CN116308615A

CN116308615A - Product recommendation method and device, electronic equipment and storage medium

Info

Publication number: CN116308615A
Application number: CN202310007245.6A
Authority: CN
Inventors: 任德鑫; 陈仁龙; 李津堂; 郭钦新
Original assignee: Industrial and Commercial Bank of China Ltd ICBC
Current assignee: Industrial and Commercial Bank of China Ltd ICBC
Priority date: 2023-01-04
Filing date: 2023-01-04
Publication date: 2023-06-23

Abstract

The disclosure provides a product recommendation method, a device, electronic equipment and a storage medium, which can be applied to the technical field of computers. The method comprises the following steps: responding to a recommendation request of a financial platform, and acquiring target historical data of a target user in a preset time period from a user behavior database, wherein the target historical data in the user behavior database are acquired through authorization of the target user, the target historical data comprise target product interaction data, and the target user is determined according to the target product interaction data; determining the portrait grading of the target user according to the target historical data; determining a target product matched with a target user from a product library according to the target product interaction data; determining a product to be recommended from a first product according to the product similarity between the target product and the first product in the product library, wherein the first product is other products except the target product in the product library; and recommending the product to be recommended to the target user according to the portrait grading.

Description

Product recommendation method and device, electronic equipment and storage medium

Technical Field

The present disclosure relates to the field of computer technology, and more particularly, to a product recommendation method, apparatus, electronic device, storage medium, and computer program product.

Background

In the related art, in order to accelerate strategic deployment of digital transformation, for a financial institution with intense competition, improving user viscosity is a key for improving market competitiveness, and providing accurate marketing for users and formulating a suitable product recommendation strategy are important strategies for increasing a customer acquisition channel. Therefore, how to discover valuable user groups from massive user transaction information is a key problem to be solved in the present urgent need. Among the numerous methods of user relationship management analysis, the RFM (Reduction-Frequency-Monetary) model is widely mentioned. The RFM model is an important tool and means for measuring the value of a user and the capability of the user to create the profit, and describes the self value condition of the user through three indexes of recent purchasing behavior of the user, the total frequency of purchasing and the total consumption.

In the process of implementing the disclosed concept, the inventor finds that at least the following problems exist in the related art: the product recommendation method in the related art cannot accurately recommend the product in the financial platform to the user, so that the user viscosity cannot be improved well.

Disclosure of Invention

In view of this, the present disclosure provides a product recommendation method, apparatus, electronic device, storage medium, and computer program product.

One aspect of the present disclosure provides a product recommendation method, comprising:

responding to a recommendation request of a financial platform, and acquiring target historical data of a target user in a preset time period from a user behavior database, wherein the target historical data in the user behavior database are acquired through authorization of the target user, the target historical data comprise target product interaction data, and the target user is determined according to the target product interaction data;

determining the portrait grading of the target user according to the target historical data;

determining a target product matched with a target user from a product library according to the target product interaction data;

determining a product to be recommended from a first product according to the product similarity between the target product and the first product in the product library, wherein the first product is other products except the target product in the product library;

and recommending the product to be recommended to the target user according to the portrait grading.

According to an embodiment of the present disclosure, determining a product to be recommended from a first product in a product library according to a product similarity of a target product and the first product includes:

according to the target product and the first product, respectively determining a target product vector of the target product and a first product vector of the first product;

Determining the product similarity of the target product and the first product according to the target product vector and the first product vector;

under the condition that the similarity of the products is larger than a preset threshold value, determining the first product as a second product, and obtaining a second product set;

sorting the second products according to the preference degree of the target users and the second products to obtain sorting results;

and determining the product to be recommended from the second product according to the sorting result.

According to an embodiment of the present disclosure, determining a portrait rating of a target user according to target history data includes:

according to the target historical data, determining a dimension index value of a target user under a plurality of dimension indexes;

determining the dimension index score value of each dimension index according to the respective dimension index weights of the plurality of dimension indexes to obtain a plurality of dimension index score values;

determining a representation score for the target user based on the plurality of dimension index score values;

based on the portrait score, a portrait score level of the target user is determined.

According to an embodiment of the present disclosure, each dimension index includes a plurality of sub-dimension indexes;

according to the target historical data, determining a dimension index value of the target user under a plurality of dimension indexes comprises:

Determining respective sub-dimension index values of a plurality of sub-dimension indexes of each dimension index of a target user according to the target historical data;

determining a sub-dimension index score of each sub-dimension index according to the sub-dimension index weight of each sub-dimension index;

and determining the dimension index value of the target user under the plurality of dimension indexes according to the plurality of sub-dimension index scores.

According to an embodiment of the disclosure, the dimension index comprises a user interaction dimension index, and the user interaction dimension index comprises a plurality of sub-user interaction dimension indexes;

the product recommendation method further comprises the following steps:

acquiring product interaction data of a user in a preset time period from a user behavior database, wherein the product interaction data in the user behavior database is acquired through authorization of a target user;

determining respective sub-user interaction dimension index values of the plurality of sub-user interaction dimension indexes according to the product interaction data;

determining respective sub-user interaction dimension index scores of the plurality of sub-user interaction dimension index values by using a chi-square box dividing method;

calculating respective sub-user interaction weight values of the plurality of sub-user interaction dimension indexes by using an XGboost algorithm;

determining a user interaction dimension index value of the user interaction dimension index according to the plurality of sub-user interaction dimension index scores and the plurality of sub-user interaction dimension index weight values;

And determining the target user from the users according to the user interaction dimension index value.

According to the embodiment of the disclosure, the sub-user interaction dimension indexes comprise a near sub-dimension index, a frequency sub-dimension index and a quota sub-dimension index, wherein the near sub-dimension index represents the latest time that a user transacts on a financial platform, the frequency sub-dimension index represents the number of times that the user transacts on the financial platform in a preset time period, and the quota sub-dimension index represents the quota of the user transacting on the financial platform in the preset time period.

Another aspect of the present disclosure provides a product recommendation device, comprising:

the first acquisition module is used for responding to a recommendation request of the financial platform and acquiring target historical data of a target user in a preset time period from the user behavior database, wherein the target historical data in the user behavior database are acquired through authorization of the target user, the target historical data comprise target product interaction data, and the target user is determined according to the target product interaction data;

the first determining module is used for determining the portrait grading of the target user according to the target historical data;

the second determining module is used for determining a target product matched with the target user from the product library according to the target product interaction data;

The third determining module is used for determining a product to be recommended from the first product according to the product similarity between the target product and the first product in the product library, wherein the first product is other products except the target product in the product library;

and the recommending module is used for recommending the product to be recommended to the target user according to the portrait grading.

Another aspect of the present disclosure provides an electronic device, comprising: one or more processors; and a memory for storing one or more programs, wherein the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the product recommendation method.

Another aspect of the present disclosure provides a computer-readable storage medium having stored thereon executable instructions that, when executed by a processor, cause the processor to perform the product recommendation method described above.

Another aspect of the present disclosure provides a computer program product comprising computer executable instructions which, when executed, are configured to implement the product recommendation method described above.

According to the embodiment of the disclosure, because the recommendation request responding to the financial platform is adopted, target historical data of a target user in a preset time period is obtained from a user behavior database, wherein the target historical data in the user behavior database is acquired through authorization of the target user, the target historical data comprises target product interaction data, and the target user is determined according to the target product interaction data; determining the portrait grading of the target user according to the target historical data; determining a target product matched with a target user from a product library according to the target product interaction data; determining a product to be recommended from a first product according to the product similarity between the target product and the first product in the product library, wherein the first product is other products except the target product in the product library; according to the portrait grading, the technical means of recommending the products to be recommended to the target users can comprehensively conduct grading management on the target users through the portrait grading, and different target products are recommended according to the target users with different portrait grading, so that the fineness of product recommendation is improved, and the user viscosity is effectively improved. Therefore, the technical problem that the product recommending method in the related art cannot accurately recommend the product in the financial platform to the user, so that the viscosity of the user cannot be improved well is at least partially solved.

Drawings

The above and other objects, features and advantages of the present disclosure will become more apparent from the following description of embodiments thereof with reference to the accompanying drawings in which:

FIG. 1 schematically illustrates an exemplary system architecture to which a product recommendation method may be applied, according to an embodiment of the present disclosure;

FIG. 2 schematically illustrates a flow chart of a product recommendation method according to an embodiment of the present disclosure;

FIG. 3 schematically illustrates a flow chart of a product recommendation method according to yet another embodiment of the present disclosure;

FIG. 4 schematically illustrates a block diagram of a product recommendation device, according to an embodiment of the present disclosure; and

fig. 5 schematically shows a block diagram of an electronic device adapted to implement the method described above, according to an embodiment of the disclosure.

Detailed Description

Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. It should be understood that the description is only exemplary and is not intended to limit the scope of the present disclosure. In the following detailed description, for purposes of explanation, numerous specific details are set forth in order to provide a thorough understanding of the embodiments of the present disclosure. It may be evident, however, that one or more embodiments may be practiced without these specific details. In addition, in the following description, descriptions of well-known structures and techniques are omitted so as not to unnecessarily obscure the concepts of the present disclosure.

The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the disclosure. The terms "comprises," "comprising," and/or the like, as used herein, specify the presence of stated features, steps, operations, and/or components, but do not preclude the presence or addition of one or more other features, steps, operations, or components.

All terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art unless otherwise defined. It should be noted that the terms used herein should be construed to have meanings consistent with the context of the present specification and should not be construed in an idealized or overly formal manner.

Where expressions like at least one of "A, B and C, etc. are used, the expressions should generally be interpreted in accordance with the meaning as commonly understood by those skilled in the art (e.g.," a system having at least one of A, B and C "shall include, but not be limited to, a system having a alone, B alone, C alone, a and B together, a and C together, B and C together, and/or A, B, C together, etc.). Where a formulation similar to at least one of "A, B or C, etc." is used, in general such a formulation should be interpreted in accordance with the ordinary understanding of one skilled in the art (e.g. "a system with at least one of A, B or C" would include but not be limited to systems with a alone, B alone, C alone, a and B together, a and C together, B and C together, and/or A, B, C together, etc.).

In the technical scheme of the disclosure, the acquisition, storage, application and the like of the related personal information of the user all conform to the regulations of related laws and regulations, necessary security measures are taken, and the public order harmony is not violated.

In the technical scheme of the disclosure, the authorization or consent of the user is obtained before the personal information of the user is obtained or acquired.

In the related art, the RFM model has two obvious defects, namely, the consumption frequency and the quantity relation in a subjective layer definition time period are used for distinguishing light and heavy users, and the user log information of which time range is intercepted in detail is reasonable as modeling sample data and is not subjected to in-depth comparison; secondly, for the RFM model under the user scoring mechanism, the index weight construction method mainly takes an Analytic Hierarchy Process (AHP) construction feature judgment matrix or entropy weight method, voting method or weighted average of human factors as a main part, but the analytic hierarchy process also has artificial factor interference based on expert experience and the like when constructing an initialization matrix, and lacks necessary theoretical verification algorithm support.

Traditionally, early financial institutions had relatively few online financial product types and low user transaction frequency, and more O2O mode (namely online and offline product linkage mode) and MGM mode (old zone new promotion mode) are adopted in product marketing strategies, so that users are essentially attracted through a certain incentive mode, forward transmission of the product really interested by the users is not provided, and the product is difficult to maintain for later continuous promotion; meanwhile, in online financial scenes such as banks, the problems of low user transaction frequency, relatively few product types, a plurality of transaction limiting factors and the like exist, and the traditional recommendation algorithm such as collaborative filtering recommendation algorithm based on users or articles cannot be directly suitable for online product recommendation in the financial scenes such as banks.

In view of the above, the embodiments of the present disclosure provide a product recommendation method. The method comprises the steps of responding to a recommendation request of a financial platform, acquiring target historical data of a target user in a preset time period from a user behavior database, wherein the target historical data in the user behavior database are acquired through authorization of the target user, the target historical data comprise target product interaction data, and the target user is determined according to the target product interaction data; determining the portrait grading of the target user according to the target historical data; determining a target product matched with a target user from a product library according to the target product interaction data; determining a product to be recommended from a first product according to the product similarity between the target product and the first product in the product library, wherein the first product is other products except the target product in the product library; and recommending the product to be recommended to the target user according to the portrait grading.

Fig. 1 schematically illustrates an exemplary system architecture 100 in which a product recommendation method may be applied according to an embodiment of the present disclosure. It should be noted that fig. 1 is only an example of a system architecture to which embodiments of the present disclosure may be applied to assist those skilled in the art in understanding the technical content of the present disclosure, but does not mean that embodiments of the present disclosure may not be used in other devices, systems, environments, or scenarios.

As shown in fig. 1, a system architecture 100 according to this embodiment may include

terminal devices

101, 102, 103, a network 104, and a server 105. The network 104 is used as a medium to provide communication links between the

terminal devices

101, 102, 103 and the server 105. The network 104 may include various connection types, such as wired and/or wireless communication links, and the like.

The user may interact with the server 105 via the network 104 using the

terminal devices

101, 102, 103 to receive or send messages or the like. Various communication client applications may be installed on the

terminal devices

101, 102, 103, such as shopping class applications, web browser applications, search class applications, instant messaging tools, mailbox clients and/or social platform software, to name a few.

The

terminal devices

101, 102, 103 may be a variety of electronic devices having a display screen and supporting web browsing, including but not limited to smartphones, tablets, laptop and desktop computers, and the like.

The server 105 may be a server providing various services, such as a background management server (by way of example only) providing support for websites browsed by users using the

terminal devices

101, 102, 103. The background management server may analyze and process the received data such as the user request, and feed back the processing result (e.g., the web page, information, or data obtained or generated according to the user request) to the terminal device.

It should be noted that, the product recommendation method provided by the embodiments of the present disclosure may be generally performed by the server 105. Accordingly, the product recommendation device provided by the embodiments of the present disclosure may be generally disposed in the server 105. The product recommendation method provided by the embodiments of the present disclosure may also be performed by a server or a server cluster that is different from the server 105 and is capable of communicating with the

terminal devices

101, 102, 103 and/or the server 105. Accordingly, the product recommendation apparatus provided by the embodiments of the present disclosure may also be provided in a server or a server cluster that is different from the server 105 and is capable of communicating with the

terminal devices

101, 102, 103 and/or the server 105. Alternatively, the product recommendation method provided by the embodiment of the present disclosure may be performed by the

terminal device

101, 102, or 103, or may be performed by another terminal device other than the

terminal device

101, 102, or 103. Accordingly, the product recommendation apparatus provided by the embodiments of the present disclosure may also be provided in the

terminal device

101, 102, or 103, or in another terminal device different from the

terminal device

101, 102, or 103.

For example, the target history data may be stored in a user behavior database and may be imported into the terminal apparatus 101. Then, the terminal device 101 may locally perform the product recommendation method provided by the embodiment of the present disclosure, or transmit the target history data to other terminal devices, servers, or server clusters, and perform the product recommendation method provided by the embodiment of the present disclosure by the other terminal devices, servers, or server clusters that receive the target history data.

It should be understood that the number of terminal devices, networks and servers in fig. 1 is merely illustrative. There may be any number of terminal devices, networks, and servers, as desired for implementation.

Fig. 2 schematically illustrates a flow chart of a product recommendation method according to an embodiment of the present disclosure.

As shown in fig. 2, the method includes operations S201 to S205.

In operation S201, in response to a recommendation request from the financial platform, target historical data of a target user in a preset time period is obtained from a user behavior database, wherein the target historical data in the user behavior database is collected through authorization of the target user, the target historical data includes target product interaction data, and the target user is determined according to the target product interaction data.

According to embodiments of the present disclosure, the financial platform may be a platform for financial transactions, which may be transactions to purchase funds, stocks, futures, etc., that are offered by a financial institution.

According to embodiments of the present disclosure, the target users may be users having different degrees of interaction at the financial platform, as determined from the target product interaction data. The number of target users may be 1 or plural.

According to an embodiment of the present disclosure, the target history data may be data generated by a user's operation in the financial platform within a preset period of time, for example, may be data of a user purchasing a target product in the financial platform, or may be data of a target user browsing a target product in the financial platform.

According to embodiments of the present disclosure, the user behavior database may collect data generated by user operations in the financial platform based on user authorization in the financial platform.

According to an embodiment of the present disclosure, the recommendation request may be a request issued according to a timing task preset in the financial platform, or may be a request issued in response to an input of the user in the financial platform.

In operation S202, a portrait rating of the target user is determined based on the target history data.

According to the embodiment of the disclosure, since the target historical data is generated based on the operation of the target user in the financial platform, the target historical data can characterize the value degree of the target user for the financial platform to a certain extent, and therefore, the portrait grading level of the target user can be determined by using the target historical data.

According to embodiments of the present disclosure, portrait scoring levels may be multi-level, and different levels of portrait scoring levels may characterize different degrees of value to a user in a financial platform. In the case where there are a plurality of target users, different target users may be different portrayal score levels, and the behavior preferences, basic information, and the like of the target users in each portrayal score level may be similar.

According to the embodiment of the disclosure, the target user may have a unique user identifier, which may be a registered account number or a registered mobile phone number of the target user on the financial platform, or may be an identity number of the target user.

In operation S203, a target product matching the target user is determined from the product library according to the target product interaction data.

According to embodiments of the present disclosure, the product library may include all of the products in the financial platform, which may be fund-based products, stock-based products, commodity-based products, and the like.

According to the embodiment of the disclosure, the target product may be a product purchased by the target user or a product that the target user browses a certain number of times.

According to the embodiment of the disclosure, since the target product interaction data is interaction data of the target user on the financial platform, the target product can be determined to be purchased or browsed for a certain number of times on the financial platform by the target user from the target product interaction data. In the case that the number of target users is plural, the target products matched with all the target users in each portrait grading level can be determined as target product sets according to the portrait grading levels corresponding to different target users, that is, each portrait grading level corresponds to one target product set.

In operation S204, a product to be recommended is determined from the first products according to the product similarity between the target product and the first products in the product library, wherein the first products are other products in the product library than the target product.

According to the embodiment of the disclosure, since the products in the product library are financial products, the product similarity between the target product and the first product can be calculated according to the product characteristics of the target product and the first product.

According to the embodiment of the disclosure, the product to be recommended can be determined from the first products with the product similarity larger than the preset threshold. In the case that the target users are multiple, corresponding target product sets in different portrait grading grades can correspond to respective product sets to be recommended.

In operation S205, a product to be recommended is recommended to the target user according to the portrait rating.

According to the embodiment of the disclosure, since the product to be recommended is determined according to the product similarity between the target product and the first product, the product to be recommended and the target product are similar, and thus the target user has a high probability of selecting the product to be recommended.

According to the embodiment of the disclosure, since different target users have respective portrait grading levels, the products to be recommended can be recommended to the target users according to the corresponding portrait grading levels of the products to be recommended.

According to the embodiment of the disclosure, the respective product parameter information of the target product and the first product can be used for determining the respective target product vector and the first product vector of the target product and the first product.

According to embodiments of the present disclosure, the product similarity may be determined by calculating a cosine similarity between the target product vector and the first product vector.

According to the embodiment of the disclosure, when the similarity of the products is determined to be greater than the preset threshold, that is, the similarity of the first product and the target product is higher, the first product may be determined to be the second product, so as to obtain the second product set, and when the similarity of the products is determined to be less than or equal to the preset threshold, that is, the similarity of the first product and the target product is lower, the first product may be filtered, and the similarity between the next product and the target product may be continuously calculated.

According to embodiments of the present disclosure, a user's preference for a second product in a resulting second product set may be calculated for the second product, as in the following equation (1):

wherein I is _pv Representing the preference of the target user p for the second product v, I _pu Representing the preference of the target user p to the target product u, sim (V, u) representing the similarity between the second product V and the target product u, V _(p) Representing a second set of products.

According to the embodiment of the disclosure, the second products can be ranked according to the preference degree of the second products in order from large to small, a ranking result of the second products is obtained, the first K second products ranked in the ranking result are determined to be the products to be recommended, and K is smaller than the number of the second products in the second product set.

According to the embodiment of the disclosure, the second product set is determined by calculating the similarity between the products, and the product to be recommended is determined by the preference of the target user to the second product in the second product set, so that the product to be recommended can be determined more accurately, and the product to be recommended to the target user can be more accurate.

According to an embodiment of the disclosure, the plurality of dimension indexes may be dimension indexes of basic information dimension indexes, credit dimension indexes, asset dimension indexes, user interaction dimension indexes and the like of the target user, and may also be indexes of other dimensions.

According to the embodiment of the disclosure, the dimension index value of the target user under a plurality of dimension indexes can be determined from the target historical data, the dimension index weights of the plurality of dimension indexes can be determined according to actual services, and specifically, the total portrait score of the target user can be set to be 0-100 points of a standard scoring card.

According to the embodiment of the disclosure, the portrait score can be divided into a plurality of score intervals, each portrait score grade corresponds to a different score interval, the portrait score of the target user is obtained by adding a plurality of dimension index score values obtained according to the dimension index, and the portrait score grade of the target is determined according to the score interval of the portrait score of the target user.

According to an embodiment of the present disclosure, each dimension index may include a plurality of sub-dimension indexes, for example, the basic information dimension index may include sub-dimension indexes of an age stage, a marital status, academic information, a job type, and the like, the credit dimension index may include sub-dimension indexes of overdue records, credit account numbers, individual credit inquiry records, and the like, the credit dimension index may include sub-dimension indexes of a tax amount due to recent two years, and the like, the property dimension index may include sub-dimension indexes of an individual month average income section, an individual flowability ratio section, an individual asset debt ratio section, a surplus ratio section, and the like, the user interaction dimension index may include sub-dimension indexes of a total number of times a target user purchases a target product, a total amount of purchases a target product, a time of recent purchase of the target product, and the like, and the plurality of dimension indexes of the target user and an evaluation table of each dimension index may be as shown in table 1.

According to the embodiment of the disclosure, each sub-dimension index has a corresponding sub-dimension weight and a sub-dimension index value, a sub-dimension index score of each sub-dimension index can be determined according to the sub-dimension weight and the sub-dimension index value, and the sub-dimension index scores corresponding to a plurality of sub-dimension indexes under each dimension index can be added to obtain the dimension index value under each dimension index.

TABLE 1 Portrait score rating for target users

According to the embodiment of the disclosure, the target user can be evaluated more comprehensively by evaluating different dimensions of the target user, the image grading grade divided by the target user is finer, and the recommendation accuracy of recommending the product to be recommended is further improved.

According to an embodiment of the disclosure, the dimension index comprises a user interaction dimension index, and the user interaction dimension index comprises a plurality of sub-user interaction dimension indexes.

Fig. 3 schematically illustrates a flow chart of a product recommendation method according to a further embodiment of the present disclosure.

As shown in fig. 3, the method includes operations S301 to S306.

In operation S301, product interaction data of a user within a preset period of time is obtained from a user behavior database, wherein the product interaction data in the user behavior database is acquired through authorization of a target user.

In operation S302, respective sub-user interaction dimension index values of the plurality of sub-user interaction dimension indexes are determined according to the product interaction data.

In operation S303, a child user interaction dimension index score of each of the plurality of child user interaction dimension index values is determined using a chi-square binning method.

In operation S304, a respective sub-user interaction weight value of the plurality of sub-user interaction dimension indexes is calculated using an XGboost algorithm.

In operation S305, a user interaction dimension index value of the user interaction dimension index is determined according to the plurality of sub-user interaction dimension index scores and the plurality of sub-user interaction dimension index weight values.

In operation S306, a target user is determined from among the users according to the user interaction dimension index value.

According to the embodiment of the disclosure, the target user is mainly determined according to the product interaction data generated by the target user on the financial platform, so that the product interaction data of all users in a preset time period can be obtained from the user behavior database before recommendation is performed.

According to the embodiment of the disclosure, from the product interaction data, a sub-user interaction dimension index value of a sub-user interaction dimension index of each user in the interaction data may be determined, wherein the sub-user interaction dimension index includes a near sub-dimension index, a frequency sub-dimension index and a limit sub-dimension index, wherein the near sub-dimension index characterizes a time when the user last transacted on the financial platform, that is, a time when the user last purchases a product, the frequency sub-dimension index characterizes a transaction number of the user on the financial platform in a preset time period, that is, a total number of times the user purchases the product, and the limit sub-dimension index characterizes a transaction amount of the user on the financial platform in the preset time period, that is, a total amount of the purchased product.

According to the embodiment of the disclosure, the RFM model scoring generally adopts a 5-score mode, a common scoring mode has equal-frequency equidistant division intervals, and the scoring is performed according to understanding division of data and services or according to the number of scores of the data, and as the financial platform has fewer data transaction times and lacks the experience support of related services, each sub-user interaction dimension index value is scored by adopting a supervised chi-square box-division mode.

According to the embodiment of the disclosure, a chi-square threshold value θ is preset, sub-user interaction dimension index values corresponding to sub-user interaction dimension indexes are initialized, and ordered, namely, a near sub-dimension index R, a frequency sub-dimension index F and a limit sub-dimension index M are initialized to respectively obtain R ₀ 、F ₀ And M ₀ Ordering the near sub-dimension index, the frequency sub-dimension index and the limit sub-dimension index to obtain { R } ₀₁ 、R ₀₂ …R _0n }、{F ₀₁ 、F ₀₂ …F _0n }、{M ₀₁ 、M ₀₂ …M _0n And (2) calculating the chi-square value of each pair of adjacent intervals by using the following formula (2), gradually merging the interval pairs with the smallest chi-square value until the chi-square value of the interval pairs is larger than the threshold value theta, stopping merging, and sequentially dividing score values after sorting according to the n finally merged interval numbers, as shown in the table 2.

Wherein X is ² Represents chi-square value, a _ij The number of child dimension index values in the i-th interval representing child user interaction dimension index,

representation a _ij Z is the number of multiple sub-user interaction dimension index values, Z _i A number of sections which is one of the plurality of sub-dimension index values, S _j Is a ratio of the number of one of the plurality of child dimension index values to the total number.

TABLE 2 child user interaction dimension index score for child user interaction dimension index values

According to the embodiment of the disclosure, for the sub-user interaction weight values, different weights are given to the general method based on expert experience and business guidance in combination with the data source condition, a voting method is adopted as a more objective method, a characteristic judgment matrix (AHP) is constructed by a weighted average method, a hierarchical analysis method and the like, and the sub-user interaction weight values of each sub-user interaction dimension index can be relatively and directly obtained by using an XGBoost algorithm according to the embodiment of the disclosure, wherein the higher the sub-user interaction weight value is represented by the higher the importance of constructing a decision tree in the prediction process, and the importance of the characteristic to predictive modeling is estimated by constructing parameters such as the depth max_depth of the decision tree and the weight threshold value min_child_weight of the minimum node by using an XGBoost library in python in the detailed implementation process. Specifically, average scores of all sub-user interaction dimension index values in a plurality of decision trees are calculated through weighted average calculation, a feature importance ranking distribution diagram is drawn, and then sub-user interaction weight values of all sub-user interaction dimension indexes, namely sub-user interaction weight values w corresponding to a near sub-dimension index, a frequency sub-dimension index and a frontal sub-dimension index respectively, are calculated through the score proportion in the total score _R 、w _F And w _M 。

According to the embodiment of the disclosure, the sub-user interaction dimension index score corresponding to each sub-user interaction dimension index and the sub-user interaction dimension index weight value are multiplied to obtain products corresponding to a plurality of sub-user interaction dimension indexes, and the products are added to obtain a final user interaction dimension index value, namely an RFM value.

According to the embodiment of the disclosure, the quartering points can be divided according to the user interaction dimension index values, the users corresponding to the user interaction dimension index values arranged before the second quantile are determined to be target users, and the target users are determined according to the quantiles, so that the number of the target users is ensured to be not unique, and the population of the target users is enriched.

According to the embodiment of the disclosure, in the process of establishing the RFM model, a preset time period may be determined, and product interaction data in three time dimensions of the last half year, the last 1 year and the last 3 years may be selected from the user behavior database as a total sample set, wherein the sample set in the last half year is used as an experimental group, the sample set in the last 1 year and the last 3 years is used as a comparison group, the first 70% of the product interaction data is used as a training set in each sample set according to time sequence, and the last 30% of the product interaction data is used as a test set. The test set verifies that the RFM model AUC (area enclosed by the ROC curve and the coordinate axis) value is the largest when the sample data of the financial product purchased by the user in the last 1 year is intercepted by the samples in three groups of different time dimensions, and the RFM model effect is the best; the RFM model corresponding to the sample data in the last 3 years has the smallest AUC value and the worst effect. Therefore, the preset time period may be determined to be approximately 1 year.

According to the embodiment of the disclosure, a supervised chi-square box division mode is adopted to analyze the frequency of the classified data, the statistical significance is high, the fact that the discretized features are insensitive to abnormal data is guaranteed, the risk of model fitting is reduced, the situation that training data are sparse values is considered by the XGBoost algorithm, the value effect of original data is reserved, meanwhile, the method of random forests is used by the XGBoost, column sampling is supported, fitting can be reduced, and calculation can be simplified.

Fig. 4 schematically illustrates a block diagram of a product recommendation device according to an embodiment of the present disclosure.

As shown in fig. 4, the product recommendation apparatus 400 includes a first acquisition module 410, a first determination module 420, a second determination module 430, a third determination module 440, and a recommendation module 450.

A first obtaining module 410, configured to obtain, in response to a recommendation request of the financial platform, target historical data of a target user in a preset time period from a user behavior database, where the target historical data in the user behavior database is collected by authorization of the target user, the target historical data includes target product interaction data, and the target user is determined according to the target product interaction data;

A first determining module 420, configured to determine, according to the target history data, a portrait score level of the target user;

a second determining module 430, configured to determine, from the product library, a target product that matches the target user according to the target product interaction data;

a third determining module 440, configured to determine a product to be recommended from the first products according to the product similarity between the target product and the first products in the product library, where the first products are other products in the product library except the target product;

and the recommending module 450 is used for recommending the product to be recommended to the target user according to the portrait grading.

According to an embodiment of the present disclosure, the third determining module 440 for determining a product to be recommended from the first products according to the product similarity between the target product and the first products in the product library includes:

the first determining unit is used for determining a target product vector of the target product and a first product vector of the first product according to the target product and the first product respectively;

the second determining unit is used for determining the product similarity of the target product and the first product according to the target product vector and the first product vector;

the third determining unit is used for determining the first product as a second product to obtain a second product set under the condition that the similarity of the products is larger than a preset threshold value;

A fourth determining unit, configured to sort the second products according to the preference degrees of the target user and the second products, so as to obtain a sorting result;

and a fifth determining unit, configured to determine a product to be recommended from the second products according to the sorting result.

According to an embodiment of the present disclosure, the first determining module 420 for determining the portrait rating of the target user according to the target history data includes:

a sixth determining unit, configured to determine, according to the target history data, a dimension index value of the target user under a plurality of dimension indexes;

a seventh determining unit, configured to determine a dimension index score value of each dimension index according to the respective dimension index weights of the plurality of dimension indexes, to obtain a plurality of dimension index score values;

an eighth determining unit configured to determine a portrait score of the target user based on the plurality of dimension index score values;

and a ninth determining unit for determining the portrait score level of the target user based on the portrait score.

the seventh determining unit for determining a dimension index value of the target user under a plurality of dimension indexes according to the target history data includes:

A first determining subunit, configured to determine, according to the target history data, a respective sub-dimension index value of a plurality of sub-dimension indexes of each dimension index of the target user;

a second determining subunit, configured to determine a sub-dimension index score of each sub-dimension index according to a sub-dimension index weight of each of the plurality of sub-dimension indexes;

and the third determining subunit is used for determining the dimension index value of the target user under the plurality of dimension indexes according to the plurality of sub-dimension index scores.

According to an embodiment of the disclosure, the dimension index comprises a user interaction dimension index, and the user interaction dimension index comprises a plurality of sub-user interaction dimension indexes; the device further comprises:

the second acquisition module is used for acquiring product interaction data of a user in a preset time period from the user behavior database, wherein the product interaction data in the user behavior database is acquired through authorization of a target user;

a fourth determining module, configured to determine, according to the product interaction data, a child user interaction dimension index value of each of the plurality of child user interaction dimension indexes;

a fifth determining module, configured to determine a child user interaction dimension index score of each of the plurality of child user interaction dimension index values by using a chi-square binning method;

The computing module is used for computing the sub-user interaction weight values of each of the plurality of sub-user interaction dimension indexes by utilizing an XGboost algorithm;

a sixth determining module, configured to determine a user interaction dimension index value of the user interaction dimension index according to the multiple sub-user interaction dimension index scores and the multiple sub-user interaction dimension index weight values;

and a seventh determining module, configured to determine a target user from the users according to the user interaction dimension index value.

Any number of modules, sub-modules, units, sub-units, or at least some of the functionality of any number of the sub-units according to embodiments of the present disclosure may be implemented in one module. Any one or more of the modules, sub-modules, units, sub-units according to embodiments of the present disclosure may be implemented as split into multiple modules. Any one or more of the modules, sub-modules, units, sub-units according to embodiments of the present disclosure may be implemented at least in part as a hardware circuit, such as a Field Programmable Gate Array (FPGA), a Programmable Logic Array (PLA), a system-on-chip, a system-on-substrate, a system-on-package, an Application Specific Integrated Circuit (ASIC), or in any other reasonable manner of hardware or firmware that integrates or encapsulates the circuit, or in any one of or a suitable combination of three of software, hardware, and firmware. Alternatively, one or more of the modules, sub-modules, units, sub-units according to embodiments of the present disclosure may be at least partially implemented as computer program modules, which when executed, may perform the corresponding functions.

For example, any of the first acquisition module 410, the first determination module 420, the second determination module 430, the third determination module 440, and the recommendation module 450 may be combined in one module/unit/sub-unit or any of the modules/units/sub-units may be split into a plurality of modules/units/sub-units. Alternatively, at least some of the functionality of one or more of these modules/units/sub-units may be combined with at least some of the functionality of other modules/units/sub-units and implemented in one module/unit/sub-unit. According to embodiments of the present disclosure, at least one of the first acquisition module 410, the first determination module 420, the second determination module 430, the third determination module 440, and the recommendation module 450 may be implemented at least in part as hardware circuitry, such as a Field Programmable Gate Array (FPGA), a Programmable Logic Array (PLA), a system on a chip, a system on a substrate, a system on a package, an Application Specific Integrated Circuit (ASIC), or as hardware or firmware in any other reasonable manner of integrating or packaging the circuitry, or as any one of or a suitable combination of any of the three. Alternatively, at least one of the first acquisition module 410, the first determination module 420, the second determination module 430, the third determination module 440, and the recommendation module 450 may be at least partially implemented as computer program modules, which when executed, may perform the respective functions.

It should be noted that, in the embodiment of the present disclosure, the system portion of the product recommendation device corresponds to the product recommendation method portion of the embodiment of the present disclosure, and the description of the product recommendation device portion specifically refers to the product recommendation method portion and is not described herein.

Fig. 5 schematically shows a block diagram of an electronic device adapted to implement the method described above, according to an embodiment of the disclosure. The electronic device shown in fig. 5 is merely an example and should not be construed to limit the functionality and scope of use of the disclosed embodiments.

As shown in fig. 5, an electronic device 500 according to an embodiment of the present disclosure includes a processor 501 that can perform various appropriate actions and processes according to a program stored in a Read Only Memory (ROM) 502 or a program loaded from a storage section 508 into a Random Access Memory (RAM) 503. The processor 501 may include, for example, a general purpose microprocessor (e.g., a CPU), an instruction set processor and/or an associated chipset and/or a special purpose microprocessor (e.g., an Application Specific Integrated Circuit (ASIC)), or the like. The processor 501 may also include on-board memory for caching purposes. The processor 501 may comprise a single processing unit or a plurality of processing units for performing different actions of the method flows according to embodiments of the disclosure.

In the RAM 503, various programs and data required for the operation of the electronic apparatus 500 are stored. The processor 501, ROM502, and RAM 503 are connected to each other by a bus 504. The processor 501 performs various operations of the method flow according to the embodiments of the present disclosure by executing programs in the ROM502 and/or the RAM 503. Note that the program may be stored in one or more memories other than the ROM502 and the RAM 503. The processor 501 may also perform various operations of the method flow according to embodiments of the present disclosure by executing programs stored in the one or more memories.

According to an embodiment of the present disclosure, the electronic device 500 may also include an input/output (I/O) interface 505, the input/output (I/O) interface 505 also being connected to the bus 504. The system 500 may also include one or more of the following components connected to the I/O interface 505: an input section 506 including a keyboard, a mouse, and the like; an output portion 507 including a Cathode Ray Tube (CRT), a Liquid Crystal Display (LCD), and the like, and a speaker, and the like; a storage portion 508 including a hard disk and the like; and a communication section 509 including a network interface card such as a LAN card, a modem, or the like. The communication section 509 performs communication processing via a network such as the internet. The drive 510 is also connected to the I/O interface 505 as needed. A removable medium 511 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like is mounted on the drive 510 as needed so that a computer program read therefrom is mounted into the storage section 508 as needed.

According to embodiments of the present disclosure, the method flow according to embodiments of the present disclosure may be implemented as a computer software program. For example, embodiments of the present disclosure include a computer program product comprising a computer program embodied on a computer readable storage medium, the computer program comprising program code for performing the method shown in the flowcharts. In such an embodiment, the computer program may be downloaded and installed from a network via the communication portion 509, and/or installed from the removable media 511. The above-described functions defined in the system of the embodiments of the present disclosure are performed when the computer program is executed by the processor 501. The systems, devices, apparatus, modules, units, etc. described above may be implemented by computer program modules according to embodiments of the disclosure.

The present disclosure also provides a computer-readable storage medium that may be embodied in the apparatus/device/system described in the above embodiments; or may exist alone without being assembled into the apparatus/device/system. The computer-readable storage medium carries one or more programs which, when executed, implement methods in accordance with embodiments of the present disclosure.

According to embodiments of the present disclosure, the computer-readable storage medium may be a non-volatile computer-readable storage medium. Examples may include, but are not limited to: a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the context of this disclosure, a computer-readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.

For example, according to embodiments of the present disclosure, the computer-readable storage medium may include ROM 502 and/or RAM 503 and/or one or more memories other than ROM 502 and RAM 503 described above.

Embodiments of the present disclosure also include a computer program product comprising a computer program containing program code for performing the methods shown in the flowcharts. The program code, when executed in a computer system, causes the computer system to implement the product recommendation method provided by embodiments of the present disclosure.

The above-described functions defined in the system/apparatus of the embodiments of the present disclosure are performed when the computer program is executed by the processor 501. The systems, apparatus, modules, units, etc. described above may be implemented by computer program modules according to embodiments of the disclosure.

In one embodiment, the computer program may be based on a tangible storage medium such as an optical storage device, a magnetic storage device, or the like. In another embodiment, the computer program may also be transmitted, distributed, and downloaded and installed in the form of a signal on a network medium, and/or installed from a removable medium 511 via the communication portion 509. The computer program may include program code that may be transmitted using any appropriate network medium, including but not limited to: wireless, wired, etc., or any suitable combination of the foregoing.

In such an embodiment, the computer program may be downloaded and installed from a network via the communication portion 509, and/or installed from the removable media 511. The above-described functions defined in the system of the embodiments of the present disclosure are performed when the computer program is executed by the processor 501. The systems, devices, apparatus, modules, units, etc. described above may be implemented by computer program modules according to embodiments of the disclosure.

According to embodiments of the present disclosure, program code for performing computer programs provided by embodiments of the present disclosure may be written in any combination of one or more programming languages, and in particular, such computer programs may be implemented in high-level procedural and/or object-oriented programming languages, and/or assembly/machine languages. Programming languages include, but are not limited to, such as Java, c++, python, "C" or similar programming languages. The program code may execute entirely on the user's computing device, partly on the user's device, partly on a remote computing device, or entirely on the remote computing device or server. In the case of remote computing devices, the remote computing device may be connected to the user computing device through any kind of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or may be connected to an external computing device (e.g., connected via the Internet using an Internet service provider).

The flowcharts and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams or flowchart illustration, and combinations of blocks in the block diagrams or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions. Those skilled in the art will appreciate that the features recited in the various embodiments of the disclosure and/or in the claims may be combined in various combinations and/or combinations, even if such combinations or combinations are not explicitly recited in the disclosure. In particular, the features recited in the various embodiments of the present disclosure and/or the claims may be variously combined and/or combined without departing from the spirit and teachings of the present disclosure. All such combinations and/or combinations fall within the scope of the present disclosure.

The embodiments of the present disclosure are described above. However, these examples are for illustrative purposes only and are not intended to limit the scope of the present disclosure. Although the embodiments are described above separately, this does not mean that the measures in the embodiments cannot be used advantageously in combination. The scope of the disclosure is defined by the appended claims and equivalents thereof. Various alternatives and modifications can be made by those skilled in the art without departing from the scope of the disclosure, and such alternatives and modifications are intended to fall within the scope of the disclosure.

Claims

1. A product recommendation method comprising:

responding to a recommendation request of a financial platform, acquiring target historical data of a target user in a preset time period from a user behavior database, wherein the target historical data in the user behavior database are acquired through authorization of the target user, the target historical data comprise target product interaction data, and the target user is determined according to the target product interaction data;

determining a target product matched with the target user from a product library according to the target product interaction data;

Determining a product to be recommended from a first product in the product library according to the product similarity between the target product and the first product, wherein the first product is other products except the target product in the product library;

2. The method of claim 1, wherein the determining a product to be recommended from the first product according to the product similarity of the target product to the first product in the product library comprises:

3. The method of claim 1, wherein the determining the portrait rating of the target user from the target history data comprises:

according to the target historical data, determining a dimension index value of the target user under a plurality of dimension indexes;

determining a dimension index score value of each dimension index according to the respective dimension index weights of the plurality of dimension indexes to obtain a plurality of dimension index score values;

determining a portrayal score for the target user based on the plurality of dimension index score values;

and determining a portrait score level of the target user based on the portrait score.

4. A method according to any one of claims 1 to 3, wherein each of the dimension indicators comprises a plurality of sub-dimension indicators;

the determining, according to the target history data, a dimension index value of the target user under a plurality of dimension indexes includes:

determining respective sub-dimension index values of the plurality of sub-dimension indexes of each dimension index of the target user according to the target historical data;

5. The method of claim 3, the dimension index comprising a user interaction dimension index comprising a plurality of sub-user interaction dimension indexes;

the method further comprises the steps of:

acquiring product interaction data of a user in the preset time period from the user behavior database, wherein the product interaction data in the user behavior database is acquired through authorization of the target user;

determining the respective sub-user interaction dimension index scores of the plurality of sub-user interaction dimension index values by using a chi-square box method;

calculating the respective sub-user interaction weight values of the sub-user interaction dimension indexes by using an XGboost algorithm;

6. The method of claim 5, the sub-user interaction dimension indicators comprising a near sub-dimension indicator, a frequency sub-dimension indicator, and a quota sub-dimension indicator, wherein the near sub-dimension indicator characterizes a time when the user last transacted on the financial platform, the frequency sub-dimension indicator characterizes a number of transactions of the user on the financial platform within the preset time period, and the quota sub-dimension indicator characterizes a quota of transactions of the user on the financial platform within the preset time period.

7. A product recommendation device, comprising:

the first acquisition module is used for responding to a recommendation request of the financial platform and acquiring target historical data of a target user in a preset time period from a user behavior database, wherein the target historical data in the user behavior database are acquired through authorization of the target user, the target historical data comprise target product interaction data, and the target user is determined according to the target product interaction data;

the second determining module is used for determining a target product matched with the target user from a product library according to the target product interaction data;

A third determining module, configured to determine a product to be recommended from a first product in the product library according to a product similarity between the target product and the first product, where the first product is another product in the product library except the target product;

8. An electronic device, comprising:

one or more processors;

a memory for storing one or more programs,

wherein the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the method of any of claims 1 to 6.

9. A computer readable storage medium having stored thereon executable instructions which when executed by a processor cause the processor to implement the method of any of claims 1 to 6.

10. A computer program product comprising computer executable instructions for implementing the method of any one of claims 1 to 6 when executed.