Product recommendation method and device, equipment and storage medium

Abstract

The invention discloses a product recommendation method and device, equipment and a storage medium, and relates to the technical field of artificial intelligence, and the method comprises the steps: determining an interested product of a user according to user behavior data of a current application; inputting the user basic information, the user behavior data and a preset cue word of the user into a large model, wherein the large model determines an interested category of the user according to the user basic information, the user behavior data and the cue word; inputting the user basic information, the user behavior data, the interested products and the interested categories into a pre-trained prediction model, wherein the prediction model determines target products corresponding to the user and interested rates of the target products; and determining a recommended product corresponding to the user according to the interestingness rate of each target product. According to the technical scheme, the target products corresponding to the user and the interested rate of each target product are accurately predicted in combination with the interested products and the interested categories of the user, and the products which are purchased by the user at a large probability are recommended to the user.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of artificial intelligence technology and are also applicable to financial technology scenarios, and in particular, to a product recommendation method, device, equipment and storage medium. Background Art

[0002] Financial products refer to various carriers of the financing process, including funds, insurance, financial management, deposits, loans, precious metals, etc. Financial products are the objects of buying and selling in the financial market. The supply and demand sides form the price of financial products, such as interest rates or yields, through the principle of market competition, and finally complete the transaction to achieve the purpose of financing. In order to improve the user experience, recommending suitable financial products to users has become a technical problem that needs to be solved urgently.

[0003] In the prior art, user characteristics are usually determined based on the user's behavioral data in financial applications and the user's portrait data, and then the user's financial products of interest are determined based on the user characteristics, and the financial products of interest are recommended to the user.

[0004] However, the financial products of interest determined in this way are not highly compatible with users. Summary of the invention

[0005] The present invention provides a product recommendation method, device, equipment and storage medium, so as to recommend products with higher adaptability to users.

[0006] In a first aspect, an embodiment of the present invention provides a product recommendation method, comprising:

[0007] Obtaining user behavior data of the user in the current application, and determining the user's interested products according to the user behavior data;

[0008] Inputting the user's basic information, the user's behavior data and a preset prompt word into the macro model, so that the macro model determines the user's interested category according to the user's basic information, the user's behavior data and the prompt word;

[0009] Inputting the user basic information, the user behavior data, the products of interest, and the categories of interest into a pre-trained prediction model, so that the prediction model determines the target products corresponding to the user and the interest rate of each target product;

[0010] The recommended product corresponding to the user is determined according to the interest rate of each of the target products.

[0011] The technical solution of an embodiment of the present invention provides a product recommendation method, including: obtaining user behavior data of a user in a current application, and determining the user's products of interest based on the user behavior data; inputting the user's basic information, the user behavior data, and a pre-set prompt word into a large model, so that the large model determines the user's categories of interest based on the user's basic information, the user behavior data, and the prompt word; inputting the user's basic information, the user behavior data, the products of interest, and the categories of interest into a pre-trained prediction model, so that the prediction model determines the target products corresponding to the user and the interest rate of each target product; and determining the recommended products corresponding to the user based on the interest rate of each target product. The above technical solution can first determine the user's products of interest in the current application based on the user's user behavior data in the current application, and can also input the user's basic information, user behavior data and pre-set prompt words into the big model, so that the big model determines the user's interest category based on the user's basic information, user behavior data and prompt words, and uses the big model knowledge to determine the user's interest category. Secondly, the user's basic information, user behavior data and the user's products of interest and interest categories determined above can be input into a pre-trained prediction model. The prediction model can combine the products of interest determined according to the user behavior data and the categories of interest determined by the big language model to determine the target products corresponding to the user and the interest rate of each target product, and achieve accurate prediction of the target products corresponding to the user and the interest rate of each target product. Then, the recommended products corresponding to the user can be determined according to the interest rate of each target product, so as to recommend products that the user is more likely to purchase, improve the promotion effect of the product, and then increase product sales.

[0012] Further, determining the user's interested products according to the user behavior data includes:

[0013] Behavior keywords are determined according to the user behavior data, and the products of interest are determined according to the behavior keywords.

[0014] Further, determining the user's interested products according to the user behavior data includes:

[0015] Determine the user's favorite products according to the user behavior data;

[0016] The associated products corresponding to the concerned product are determined according to the historical behavior data of each user, and the associated products are determined as the interested products.

[0017] Further, determining the user's interested products according to the user behavior data includes:

[0018] Determine a user vector and a product vector according to the user behavior data;

[0019] The user vector or the product vector is processed based on a vector recall algorithm to determine a target product vector, and the product of interest is determined according to the target product vector.

[0020] Furthermore, before determining the recommended product corresponding to the user according to the interest rate of each target product, the method further includes:

[0021] An adopted product for the user is determined.

[0022] Further, determining the recommended product corresponding to the user according to the interest rate of each target product includes:

[0023] When it is determined that the adopted product exists in the target product, the adopted product in the target product is eliminated to obtain the remaining target product;

[0024] The remaining target products are sorted based on the interest rate, and the remaining target product with the highest interest rate is determined as the recommended product corresponding to the user.

[0025] In a second aspect, an embodiment of the present invention further provides a product recommendation device, including:

[0026] An acquisition module, used to acquire user behavior data of a user in a current application, and determine a product of interest to the user based on the user behavior data;

[0027] A first input module is used to input the user's basic information, the user's behavior data and a preset prompt word into the macro model, so that the macro model determines the user's interested category according to the user's basic information, the user's behavior data and the prompt word;

[0028] A second input module is used to input the user basic information and the user behavior data as well as the interested products and the interested categories into a pre-trained prediction model, so that the prediction model determines the target product corresponding to the user and the interest rate of each target product;

[0029] A recommendation module is used to determine the recommended product corresponding to the user according to the interest rate of each target product.

[0030] In a third aspect, an embodiment of the present invention further provides a computer device, the computer device comprising:

[0031] at least one processor; and a memory communicatively coupled to the at least one processor;

[0032] The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the product recommendation method as described in any one of the first aspects.

[0033] In a fourth aspect, an embodiment of the present invention further provides a storage medium comprising computer executable instructions, wherein the computer executable instructions, when executed by a computer processor, are used to execute a product recommendation method as described in any one of the first aspects.

[0034] In a fifth aspect, the present application provides a computer program product, which includes computer instructions. When the computer instructions are executed on a computer, the computer executes the product recommendation method provided in the first aspect.

[0035] It should be noted that the above computer instructions may be stored in whole or in part on a computer-readable storage medium. The computer-readable storage medium may be packaged together with the processor of the product recommendation device, or may be packaged separately from the processor of the product recommendation device, and this application does not limit this.

[0036] The description of the second, third, fourth and fifth aspects in this application can refer to the detailed description of the first aspect; and the beneficial effects of the description of the second, third, fourth and fifth aspects can refer to the beneficial effect analysis of the first aspect, which will not be repeated here.

[0037] In this application, the name of the above-mentioned product recommendation device does not limit the device or functional module itself. In actual implementation, these devices or functional modules may appear with other names. As long as the functions of each device or functional module are similar to those of this application, they fall within the scope of the claims of this application and their equivalent technologies.

[0038] These and other aspects of the present application will become more apparent from the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0040] Figure 1 A flowchart of a product recommendation method provided by an embodiment of the present invention;

[0041] Figure 2A flowchart of another product recommendation method provided by an embodiment of the present invention;

[0042] Figure 3 A schematic diagram of the structure of a product recommendation device provided by an embodiment of the present invention;

[0043] Figure 4 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0044] The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the present invention, rather than to limit the present invention. It should also be noted that, for ease of description, only parts related to the present invention, rather than all structures, are shown in the accompanying drawings.

[0045] The term "and / or" in this article is merely a description of the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone.

[0046] The terms "first" and "second" and the like in the specification and drawings of this application are used to distinguish different objects, or to distinguish different processing of the same object, rather than to describe a specific order of objects.

[0047] In addition, the terms "including" and "having" and any variations thereof mentioned in the description of the present application are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include other steps or units that are not listed, or may optionally include other steps or units that are inherent to these processes, methods, products or devices.

[0048] It should be mentioned before discussing the exemplary embodiments in more detail that some exemplary embodiments are described as processes or methods depicted as flow charts. Although the flow charts describe the operations (or steps) as sequential processes, many of the operations therein can be implemented in parallel, concurrently or simultaneously. In addition, the order of the operations can be rearranged. The process can be terminated when its operation is completed, but can also have additional steps not included in the accompanying drawings. The process can correspond to methods, functions, procedures, subroutines, subprograms, etc. In addition, the embodiments in the present invention and the features in the embodiments can be combined with each other without conflict.

[0049] It should be noted that, in the embodiments of the present application, words such as "exemplary" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way.

[0050] In the description of the present application, unless otherwise specified, “plurality” means two or more.

[0051] In the prior art, user features are usually input into a pre-trained recommendation model so that the recommendation model can determine the user's financial products of interest based on the user features. Since user features contain a large amount of semantic information, such as the user's hobbies and consumption preferences, when the user features are used as training data for model training, the inherent semantic information of the user features may be lost due to the formal representation of the data. For example, a user clicks on an article about loans. This behavior data contains the user's interest in loans. If this behavior data is simply input into the recommendation model as a user feature, this semantic information may be lost, resulting in the recommendation model being unable to accurately infer the user's interest in loan products.

[0052] Therefore, the present application proposes a product recommendation method, which combines the semantic information in the user features to improve the accuracy of recommending products to users.

[0053] The product recommendation method proposed in this application will be described in detail below with reference to diagrams and embodiments.

[0054] Figure 1 This is a flow chart of a product recommendation method provided by an embodiment of the present invention. This embodiment is applicable to situations where it is necessary to recommend products with a high purchase probability to users. The method can be executed by a product recommendation device, such as Figure 1 As shown, the specific steps include:

[0055] Step 110: Obtain user behavior data of the user in the current application, and determine the product of interest to the user based on the user behavior data.

[0056] The current application may be a financial application, and user behavior data may include the user's clicks, purchases, follows, favorites, and other behaviors in the financial application. Products of interest may be understood as products that the user may be interested in within the current application, such as financial product A and insurance product G.

[0057] Specifically, the user behavior data of the user in the current application indicates the user's interest in the current application, for example, it can indicate the user's interest in the product in the current application. In order to make accurate product recommendations to users, the user's interest in the product in the current application can be determined based on the user's user behavior data in the current application. Specifically, first, the behavior keyword can be determined based on the user behavior data. For example, when the user behavior data indicates that the user has behavior for "financial management", the behavior keyword can be determined to be "financial management". Secondly, the product of interest can be determined based on the behavior keyword. For example, when the behavior keyword is "financial management", the financial products under "financial management" can be determined as the user's product of interest.

[0058] In the embodiment of the present invention, the user's behavior keywords are determined through the user's behavior data in the current application, and the user's products of interest are determined based on the user's behavior keywords, thereby achieving the determination of the user's products of interest.

[0059] Step 120: input the user's basic information, the user behavior data and the preset prompt words into the macro model, so that the macro model determines the user's interested categories according to the user's basic information, the user behavior data and the prompt words.

[0060] Among them, the user's basic information may include gender, age, education, and interests, etc. The category of interest can be understood as the product that the user may be interested in in the current application, which can be specifically at least one of funds, insurance, financial management, deposits, loans, and precious metals. It can be understood that the product of interest can be at least one product under at least one of funds, insurance, financial management, deposits, loans, and precious metals. The large model can be understood as a large language model, which can determine the corresponding output information based on the input information. The input information may include the user's basic information and user behavior data and pre-set prompt words. The pre-set prompt words are used to prompt the large language model to determine the output information required by the user based on the input information. Since the large model is required to determine the user's category of interest, the pre-set prompt words can be "What category of products do users like most? Please choose a word from funds, insurance, financial management, deposits, loans, and precious metals to answer."

[0061] Specifically, after determining the user's basic information and user behavior data, the user's basic information and user behavior data and the prompt word "What category of products do the user like most? Please choose a word from funds, insurance, wealth management, deposits, loans and precious metals to answer" can be input into the large language model as input information. The large language model can determine the output information by analyzing the user's basic information and user behavior data and referring to the prompt word, that is, it can determine that the user's category of interest is at least one of funds, insurance, wealth management, deposits, loans and precious metals.

[0062] In an embodiment of the present invention, by inputting the user's basic information, user behavior data and pre-set prompt words into the big model, the big model can determine the user's interest category based on the user's basic information, user behavior data and prompt words, and use the big model knowledge to determine the user's interest category.

[0063] Step 130: Input the user basic information, the user behavior data, the products of interest, and the categories of interest into a pre-trained prediction model, so that the prediction model determines the target products corresponding to the user and the interest rate of each target product.

[0064] The target product corresponding to the user can be understood as the product that the user may purchase in the current application, and the interest rate of the target product can be understood as the probability that the user may purchase the product in the current application.

[0065] The prediction model is obtained by training a neural network model based on the basic user information, user behavior data, products of interest and categories of interest of a large number of users, as well as the actual purchase of products by each user and the probability of purchasing each product. It can be used to predict the products that users may purchase and the probability of purchasing each product based on the basic user information, user behavior data, products of interest and categories of interest of users.

[0066] The user's products of interest input into the prediction model are determined based on step 110 , and the categories of interest are determined based on step 120 .

[0067] Specifically, the user's basic information, user behavior data, products of interest and categories of interest are input into a pre-trained prediction model. The prediction model can predict the products that the user may purchase and the probability of purchasing each product based on the user's basic information, user behavior data, products of interest and categories of interest, that is, determine the target products corresponding to the user and the interest rate of each target product.

[0068] In the embodiment of the present invention, accurate prediction of the target product corresponding to the user and the interest rate of each target product is achieved.

[0069] Step 140: Determine the recommended product corresponding to the user according to the interest rate of each target product.

[0070] Specifically, after determining the target product corresponding to the user, the target products can be sorted according to the interest rate of the target products. Specifically, the target products can be sorted from large to small according to the interest rate, and the top N target products are determined as recommended products corresponding to the user, where N is a natural number, and the specific value of N can be set according to actual needs, and no specific limitation is made here.

[0071] In actual applications, if the recommended products corresponding to the determined user are the user's purchased products, if these products continue to be recommended to the user, the user's purchase probability is low. Therefore, if the recommended products corresponding to the determined user are the user's purchased products, the target products with the highest ranking other than the recommended products can be added to the recommended products to recommend products to the user that he is more likely to buy, thereby improving the product promotion effect and thereby increasing product sales.

[0072] In the embodiment of the present invention, the recommended products corresponding to the user are determined according to the interest rate of each target product, thereby recommending products with a high probability of purchase to the user, improving the promotion effect of the product, and further increasing the sales volume of the product.

[0073] The product recommendation method provided by the embodiment of the present invention includes: obtaining user behavior data of the user in the current application, and determining the user's products of interest based on the user behavior data; inputting the user's basic information, the user behavior data and pre-set prompt words into a large model, so that the large model determines the user's categories of interest based on the user's basic information, the user behavior data and the prompt words; inputting the user's basic information and the user behavior data as well as the products of interest and the categories of interest into a pre-trained prediction model, so that the prediction model determines the target products corresponding to the user and the interest rate of each target product; and determining the recommended products corresponding to the user based on the interest rate of each target product. The above technical solution can first determine the user's products of interest in the current application based on the user's user behavior data in the current application, and can also input the user's basic information, user behavior data and pre-set prompt words into the big model, so that the big model determines the user's interest category based on the user's basic information, user behavior data and prompt words, and uses the big model knowledge to determine the user's interest category. Secondly, the user's basic information, user behavior data and the user's products of interest and interest categories determined above can be input into a pre-trained prediction model. The prediction model can combine the products of interest determined according to the user behavior data and the categories of interest determined by the big language model to determine the target products corresponding to the user and the interest rate of each target product, and achieve accurate prediction of the target products corresponding to the user and the interest rate of each target product. Then, the recommended products corresponding to the user can be determined according to the interest rate of each target product, so as to recommend products that the user is more likely to purchase, improve the promotion effect of the product, and then increase product sales.

[0074] Figure 2 This is a flowchart of another product recommendation method provided by an embodiment of the present invention. This embodiment is specific based on the above embodiment. Figure 2As shown, in this embodiment, the method may further include:

[0075] Step 210: Obtain user behavior data of the user in the current application.

[0076] Specifically, first, the user behavior data of the user in the current application can be obtained based on the current application, specifically, the products clicked by the user in the last 3 days, the categories clicked in the last 3 days, the most recently purchased product, and the most recently purchased category can be obtained.

[0077] It should be noted that the category can be at least one of funds, insurance, wealth management, deposits, loans and precious metals, and the product can be at least one product under at least one of funds, insurance, wealth management, deposits, loans and precious metals, such as wealth management product A and wealth management product G under wealth management, and loan product C and loan product D under loans.

[0078] In practical applications, after obtaining the user behavior data, the user behavior data can be preprocessed. For example, the user behavior data can be subjected to data cleaning, feature selection, feature extraction, feature construction, processing of missing values, encoding classification (converting non-numeric features into numeric types), and other data processing.

[0079] In the embodiment of the present invention, it is achieved to obtain the user behavior data of the user in the current application.

[0080] Step 220: Determine the products of interest to the user based on the user behavior data.

[0081] In one implementation, step 220 may specifically include:

[0082] Behavior keywords are determined according to the user behavior data, and the products of interest are determined according to the behavior keywords.

[0083] Specifically, first, behavioral keywords can be determined based on user behavior data. For example, when the categories in "products clicked by the user in the last 3 days, categories clicked in the last 3 days, products purchased most recently, and categories purchased most recently" include "wealth management" and "precious metals", and the products include products under "wealth management" and "deposits", the behavioral keywords can be determined to be "wealth management", "deposits" and "precious metals". Secondly, products of interest can be determined based on the behavioral keywords. For example, when the behavioral keywords are "wealth management", "deposits" and "precious metals", wealth management products under "wealth management", deposit wealth management products under "deposits" and precious metal products under "precious metals" can be determined as the user's products of interest.

[0084] In another implementation manner, step 220 may specifically include:

[0085] Determine the user's concerned products according to the user behavior data; determine the associated products corresponding to the concerned products according to the historical behavior data of each user, and determine the associated products as the interested products.

[0086] Specifically, first, the user's products of interest can be determined based on the user's behavior data. For example, when the categories in "financial management" and "precious metals" include "financial management" and "precious metals" and the products include products under "financial management" and "deposits", it can be determined that the user's products of interest are all products under "financial management", all products under "precious metals", and products under "deposits" that are clicked and purchased in the user's behavior data. Secondly, the associated products corresponding to the products of interest can be determined. Specifically, the associated products corresponding to the products of interest can be determined based on the historical behavior data of all users in the current application. Specifically, if the user purchases a combination product consisting of multiple products (such as a combination product consisting of A financial management and B insurance), it is determined that A financial management and B insurance are related products. Furthermore, the related products can be determined as products of interest, that is, the associated products corresponding to the products of interest to the user can be determined as the products of interest to the user.

[0087] In another implementation manner, step 220 may specifically include:

[0088] A user vector and a product vector are determined according to the user behavior data; a target product vector is determined by processing the user vector or the product vector based on a vector recall algorithm, and the product of interest is determined according to the target product vector.

[0089] U2I vector recall is a technology for information retrieval based on vector representation of users and products. It is used to quickly find the products that best match the user's interests in a large collection of products. I2I vector recall is a technology for information retrieval based on vector representation of products. Users can quickly find other products that best match the user's product in a large collection of products.

[0090] Specifically, the user's products of interest can be determined based on U2I vector recall or I2I vector recall. Therefore, the user vector and product vector can be determined based on the user behavior data first. Specifically, the user behavior data can be input into a network model for vector conversion so that the network model determines the user vector and product vector based on the user behavior data. When determining the user's products of interest based on U2I vector recall, similar users can be first determined in the user database based on the user vector, and then the products corresponding to the product vectors corresponding to the similar users are determined as the products of interest corresponding to the user; when determining the user's products of interest based on I2I vector recall, similar product vectors can be determined in the product database based on the product vector, and the products corresponding to the similar product vectors are determined as the products of interest corresponding to the user.

[0091] In practical applications, all the aforementioned products of interest may be determined as the user's products of interest.

[0092] In the embodiment of the present invention, it is implemented to determine the user's products of interest based on the user behavior data of the user in the current application, and to determine a richer range of products of interest to the user in a variety of ways.

[0093] Step 230: input the user's basic information, the user behavior data and the preset prompt words into the macro model, so that the macro model determines the user's interested categories according to the user's basic information, the user behavior data and the prompt words.

[0094] As mentioned above, the user's basic information and user behavior data and the prompt word "What category of products do you like most? Please choose a word from funds, insurance, wealth management, deposits, loans and precious metals to answer" are input into the large language model as input information. The large language model can determine the output information by analyzing the user's basic information and user behavior data and referring to the prompt word, that is, it can determine that the user's category of interest is at least one of funds, insurance, wealth management, deposits, loans and precious metals.

[0095] For example, the input information of the large language model is "a user, occupation is professor, gender is male, age is 60, clicked and browsed four products A fund, B financial management, C financial management, and D financial management on the page in turn. What category of products does the user like most? Please choose a word from funds, insurance, financial management, deposits, loans, and precious metals to answer", and the output information of the large language model can be "financial management", so as to determine the user's category of interest.

[0096] In the embodiment of the present invention, the user's interested categories are determined through a large model, thereby realizing the determination of the user's interested categories by utilizing the large model knowledge.

[0097] Step 240: input the user basic information, the user behavior data, the products of interest, and the categories of interest into a pre-trained prediction model, so that the prediction model determines the target products corresponding to the user and the interest rate of each target product.

[0098] As mentioned above, the user's basic information, user behavior data, products of interest and categories of interest are input into a pre-trained prediction model. The prediction model can predict the products that the user may purchase and the probability of purchasing each product based on the user's basic information, user behavior data, products of interest and categories of interest, that is, determine the target products corresponding to the user and the interest rate of each target product.

[0099] For example, the prediction model may output product 1 and interest rate 1, product 2 and interest rate 2, product 3 and interest rate 3, product 4 and interest rate 4, product 5 and interest rate 5, product 6 and interest rate 6, product 7 and interest rate 7.

[0100] The prediction model can combine the aforementioned products of interest determined based on user behavior data and the categories of interest determined by the large language model to determine the target products corresponding to the user and the interest rate of each target product, thereby achieving accurate prediction of the products that the user may purchase and the probability of purchasing each product.

[0101] By using the output of the large model as the enhanced input of the prediction model, some enhanced user behavior features can be subsequently integrated into the input of the large model in a parameterized manner, eliminating the need for secondary training of the prediction model, effectively reducing the cost of training.

[0102] In the embodiment of the present invention, accurate prediction of the target product corresponding to the user and the interest rate of each target product is achieved.

[0103] Step 250: Determine the product adopted by the user.

[0104] When the target products corresponding to the user include products that the user has purchased, if these products continue to be recommended to the user, the user's experience will decrease. Therefore, before determining the recommended products for the user, the user's adopted products can be determined.

[0105] Specifically, the user's adopted products can be determined based on the products purchased the most recently P times in the user behavior data, where P is a natural number. The specific value of P can also be set according to actual needs and is not specifically limited here.

[0106] In the embodiment of the present invention, by determining the products adopted by the user, the adopted products in the target products are eliminated before determining the recommended products in the target products of the user, thereby improving the accuracy of the recommended products.

[0107] Step 260: Determine the recommended product corresponding to the user according to the interest rate of each target product.

[0108] In one implementation, step 260 may specifically include:

[0109] When it is determined that the adopted product exists in the target products, the adopted product in the target products is eliminated to obtain the remaining target products; the remaining target products are sorted based on the interest rate, and the remaining target product with the highest interest rate is determined as the recommended product corresponding to the user.

[0110] Specifically, we can first determine whether there are adopted products in the target products. When it is determined that there are adopted products in the target products, we can eliminate the adopted products in the target products, that is, we can delete the adopted products in the target products to obtain the remaining target products. Then, we can sort the remaining target products according to the interest rates of the remaining target products. Specifically, we can sort the remaining target products from large to small according to the interest rates, and determine the top N remaining target products as recommended products corresponding to the user.

[0111] For example, if product 1 has been purchased, after removing product 1 from the target products, recommended products can be selected from the remaining products 2, 3, 4, 5, 6 and 7. In order to recommend products that the user is more likely to purchase, the product that the user is most interested in can be determined as the recommended product corresponding to the user.

[0112] In the embodiment of the present invention, the recommended product corresponding to the user is determined according to the interest rate of the target product excluding the adopted product, thereby recommending the product with the highest probability of purchase to the user, improving the promotion effect of the product, and further increasing the sales volume of the product.

[0113] The product recommendation method provided by the embodiment of the present invention includes: obtaining user behavior data of a user in a current application; determining the user's products of interest based on the user behavior data; inputting the user's basic information, the user behavior data and a pre-set prompt word into a large model, so that the large model determines the user's categories of interest based on the user's basic information, the user behavior data and the prompt word; inputting the user's basic information and the user behavior data as well as the products of interest and the categories of interest into a pre-trained prediction model, so that the prediction model determines the target products corresponding to the user and the interest rate of each target product; determining the user's adopted products; and determining the recommended products corresponding to the user based on the interest rate of each target product. The above technical solution, after obtaining the user behavior data of the user in the current application, can determine the user's products of interest in the current application in a variety of ways based on the user behavior data of the user in the current application, determine the user's more abundant products of interest in a variety of ways, and can also input the user's basic information and user behavior data and pre-set prompt words into the big model, so that the big model determines the user's category of interest based on the user's basic information and user behavior data and prompt words, and use the knowledge of the big model to determine the user's category of interest, and then the user's basic information and user behavior data as well as the user's products of interest and categories of interest determined above can be input into a pre-trained prediction model, and the prediction model can combine the products of interest determined based on the user behavior data and the categories of interest determined by the big language model to determine the target products corresponding to the user and the interest rate of each target product, and realize accurate prediction of the target products corresponding to the user and the interest rate of each target product. When it is determined that there are products that the user has adopted in the target products corresponding to the user, the adopted products in the target products are eliminated to obtain the remaining target products, and the recommended products corresponding to the user are determined according to the interest rates of the remaining target products, so as to recommend the products that the user has the highest probability of purchasing to the user, improve the promotion effect of the product, and then increase the sales of the product.

[0114] The product recommendation method combining big models and predictive models provides an effective way to utilize big model knowledge, and recommends recommended products determined by the product recommendation method combining big models and predictive models to users, which significantly improves the user's product purchase rate and is practical.

[0115] Figure 3 This is a schematic diagram of the structure of a product recommendation device provided by an embodiment of the present invention, which can be applied to situations where it is necessary to recommend products with a high purchase probability to users, thereby improving the promotion effect of the products. The device can be implemented by software and / or hardware, and is generally integrated in an electronic device, such as a computer device.

[0116] like Figure 3 As shown, the device comprises:

[0117] An acquisition module 310 is used to acquire user behavior data of a user in a current application, and determine a product of interest to the user according to the user behavior data;

[0118] A first input module 320, for inputting the user's basic information, the user's behavior data and a preset prompt word into the macro model, so that the macro model determines the user's interested category according to the user's basic information, the user's behavior data and the prompt word;

[0119] The second input module 330 is used to input the user basic information and the user behavior data as well as the interested products and the interested categories into a pre-trained prediction model, so that the prediction model determines the target product corresponding to the user and the interest rate of each target product;

[0120] The recommendation module 340 is used to determine the recommended product corresponding to the user according to the interest rate of each target product.

[0121] The product recommendation device provided in this embodiment obtains the user behavior data of the user in the current application, and determines the user's products of interest based on the user behavior data; inputs the user's basic information, the user behavior data and a pre-set prompt word into a large model, so that the large model determines the user's categories of interest based on the user's basic information, the user behavior data and the prompt word; inputs the user's basic information and the user behavior data as well as the products of interest and the categories of interest into a pre-trained prediction model, so that the prediction model determines the target products corresponding to the user and the interest rate of each target product; and determines the recommended products corresponding to the user based on the interest rate of each target product. The above technical solution can first determine the user's products of interest in the current application based on the user's user behavior data in the current application, and can also input the user's basic information, user behavior data and pre-set prompt words into the big model, so that the big model determines the user's interest category based on the user's basic information, user behavior data and prompt words, and uses the big model knowledge to determine the user's interest category. Secondly, the user's basic information, user behavior data and the user's products of interest and interest categories determined above can be input into a pre-trained prediction model. The prediction model can combine the products of interest determined according to the user behavior data and the categories of interest determined by the big language model to determine the target products corresponding to the user and the interest rate of each target product, and achieve accurate prediction of the target products corresponding to the user and the interest rate of each target product. Then, the recommended products corresponding to the user can be determined according to the interest rate of each target product, so as to recommend products that the user is more likely to purchase, improve the promotion effect of the product, and then increase product sales.

[0122] Based on the above embodiment, the acquisition module 310 is specifically used for:

[0123] Obtain user behavior data of the user in the current application, determine behavior keywords based on the user behavior data, and determine the product of interest based on the behavior keywords.

[0124] Based on the above embodiment, the acquisition module 310 is specifically used for:

[0125] Obtain user behavior data of the user in the current application, and determine the user's concerned products according to the user behavior data; determine the associated products corresponding to the concerned products according to the historical behavior data of each user, and determine the associated products as the interested products.

[0126] Based on the above embodiment, the acquisition module 310 is specifically used for:

[0127] Obtain user behavior data of the user in the current application, and determine a user vector and a product vector based on the user behavior data; process the user vector or the product vector based on a vector recall algorithm to determine a target product vector, and determine the product of interest based on the target product vector.

[0128] Based on the above embodiment, the device further includes:

[0129] The determination module is used to determine the products adopted by the user before determining the recommended products corresponding to the user according to the interest rates of the target products.

[0130] Based on the above embodiment, the recommendation module 340 is specifically used for:

[0131] When it is determined that the adopted product exists in the target products, the adopted product in the target products is eliminated to obtain the remaining target products; the remaining target products are sorted based on the interest rate, and the remaining target product with the highest interest rate is determined as the recommended product corresponding to the user.

[0132] The product recommendation device provided in the embodiment of the present invention can execute the product recommendation method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of executing the product recommendation method.

[0133] It is worth noting that in the embodiment of the above-mentioned product recommendation device, the various units and modules included are only divided according to functional logic, but are not limited to the above-mentioned division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention.

[0134] Figure 4 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. Figure 4 A block diagram of an exemplary electronic device 4 suitable for implementing embodiments of the present invention is shown. Figure 4 The electronic device 4 shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0135] like Figure 4 As shown, the electronic device 4 is in the form of a general purpose computing electronic device. The components of the electronic device 4 may include, but are not limited to: one or more processors or processing units 16, a system memory 28, and a bus 18 connecting different system components (including the system memory 28 and the processing unit 16).

[0136] Bus 18 represents one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, a processor or a local bus using any of a variety of bus architectures. By way of example, these architectures include, but are not limited to, an Industry Standard Architecture (ISA) bus, a Micro Channel Architecture (MAC) bus, an Enhanced ISA bus, a Video Electronics Standards Association (VESA) local bus, and a Peripheral Component Interconnect (PCI) bus.

[0137] The electronic device 4 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by the electronic device 4, including volatile and non-volatile media, removable and non-removable media.

[0138] The system memory 28 may include computer system readable media in the form of volatile memory, such as random access memory (RAM) 30 and / or cache memory 32. The electronic device 4 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, the storage system 34 may be used to read and write non-removable, non-volatile magnetic media ( Figure 4 not shown, usually called a "hard drive"). Although Figure 4 Not shown in the figure, a disk drive for reading and writing to a removable non-volatile disk (e.g., a "floppy disk"), and an optical disk drive for reading and writing to a removable non-volatile optical disk (e.g., a CD-ROM, a DVD-ROM, or other optical media) may be provided. In these cases, each drive may be connected to the bus 18 via one or more data medium interfaces. The system memory 28 may include at least one program product having a set (e.g., at least one) of program modules that are configured to perform the functions of the various embodiments of the present invention.

[0139] A program / utility 40 having a set (at least one) of program modules 42 may be stored, for example, in system memory 28, such program modules 42 including, but not limited to, an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment. Program modules 42 generally perform the functions and / or methods of the embodiments described herein.

[0140] The electronic device 4 may also communicate with one or more external devices 14 (e.g., keyboards, pointing devices, displays 24, etc.), one or more devices that enable a user to interact with the electronic device 4, and / or any device that enables the electronic device 4 to communicate with one or more other computing devices (e.g., network cards, modems, etc.). Such communication may be performed via an input / output (I / O) interface 22. Furthermore, the electronic device 4 may also communicate with one or more networks (e.g., a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) via a network adapter 20. Figure 4 As shown, the network adapter 20 communicates with other modules of the electronic device 4 via the bus 18. It should be understood that although Figure 4 Not shown, other hardware and / or software modules may be used in conjunction with the electronic device 4, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0141] The processing unit 16 executes various functional applications and page displays by running the programs stored in the system memory 28, for example, implementing the product recommendation method provided in the embodiment of the present invention, which includes:

[0142] Obtaining user behavior data of the user in the current application, and determining the user's interested products according to the user behavior data;

[0143] Inputting the user's basic information, the user's behavior data and a preset prompt word into the macro model, so that the macro model determines the user's interested category according to the user's basic information, the user's behavior data and the prompt word;

[0144] Inputting the user basic information, the user behavior data, the products of interest, and the categories of interest into a pre-trained prediction model, so that the prediction model determines the target products corresponding to the user and the interest rate of each target product;

[0145] The recommended product corresponding to the user is determined according to the interest rate of each of the target products.

[0146] Of course, those skilled in the art can understand that the processor can also implement the technical solution of the product recommendation method provided by any embodiment of the present invention.

[0147] An embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon. When the program is executed by a processor, for example, a product recommendation method provided by an embodiment of the present invention is implemented. The method includes:

[0148] Obtaining user behavior data of the user in the current application, and determining the user's interested products according to the user behavior data;

[0149] Inputting the user's basic information, the user's behavior data and a preset prompt word into the macro model, so that the macro model determines the user's interested category according to the user's basic information, the user's behavior data and the prompt word;

[0150] Inputting the user basic information, the user behavior data, the products of interest, and the categories of interest into a pre-trained prediction model, so that the prediction model determines the target products corresponding to the user and the interest rate of each target product;

[0151] The recommended product corresponding to the user is determined according to the interest rate of each of the target products.

[0152] The computer storage medium of the embodiment of the present invention can adopt any combination of one or more computer-readable media. The computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium can be, for example, but not limited to: an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples (non-exhaustive list) of computer-readable storage media include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, a computer-readable storage medium can be any tangible medium containing or storing a program, which can be used by an instruction execution system, device or device or used in combination with it.

[0153] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, which carry computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0154] The program code embodied on the computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

[0155] Computer program code for performing the operations of the present invention may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0156] It should be understood by those skilled in the art that the modules or steps of the present invention described above can be implemented by a general-purpose computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, optionally, they can be implemented by a program code executable by a computer device, so that they can be stored in a storage device and executed by the computing device, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software.

[0157] In addition, the information collected in the technical solution of the present invention is information and data authorized by the user or fully authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of relevant data comply with the relevant laws, regulations and standards of relevant countries and regions, take necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0158] For example, after collecting basic user information and user behavior data, we will de-identify the data through technical means.

[0159] Note that the above are only preferred embodiments of the present invention and the technical principles used. Those skilled in the art will understand that the present invention is not limited to the specific embodiments herein, and that various obvious changes, readjustments and substitutions can be made by those skilled in the art without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in more detail through the above embodiments, the present invention is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present invention, and the scope of the present invention is determined by the scope of the appended claims.

Claims

1. A product recommendation method, characterized in that: include: Obtaining user behavior data of the user in the current application, and determining the user's interested products according to the user behavior data; Inputting the user's basic information, the user's behavior data and a preset prompt word into the macro model, so that the macro model determines the user's interested category according to the user's basic information, the user's behavior data and the prompt word; Inputting the user basic information, the user behavior data, the products of interest, and the categories of interest into a pre-trained prediction model, so that the prediction model determines the target products corresponding to the user and the interest rate of each target product; The recommended product corresponding to the user is determined according to the interest rate of each of the target products.

2. The product recommendation method according to claim 1, characterized in that: Determining the user's interested products according to the user behavior data includes: Behavior keywords are determined according to the user behavior data, and the products of interest are determined according to the behavior keywords.

3. The product recommendation method according to claim 1, characterized in that: Determining the user's interested products according to the user behavior data includes: Determine the user's favorite products according to the user behavior data; The associated products corresponding to the concerned product are determined according to the historical behavior data of each user, and the associated products are determined as the interested products.

4. The product recommendation method according to claim 1, characterized in that: Determining the user's interested products according to the user behavior data includes: Determine a user vector and a product vector according to the user behavior data; The user vector or the product vector is processed based on a vector recall algorithm to determine a target product vector, and the product of interest is determined according to the target product vector.

5. The product recommendation method according to claim 1, characterized in that: Before determining the recommended product corresponding to the user according to the interest rate of each target product, the method further includes: An adopted product for the user is determined.

6. The product recommendation method according to claim 5, characterized in that: Determining the recommended product corresponding to the user according to the interest rate of each target product includes: When it is determined that the adopted product exists in the target product, the adopted product in the target product is eliminated to obtain the remaining target product; The remaining target products are sorted based on the interest rate, and the remaining target product with the highest interest rate is determined as the recommended product corresponding to the user.

7. A product recommendation device, characterized in that: include: An acquisition module, used to acquire user behavior data of a user in a current application, and determine a product of interest to the user based on the user behavior data; A first input module is used to input the user's basic information, the user's behavior data and a preset prompt word into the macro model, so that the macro model determines the user's interested category according to the user's basic information, the user's behavior data and the prompt word; A second input module is used to input the user basic information and the user behavior data as well as the interested products and the interested categories into a pre-trained prediction model, so that the prediction model determines the target product corresponding to the user and the interest rate of each target product; A recommendation module is used to determine the recommended product corresponding to the user according to the interest rate of each target product.

8. A computer device, characterized in that: The computer device comprises: at least one processor; and a memory communicatively coupled to the at least one processor; The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the product recommendation method as described in any one of claims 1-6.

9. A storage medium containing computer executable instructions, characterized in that: The computer executable instructions are used to execute the product recommendation method as described in any one of claims 1-6 when executed by a computer processor.

10. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the computer program implements the product recommendation method according to any one of claims 1 to 6.

Images