Product Recommendation Method and Device

Abstract

The present disclosure provides a product recommendation method and apparatus. The method includes: obtaining a user to be recommended and a product investment strategy matched to the user to be recommended, where the product investment strategy is used to indicate the recommended investment proportion corresponding to each product category; determining a target product category for which product investment is to be made based on the product investment strategy, and the target product category corresponds to a target product label; obtaining preference data of the user to be recommended for the target product label; and determining a target product matched to the user to be recommended based on the preference data of the user to be recommended for the target product label. Among them, by obtaining the target product category and the preference data of the user to be recommended for the target product label, the personalized needs and preferences of the user can be captured more accurately, and then product matching can be performed under the target product category based on the preference data, so as to achieve personalized product recommendation. In addition, compared with manual decision-making, the present disclosure has higher efficiency and the product recommendation result is more objective and accurate.

Description

Technical Field

[0001] The present disclosure relates to the technical field of data processing, and particularly to a product recommendation method and apparatus. Background Art

[0002] Currently, the product investment plans given to customers in financial investment services are all product plans selected based on manual experience, that is, through regular decision-making by an expert team and investment advisors, standard product investment plans are formulated for customers. Although this method ensures the professionalism and stability of the plan to a certain extent, it also has significant limitations, such as relying too much on the experience and subjective judgment of professionals, and it is difficult to fully cover the personalized needs and preferences of each customer. Summary of the Invention

[0003] The present disclosure provides a product recommendation method and apparatus to at least solve one of the technical problems in the related art to a certain extent. The technical solutions of the present disclosure are as follows:

[0004] According to a first aspect of an embodiment of the present disclosure, a product recommendation method is provided, including: obtaining a user to be recommended and a product investment strategy matched with the user to be recommended, where the product investment strategy is used to indicate the recommended investment proportion corresponding to each product category; based on the product investment strategy, determining a target product category for which product investment is to be made, where the target product category corresponds to a target product label; obtaining preference data of the user to be recommended for the target product label; and based on the preference data of the user to be recommended for the target product label, determining a target product matched with the user to be recommended.

[0005] According to a second aspect of an embodiment of the present disclosure, a product recommendation apparatus is provided, including: a first obtaining module, configured to obtain a user to be recommended and a product investment strategy matched with the user to be recommended, where the product investment strategy is used to indicate the recommended investment proportion corresponding to each product category; a first determining module, configured to determine a target product category for which product investment is to be made based on the product investment strategy, where the target product category corresponds to a target product label; a second obtaining module, configured to obtain preference data of the user to be recommended for the target product label; and a second determining module, configured to determine a target product matched with the user to be recommended based on the preference data of the user to be recommended for the target product label.

[0006] According to a third aspect of an embodiment of the present disclosure, an electronic device is provided, including: a processor, and a memory communicatively connected to the processor; the memory stores computer-executable instructions; and the processor executes the computer-executable instructions stored in the memory to implement a product recommendation method as described in the first aspect of the embodiments of the present disclosure.

[0007] According to a fourth aspect of the embodiments of the present disclosure, there is provided a computer-readable storage medium storing computer-executable instructions that, when executed by a processor, are used to implement a product recommendation method as described in the first aspect of the embodiments of the present disclosure.

[0008] According to a fifth aspect of the embodiments of the present disclosure, there is provided a computer program product including a computer program that, when executed by a processor, implements a product recommendation method as described in the first aspect of the embodiments of the present disclosure.

[0009] The technical solutions provided by the embodiments of the present disclosure at least bring the following beneficial effects:

[0010] Obtain a user to be recommended and a product investment strategy matched by the user to be recommended, where the product investment strategy is used to indicate the recommended investment proportion corresponding to each product category; based on the product investment strategy, determine a target product category for product investment, and the target product category corresponds to a target product label; obtain preference data of the user to be recommended for the target product label; based on the preference data of the user to be recommended for the target product label, determine a target product matched with the user to be recommended. Among them, by obtaining the target product category and the preference data of the user to be recommended for the target product label, the personalized needs and preferences of the user can be captured more accurately, and then product matching can be performed based on the preference data under the target product category, so as to achieve personalized product recommendation. In addition, the present disclosure uses an automated and intelligent method for product recommendation. Compared with manual decision-making, the efficiency is higher and the product recommendation result is more objective and accurate. In addition, when determining the target user label, through the first verification result, user labels highly relevant to the user's product investment intention can be screened out, through the second verification result, user labels with rich information and important influence on the user's investment behavior can be screened out, and through the third verification result, user labels with higher representativeness can be screened out from many similar user labels. Therefore, screening the target user label by combining at least one of the first verification result, the second verification result, and the third verification result can screen out target user labels with higher quality. Then, by combining the target user label, the user's needs can be understood more accurately, thereby improving the accuracy of product recommendation and user satisfaction.

[0011] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] The above and / or additional aspects and advantages of the present disclosure will become apparent and easier to understand from the following description of the embodiments in conjunction with the drawings, where:

[0013] Figure 1 It is a schematic flowchart of a product recommendation method provided by the first embodiment of the present disclosure;

[0014] Figure 2 A flowchart of a product recommendation method provided by the second embodiment of the present disclosure;

[0015] Figure 3 A flowchart of a product recommendation method provided by the third embodiment of the present disclosure;

[0016] Figure 4 A flowchart of a product recommendation method provided by the fourth embodiment of the present disclosure;

[0017] Figure 5 A flowchart of a product recommendation method provided by the fifth embodiment of the present disclosure;

[0018] Figure 6 A structural diagram of a product recommendation device provided by the sixth embodiment of the present disclosure;

[0019] Figure 7 A block diagram of an electronic device provided by the seventh embodiment of the present disclosure. Detailed implementation manners

[0020] In order to enable those of ordinary skill in the art to better understand the technical solutions of the present disclosure, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below with reference to the accompanying drawings.

[0021] It should be noted that the terms "first", "second", etc. in the specification and claims of the present disclosure and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily need to describe a specific order or sequence. It should be understood that such used data can be interchanged under appropriate circumstances so that the embodiments of the present disclosure described herein can be implemented in an order different from those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. On the contrary, they are merely examples of devices and methods consistent with some aspects of the present disclosure as detailed in the appended claims.

[0022] It should be noted that in the technical solutions of the present disclosure, the collection, storage, use, processing, transmission, provision, and disclosure of user personal information and other processing are all carried out on the premise of obtaining the user's consent, and all comply with the provisions of relevant laws and regulations and do not violate public order and good customs.

[0023] The following describes the product recommendation methods and devices of the embodiments of the present disclosure with reference to the drawings.

[0024] Among them, it should be noted that the execution subject of the product recommendation method in this embodiment is a product recommendation device, which can be implemented in software and / or hardware and can be configured in an electronic device. The electronic device can refer to a terminal, a server, etc.

[0025] Figure 1 It is a schematic flowchart of a product recommendation method provided by the first embodiment of the present disclosure.

[0026] As Figure 1 shown, the product recommendation method includes the following steps:

[0027] Step 101, obtain the user to be recommended and the product investment strategy matched by the user to be recommended. The product investment strategy is used to indicate the recommended investment proportion corresponding to each product category.

[0028] Among them, the user to be recommended refers to the user for whom product recommendation is to be carried out. The product investment strategy matched by the user to be recommended refers to the investment strategy that matches the investment style or risk tolerance of the user to be recommended. The product investment strategy includes the recommended investment proportion corresponding to each product category.

[0029] For example, the investment styles include conservative, stable, balanced, aggressive, radical, etc.; the product categories include equity, hybrid, fixed income, cash management, alternative assets, etc.

[0030] In order to match a suitable product recommendation strategy, as an example, obtain the investment style or risk tolerance of the user to be recommended, and based on the investment style or risk tolerance of the user to be recommended, obtain the product recommendation strategy that matches the user to be recommended from a set of multiple product recommendation strategies.

[0031] Step 102, based on the product investment strategy, determine the target product category for which product investment is to be carried out. The target product category corresponds to a target product label.

[0032] Among them, the target product category refers to one or more of each product category. Each product category has a corresponding product label, and the product labels corresponding to each product category may be different. The target product label refers to the product label corresponding to the target product category. It should be noted that the product label is used to describe the product characteristics. For example, the product label includes but is not limited to yield, maximum drawdown, risk level, etc.

[0033] When making product recommendations, in order to improve the accuracy of recommendations, first determine the target product categories for which product recommendations are to be made. As an example, obtain the actual investment proportions of the user to be recommended in each product category; based on the actual investment proportions and the recommended investment proportions, determine the target product categories for which product investments are to be made. Among them, by comparing the actual investment proportions and the recommended investment proportions in each product category, the product categories with actual investment proportions less than the recommended investment proportions are determined as the target product categories.

[0034] It should be noted that if there are products that are about to expire in a product category, then such products that are about to expire need to be excluded when obtaining the actual investment proportion.

[0035] Step 103, obtain the preference data of the user to be recommended for the target product labels.

[0036] Among them, the preference data of the user to be recommended for the target product labels is used to indicate the preference situation of the user to be recommended for each product feature. Assuming the target product label is the rate of return, the preference data refers to the specific rate of return value, such as 5%.

[0037] As a possible implementation method, obtain the preference data input by the user to be recommended for the target product labels.

[0038] Step 104, based on the preference data of the user to be recommended for the target product labels, determine the target products that match the user to be recommended.

[0039] As a possible implementation method, based on the preference data of the user to be recommended for the target product labels, screen the target products from the product pool whose product label data conforms to the preference data, and recommend the target products to the user to be recommended.

[0040] In summary, obtain the user to be recommended and the product investment strategies that match the user to be recommended. The product investment strategies are used to indicate the recommended investment proportions corresponding to each product category; based on the product investment strategies, determine the target product categories for which product investments are to be made. The target product categories correspond to target product labels; obtain the preference data of the user to be recommended for the target product labels; based on the preference data of the user to be recommended for the target product labels, determine the target products that match the user to be recommended. Among them, by obtaining the target product categories and the preference data of the user to be recommended for the target product labels, the personalized needs and preferences of the user can be captured more accurately. Furthermore, based on the preference data, product matching is performed under the target product categories, and personalized product recommendations can be realized. In addition, the present disclosure uses an automated and intelligent method for product recommendation. Compared with manual decision-making, the efficiency is higher and the product recommendation results are more objective and accurate.

[0041] To clearly illustrate how to obtain the preference data of the user to be recommended for the target product label in the above embodiments, the present disclosure proposes another product recommendation method.

[0042] Figure 2 It is a schematic flowchart of a product recommendation method provided by the second embodiment of the present disclosure.

[0043] As Figure 2 shown, the product recommendation method may include the following steps:

[0044] Step 201, obtain the user to be recommended and the product investment strategy matched by the user to be recommended.

[0045] Step 202, based on the product investment strategy, determine the target product category for which product investment is to be made.

[0046] Step 203, obtain the product investment records of the user to be recommended under the target product category.

[0047] Among them, the product investment record refers to the product purchase record, and the product investment record includes the investment amount, investment date, product maturity date, product term, etc.

[0048] Step 204, in the case of obtaining the product investment record, parse the product investment record to obtain the preference data of the user to be recommended for the target product label.

[0049] Among them, obtaining the product investment record indicates that the user to be recommended has purchased relevant products in the past, that is, the user to be recommended is an old user. At this time, based on the product investment record, parse the product label data of the products purchased in the past to obtain the preference data of the user to be recommended for the target product label.

[0050] Step 205, in the case of not obtaining the product investment record, obtain the user label data of the user to be recommended according to the set target user label, and based on the user label data, obtain the preference data of the user to be recommended for the target product label.

[0051] Among them, not obtaining the product investment record indicates that the user to be recommended has not purchased relevant products in the past, that is, the user to be recommended is a new user. At this time, based on the user label data of the user to be recommended, infer the preference data of the user to be recommended for the target product label.

[0052] Among them, the target user label refers to the user label related to the product investment intention, and the target user label includes basic attribute labels, behavioral attribute labels, preference attribute labels, etc. For example, the target user label includes age, income level, asset size, risk tolerance, whether to purchase financial products, etc.

[0053] To obtain accurate preference data and achieve precise product recommendations, as an example, based on user tag data, similar users of the user to be recommended are obtained; preference data of the similar users for the target product tags is obtained; based on the preference data of the similar users for the target product tags, the preference data of the user to be recommended for the target product tags is determined.

[0054] It should be noted that, to ensure the accuracy and reliability of product recommendations and avoid interference of missing data or abnormal data on the recommendation results, before obtaining the similar users of the user to be recommended, operations such as missing value processing, outlier processing, and extreme value processing are performed on the relevant data of existing users.

[0055] As a possible implementation manner, based on user tag data, the user similarity between the user to be recommended and old users is obtained, and based on the user similarity, the similar users of the user to be recommended are determined.

[0056] As a possible implementation manner, for each target product tag, based on the preference data of each similar user for this tag, mode statistics or mean calculation is performed to obtain the preference data of the user to be recommended for the target product tag.

[0057] Step 206, based on the preference data of the user to be recommended for the target product tag, determine the target products that match the user to be recommended.

[0058] It should be noted that the execution processes of Step 201, Step 202, and Step 206 can be respectively implemented in any manner in the various embodiments of the present disclosure. The embodiments of the present disclosure do not make any limitations in this regard and will not be elaborated further.

[0059] In summary, obtain the product investment records of the user to be recommended under the target product category; in the case of obtaining the product investment records, parse the product investment records to obtain the preference data of the user to be recommended for the target product tag; in the case of not obtaining the product investment records, according to the set target user tags, obtain the user tag data of the user to be recommended, and based on the user tag data, obtain the preference data of the user to be recommended for the target product tag. Among them, whether the product investment records are obtained or not, that is, whether the user to be recommended is a new user or an old user, the preference data of the user to be recommended for the target product tag can be obtained, thereby realizing personalized product recommendations.

[0060] To clearly illustrate how the target user tags are obtained in the above embodiments, the present disclosure proposes another product recommendation method.

[0061] Figure 3 It is a schematic flowchart of a product recommendation method provided by the third embodiment of the present disclosure.

[0062] AsFigure 3 As shown in the figure, the product recommendation method may include the following steps:

[0063] Step 301: Obtain multiple user tags to be screened.

[0064] Among them, the user tags to be screened refer to all or part of the user tags in the user tag library. For any user tag, the user tag may be related to the product investment intention or may not be related to the product investment intention.

[0065] Step 302: For any user tag, determine the first verification result of the user tag according to the correlation coefficient between the user tag and the product investment intention, and determine the second verification result of the user tag according to the information entropy and / or information value of the user tag.

[0066] Among them, the first verification result is used to indicate whether the user tag passes the correlation relationship verification, and the second verification result is used to indicate whether the user tag passes the information entropy / information value analysis verification. Among them, the information entropy is used to indicate the amount of information contained in the user tag, and the information value is used to indicate the prediction ability / importance degree of the user tag for the product investment intention.

[0067] To improve the acquisition efficiency of the first verification result, as an example, if the correlation coefficient between the user tag and the product investment intention is greater than or equal to the set correlation coefficient threshold, the first verification result is that the user tag passes the correlation relationship verification; if the correlation coefficient between the user tag and the product investment intention is less than the set correlation coefficient threshold, the first verification result is that the user tag fails to pass the correlation relationship verification.

[0068] As a possible implementation manner, the correlation coefficient between the user tag and the product investment intention is the Pearson correlation coefficient.

[0069] To improve the acquisition efficiency of the second verification result, as an example, if the information entropy of the user tag is greater than or equal to the set information entropy threshold, and / or if the information value (Information Value, IV) of the user tag is greater than or equal to the set information value threshold, the second verification result is that the user tag passes the information entropy / information value analysis verification; if the information entropy of the user tag is less than the set information entropy threshold, and / or if the information value of the user tag is less than the set information value threshold, the second verification result is that the user tag fails to pass the information entropy / information value analysis verification.

[0070] Step 303: Cluster multiple user tags to obtain multiple clusters.

[0071] Among them, according to the variable clustering (Varclus clustering) algorithm, cluster multiple user tags to obtain multiple clusters.

[0072] The clustering process of the Varclus clustering algorithm is as follows. First, all user tags (variables) are regarded as a single cluster. Then, based on principal component analysis and covariance rotation, the cluster is recursively split into two sub-clusters until a stopping condition is met (such as the percentage of variance explained by the cluster components being below a threshold). In each split, the variables are assigned to the cluster with the highest correlation with the principal component to optimize the composition of the cluster. Finally, the variables within each cluster are highly correlated, while the correlation between different clusters is low, thus achieving effective clustering of the variables.

[0073] Step 304: For any user tag, determine the third verification result of the user tag according to the tightness between the user tag and the clustering cluster.

[0074] Among them, the third verification result is used to indicate whether the user tag passes the clustering analysis verification.

[0075] To make the tightness accurately reflect the tight relationship between the user tag and the clustering cluster, as an example, the tightness between the user tag and the clustering cluster is determined by the following formula:

[0076]

[0077] Among them, r represents the tightness between the user tag and the clustering cluster; R within the cluster 2 represents the square of the correlation coefficient between the user tag and the variables within the cluster to which it belongs (the first principal component of the cluster), and the larger its value, the closer the user tag is to the cluster to which it belongs; R between clusters 2 represents the square of the correlation coefficient between the user tag and the variables within the adjacent cluster, and the smaller its value, the more separated the user tag is from the adjacent cluster; the larger R within the cluster 2 , the smaller R between clusters 2 , the smaller 1 - R 2 Ratio will be, and the smaller the value of 1 - R 2 Ratio, the closer the user tag is to the cluster to which it belongs and the more separated it is from the adjacent cluster. Among them, the adjacent cluster refers to the clustering clusters other than the cluster to which the user tag belongs.

[0078] To improve the acquisition efficiency of the third verification result, as an example, if the tightness between the user tag and the clustering cluster is less than or equal to the set tightness threshold, the third verification result is that the user tag passes the clustering analysis verification; if the tightness between the user tag and the clustering cluster is greater than the set tightness threshold, the third verification result is that the user tag fails the clustering analysis verification.

[0079] Step 305: Based on at least one of the first verification result, the second verification result, and the third verification result, perform label screening to obtain the target user tag.

[0080] To obtain high-quality target user tags, as an example, based on the first verification result, the second verification result, and the third verification result, the user tags that pass the correlation verification, information entropy / information value analysis verification, and clustering analysis verification simultaneously are used as the target user tags.

[0081] It should be noted that steps 301 - 305 can be executed before step 101 or step 201.

[0082] In summary, obtain multiple user tags to be screened; for any user tag, determine the first verification result of the user tag according to the correlation coefficient between the user tag and the product investment intention, and determine the second verification result of the user tag according to the information entropy and / or information value of the user tag; cluster the multiple user tags to obtain multiple clusters; for any user tag, determine the third verification result of the user tag according to the closeness between the user tag and the cluster; based on at least one of the first verification result, the second verification result, and the third verification result, perform tag screening to obtain the target user tags. Among them, through the first verification result, user tags highly correlated with the user's product investment intention can be screened out, through the second verification result, user tags with rich information and important influence on the user's investment behavior can be screened out, and through the third verification result, user tags with higher representativeness can be screened out from many similar user tags. Therefore, combining at least one of the first verification result, the second verification result, and the third verification result for tag screening can screen out target user tags with higher quality. Furthermore, by combining the target user tags, the user needs can be understood more accurately, thereby improving the accuracy of product recommendation and user satisfaction.

[0083] To clearly illustrate how to perform tag screening based on the verification results to obtain the target user tags in the above embodiments, the present disclosure proposes another product recommendation method.

[0084] Figure 4 It is a flowchart of a product recommendation method provided by the fourth embodiment of the present disclosure.

[0085] As Figure 4 shown, based on the embodiment shown in Figure 3 step 305 includes the following steps:

[0086] Step 401, group the user tags based on the correlation coefficient between the user tags to obtain a first tag group and at least one second tag group; wherein, the correlation coefficient between any two user tags in the second tag group is greater than or equal to the set coefficient threshold, and the first tag group consists of user tags not included in each second tag group.

[0087] Among them, a set coefficient threshold is used to determine whether any two user tags are highly correlated variables.

[0088] To ensure the completeness and accuracy of grouping, as an example, traverse the user tags. For any two user tags, obtain the correlation coefficient between the two user tags. If the correlation coefficient is greater than or equal to the set coefficient threshold, the two user tags form a second tag group. After the traversal is completed, the user tags not included in each second tag group form the first tag group. It should be noted that if any one of the two user tags is already included in an existing second tag group, the other user tag is added to the existing second tag group.

[0089] Step 402: Based on at least one of the first verification result, the second verification result, and the third verification result, perform tag screening in the first tag group to obtain the first user tag.

[0090] To obtain high-quality target user tags, as an example, based on the first verification result, the second verification result, and the third verification result, perform tag screening in the first tag group in sequence to obtain the first user tag.

[0091] Step 403: For any second tag group, based on the first verification result, the second verification result, and the third verification result, determine the number of times each user tag in the second tag group passes the verification, and based on the number of times of passing the verification, perform tag screening in the second tag group to obtain the second user tag.

[0092] To obtain high-quality target user tags, as an example, for any second tag group, use the user tag corresponding to the maximum number of times of passing the verification in the second tag group as the second user tag.

[0093] There may be multiple user tags corresponding to the maximum number of times of passing the verification in the second tag group. To more accurately understand the user's needs and thus improve the accuracy of product recommendations, as another example, in the second tag group, obtain the user tags corresponding to the maximum number of times of passing the verification; in the case where there are multiple user tags corresponding to the maximum number of times of passing the verification, based on the information entropy and / or information value of the user tags, select the second user tag from the multiple user tags corresponding to the maximum number of times of passing the verification.

[0094] As a possible implementation, select the user tag with the highest information entropy or the highest information value from the multiple user tags corresponding to the maximum number of times of passing the verification as the second user tag.

[0095] As another possible implementation, based on the information entropy and information value of user tags, determine the tag score of the user tag corresponding to the maximum number of successful validations; according to the tag score, select the second user tag from the multiple user tags corresponding to the maximum number of successful validations.

[0096] Step 404, determine the target user tag based on the first user tag and the second user tag.

[0097] Among them, use the first user tag and the second user tag as the target user tag, or use some of the first user tag and the second user tag as the target user tag.

[0098] In summary, group the user tags based on the correlation coefficient between user tags to obtain the first tag group and at least one second tag group; among them, the correlation coefficient between any two user tags in the second tag group is greater than or equal to the set coefficient threshold, and the first tag group consists of user tags not included in each second tag group; based on at least one of the first verification result, the second verification result, and the third verification result, perform tag screening in the first tag group to obtain the first user tag; for any second tag group, based on the first verification result, the second verification result, and the third verification result, determine the number of successful validations corresponding to each user tag in the second tag group, and according to the number of successful validations, perform tag screening in the second tag group to obtain the second user tag; determine the target user tag based on the first user tag and the second user tag. Among them, through the first tag group, the first user tag with high correlation with the product investment intention, rich information volume, important influence on the user investment intention, and more representative among similar user tags can be screened out. Through the second tag group, the second user tag with better quality can be selected from the highly correlated user tags, and the redundant tags with poor quality can be removed, so as to more accurately understand the user needs and further improve the accuracy of product recommendation.

[0099] To clearly illustrate how to obtain the users to be recommended and how to determine the target products in the above embodiments, the present disclosure proposes another product recommendation method.

[0100] Figure 5 It is a schematic flowchart of a product recommendation method provided by the fifth embodiment of the present disclosure.

[0101] As Figure 5 shown, the product recommendation method may include the following steps:

[0102] Step 501: In response to receiving a product recommendation request from a user, determine the user as a user to be recommended; or, obtain the user's product investment records in each product category, and in response to the time interval between the product maturity date of the corresponding product indicated by the product investment record and the current date being less than a set interval threshold, determine the user as a user to be recommended.

[0103] Among them, the users to be recommended include not only the users who actively request product recommendations, but also the users who may need product recommendations detected based on product investment records.

[0104] It should be noted that determining whether a user is a user to be recommended based on the product maturity date is a strategy to prevent customer churn. By recommending matching products to such users, the customer churn rate can be reduced.

[0105] Step 502: Obtain the product investment strategies matched by the users to be recommended; based on the product investment strategies, determine the target product categories for which product investments are to be made.

[0106] Step 503: Obtain the preference data of the users to be recommended for the target product tags.

[0107] Step 504: Based on the preference data of the users to be recommended for the target product tags, screen candidate products from the product pool whose product tag data conforms to the preference data.

[0108] To quickly screen out candidate products from the product pool, as an example, match the product tag data of each product in the product pool with the preference data of the users to be recommended for the target product tags, and screen out the candidate products whose product tag data conforms to the preference data.

[0109] Step 505: Using the selected parameters corresponding to multiple candidate products as variables, construct an objective function based on the set recommendation control parameters, where the selected parameters are used to indicate whether a candidate product is selected as a target product.

[0110] Among them, the recommendation control parameters include turnover rate control parameters, allocation ratio control parameters for key recommended products, product rating control parameters, user interest score control parameters, etc. The objective function is used to indicate the difference between the actual parameter values of the selected products in the recommendation control parameters and the set parameter values.

[0111] Step 506: Solve the objective function in combination with the set constraint conditions to obtain the values of the selected parameters for multiple candidate products, and based on the values of the selected parameters for multiple candidate products, determine the target product from the candidate products.

[0112] Among them, the set constraint conditions include the product configuration ratio constraint condition under the target product category, the product quantity constraint condition, the portfolio duration constraint condition, the product purchase amount constraint condition, the product position amount constraint condition, etc. For example, the product configuration ratio constraint condition under the target product category is that the ratio deviation between the product configuration ratio under the target product category and the recommended ratio is less than or equal to the set ratio deviation value; the product quantity constraint condition is that the total quantity of the recommended products is less than or equal to the set quantity.

[0113] Among them, the process of solving the objective function refers to the process of minimizing or maximizing the difference between the actual parameter value and the set parameter value in combination with the constraint conditions.

[0114] In a possible implementation manner, when the value of the candidate product on the selected parameter is the first value, it indicates that the candidate product is selected as the target product; when the value of the candidate product on the selected parameter is the second value, it indicates that the candidate product is not selected as the target product. For example, the first value is 1 and the second value is 0.

[0115] By constructing the objective function and combining the constraint conditions, it is possible to consider in multiple aspects whether the candidate product is selected as the target product, ensuring that the selection of the target product not only meets the set constraints but also satisfies the actual needs. In addition, the recommended control parameters and constraint conditions can be adjusted according to the actual needs, with good flexibility.

[0116] Step 507: Based on the set product evaluation indicators, determine the product scores of each candidate product.

[0117] Among them, the product evaluation indicators include product performance trend, product risk level, product dispersion, product volatility, product liquidity, etc.

[0118] To improve the accuracy and acquisition efficiency of the scores, as an example, the indicator values of the candidate products on the product evaluation indicators are input into the scoring model to obtain the product scores of the candidate products.

[0119] Step 508: According to the product scores, select the target products that match the user to be recommended from the candidate products.

[0120] In a possible implementation manner, a certain number of candidate products are selected in descending order of the product scores as the target products.

[0121] The method of product recommendation based on product scores simplifies the recommendation process, requires less computational effort, and improves the recommendation efficiency.

[0122] As a possible implementation, the target products obtained in step 506 and the target products obtained in step 508 can be combined to determine the products finally recommended to the user to be recommended. For example, for the target products obtained in step 506, obtain the product scores of the target products, and based on the product scores, select the products finally recommended to the user to be recommended from the target products obtained in step 506.

[0123] It should be noted that the execution processes of steps 502 - 503 can be implemented in any of the ways in the various embodiments of the present disclosure. The embodiments of the present disclosure do not make any limitations on this and will not be elaborated further.

[0124] In summary, based on the preference data of the user to be recommended for the target product tags, candidate products whose product tag data conforms to the preference data are screened from the product pool; taking the selected parameters corresponding to the multiple candidate products as variables, a target function is constructed based on the set recommendation control parameters, and the selected parameters are used to indicate whether the candidate products are selected as target products; the target function is solved in combination with the set constraint conditions to obtain the values of the selected parameters for the multiple candidate products, and based on the values of the selected parameters for the multiple candidate products, target products are determined from the candidate products; based on the set product evaluation indicators, the product scores of each candidate product are determined; according to the product scores, target products that match the user to be recommended are selected from the candidate products. Among them, the present disclosure provides two ways to obtain target products from candidate products. The method based on the target function can not only consider in multiple aspects whether the candidate products are selected as target products, ensure that the selection of target products not only conforms to the set constraints but also meets the actual needs, but also can flexibly adjust the recommendation control parameters and constraint conditions according to the actual needs; the method based on the product score simplifies the recommendation process, requires less computational effort, and improves the recommendation efficiency.

[0125] To implement the above embodiments, the embodiments of the present disclosure also propose a product recommendation device.

[0126] Figure 6 FIG. is a schematic structural diagram of a product recommendation device provided in the sixth embodiment of the present disclosure.

[0127] As Figure 6 shown, the product recommendation device 600 includes: a first acquisition module 601, a first determination module 602, a second acquisition module 603, and a second determination module 604.

[0128] Among them, the first acquisition module 601 is configured to acquire a user to be recommended and a product investment strategy matched by the user to be recommended, where the product investment strategy is used to indicate the recommended investment ratio corresponding to each product category; the first determination module 602 is configured to determine, based on the product investment strategy, a target product category for which product investment is to be made, and the target product category corresponds to a target product label; the second acquisition module 603 is configured to acquire preference data of the user to be recommended for the target product label; the second determination module 604 is configured to determine a target product matched by the user to be recommended based on the preference data of the user to be recommended for the target product label.

[0129] As a possible implementation manner of the present disclosure, the second acquisition module 603 is configured to: acquire a product investment record of the user to be recommended under the target product category; in the case of acquiring the product investment record, parse the product investment record to obtain the preference data of the user to be recommended for the target product label; in the case of not acquiring the product investment record, acquire the user label data of the user to be recommended according to the set target user label, and based on the user label data, acquire the preference data of the user to be recommended for the target product label.

[0130] As a possible implementation manner of the present disclosure, the second acquisition module 603 is configured to: acquire similar users of the user to be recommended based on the user label data; acquire the preference data of the similar users for the target product label; and determine the preference data of the user to be recommended for the target product label based on the preference data of the similar users for the target product label.

[0131] As a possible implementation manner of the present disclosure, the apparatus further includes: a third acquisition module, configured to acquire a plurality of user labels to be screened; a third determination module, configured to, for any user label, determine a first verification result of the user label according to the correlation coefficient between the user label and the product investment intention, and determine a second verification result of the user label according to the information entropy and / or information value of the user label; a clustering module, configured to cluster the plurality of user labels to obtain a plurality of clustering clusters; a fourth determination module, configured to, for any user label, determine a third verification result of the user label according to the tightness between the user label and the clustering cluster; and a label screening module, configured to perform label screening based on at least one of the first verification result, the second verification result, and the third verification result to obtain the target user label.

[0132] As a possible implementation manner of the present disclosure, the label screening module is configured to: group user labels based on the correlation coefficients between user labels to obtain a first label group and at least one second label group, where the correlation coefficient between any two user labels in the second label group is greater than or equal to a set coefficient threshold, and the first label group consists of user labels not included in each second label group; perform label screening in the first label group based on at least one of a first verification result, a second verification result, and a third verification result to obtain a first user label; for any second label group, determine the number of times of passing verification corresponding to each user label in the second label group based on the first verification result, the second verification result, and the third verification result, and perform label screening in the second label group according to the number of times of passing verification to obtain a second user label; determine a target user label based on the first user label and the second user label.

[0133] As a possible implementation manner of the present disclosure, the label screening module is configured to: in the second label group, obtain the user label corresponding to the maximum value of the number of times of passing verification; in the case where the maximum value of the number of times of passing verification corresponds to multiple user labels, select a second user label from the multiple user labels corresponding to the maximum value of the number of times of passing verification based on the information entropy and / or information value of the user labels.

[0134] As a possible implementation manner of the present disclosure, the first obtaining module 601 is configured to: determine the user as a user to be recommended in response to receiving a product recommendation request of the user; or, obtain the product investment records of the user in each product category, and determine the user as a user to be recommended in response to that the time interval between the product maturity date corresponding to the product investment record and the current date is less than a set interval threshold.

[0135] As a possible implementation manner of the present disclosure, the second determining module 604 is configured to: screen candidate products whose product label data conforms to the preference data from the product pool based on the preference data of the user to be recommended for the target product label; construct an objective function based on the set product recommendation control parameters with the selected parameters corresponding to the multiple candidate products as variables, where the selected parameters are used to indicate whether the candidate products are selected as target products; solve the objective function in combination with the set constraint conditions to obtain the values of the selected parameters of the multiple candidate products, and determine the target product from the candidate products based on the values of the selected parameters of the multiple candidate products.

[0136] As a possible implementation manner of the present disclosure, the second determining module 604 is configured to: screen candidate products whose product label data conforms to the preference data from the product pool based on the preference data of the user to be recommended for the target product label; determine the product scores of the candidate products based on the set product evaluation indicators; select the target product that matches the user to be recommended from the candidate products according to the product scores.

[0137] It should be noted that the foregoing explanation of the embodiments of the product recommendation method also applies to the product recommendation device of this embodiment, and will not be repeated here.

[0138] Figure 7 FIG. is a schematic structural diagram of an electronic device provided in the seventh embodiment of the present disclosure. Among them, the electronic device 700 in this embodiment is intended to represent various forms of digital computers, such as, laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as, personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0139] As Figure 7 shown, the above-mentioned electronic device 700 includes:

[0140] A memory 701 and a processor 702, a bus 703 connecting different components (including the memory 701 and the processor 702), and the memory 701 stores a computer program, and when the processor 702 executes the program, the product recommendation method of the embodiment of the present disclosure is implemented.

[0141] The bus 703 represents one or more of several types of bus structures, including a memory bus or a memory controller, a peripheral bus, a graphics acceleration port, a processor, or a local bus using any bus structure in a variety of bus structures. By way of example, these architectures include, but are not limited to, Industry Standard Architecture (ISA) bus, Micro Channel Architecture (MAC) bus, Enhanced ISA bus, Video Electronics Standards Association (VESA) local bus, and Peripheral Component Interconnect (PCI) bus.

[0142] The electronic device 700 typically includes a variety of electronic device-readable media. These media can be any available media that can be accessed by the electronic device 700, including volatile and non-volatile media, removable and non-removable media.

[0143] The memory 701 may also include a computer system-readable medium in the form of volatile memory, such as random access memory (RAM) 704 and / or cache memory 705. The electronic device 700 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, the storage system 706 can be used to read and write non-removable, non-volatile magnetic media ( Figure 7 not shown, commonly referred to as a "hard disk drive"). Although Figure 7Not shown in the figure, a disk drive for reading and writing to a removable non-volatile disk (such as a "floppy disk") and an optical disk drive for reading and writing to a removable non-volatile optical disk (such as a CD-ROM, DVD-ROM or other optical medium) can be provided. In these cases, each drive can be connected to the bus 703 through one or more data medium interfaces. The memory 701 may include at least one program product, which has a set (such as at least one) of program modules, and these program modules are configured to execute the functions of the embodiments of the present disclosure.

[0144] A program / utility 708 having a set (at least one) of program modules 707 can be stored in, for example, the memory 701. Such program modules 707 include, but are not limited to, an operating system, one or more application programs, other program modules, and program data. Each or some combination of these examples may include the implementation of a network environment. The program modules 707 generally execute the functions and / or methods in the embodiments described in the present disclosure.

[0145] The electronic device 700 can also communicate with one or more external devices 709 (such as a keyboard, a pointing device, a display 711, etc.), and can also communicate with one or more devices that enable a user to interact with the electronic device 700, and / or communicate with any device that enables the electronic device 700 to communicate with one or more other computing devices (such as a network card, a modem, etc.). Such communication can be carried out through an input / output (I / O) interface 712. And, the electronic device 700 can also communicate with one or more networks (such as a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) through a network adapter 713. As Figure 7 shown, the network adapter 713 communicates with other modules of the electronic device 700 through the bus 703. It should be understood that although not shown in the figure, other hardware and / or software modules can be used in combination with the electronic device 700, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems, etc.

[0146] The processor 702 executes various functional applications and data processing by running the programs stored in the memory 701.

[0147] It should be noted that for the implementation process and technical principle of the electronic device in this embodiment, refer to the foregoing explanation of the product recommendation method of the embodiments of the present disclosure, and details will not be repeated here.

[0148] To implement the above embodiments, the present disclosure also provides a computer-readable storage medium storing computer-executable instructions, which are used to implement the method provided by the foregoing embodiments when executed by a processor. Among them, the computer-readable storage medium may be ROM, random access memory (RAM), CD-ROM, magnetic tape, floppy disk, and optical data storage device, etc.

[0149] To implement the above embodiments, the present disclosure also provides a computer program product, including a computer program, which implements the method provided by the foregoing embodiments when executed by a processor.

[0150] Those skilled in the art will readily conceive of other embodiments of the present disclosure after considering the specification and practicing the invention disclosed herein. The present disclosure is intended to cover any variations, uses, or adaptations of the present disclosure, which follow the general principles of the present disclosure and include known common knowledge or conventional technical means in the technical field not disclosed by the present disclosure. The specification and embodiments are only regarded as exemplary, and the true scope and spirit of the present disclosure are pointed out by the following claims.

[0151] It should be understood that the present disclosure is not limited to the exact structures described above and shown in the drawings, and various modifications and changes can be made without departing from its scope. The scope of the present disclosure is only limited by the appended claims.

Claims

1. A product recommendation method, characterized in that, Including the following steps: Obtain the user to be recommended and the product investment strategy matched by the user to be recommended, where the product investment strategy is used to indicate the recommended investment proportion corresponding to each product category; Based on the product investment strategy, determine the target product category for which product investment is to be made, and the target product category corresponds to a target product label; Obtain the preference data of the user to be recommended for the target product label; wherein, the obtaining of the preference data of the user to be recommended for the target product label includes: obtaining the product investment record of the user to be recommended under the target product category; in the case where the product investment record is not obtained, obtaining the user label data of the user to be recommended according to the set target user label, and based on the user label data, obtaining the preference data of the user to be recommended for the target product label; Based on the preference data of the user to be recommended for the target product label, determine the target product matched with the user to be recommended; Wherein, the method further includes: Obtain a plurality of user labels to be screened; For any one of the user labels, determine the first verification result of the user label according to the correlation coefficient between the user label and the product investment intention, and determine the second verification result of the user label according to the information entropy and / or information value of the user label; Cluster the plurality of user labels to obtain a plurality of clustering clusters; For any one of the user labels, determine the third verification result of the user label according to the tightness between the user label and the clustering cluster; Based on at least one of the first verification result, the second verification result, and the third verification result, perform label screening to obtain the target user label.

2. The method according to claim 1, characterized in that, The obtaining of the preference data of the user to be recommended for the target product label further includes: In the case where the product investment record is obtained, parse the product investment record to obtain the preference data of the user to be recommended for the target product label.

3. The method according to claim 1, characterized in that, The obtaining of the preference data of the user to be recommended for the target product label based on the user label data includes: Based on the user label data, obtain the similar users of the user to be recommended; Obtain the preference data of the similar users for the target product label; Based on the preference data of the similar users for the target product label, determine the preference data of the user to be recommended for the target product label.

4. The method according to claim 1, characterized in that, The performing of label screening based on at least one of the first verification result, the second verification result, and the third verification result to obtain the target user label includes: Based on the correlation coefficient between the user labels, group the user labels to obtain a first label group and at least one second label group, wherein the correlation coefficient between any two user labels in the second label group is greater than or equal to the set coefficient threshold, and the first label group is composed of user labels not included in each of the second label groups; Based on at least one of the first verification result, the second verification result, and the third verification result, perform label screening in the first label group to obtain first user labels; For any one of the second label groups, based on the first verification result, the second verification result, and the third verification result, determine the number of times each user label in the second label group passes the verification, and according to the number of times of passing the verification, perform label screening in the second label group to obtain second user labels; Based on the first user labels and the second user labels, determine the target user labels.

5. The method according to claim 4, characterized in that, The performing label screening in the second label group according to the number of times of passing the verification to obtain second user labels includes: In the second label group, obtain the user label corresponding to the maximum value of the number of times of passing the verification; In the case where the maximum value of the number of times of passing the verification corresponds to multiple user labels, based on the information entropy and / or information value of the user labels, select the second user labels from the multiple user labels corresponding to the maximum value of the number of times of passing the verification.

6. The method according to claim 1, characterized in that The obtaining the user to be recommended includes: In response to receiving a product recommendation request from a user, determine the user as the user to be recommended; Or, obtain the product investment records of the user in each product category, and in response to the time interval between the product maturity date of the corresponding product indicated by the product investment record and the current date being less than the set interval threshold, determine the user as the user to be recommended.

7. The method according to any one of claims 1-6, characterized in that, The determining the target product that matches the user to be recommended based on the preference data of the user to be recommended for the target product label includes: Based on the preference data of the user to be recommended for the target product label, screen candidate products whose product label data in the product pool conforms to the preference data; Taking the selected parameters corresponding to the multiple candidate products as variables, construct an objective function based on the set product recommendation control parameters, where the selected parameters are used to indicate whether a candidate product is selected as a target product; Solve the objective function in combination with the set constraint conditions to obtain the values of the multiple candidate products on the selected parameters, and based on the values of the multiple candidate products on the selected parameters, determine the target product from the candidate products.

8. The method according to any one of claims 1 to 6, characterized in that, The determining the target product that matches the user to be recommended based on the preference data of the user to be recommended for the target product label includes: Based on the preference data of the user to be recommended for the target product label, screen candidate products whose product label data in the product pool conforms to the preference data; Based on the set product evaluation indicators, determine the product scores of the candidate products; According to the product scores, select the target product that matches the user to be recommended from the candidate products.

9. A product recommendation device, characterized in that, including: A first acquisition module, configured to acquire a user to be recommended and the product investment strategy matched by the user to be recommended, where the product investment strategy is used to indicate the recommended investment ratio corresponding to each product category; A first determination module, configured to determine a target product category for which product investment is to be made based on the product investment strategy, and the target product category corresponds to a target product label; A second acquisition module, configured to acquire preference data of the to-be-recommended user for the target product label; wherein, acquiring the preference data of the to-be-recommended user for the target product label includes: acquiring the product investment record of the to-be-recommended user under the target product category; in the case where the product investment record is not acquired, acquiring the user label data of the to-be-recommended user according to the set target user label, and based on the user label data, acquiring the preference data of the to-be-recommended user for the target product label; A second determination module, configured to determine a target product matching the to-be-recommended user based on the preference data of the to-be-recommended user for the target product label; Wherein, the device further includes: A third acquisition module, configured to acquire a plurality of user labels to be screened; A third determination module, configured to, for any one of the user labels, determine a first verification result of the user label according to the correlation coefficient between the user label and the product investment intention, and determine a second verification result of the user label according to the information entropy and / or information value of the user label; A clustering module, configured to cluster the plurality of user labels to obtain a plurality of clustering clusters; A fourth determination module, configured to, for any one of the user labels, determine a third verification result of the user label according to the tightness between the user label and the clustering cluster; A label screening module, configured to perform label screening based on at least one of the first verification result, the second verification result, and the third verification result to obtain the target user label.

Images