module of a financial product recommendation program in an embodiment of the financial product recommendation device of the present application.
  • the financial product recommendation program may be divided into an acquisition module 10 and a calculation module 20,
  • the selection module 30 and the recommendation module 40 are exemplarily:
  • the obtaining module 10 is configured to: extract, from the item information table, a feature feature corresponding to the target customer;
  • the calculating module 20 is configured to: according to a preset classification model, calculate a satisfaction level of the target customer to the held financial product according to the target feature of the target customer;
  • the selecting module 30 is configured to: obtain product data of the financial product held by the target customer, and according to the product data of the financial product held by the target customer and the satisfaction level of the held financial product, the target is The customer selects the financial product to be recommended;
  • the obtaining module 10 is further configured to: acquire a contact medium used by an existing customer corresponding to the financial product to be recommended, and predict a recommended contact medium of the customer according to the acquired contact medium;
  • the recommendation module 40 is configured to: recommend the financial product to the target customer through the common contact medium.
  • the present application also provides a financial product recommendation method.
  • FIG. 3 it is a flowchart of the first embodiment of the financial product recommendation method of the present application.
  • the financial product recommendation method includes:
  • step S10 the item characteristics corresponding to the target customer are extracted from the item information table.
  • step S20 according to the preset classification model, the satisfaction level of the target customer to the held financial product is calculated according to the characteristics of the target customer.
  • the method of the embodiments of the present application may be performed by a device, which may be implemented by software and/or hardware.
  • a device information table for storing each customer transaction item and other various items occurring based on the transaction item, such as a complaint item, a compensation item, a comment item, and a surrender item, are pre-established in the apparatus.
  • the above transactions include transactions for fund purchases, insurance purchases, etc. for various financial products.
  • various financial product-related matters will be recorded in the above information table.
  • the contact medium at the time of the event is also recorded.
  • the contact medium is, in this embodiment, mainly includes the following channels: a PC client issued by a bank, an APP client, and a telemarketing channel.
  • the identity information of the target customer may be determined when receiving the request for product recommendation to the target customer, wherein the identity information may be identification information of the unique customer in the database by the ID card number or the mobile phone number. According to its identity information, all matters of the customer or items recorded in the past period of time are extracted from the item information table.
  • the step of extracting the item feature corresponding to the target customer from the item information table includes: extracting all items of the target customer from the item information table according to the identity information of the target customer, and filtering out the pre-predetermined items from the extracted items.
  • a matter of the item category; the corresponding item feature is extracted from the item belonging to the preset item category.
  • the items are subjected to dimensionality reduction processing, and the items belonging to the preset category are retained, and the items are not filtered out.
  • a statement that reflects the customer's satisfaction with financial products where one or more items may correspond to a financial product that a user is holding or has held.
  • the following categories of items are set in advance: an account opening category, a buying category, an insurance class, a surrender item, a complaint category, a compensation category, and the like.
  • You can set up more event categories in advance as needed.
  • a record of the surrender item is generated in the item information table, and it can be inferred that the customer has low satisfaction with the insurance product; or If a customer purchases the fund products and insurance products of the bank after opening an account with the bank, an account opening item and two purchase items will be generated correspondingly in the item information table, thereby inferring the customer.
  • High satisfaction with related financial products It can be seen that different issues reflect the different levels of satisfaction of customers with existing financial products. Therefore, in the comprehensive evaluation of the target customers' satisfaction with the financial products they are currently holding or the financial products they have held. When evaluating all historical matters of the target customer.
  • the item feature extraction process in the embodiment mainly extracts content having relevance from the customer satisfaction content as a matter feature from the recorded content content, for example, for the item for purchasing the insurance product, the item feature may be the insurance item.
  • Each item recorded in the item information table has corresponding information items, such as the item name, the item of the item, and the attribute of the item. Therefore, the information items to be extracted in the items under each item category can be set in advance, and the content of the information items set in advance can be extracted when the item features are extracted.
  • the classification model needs to be trained when calculating the satisfaction level corresponding to the target customer by using the preset classification model.
  • the support vector machine classification model is selected as a preset classification model. Obtaining a feature training set, each item feature in the item feature training set has a corresponding satisfaction level, that is, the feature in the training set needs to be artificially pre-determined to reflect the characteristics of each item; The support vector machine classification model is trained to obtain model parameters of the classification model.
  • the satisfaction degree is set to a plurality of levels in advance.
  • the satisfaction degree is set to five levels. The higher the level, the user is interested in the financial product currently held or has been held. The higher the satisfaction of financial products. Extract all the items belonging to the preset categories in the item information table and extract the item characteristics from them, and manually evaluate the satisfaction level of the financial products held by each user according to the characteristics of the items, and set the characteristics of the items. After the label of the satisfaction level is associated, the item feature database is created, and 80% of the item features are selected as the training set for training the model, and the remaining 20% of the item features are used as the verification set.
  • the training set is input into the support vector machine classification model to train the model, and the model parameters are obtained.
  • the training results are evaluated through the verification set.
  • the more users in the training set the more accurate the training parameters are.
  • the resulting model parameters reflect the correlation between the user's default type of matter and its level of satisfaction with the financial products currently held or financial products that have been held.
  • the training set can be continuously adjusted, and the model optimal parameters are obtained after multiple iterations.
  • step S30 the extracted target feature of the target customer is input into the trained support vector machine classification model, and the satisfaction level of the target customer to the held financial product is calculated.
  • pre-establishing a mapping relationship between each financial product and other one or more financial products under different satisfaction levels when selecting a recommended product, selecting a suitable financial product for the user according to the mapping relationship .
  • Step S40 Acquire a contact medium used by an existing customer corresponding to the financial product to be recommended, and predict a recommended contact medium of the customer according to the acquired contact medium.
  • Step S50 recommending the financial product to the target customer through the common contact medium.
  • the contact medium used by the customer After determining the financial product to be recommended, analyzing the existing customer group of the financial product, obtaining the contact medium used by each existing customer corresponding to the financial product to be recommended;
  • the contact medium used by the customer performs statistics to determine the probability of distribution of the product to be recommended on each contact medium, and the distribution probability of the contact medium is large, indicating that the probability of the customer purchasing the financial product through the contact medium is higher, and the distribution will be distributed.
  • the most probable contact medium serves as a common contact medium for the target customer. To improve the recommendation success rate.
  • the financial product recommendation method proposed in this embodiment calculates the satisfaction level of the target customer's financial products according to the acquired classification characteristics according to the characteristics of the target customer according to the predetermined classification model, and then according to the target customer's holding level.
  • the product data of the financial products and the level of satisfaction with the financial products held select the financial products to be recommended for the target customers.
  • the target customer is selected with the appropriate contact medium. In this way, all the financial products of the bank can be integrated, and targeted recommendations can be made for a certain customer to improve the recommendation success rate of the product.
  • the embodiment of the present application further provides a computer readable storage medium, where the financial product recommendation program is stored, and the financial product recommendation program can be executed by one or more processors to implement the following operating:
  • the financial product is recommended to the target customer via the common contact medium.
  • the step of extracting a feature feature corresponding to the target customer from the item information table includes:
  • Extracting corresponding event features from the items belonging to the preset item category Extracting corresponding event features from the items belonging to the preset item category.
  • the step of obtaining the contact medium used by the existing customer corresponding to the financial product to be recommended, and predicting the recommended contact medium of the customer according to the acquired contact medium includes:
  • the contact medium used by each existing customer is counted, the distribution probability of the product to be recommended on each contact medium is determined, and the contact medium with the highest distribution probability is used as the common contact medium of the target customer.
  • the preset classification model is a support vector machine classification model
  • the processor is further configured to execute the financial product recommendation program to further extract the item feature corresponding to the target customer from the item information table.
  • the technical solution of the present application which is essential or contributes to the prior art, may be embodied in the form of a software product stored in a storage medium (such as ROM/RAM as described above). , a disk, an optical disk, including a number of instructions for causing a terminal device (which may be a mobile phone, a computer, a server, or a network device, etc.) to perform the methods described in the various embodiments of the present application.
  • a terminal device which may be a mobile phone, a computer, a server, or a network device, etc.

Landscapes

  • Business, Economics & Management (AREA)
  • Accounting & Taxation (AREA)
  • Finance (AREA)
  • Engineering & Computer Science (AREA)
  • Marketing (AREA)
  • Economics (AREA)
  • Development Economics (AREA)
  • Strategic Management (AREA)
  • Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Technology Law (AREA)
  • Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)

Abstract

本申请公开了一种金融产品推荐装置，包括存储器和处理器，存储器上存储有可在处理器上运行的金融产品推荐程序，该程序被处理器执行时实现如下步骤：从事项信息表中提取目标客户对应的事项特征；按照预设的分类模型，根据目标客户的事项特征计算目标客户对持有的金融产品的满意度级别；目标客户的持有的金融产品的产品数据和对持有的金融产品的满意度级别，为目标客户选择待推荐的金融产品；获取目标客户在事项信息表中的每一事项对应的接触媒介，根据获取的接触媒介预测客户的常用接触媒介；将金融产品通过常用接触媒介推荐给目标客户。本申请还提出一种金融产品推荐方法以及一种计算机可读存储介质。本申请提高了金融产品的推荐成功率。

Description

金融产品推荐装置、方法及计算机可读存储介质
本申请基于巴黎公约申明享有2017年10月9日递交的申请号为201710930686.8、名称为“金融产品推荐装置、方法及计算机可读存储介质”的中国专利申请的优先权，该中国专利申请的整体内容以参考的方式结合在本申请中。
技术领域
本申请涉及信息处理技术领域，尤其涉及一种金融产品推荐装置、方法及计算机可读存储介质。
背景技术
目前，大多数银行采用的营销方案是建立在传统的营销***之上，对客户的交易数据进行统计，筛选出一定量的客户作为潜在客户，通过电话或者短信等向这些客户推荐银行指定的金融产品，而随着金融行业以及互联网行业的发展，金融产品的种类越来越多，有基金、理财、贵金属、保险等各类产品，而每一类产品下又包含有若干个产品，这就导致对于客户来说，面对数量众多的金融产品，难以选择出适合自己的产品，而对于银行来说，由于人手不足等原因，无法实现对产品进行全面的推销，一般会选择数量较少的较为热门的金融产品的推销，此外这种推销模式不具有针对性，往往是结合产品的业务规则批量选择一些客户，进行批量营销，没有对客户的交易行为等进行深入挖掘而进行个性化精准营销。
综上所述，对于现有的营销模式来说，由于上述各种缺陷的存在，导致将金融产品推荐给客户的成功率较低。
发明内容
本申请提供一种金融产品推荐装置、方法及计算机可读存储介质，其主要目的在于提高金融产品的推荐成功率。
为实现上述目的，本申请提供一种金融产品推荐装置，该装置包括存储器和处理器，所述存储器中存储有可在所述处理器上运行的金融产品推荐程 序，所述金融产品推荐程序被所述处理器执行时实现如下步骤：
从事项信息表中提取目标客户对应的事项特征；
按照预设的分类模型，根据目标客户的事项特征计算目标客户对持有的金融产品的满意度级别；
获取目标客户的持有的金融产品的产品数据，并根据所述目标客户的持有的金融产品的产品数据和对持有的金融产品的满意度级别，为所述目标客户选择待推荐的金融产品；
获取所述待推荐的金融产品对应的现有客户所使用的接触媒介，并根据获取的接触媒介预测客户的推荐接触媒介；
将所述金融产品通过所述常用接触媒介推荐给所述目标客户。
此外，为实现上述目的，本申请还提供一种金融产品推荐方法，该方法包括：
从事项信息表中提取目标客户对应的事项特征；
按照预设的分类模型，根据目标客户的事项特征计算目标客户对持有的金融产品的满意度级别；
获取目标客户的持有的金融产品的产品数据，并根据所述目标客户的持有的金融产品的产品数据和对持有的金融产品的满意度级别，为所述目标客户选择待推荐的金融产品；
获取所述待推荐的金融产品对应的现有客户所使用的接触媒介，并根据获取的接触媒介预测客户的推荐接触媒介；
将所述金融产品通过所述常用接触媒介推荐给所述目标客户。
此外，为实现上述目的，本申请还提供一种计算机可读存储介质，所述计算机可读存储介质上存储有金融产品推荐程序，所述金融产品推荐程序可被一个或者多个处理器执行，以实现如上所述的金融产品推荐方法的步骤。
本申请提出的金融产品推荐装置、方法及计算机可读存储介质，根据目标客户的事项特征，按照预设的分类模型根据获取到的事项特征计算目标客户对持有的金融产品的满意度级别，然后根据目标客户的持有的金融产品的产品数据和对持有的金融产品的满意度级别，为目标客户选择待推荐的金融产品。通过上述流程能够实现自动对目标客户的事项进行分析，并根据分类算法计算出客户对持有的金融产品的满意度级别，将满意度级别作为推荐新 的金融产品的依据，并结合该金融产品的已有客户使用的接触媒介情况，为目标客户选择合适的接触媒介，通过这种方式，能够综合银行的所有金融产品，并针对某一客户进行针对性的推荐，提高产品的推荐成功率。
附图说明
图1为本申请金融产品推荐装置较佳实施例的示意图；
图2为本申请金融产品推荐装置一实施例中金融产品推荐程序的功能模块示意图；
图3为本申请金融产品推荐方法较佳实施例的流程图。
本申请目的的实现、功能特点及优点将结合实施例，参照附图做进一步说明。
具体实施方式
应当理解，此处所描述的具体实施例仅仅用以解释本申请，并不用于限定本申请。
本申请提供一种金融产品推荐装置。参照图1所示，为本申请金融产品推荐装置较佳实施例的示意图。
在本实施例中，金融产品推荐装置可以是PC(Personal Computer，个人电脑)，也可以是智能手机、平板电脑、便携计算机等具有显示功能的可移动式终端设备。
该金融产品推荐装置包括存储器11、处理器12，通信总线13，以及网络接口14。
其中，存储器11至少包括一种类型的可读存储介质，所述可读存储介质包括闪存、硬盘、多媒体卡、卡型存储器(例如，SD或DX存储器等)、磁性存储器、磁盘、光盘等。存储器11在一些实施例中可以是金融产品推荐装置的内部存储单元，例如该金融产品推荐装置的硬盘。存储器11在另一些实施例中也可以是金融产品推荐装置的外部存储设备，例如金融产品推荐装置上配备的插接式硬盘，智能存储卡(Smart Media Card,SMC)，安全数字(Secure Digital,SD)卡，闪存卡(Flash Card)等。进一步地，存储器11还可以既包括 金融产品推荐装置的内部存储单元也包括外部存储设备。存储器11不仅可以用于存储安装于金融产品推荐装置的应用软件及各类数据，例如金融产品推荐程序的代码等，还可以用于暂时地存储已经输出或者将要输出的数据。
处理器12在一些实施例中可以是一中央处理器(Central Processing Unit,CPU)、控制器、微控制器、微处理器或其他数据处理芯片，用于运行存储器11中存储的程序代码或处理数据，例如执行金融产品推荐程序等。
通信总线13用于实现这些组件之间的连接通信。
网络接口14可选的可以包括标准的有线接口、无线接口(如WI-FI接口)，通常用于在该装置与其他电子设备之间建立通信连接。
图1仅示出了具有组件11-14以及金融产品推荐程序的金融产品推荐装置，但是应理解的是，并不要求实施所有示出的组件，可以替代的实施更多或者更少的组件。
可选地，该装置还可以包括用户接口，用户接口可以包括显示器(Display)、输入单元比如键盘(Keyboard)，可选的用户接口还可以包括标准的有线接口、无线接口。可选地，在一些实施例中，显示器可以是LED显示器、液晶显示器、触控式液晶显示器以及OLED(Organic Light-Emitting Diode，有机发光二极管)触摸器等。其中，显示器也可以适当的称为显示屏或显示单元，用于显示在金融产品推荐装置中处理的信息以及用于显示可视化的用户界面。
在图1所示的装置实施例中，存储器11中存储有金融产品推荐程序；处理器12执行存储器11中存储的金融产品推荐程序时实现如下步骤：
从事项信息表中提取目标客户对应的事项特征。
按照预设的分类模型，根据目标客户的事项特征计算目标客户对持有的金融产品的满意度级别。
本申请实施例中，预先建立有事项信息表，该事项信息表用于存储各个客户交易事项，以及基于交易事项发生的其他各种事项，例如，投诉事项、赔偿事项、评论事项、退保事项等。上述交易事项包括基金买入事项、保险买入事项等等针对各种金融产品的交易事项。也就是说，只要在本银行开户的客户，发生的各种与金融产品相关的事项都会记录在上述事项信息表中。同时，在记录一个事项时，同时还记录该事项发生时的接触媒介。接触媒介为，在该实施例中主要包括以下渠道：银行发布的PC客户端、APP客户端以 及电话营销渠道。
可以在接收到对目标客户进行产品推荐的请求时，确定目标客户的身份信息，其中，身份信息可以是身份证号码或者手机号码等都能够在上述数据库中识别唯一客户的标识信息。根据其身份信息从事项信息表中提取出该客户的所有事项或者在过去一段时间内记录的事项。
作为一种实施方式，从事项信息表中提取目标客户对应的事项特征的步骤包括：根据目标客户的身份信息从事项信息表中提取目标客户的所有事项，并从提取的事项中过滤出属于预设事项类别的事项；从所述属于预设事项类别的事项中提取出对应的事项特征。
在该实施方式中，为了提高对于客户满意度级别计算的准确性，在获取到与目标客户对应的所有事项后，对这些事项进行降维处理，保留属于预设类别的事项，过滤掉其中没有体现出客户对金融产品的满意度的事项，其中，一个或者多个事项可以对应于一个用户正在持有或者曾经持有过的金融产品。可选地，在该实施例中，预先设置如下几个类别的事项：开户类、买入类、加保类、退保类事项、投诉类、赔偿类等事项类别，在其他实施例中，可以根据需要预先设置更多的事项类别。例如，某客户购买某保险产品一段时间后，进行了退保，则会在事项信息表中生成一条退保事项的记录，由此可以推断出该客户对于该保险产品的满意度较低；或者，如果某客户在本银行开户后，又购买本银行的基金产品、保险产品，则会在事项信息表中对应地生成一个开户类事项和两个买入类事项，由此可以推断出该客户对相关金融产品的满意度较高。由此可见，不同的事项反映出客户对已有的金融产品的不同的满意度级别，因此，在综合评价目标客户对当前正在持有的金融产品或者曾经持有过的金融产品的满意度情况时，结合目标客户的所有历史事项进行评估。
从事项信息表中提取目标客户的所有事项，并从提取的事项中过滤出属于预设事项类别的事项后，从这些事项中提取事项特征。本实施例中的事项特征提取过程主要是从记录的事项内容中提取出与客户满意度之间具有相关性的内容作为事项特征，例如，对于购买保险产品的事项，则事项特征可以为该保险产品的属性信息；或者，若记录的事项为：客户通过APP客户端对某一理财产品进行了评论，则可以对其评论内容进行分析，提取评价中能够反 映客户满意度的关键词作为事项特征。事项信息表中记录的每一条事项都有对应的各个信息项，例如事项名称，事项对象，事项属性等。因此，可以预先设置每个事项类别下的事项要提取的信息项，在提取事项特征时，提取预先设置的信息项的内容即可。
可以理解的是，在使用预设的分类模型计算目标客户对应的满意度级别时，需要对分类模型进行训练。可选地，在一些实施例中，选择支持向量机分类模型作为预设的分类模型。获取事项特征训练集，所述事项特征训练集中的每一个事项特征有对应的满意度级别，也就是说，训练集中的特征需要人工预先判断各个事项特征体现出的；根据所述事项特征训练集训练所述支持向量机分类模型，以获取所述分类模型的模型参数。
预先将满意度设为多个等级，可选地，在本实施例中，将满意度设为五个级别，级别越高，则说明用户对当前正在持有的金融产品或者曾经持有过的金融产品的满意度越高。将事项信息表中所有客户的属于预设类别的事项提取出来并从中提取事项特征，并针对每个用户，根据其事项特征人工评估其对持有的金融产品的满意度级别，并将事项特征关联满意度级别的标签后，建立事项特征库，从中选择80％的事项特征作为训练集，用于训练模型，剩余的20％的事项特征作为验证集。将训练集输入到支持向量机分类模型中对模型进行训练，得到模型参数，并通过验证集对训练结果进行评估，其中，训练集中的用户越多，则训练得到模型参数越精确。得到的模型参数反映出用户的预设类型的事项与其对当前正在持有的金融产品或者曾经持有过的金融产品的满意度级别之间的相关关系。此外，可以理解的是，在对模型的训练过程中，可以不断地调整训练集，经过多次迭代获取模型最优参数。
将提取得到的目标客户的事项特征输入到上述训练好的支持向量机分类模型中，计算得到目标客户对持有的金融产品的满意度级别。
获取目标客户的持有的金融产品的产品数据，并根据所述目标客户的持有的金融产品的产品数据和对持有的金融产品的满意度级别，为所述目标客户选择待推荐的金融产品。
本实施例提出的装置中，预先建立每一金融产品在不同的满意度级别下，与其他一个或者多个金融产品之间的映射关系，则在选择推荐产品时，根据映射关系为用户选择合适的金融产品。
获取所述待推荐的金融产品对应的现有客户所使用的接触媒介，并根据获取的接触媒介预测客户的推荐接触媒介。
将所述金融产品通过所述常用接触媒介推荐给所述目标客户。
在确定待推荐的金融产品后，对该金融产品的现有的客户群进行分析，获取所述待推荐的金融产品对应的每一现有客户所使用的接触媒介；对获取的每一现有客户所使用的接触媒介进行统计，确定所述待推荐的产品在各个接触媒介上的分布概率，接触媒介的分布概率大，说明客户通过该接触媒介购买该金融产品的概率更高，则将分布概率最大的接触媒介作为所述目标客户的常用接触媒介。以提高推荐成功率。
本实施例提出的金融产品推荐装置，根据目标客户的事项特征，按照预设的分类模型根据获取到的事项特征计算目标客户对持有的金融产品的满意度级别，然后根据目标客户的持有的金融产品的产品数据和对持有的金融产品的满意度级别，为目标客户选择待推荐的金融产品。通过上述流程能够实现自动对目标客户的事项进行分析，并根据分类算法计算出客户对持有的金融产品的满意度级别，将满意度级别作为推荐新的金融产品的依据，并结合该金融产品的已有客户使用的接触媒介情况，为目标客户选择合适的接触媒介，通过这种方式，能够综合银行的所有金融产品，并针对某一客户进行针对性的推荐，提高产品的推荐成功率。
可选地，在其他的实施例中，金融产品推荐程序还可以被分割为一个或者多个模块，一个或者多个模块被存储于存储器11中，并由一个或多个处理器(本实施例为处理器12)所执行以完成本申请，本申请所称的模块是指能够完成特定功能的一系列计算机程序指令段，用于描述金融产品推荐程序在金融产品推荐装置中的执行过程。
例如，参照图2所示，为本申请金融产品推荐装置一实施例中的金融产品推荐程序的功能模块示意图，该实施例中，金融产品推荐程序可以被分割为获取模块10、计算模块20、选择模块30和推荐模块40，示例性地：
获取模块10用于：从事项信息表中提取目标客户对应的事项特征；
计算模块20用于：按照预设的分类模型，根据目标客户的事项特征计算目标客户对持有的金融产品的满意度级别；
选择模块30用于：获取目标客户的持有的金融产品的产品数据，并根据 所述目标客户的持有的金融产品的产品数据和对持有的金融产品的满意度级别，为所述目标客户选择待推荐的金融产品；
获取模块10还用于：获取所述待推荐的金融产品对应的现有客户所使用的接触媒介，并根据获取的接触媒介预测客户的推荐接触媒介；
推荐模块40用于：将所述金融产品通过所述常用接触媒介推荐给所述目标客户。
上述获取模块10、计算模块20、选择模块30和推荐模块40被执行所实现的功能或操作步骤与上述实施例大体相同，在此不再赘述。
此外，本申请还提供一种金融产品推荐方法。参照图3所示，为本申请金融产品推荐方法第一实施例的流程图。
在本实施例中，金融产品推荐方法包括：
步骤S10，从事项信息表中提取目标客户对应的事项特征。
步骤S20，按照预设的分类模型，根据目标客户的事项特征计算目标客户对持有的金融产品的满意度级别。
本申请实施例的方法可以由一个装置执行，该装置可以由软件和/或硬件实现。该装置中预先建立有事项信息表，该事项信息表用于存储各个客户交易事项，以及基于交易事项发生的其他各种事项，例如，投诉事项、赔偿事项、评论事项、退保事项等。上述交易事项包括基金买入事项、保险买入事项等等针对各种金融产品的交易事项。也就是说，只要在本银行开户的客户，发生的各种与金融产品相关的事项都会记录在上述事项信息表中。同时，在记录一个事项时，同时还记录该事项发生时的接触媒介。接触媒介为，在该实施例中主要包括以下渠道：银行发布的PC客户端、APP客户端以及电话营销渠道。
可以在接收到对目标客户进行产品推荐的请求时，确定目标客户的身份信息，其中，身份信息可以是身份证号码或者手机号码等都能够在上述数据库中识别唯一客户的标识信息。根据其身份信息从事项信息表中提取出该客户的所有事项或者在过去一段时间内记录的事项。
作为一种实施方式，从事项信息表中提取目标客户对应的事项特征的步骤包括：根据目标客户的身份信息从事项信息表中提取目标客户的所有事项，并从提取的事项中过滤出属于预设事项类别的事项；从所述属于预设事项类 别的事项中提取出对应的事项特征。
在该实施方式中，为了提高对于客户满意度级别计算的准确性，在获取到与目标客户对应的所有事项后，对这些事项进行降维处理，保留属于预设类别的事项，过滤掉其中没有体现出客户对金融产品的满意度的事项，其中，一个或者多个事项可以对应于一个用户正在持有或者曾经持有过的金融产品。可选地，在该实施例中，预先设置如下几个类别的事项：开户类、买入类、加保类、退保类事项、投诉类、赔偿类等事项类别，在其他实施例中，可以根据需要预先设置更多的事项类别。例如，某客户购买某保险产品一段时间后，进行了退保，则会在事项信息表中生成一条退保事项的记录，由此可以推断出该客户对于该保险产品的满意度较低；或者，如果某客户在本银行开户后，又购买本银行的基金产品、保险产品，则会在事项信息表中对应地生成一个开户类事项和两个买入类事项，由此可以推断出该客户对相关金融产品的满意度较高。由此可见，不同的事项反映出客户对已有的金融产品的不同的满意度级别，因此，在综合评价目标客户对当前正在持有的金融产品或者曾经持有过的金融产品的满意度情况时，结合目标客户的所有历史事项进行评估。
从事项信息表中提取目标客户的所有事项，并从提取的事项中过滤出属于预设事项类别的事项后，从这些事项中提取事项特征。本实施例中的事项特征提取过程主要是从记录的事项内容中提取出与客户满意度之间具有相关性的内容作为事项特征，例如，对于购买保险产品的事项，则事项特征可以为该保险产品的属性信息；或者，若记录的事项为：客户通过APP客户端对某一理财产品进行了评论，则可以对其评论内容进行分析，提取评价中能够反映客户满意度的关键词作为事项特征。事项信息表中记录的每一条事项都有对应的各个信息项，例如事项名称、事项对象、事项属性等。因此，可以预先设置每个事项类别下的事项要提取的信息项，在提取事项特征时，提取预先设置的信息项的内容即可。
可以理解的是，在使用预设的分类模型计算目标客户对应的满意度级别时，需要对分类模型进行训练。可选地，在一些实施例中，选择支持向量机分类模型作为预设的分类模型。获取事项特征训练集，所述事项特征训练集中的每一个事项特征有对应的满意度级别，也就是说，训练集中的特征需要 人工预先判断各个事项特征体现出的；根据所述事项特征训练集训练所述支持向量机分类模型，以获取所述分类模型的模型参数。
预先将满意度设为多个等级，可选地，在本实施例中，将满意度设为五个级别，级别越高，则说明用户对当前正在持有的金融产品或者曾经持有过的金融产品的满意度越高。将事项信息表中所有客户的属于预设类别的事项提取出来并从中提取事项特征，并针对每个用户，根据其事项特征人工评估其对持有的金融产品的满意度级别，并将事项特征关联满意度级别的标签后，建立事项特征库，从中选择80％的事项特征作为训练集，用于训练模型，剩余的20％的事项特征作为验证集。将训练集输入到支持向量机分类模型中对模型进行训练，得到模型参数，并通过验证集对训练结果进行评估，其中，训练集中的用户越多，则训练得到模型参数越精确。得到的模型参数反映出用户的预设类型的事项与其对当前正在持有的金融产品或者曾经持有过的金融产品的满意度级别之间的相关关系。此外，可以理解的是，在对模型的训练过程中，可以不断地调整训练集，经过多次迭代获取模型最优参数。
步骤S30，将提取得到的目标客户的事项特征输入到上述训练好的支持向量机分类模型中，计算得到目标客户对持有的金融产品的满意度级别。
获取目标客户的持有的金融产品的产品数据，并根据所述目标客户的持有的金融产品的产品数据和对持有的金融产品的满意度级别，为所述目标客户选择待推荐的金融产品。
本实施例中，预先建立每一金融产品在不同的满意度级别下，与其他一个或者多个金融产品之间的映射关系，则在选择推荐产品时，根据映射关系为用户选择合适的金融产品。
步骤S40，获取所述待推荐的金融产品对应的现有客户所使用的接触媒介，并根据获取的接触媒介预测客户的推荐接触媒介。
步骤S50，将所述金融产品通过所述常用接触媒介推荐给所述目标客户。
在确定待推荐的金融产品后，对该金融产品的现有的客户群进行分析，获取所述待推荐的金融产品对应的每一现有客户所使用的接触媒介；对获取的每一现有客户所使用的接触媒介进行统计，确定所述待推荐的产品在各个接触媒介上的分布概率，接触媒介的分布概率大，说明客户通过该接触媒介购买该金融产品的概率更高，则将分布概率最大的接触媒介作为所述目标客 户的常用接触媒介。以提高推荐成功率。
本实施例提出的金融产品推荐方法，根据目标客户的事项特征，按照预设的分类模型根据获取到的事项特征计算目标客户对持有的金融产品的满意度级别，然后根据目标客户的持有的金融产品的产品数据和对持有的金融产品的满意度级别，为目标客户选择待推荐的金融产品。通过上述流程能够实现自动对目标客户的事项进行分析，并根据分类算法计算出客户对持有的金融产品的满意度级别，将满意度级别作为推荐新的金融产品的依据，并结合该金融产品的已有客户使用的接触媒介情况，为目标客户选择合适的接触媒介，通过这种方式，能够综合银行的所有金融产品，并针对某一客户进行针对性的推荐，提高产品的推荐成功率。
此外，本申请实施例还提出一种计算机可读存储介质，所述计算机可读存储介质上存储有金融产品推荐程序，所述金融产品推荐程序可被一个或多个处理器执行，以实现如下操作：
从事项信息表中提取目标客户对应的事项特征；
按照预设的分类模型，根据目标客户的事项特征计算目标客户对持有的金融产品的满意度级别；
获取目标客户的持有的金融产品的产品数据，并根据所述目标客户的持有的金融产品的产品数据和对持有的金融产品的满意度级别，为所述目标客户选择待推荐的金融产品；
获取所述待推荐的金融产品对应的现有客户所使用的接触媒介，并根据获取的接触媒介预测客户的推荐接触媒介；
将所述金融产品通过所述常用接触媒介推荐给所述目标客户。
进一步地，所述从事项信息表中提取目标客户对应的事项特征的步骤包括：
根据目标客户的身份信息从事项信息表中提取目标客户的所有事项，并从提取的事项中过滤出属于预设事项类别的事项；
从所述属于预设事项类别的事项中提取出对应的事项特征。
进一步地，所述获取所述待推荐的金融产品对应的现有客户所使用的接触媒介，并根据获取的接触媒介预测客户的推荐接触媒介的步骤包括：
获取所述待推荐的金融产品对应的每一现有客户所使用的接触媒介；
对获取的每一现有客户所使用的接触媒介进行统计，确定所述待推荐的产品在各个接触媒介上的分布概率，将分布概率最大的接触媒介作为所述目标客户的常用接触媒介。
进一步地，所述预设的分类模型为支持向量机分类模型，所述处理器还用于执行所述金融产品推荐程序，以在从事项信息表中提取目标客户对应的事项特征的步骤之前还实现如下步骤：
获取事项特征训练集，所述事项特征训练集中的每一个事项特征有对应的满意度级别；
根据所述事项特征训练集训练所述支持向量机分类模型，以获取所述分类模型的模型参数。
本发明本申请计算机可读存储介质具体实施方式与上述金融产品推荐装置和方法各实施例基本相同，在此不作累述。
需要说明的是，上述本申请实施例序号仅仅为了描述，不代表实施例的优劣。并且本文中的术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含，从而使得包括一系列要素的过程、装置、物品或者方法不仅包括那些要素，而且还包括没有明确列出的其他要素，或者是还包括为这种过程、装置、物品或者方法所固有的要素。在没有更多限制的情况下，由语句“包括一个……”限定的要素，并不排除在包括该要素的过程、装置、物品或者方法中还存在另外的相同要素。
通过以上的实施方式的描述，本领域的技术人员可以清楚地了解到上述实施例方法可借助软件加必需的通用硬件平台的方式来实现，当然也可以通过硬件，但很多情况下前者是更佳的实施方式。基于这样的理解，本申请的技术方案本质上或者说对现有技术做出贡献的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如上所述的一个存储介质(如ROM/RAM、磁碟、光盘)中，包括若干指令用以使得一台终端设备(可以是手机，计算机，服务器，或者网络设备等)执行本申请各个实施例所述的方法。
以上仅为本申请的优选实施例，并非因此限制本申请的专利范围，凡是利用本申请说明书及附图内容所作的等效结构或等效流程变换，或直接或间接运用在其他相关的技术领域，均同理包括在本申请的专利保护范围内。

Claims (20)

  1. 一种金融产品推荐装置，其特征在于，所述装置包括存储器和处理器，所述存储器上存储有可在所述处理器上运行的金融产品推荐程序，所述金融产品推荐程序被所述处理器执行时实现如下步骤：
    从事项信息表中提取目标客户对应的事项特征；
    按照预设的分类模型，根据目标客户的事项特征计算目标客户对持有的金融产品的满意度级别；
    获取目标客户的持有的金融产品的产品数据，并根据所述目标客户的持有的金融产品的产品数据和对持有的金融产品的满意度级别，为所述目标客户选择待推荐的金融产品；
    获取所述待推荐的金融产品对应的现有客户所使用的接触媒介，并根据获取的接触媒介预测客户的推荐接触媒介；
    将所述金融产品通过所述常用接触媒介推荐给所述目标客户。
  2. 根据权利要求1所述的金融产品推荐装置，其特征在于，所述从事项信息表中提取目标客户对应的事项特征的步骤包括：
    根据目标客户的身份信息从事项信息表中提取目标客户的所有事项，并从提取的事项中过滤出属于预设事项类别的事项；
    从所述属于预设事项类别的事项中提取出对应的事项特征。
  3. 根据权利要求1所述的金融产品推荐装置，其特征在于，所述获取所述待推荐的金融产品对应的现有客户所使用的接触媒介，并根据获取的接触媒介预测客户的推荐接触媒介的步骤包括：
    获取所述待推荐的金融产品对应的每一现有客户所使用的接触媒介；
    对获取的每一现有客户所使用的接触媒介进行统计，确定所述待推荐的产品在各个接触媒介上的分布概率，将分布概率最大的接触媒介作为所述目标客户的常用接触媒介。
  4. 根据权利要求1所述的金融产品推荐装置，其特征在于，所述预设的分类模型为支持向量机分类模型，所述处理器还用于执行所述金融产品推荐程序，以在从事项信息表中提取目标客户对应的事项特征的步骤之前还实现如下步骤：
    获取事项特征训练集，所述事项特征训练集中的每一个事项特征有对应的满意度级别；
    根据所述事项特征训练集训练所述支持向量机分类模型，以获取所述分类模型的模型参数。
  5. 根据权利要求2所述的金融产品推荐装置，其特征在于，所述预设的分类模型为支持向量机分类模型，所述处理器还用于执行所述金融产品推荐程序，以在从事项信息表中提取目标客户对应的事项特征的步骤之前还实现如下步骤：
    获取事项特征训练集，所述事项特征训练集中的每一个事项特征有对应的满意度级别；
    根据所述事项特征训练集训练所述支持向量机分类模型，以获取所述分类模型的模型参数。
  6. 根据权利要求3所述的金融产品推荐装置，其特征在于，所述预设的分类模型为支持向量机分类模型，所述处理器还用于执行所述金融产品推荐程序，以在从事项信息表中提取目标客户对应的事项特征的步骤之前还实现如下步骤：
    获取事项特征训练集，所述事项特征训练集中的每一个事项特征有对应的满意度级别；
    根据所述事项特征训练集训练所述支持向量机分类模型，以获取所述分类模型的模型参数。
  7. 根据权利要求4所述的金融产品推荐装置，其特征在于，所述接触媒介包括PC客户端、APP客户端以及电话营销渠道。
  8. 一种金融产品推荐方法，其特征在于，所述金融产品推荐方法包括：
    从事项信息表中提取目标客户对应的事项特征；
    按照预设的分类模型，根据目标客户的事项特征计算目标客户对持有的金融产品的满意度级别；
    获取目标客户的持有的金融产品的产品数据，并根据所述目标客户的持有的金融产品的产品数据和对持有的金融产品的满意度级别，为所述目标客户选择待推荐的金融产品；
    获取所述待推荐的金融产品对应的现有客户所使用的接触媒介，并根据获取的接触媒介预测客户的推荐接触媒介；
    将所述金融产品通过所述常用接触媒介推荐给所述目标客户。
  9. 根据权利要求8所述的金融产品推荐方法，其特征在于，所述从事项信息表中提取目标客户对应的事项特征的步骤包括：
    根据目标客户的身份信息从事项信息表中提取目标客户的所有事项，并从提取的事项中过滤出属于预设事项类别的事项；
    从所述属于预设事项类别的事项中提取出对应的事项特征。
  10. 根据权利要求8所述的金融产品推荐方法，其特征在于，所述获取所述待推荐的金融产品对应的现有客户所使用的接触媒介，并根据获取的接触媒介预测客户的推荐接触媒介的步骤包括：
    获取所述待推荐的金融产品对应的每一现有客户所使用的接触媒介；
    对获取的每一现有客户所使用的接触媒介进行统计，确定所述待推荐的产品在各个接触媒介上的分布概率，将分布概率最大的接触媒介作为所述目标客户的常用接触媒介。
  11. 根据权利要求8所述的金融产品推荐方法，其特征在于，所述预设的分类模型为支持向量机分类模型，所述从事项信息表中提取目标客户对应的事项特征的步骤之前，所述方法还包括如下步骤：
    获取事项特征训练集，所述事项特征训练集中的每一个事项特征有对应的满意度级别；
    根据所述事项特征训练集训练所述支持向量机分类模型，以获取所述分类模型的模型参数。
  12. 根据权利要求9所述的金融产品推荐方法，其特征在于，所述预设的分类模型为支持向量机分类模型，所述从事项信息表中提取目标客户对应的事项特征的步骤之前，所述方法还包括如下步骤：
    获取事项特征训练集，所述事项特征训练集中的每一个事项特征有对应的满意度级别；
    根据所述事项特征训练集训练所述支持向量机分类模型，以获取所述分类模型的模型参数。
  13. 根据权利要求10所述的金融产品推荐方法，其特征在于，所述预设的分类模型为支持向量机分类模型，所述从事项信息表中提取目标客户对应的事项特征的步骤之前，所述方法还包括如下步骤：
    获取事项特征训练集，所述事项特征训练集中的每一个事项特征有对应 的满意度级别；
    根据所述事项特征训练集训练所述支持向量机分类模型，以获取所述分类模型的模型参数。
  14. 根据权利要求11所述的金融产品推荐方法，其特征在于，所述接触媒介包括PC客户端、APP客户端以及电话营销渠道。
  15. 一种计算机可读存储介质，其特征在于，所述计算机可读存储介质上存储有金融产品推荐程序，所述金融产品推荐程序可被一个或者多个处理器执行，以实现如下步骤：
    从事项信息表中提取目标客户对应的事项特征；
    按照预设的分类模型，根据目标客户的事项特征计算目标客户对持有的金融产品的满意度级别；
    获取目标客户的持有的金融产品的产品数据，并根据所述目标客户的持有的金融产品的产品数据和对持有的金融产品的满意度级别，为所述目标客户选择待推荐的金融产品；
    获取所述待推荐的金融产品对应的现有客户所使用的接触媒介，并根据获取的接触媒介预测客户的推荐接触媒介；
    将所述金融产品通过所述常用接触媒介推荐给所述目标客户。
  16. 根据权利要求15所述的计算机可读存储介质，其特征在于，所述从事项信息表中提取目标客户对应的事项特征的步骤包括：
    根据目标客户的身份信息从事项信息表中提取目标客户的所有事项，并从提取的事项中过滤出属于预设事项类别的事项；
    从所述属于预设事项类别的事项中提取出对应的事项特征。
  17. 根据权利要求15所述的计算机可读存储介质，其特征在于，所述获取所述待推荐的金融产品对应的现有客户所使用的接触媒介，并根据获取的接触媒介预测客户的推荐接触媒介的步骤包括：
    获取所述待推荐的金融产品对应的每一现有客户所使用的接触媒介；
    对获取的每一现有客户所使用的接触媒介进行统计，确定所述待推荐的产品在各个接触媒介上的分布概率，将分布概率最大的接触媒介作为所述目标客户的常用接触媒介。
  18. 根据权利要求15所述的计算机可读存储介质，其特征在于，所述预 设的分类模型为支持向量机分类模型，所述金融产品推荐程序可被一个或者多个处理器执行，以在从事项信息表中提取目标客户对应的事项特征的步骤之前，还实现如下步骤：
    获取事项特征训练集，所述事项特征训练集中的每一个事项特征有对应的满意度级别；
    根据所述事项特征训练集训练所述支持向量机分类模型，以获取所述分类模型的模型参数。
  19. 根据权利要求16所述的计算机可读存储介质，其特征在于，所述预设的分类模型为支持向量机分类模型，所述金融产品推荐程序可被一个或者多个处理器执行，以在从事项信息表中提取目标客户对应的事项特征的步骤之前，还实现如下步骤：
    获取事项特征训练集，所述事项特征训练集中的每一个事项特征有对应的满意度级别；
    根据所述事项特征训练集训练所述支持向量机分类模型，以获取所述分类模型的模型参数。
  20. 根据权利要求17所述的计算机可读存储介质，其特征在于，所述预设的分类模型为支持向量机分类模型，所述金融产品推荐程序可被一个或者多个处理器执行，以在从事项信息表中提取目标客户对应的事项特征的步骤之前，还实现如下步骤：
    获取事项特征训练集，所述事项特征训练集中的每一个事项特征有对应的满意度级别；
    根据所述事项特征训练集训练所述支持向量机分类模型，以获取所述分类模型的模型参数。
PCT/CN2018/077631 2017-10-09 2018-02-28 金融产品推荐装置、方法及计算机可读存储介质 WO2019071906A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201710930686.8 2017-10-09
CN201710930686.8A CN108399565A (zh) 2017-10-09 2017-10-09 金融产品推荐装置、方法及计算机可读存储介质

Publications (1)

Publication Number Publication Date
WO2019071906A1 true WO2019071906A1 (zh) 2019-04-18

Family

ID=63094507

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/077631 WO2019071906A1 (zh) 2017-10-09 2018-02-28 金融产品推荐装置、方法及计算机可读存储介质

Country Status (2)

Country Link
CN (1) CN108399565A (zh)
WO (1) WO2019071906A1 (zh)

Families Citing this family (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109410027B (zh) * 2018-08-29 2022-09-09 中国建设银行股份有限公司 一种基于特征识别的金融信息处理方法、智能终端及介质
CN109255715A (zh) * 2018-09-03 2019-01-22 平安科技（深圳）有限公司 电子装置、产品推荐方法和计算机可读存储介质
CN109447728A (zh) * 2018-09-07 2019-03-08 平安科技（深圳）有限公司 金融产品推荐方法、装置、计算机设备及存储介质
CN110009159A (zh) * 2019-04-11 2019-07-12 湖北风口网络科技有限公司 基于网络大数据的金融借贷需求预测方法及***
CN110135942A (zh) * 2019-04-12 2019-08-16 深圳壹账通智能科技有限公司 产品推荐方法、装置及计算机可读存储介质
CN110415123A (zh) * 2019-06-06 2019-11-05 财付通支付科技有限公司 金融产品推荐方法、装置和设备及计算机存储介质
CN111429232A (zh) * 2020-04-12 2020-07-17 中信银行股份有限公司 产品推荐方法、装置和电子设备及计算机可读存储介质
CN111858686B (zh) 2020-07-08 2024-05-28 深圳市富途网络科技有限公司 数据显示方法、装置、终端设备及存储介质
CN112669136A (zh) * 2020-12-10 2021-04-16 前海飞算科技(深圳)有限公司 基于大数据的金融产品推荐方法、***、设备及存储介质
CN113468421A (zh) * 2021-06-29 2021-10-01 平安信托有限责任公司 基于向量匹配技术的产品推荐方法、装置、设备及介质
CN115619436A (zh) * 2022-11-14 2023-01-17 平安银行股份有限公司 一种金融产品的推荐方法、装置、计算机设备及存储介质

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20140114674A1 (en) * 2012-10-22 2014-04-24 Robert M. Krughoff Health Insurance Plan Comparison Tool
CN104424247A (zh) * 2013-08-28 2015-03-18 北京闹米科技有限公司 一种产品信息过滤推荐方法和装置
CN106991609A (zh) * 2016-01-21 2017-07-28 阿里巴巴集团控股有限公司 投资产品的推荐方法和装置
CN107194754A (zh) * 2017-04-11 2017-09-22 美林数据技术股份有限公司 基于混合协同过滤的券商产品推荐方法

Family Cites Families (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104008109B (zh) * 2013-02-26 2017-11-14 南京邮电大学 基于用户兴趣的Web信息推送服务***
US20150032523A1 (en) * 2013-07-29 2015-01-29 Bank Of America Corporation Credit source recommendation based on product level data analysis
CN104463630B (zh) * 2014-12-11 2015-08-26 新一站保险代理有限公司 一种基于网购保险产品特性的产品推荐方法及***
CN104835066A (zh) * 2015-05-25 2015-08-12 北京京东尚科信息技术有限公司 一种投放渠道的选择方法及***
CN106339469A (zh) * 2016-08-29 2017-01-18 乐视控股（北京）有限公司 数据推荐方法及装置
CN106600369A (zh) * 2016-12-09 2017-04-26 广东奡风科技股份有限公司 基于朴素贝叶斯分类的银行金融产品实时推荐***和方法
TWM543416U (zh) * 2016-12-29 2017-06-11 First Commercial Bank Co Ltd 智能產品行銷系統
CN106682949A (zh) * 2016-12-31 2017-05-17 ***通信集团江苏有限公司 业务推荐方法、业务信息接收方法、装置及***
CN106780052A (zh) * 2017-01-10 2017-05-31 上海诺悦智能科技有限公司 基于分类客户行为分析的保险服务推荐方法及***
CN106815364A (zh) * 2017-01-24 2017-06-09 百度在线网络技术（北京）有限公司 内容推送方法和装置
CN106991598A (zh) * 2017-04-07 2017-07-28 北京百分点信息科技有限公司 数据推送方法及其***
CN107025310A (zh) * 2017-05-17 2017-08-08 长春嘉诚信息技术股份有限公司 一种自动实时新闻推荐方法

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20140114674A1 (en) * 2012-10-22 2014-04-24 Robert M. Krughoff Health Insurance Plan Comparison Tool
CN104424247A (zh) * 2013-08-28 2015-03-18 北京闹米科技有限公司 一种产品信息过滤推荐方法和装置
CN106991609A (zh) * 2016-01-21 2017-07-28 阿里巴巴集团控股有限公司 投资产品的推荐方法和装置
CN107194754A (zh) * 2017-04-11 2017-09-22 美林数据技术股份有限公司 基于混合协同过滤的券商产品推荐方法

Also Published As

Publication number Publication date
CN108399565A (zh) 2018-08-14

Similar Documents

Publication Publication Date Title
WO2019071906A1 (zh) 金融产品推荐装置、方法及计算机可读存储介质
US11977995B2 (en) Machine learning artificial intelligence system for predicting hours of operation
US20200272917A1 (en) Method, apparatus, and computer program product for determining a provider return rate
WO2019061994A1 (zh) 电子装置、保险产品推荐方法、***及计算机可读存储介质
US20220180379A1 (en) Transaction-based information processing system, method, and article
CN108256537A (zh) 一种用户性别预测方法和***
WO2019179030A1 (zh) 产品购买预测方法、服务器及存储介质
CN108509458B (zh) 一种业务对象识别方法及装置
CN110457364B (zh) 用户信息视图生成的方法及装置
US20220382794A1 (en) System and method for programmatic generation of attribute descriptors
CN115391669B (zh) 一种智能推荐方法、装置、电子设备
JP2018197985A (ja) レシートを用いたプロジェクトのためのレシート解析システム、方法、及びプログラム
US20180075468A1 (en) Systems and methods for merchant business intelligence tools
US9830584B2 (en) Display an item detail with a receipt snippet
CN112511632B (zh) 基于多源数据的对象推送方法、装置、设备及存储介质
CN104331395A (zh) 从文本中识别中文商品名称的方法和装置
US20190034943A1 (en) Spend engagement relevance tools
CN113807066A (zh) 一种图表生成方法、装置及电子设备
CN108959289B (zh) 网站类别获取方法和装置
US20210377201A1 (en) System and Method for Tagging Data
US20150331878A1 (en) Ranking autocomplete results based on a business cohort
US10152754B2 (en) System and method for small business owner identification
CN107357847B (zh) 数据处理方法及其装置
Dorokhova et al. Comparison of Pharmacy Websites: An Integrated Approach Based on Consumer Perception and Technical Parameters.
US20140019205A1 (en) Impact measurement based on data distributions

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: 18865390

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 28/09/2020)

122 Ep: pct application non-entry in european phase

Ref document number: 18865390

Country of ref document: EP

Kind code of ref document: A1