CN112015970A - Product recommendation method, related equipment and computer storage medium - Google Patents
Product recommendation method, related equipment and computer storage medium Download PDFInfo
- Publication number
- CN112015970A CN112015970A CN201910461563.3A CN201910461563A CN112015970A CN 112015970 A CN112015970 A CN 112015970A CN 201910461563 A CN201910461563 A CN 201910461563A CN 112015970 A CN112015970 A CN 112015970A
- Authority
- CN
- China
- Prior art keywords
- product
- products
- user
- list
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0631—Item recommendations
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Physics & Mathematics (AREA)
- Economics (AREA)
- General Business, Economics & Management (AREA)
- Strategic Management (AREA)
- Marketing (AREA)
- Development Economics (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The application provides a product recommendation method and related equipment, wherein the method comprises the following steps: obtaining user data of at least one user and product data of a plurality of products from a database; performing characteristic engineering on the user data and the product data of the plurality of products to obtain characteristic information; obtaining similarity information of the products based on the product characteristic data of the products; and obtaining a product recommendation result of a target user in the at least one user by using a deep learning model based on the feature information, the similarity information of the plurality of products and a current popular product list, wherein the current popular product list comprises at least one product in the plurality of products. Through analyzing user data and product data, can recommend different products for the user to the preference degree or the degree of concern of different users to different products to the realization is promoted the precision of product, improves product popularization efficiency.
Description
Technical Field
The present invention relates to the field of computer technologies, and in particular, to a product recommendation method, related devices, and a computer storage medium.
Background
With the development of network technology, a network becomes one of main platforms for product promotion, traditional marketing is market-oriented, is not data-driven, needs a large amount of human resources to carry out market research, and has no pertinence when promoting products, for example, when product recommendation is carried out through the network at present, what is recommended for different users is the current main product, for example, when new products appear, new products can be recommended to all users at the same time. However, different users have different interests and attention degrees for different products, and the effect of performing unified product recommendation needs to be improved.
Disclosure of Invention
The embodiment of the invention discloses a product recommendation scheme.
In a first aspect, an embodiment of the present application provides a product recommendation method, where the method includes: obtaining user data of at least one user and product data of a plurality of products from a database; performing feature engineering on the user data of the at least one user and the product data of the plurality of products to obtain feature information, wherein the feature information comprises the user feature data of the at least one user and the product feature data of the plurality of products; obtaining similarity information of the products based on the product characteristic data of the products; and obtaining a product recommendation result of a target user in the at least one user by using a deep learning model based on the feature information, the similarity information of the plurality of products and a current popular product list, wherein the current popular product list comprises at least one product in the plurality of products.
In a possible implementation manner, the obtaining, by using a deep learning model, a product recommendation result of a target user of the at least one user based on the feature information, the similarity information of the plurality of products, and the current popular product list includes: screening the products according to the user characteristic data of the target user and the product characteristic data of the products to obtain a candidate product list of the target user, wherein the candidate product list comprises at least one candidate product in the products; and obtaining a product recommendation result of the target user by utilizing a deep learning model based on the feature data of the target user, the similarity corresponding to the at least one candidate product contained in the similarity information of the products and the current popular product list.
In a possible implementation manner, the screening the multiple products according to the user feature data of the target user and the product feature data of the multiple products to obtain the candidate product list of the target user includes: screening the plurality of products according to feature data which is included in the user feature data of the target user and used for representing the product requirements of the target user and feature data which is included in the product feature data of the plurality of products and used for representing the product attributes of the products to obtain a candidate product list of the target user, wherein the product requirements include at least one of the following items: the type of the product desired to be purchased by the target user, the risk acceptance information of the target user, the fund information of the target user and the desired product income information of the target user, wherein the product attributes comprise: product type, product risk level, product open purchase time, minimum purchase funds for the product, and revenue information for the product.
In a possible implementation manner, the similarity corresponding to the at least one candidate product includes a similarity between the at least one candidate product and at least one current popular product included in the current popular product list; the obtaining, by using a deep learning model, a product recommendation result of the target user based on the feature information, the similarity corresponding to the at least one candidate product included in the similarity information of the plurality of products, and the current popular product list, includes: updating the candidate product list of the target user based on the similarity between the at least one candidate product and at least one current popular product contained in the current popular product list to obtain an updated candidate product list; and obtaining a product recommendation result of the target user by utilizing a deep learning model based on the user characteristic data of the target user and the product characteristic data of at least one candidate product contained in the updated candidate product list.
In a possible implementation manner, the updating the candidate product list of the target user based on the similarity between the at least one candidate product and the at least one current top product included in the current top product list to obtain an updated candidate product list includes: determining whether a target candidate product with the similarity between the target candidate product and a target popular product in the current popular product list being greater than a first preset threshold exists in the candidate product list of the target user; and responding to the target candidate product in the candidate product list of the target user, and adding the target popular product to the candidate product list of the target user to obtain an updated candidate product list.
In one possible implementation, the adding the target trending product to the candidate product list of the target user includes: replacing the target candidate product in the candidate product list of the target user with the target popular product.
In a possible implementation manner, the obtaining, by using a deep learning model, a product recommendation result of a target user based on the feature information, the similarity information of the plurality of products, and the current popular product list includes: based on the feature information, the similarity information of the products and the current popular product list, obtaining a product recommendation list of a target user by using a deep learning model; and in response to the first product and the second product with the similarity reaching a second preset threshold exist in the product recommendation list and the first product included in the current popular product list is ranked behind the second product in the product recommendation list, interchanging the positions of the first product and the second product in the product recommendation list to obtain a product recommendation result.
In one possible implementation manner, the user feature data of the target user includes user static information and user dynamic information, where the user static information includes at least one of the following: the user identification, the occupation of the user, the age of the user and the area where the user is located; the user dynamic information comprises at least one of the following: the method comprises the steps of obtaining information of products purchased by a user, information of the number of the products purchased by the user, the buried point data of the products browsed by the user, the path of the products purchased by the user, the profit demand of the user, the fund state of the user and the risk threshold set by the user.
In one possible implementation, the feature information further includes interaction features between the at least one user and the plurality of products.
In one possible implementation, the product characteristic data of the plurality of products includes at least one of: product identification, product category, product open purchase time, product risk level, whether the product is a popular product, revenue information of the product, maximum withdrawal of the product, and minimum purchase funds of the product.
In one possible implementation manner, the interactive features include the number of times that the target user purchases different products, one or more ways that the target user purchases different products, the number of times that the target user purchases different products through the first way, the attention information of the target user for different products, the time and frequency that the target user purchases different products, and the web page ranking pagerank of the products relative to the target user.
In one possible implementation manner, the plurality of products include a plurality of existing products and at least one added product, and the similarity information includes a similarity between every two products in the plurality of products.
In a possible implementation manner, after obtaining a product recommendation result of a target user of the at least one user by using a deep learning model based on the feature information, the similarity information of the plurality of products, and the current popular product list, the method further includes: and sending the product recommendation result to a terminal device, or displaying the product recommendation result.
In a possible implementation manner, after obtaining the product recommendation result of the target user by using a deep learning model based on the feature information, the similarity information of the plurality of products, and the current popular product list, the method further includes: and adjusting the network parameters of the deep learning model according to the product recommendation result and the product purchase information and/or the product browsing information of the target user after the product recommendation result is obtained, so as to obtain the deep learning model for recommending the product in the next time period.
By implementing the method provided by the embodiment of the application, the user characteristic data, the product characteristic data and the interactive characteristics implicit between the user and the products are obtained by performing characteristic engineering on the user data of at least one user and the product data of the products, and the data are analyzed based on the deep learning model according to the user characteristic data, the product characteristic data and the interactive characteristics, so that different products can be recommended to different users according to the user requirements embodied in the user characteristic data, the products can be accurately promoted according to the user preferences, and the product promotion efficiency is improved. In addition, by calculating the similarity information among the products, the product recommendation can be carried out according to the similarity among the products, so that the influence caused by the large number of users, the large number of products, the sparse data between the user requirements and the product attributes in the product recommendation can be reduced.
In a second aspect, an embodiment of the present application provides a product recommendation device, where the device includes:
an acquisition unit for acquiring user data of at least one user and product data of a plurality of products from a database; a feature unit, configured to perform feature engineering on the user data of the at least one user and the product data of the multiple products to obtain feature information, where the feature information includes the user feature data of the at least one user and the product feature data of the multiple products; the similarity calculation unit is used for obtaining similarity information of the products based on the product characteristic data of the products; and the processing unit is used for obtaining a product recommendation result of a target user in the at least one user by utilizing a deep learning model based on the feature information, the similarity information of the plurality of products and a current popular product list, wherein the current popular product list comprises at least one product in the plurality of products.
In one possible implementation, the apparatus further includes: the screening unit is used for screening the products according to the user characteristic data of the target user and the product characteristic data of the products to obtain a candidate product list of the target user, wherein the candidate product list comprises at least one candidate product in the products; the processing unit is specifically configured to obtain a product recommendation result of the target user by using a deep learning model based on the feature data of the target user, the similarity corresponding to the at least one candidate product included in the similarity information of the plurality of products, and the current popular product list.
In a possible implementation manner, the screening unit is specifically configured to: screening the plurality of products according to feature data which is included in the user feature data of the target user and used for representing the product requirements of the target user and feature data which is included in the product feature data of the plurality of products and used for representing the product attributes of the products to obtain a candidate product list of the target user, wherein the product requirements include at least one of the following items: the type of the product desired to be purchased by the target user, the risk acceptance information of the target user, the fund information of the target user and the desired product income information of the target user, wherein the product attributes comprise: product type, product risk level, product open purchase time, minimum purchase funds for the product, and revenue information for the product.
In a possible implementation manner, the similarity corresponding to the at least one candidate product includes a similarity between the at least one candidate product and at least one current popular product included in the current popular product list; the processing unit is specifically configured to: updating the candidate product list of the target user based on the similarity between the at least one candidate product and at least one current popular product contained in the current popular product list to obtain an updated candidate product list; and obtaining a product recommendation result of the target user by utilizing a deep learning model based on the user characteristic data of the target user and the product characteristic data of at least one candidate product contained in the updated candidate product list.
In a possible implementation manner, the processing unit is specifically configured to: determining whether a target candidate product with the similarity between the target candidate product and a target popular product in the current popular product list being greater than a first preset threshold exists in the candidate product list of the target user; and responding to the target candidate product in the candidate product list of the target user, and adding the target popular product to the candidate product list of the target user to obtain an updated candidate product list.
In one possible implementation, the adding the target trending product to the candidate product list of the target user includes: replacing the target candidate product in the candidate product list of the target user with the target popular product.
In a possible implementation manner, the processing unit is specifically configured to: based on the feature information, the similarity information of the products and the current popular product list, obtaining a product recommendation list of a target user by using a deep learning model; and in response to the first product and the second product with the similarity reaching a second preset threshold exist in the product recommendation list and the first product included in the current popular product list is ranked behind the second product in the product recommendation list, interchanging the positions of the first product and the second product in the product recommendation list to obtain a product recommendation result.
In one possible implementation manner, the user feature data of the target user includes user static information and user dynamic information, where the user static information includes at least one of the following: the user identification, the occupation of the user, the age of the user and the area where the user is located; the user dynamic information comprises at least one of the following: the method comprises the steps of obtaining information of products purchased by a user, information of the number of the products purchased by the user, the buried point data of the products browsed by the user, the path of the products purchased by the user, the profit demand of the user, the fund state of the user and the risk threshold set by the user.
In one possible implementation, the feature information further includes interaction features between the at least one user and the plurality of products.
In one possible implementation, the product characteristic data of the plurality of products includes at least one of: product identification, product category, product open purchase time, product risk level, whether the product is a popular product, revenue information of the product, maximum withdrawal of the product, and minimum purchase funds of the product.
In one possible implementation manner, the interactive features include the number of times that the target user purchases different products, one or more ways that the target user purchases different products, the number of times that the target user purchases different products through the first way, the attention information of the target user for different products, the time and frequency that the target user purchases different products, and the web page ranking pagerank of the products relative to the target user.
In one possible implementation manner, the plurality of products include a plurality of existing products and at least one added product, and the similarity information includes a similarity between every two products in the plurality of products.
In one possible implementation, the apparatus further includes: and the sending unit is used for sending the product recommendation result to the terminal equipment, or the display unit displays the product recommendation result.
In one possible implementation, the processing unit is further configured to: and adjusting the network parameters of the deep learning model according to the product recommendation result and the product purchase information and/or the product browsing information of the target user after the product recommendation result is obtained, so as to obtain the deep learning model for recommending the product in the next time period.
In a third aspect, the present application provides an electronic device comprising a processor, a communication interface, and a memory; the memory is configured to store instructions, the processor is configured to execute the instructions, and the communication interface is configured to receive or transmit data; wherein the processor executes the instructions to perform the method as described in the first aspect and any possible implementation manner of the first aspect.
In a fourth aspect, the present application provides a computer-readable storage medium having stored thereon computer program instructions which, when executed by a processor, implement the method described in the first aspect above and any possible implementation manner of the first aspect.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.
Fig. 1 is a schematic flowchart of a product recommendation method provided in an embodiment of the present application;
FIG. 2 is a schematic flow chart diagram illustrating another product recommendation method provided in an embodiment of the present application;
FIG. 3 is a schematic structural diagram of a product recommendation device according to an embodiment of the present application;
fig. 4 is a schematic block diagram of an electronic device according to an embodiment of the present application.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
It will be understood that the terms "comprises" and/or "comprising," when used in this specification and the appended claims, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.
It is also to be understood that the terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used in the specification of the present invention and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise.
It should be further understood that the term "and/or" as used in this specification and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.
As used in this specification and the appended claims, the term "if" may be interpreted contextually as "when", "upon" or "in response to a determination" or "in response to a detection". Similarly, the phrase "if it is determined" or "if a [ described condition or event ] is detected" may be interpreted contextually to mean "upon determining" or "in response to determining" or "upon detecting [ described condition or event ]" or "in response to detecting [ described condition or event ]".
The embodiment of the application provides a product recommendation method, which can be applied to recommending products to users on a network platform. Referring to fig. 1, fig. 1 is a schematic flowchart of a product recommendation method provided in an embodiment of the present application, and as shown in fig. 1, the method includes:
101. user data for at least one user and product data for a plurality of products are obtained from a database.
Wherein the plurality of products comprise a plurality of existing products and at least one added product. The user data comprises user static information and user dynamic information, and the user static information comprises at least one of the following items: a user identification, a user's occupation, a user's age, a user's income, and a user's location; the user dynamic information comprises at least one of the following: the method comprises the steps of obtaining information of products purchased by a user, information of the number of the products purchased by the user, the buried point data of the products browsed by the user, the path of the products purchased by the user, the profit demand of the user, the fund state of the user and the risk threshold set by the user.
The information of the purchased products comprises the name, the category, the route, the quantity and the like of the base products purchased by the user each time; the way for the user to purchase the product includes purchase through an exchange, purchase through an application, purchase through a phone, and the like; the product information browsed by the user includes the number of times each product is browsed by the user, the stay time when each product is browsed, and the like.
The product data includes a name or code of the product, type information of the product, income information of the product, open purchase time of the product, a risk level of the product, a product size of the product, maximum withdrawal of the product, minimum purchase fund of the product, and the like. For example, in the case of a fund product, the product categories include stock type, index type, mix type, bond type, currency type, and the like; the return rate of the product comprises seven-day-year return of one month in the past of fund, seven-day-year return of three months in the past, and the like; the product risk grades comprise five risk grades of high risk, high and medium risk, medium and low risk; the open time of the product is the open purchase time of the fund product; the minimum purchase fund of the product is the minimum cost for each fund purchase of the user. It is to be understood that the above examples are intended to be illustrative only and are not to be construed as limiting in any way.
102. And performing characteristic engineering on the user data of the at least one user and the product data of the plurality of products to obtain characteristic information.
The characteristic information includes user characteristic data of the at least one user and product characteristic data of the plurality of products. And performing data cleaning, feature construction, feature extraction, feature selection and other processing on the user data and the product data through feature engineering, and converting the original data such as the user data and the product data into feature data expressed by feature vectors. The feature vector may be an embedding vector.
103. And obtaining the similarity information of the products based on the product characteristic data of the products.
After the product characteristic data of each product is obtained, calculating similarity information between each two products in the plurality of products according to the product characteristic data, wherein the similarity information may be euclidean metric (euclidean metric) between the product characteristic data of the two products, may also be cosine similarity, and may also be Pearson (Pearson) correlation coefficient, which is not specifically limited in the application.
104. And obtaining a product recommendation result of a target user in the at least one user by utilizing a deep learning model based on the feature information, the similarity information of the products and the current popular product list.
And obtaining a product recommendation result of a target user in the at least one user by utilizing a deep learning model according to the user characteristic data of each user in the at least one user, the product characteristic data of the plurality of products, the similarity information between every two products in the plurality of products, the current popular product and the like. The current popular product list comprises at least one product in the plurality of products, and the current popular product is a recommended product designated in the period and/or a hot sale product obtained through data analysis of each product purchased by the user in the previous period.
The product recommendation result is a product ranking corresponding to the target user obtained through a deep learning model, and products ranked more forward in the product recommendation result represent that the target user has a higher possibility of purchasing the products, wherein the period refers to a period from when the server acquires user data and product data from the database to when the server acquires the user data and the product data from the database next time.
In some embodiments, the deep learning model may be any one of the following learning models: residual network (ResNet) model, VGG16 model, VGGNet model, inclusion model, Full Convolutional Network (FCN) model, multitask network cascade MNC model, and Mask-RCNN model.
In some embodiments, the characteristic information may further include interaction characteristics between the user and the plurality of products. The server may mine more implicit interaction features between each user and each product based on the user characteristic data of at least one user and the product characteristic data of the plurality of products. For example, by counting the information of the products purchased by the user, the frequency of purchasing the product a by the user a, the path of purchasing the product a by the user a each time, the proportion of the times of purchasing the product a by the user a through the application program, the web page ranking pagerank of each product of the plurality of products relative to the user, the average value of the ages of the users who purchase the product a, the proportion of the number of users who purchase the product a to the number of all users, and the like can be obtained; the relationship between the age of the user and the risk level of the product expected to be purchased can be obtained according to the age of each user and the risk level selected by each user; the relationship between the user income and the purchased product type can be obtained by analyzing the product type purchased by the user in history and the income of the user; the number and frequency of different products purchased by the user or the frequency and duration of browsing different product information can be trained and analyzed to obtain the attention degree of different users to different products, and the like.
The interaction characteristics are obtained by analyzing the user data and the product data of the products, the attention degree of the user to different products or users suitable for different products and the like can be better described, so that the deep learning model can obtain more information, and a product recommendation list finally obtained by the deep learning model is more in line with the requirements of the user on the products. For example, the deep learning model may analyze and obtain people aged over thirty-five years by learning the interactive features, the proportion of the low-risk products to be purchased is large, and if the target user is aged over thirty-five years, the deep learning model may rank the product with high risk at a position behind the product recommendation result corresponding to the target user.
In some embodiments, as shown in fig. 2, fig. 2 is a schematic diagram of another product recommendation method provided in the embodiments of the present application. The server obtains user data of the at least one user and product data of the products from a database, performs feature engineering on the user data of the at least one user and the product data of the products to obtain feature information, then screens the products according to the user feature data of the target user and the product feature data of the products to obtain a candidate product list of the target user, and then obtains a product recommendation result of the target user by using a deep learning model based on the feature data of the target user, the product feature information of the candidate products in the candidate product list, the similarity information corresponding to the candidate products in the candidate list and the current popular product list. The candidate product list comprises at least one candidate product in the plurality of products, and the products are screened, so that the data volume of subsequent processing of the server can be reduced, the efficiency is improved, and more accurate recommendation for different users is realized.
Specifically, when the multiple products are screened according to the user feature data of the target user and the product feature data of the multiple products, the server may determine feature information used for characterizing product requirements of the target user based on the user feature data of the target user, determine feature information used for characterizing product attributes of the multiple products based on the product feature data of the multiple products, and further screen the multiple products according to the feature information used for characterizing the product requirements of the target user and the feature information used for characterizing the product attributes of the multiple products to obtain a candidate product list of the target user. Wherein the product requirements include at least one of: the type of a product which the target user desires to purchase, the risk acceptance information of the target user, the fund information of the target user and the expected product income information of the target user; the product attributes include: product type, product risk level, product open purchase time, minimum purchase funds for the product, and revenue information for the product. For example, if the product demand of the target user requires that the fund desired to be purchased is greater than or equal to one hundred million in size, the fund with the fund size less than one hundred million in the plurality of products is deleted from the recommended products of the target user; for another example, if the time when a certain product is opened for purchase does not include the current time, the product whose opening time does not include the current time is deleted from the recommended product; for another example, if the risk level of the product desired to be purchased in the customization request of the user is middle risk, the product with higher risk level than middle risk is deleted from the recommended product of the user.
In some embodiments, after the candidate product list is obtained by screening the plurality of products in the above manner, the popular product may be screened from recommended products of a plurality of users, and in order to recommend the popular product to as many users as possible, as shown in fig. 2, the server may update the candidate product list of the target user based on a similarity between the at least one candidate product and a current popular product included in the current popular product list, so as to obtain an updated candidate product list; specifically, the server determines whether a target candidate product with a similarity greater than a first preset threshold with a target popular product in the current popular product list exists in the candidate product list of the target user; and if the target candidate product exists in the candidate product list of the target user, adding the target popular product to the candidate product list of the target user, or replacing the target candidate product with the target popular product to obtain an updated candidate product list. And the server obtains a product recommendation result of the target user by utilizing a deep learning model according to the user characteristic data of the target user and the product characteristic data of at least one candidate product contained in the updated candidate product list.
In some embodiments, since the popular product may be a popular product obtained by counting according to the information of the product purchased by the user in the previous period, the possibility of being purchased by the user in the period is high; the popular product may also be a product that a customer (product provider) specifies needs to be recommended or a new product that the customer has launched, and therefore needs to be displayed at the front in the product recommendation results of as many users as possible. Therefore, in order to meet the needs of customers or more products for sale, the server processes the user feature data of the target user, the product feature data of the products, the similarity information of the products, and the current popular product list by using a deep learning model to obtain a product recommendation list of the target user of the at least one user, and if a first product and a second product with a similarity larger than a second preset threshold exist in the product recommendation list and the first product included in the current popular product list is ranked behind the second product in the product recommendation list, interchanges the positions of the first product and the second product in the product recommendation list to obtain a product recommendation result.
The user data and the product data are subjected to learning analysis to obtain the implicit interactive characteristics between the user data and the product data, the preference degrees or the attention degrees of different users to different products are obtained according to the user data, the product data and the interactive characteristics between the user data and the product data, then the product similarity information and the popular product information are combined, different products are recommended to the different users, products are accurately promoted according to the requirements of the users, and the product popularization efficiency is improved.
In some embodiments, after obtaining the product recommendation result corresponding to the target user, the server may send the product recommendation result to a user terminal device, or display the product recommendation result on a display device of the server.
In some embodiments, the server may adjust network parameters of the deep learning model according to the product recommendation result, product purchase information and/or product browsing information of the target user in the period after obtaining the product recommendation result, and obtain the deep learning model for product recommendation in a next time period.
In some embodiments, the server may further update the user data in the database and the product data of the multiple products according to the product purchase information and/or the product browsing information in the period, for example, update browsing times corresponding to the goods browsed by the target user, increase a product name, a purchase amount, a purchase manner, and the like of a certain product purchased by the target user, update browsing times of a certain product, update an accumulated sales volume of a certain product, and the like. In addition, the server needs to obtain revenue information of each product, open time of each product, and the like, and update data information of each product in the database. And taking the updated user data and product data in the database as data acquired from the database in the next period service, thereby overcoming the problem of weak timeliness of product recommendation results caused by large change and strong time sequence of financial product information.
By implementing the method provided by the embodiment of the application, the user characteristic data, the product characteristic data and the interactive characteristics implicit between the user and the products are obtained by performing characteristic engineering on the user data of at least one user and the product data of the products, and the data are analyzed based on the deep learning model according to the user characteristic data, the product characteristic data and the interactive characteristics, so that different products can be recommended to different users according to the user requirements embodied in the user characteristic data, the products can be accurately promoted according to the preference of the users, and the product recommendation efficiency is improved. In addition, by calculating the similarity information among the products, the product recommendation can be carried out according to the similarity among the products, so that the influence caused by huge number of users, numerous financial products and sparse data between user requirements and product attributes in the financial product recommendation can be reduced; further, the server updates the data information of each product in the database by acquiring the purchase information of the product after the user obtains the product recommendation result, and the like. And taking the updated user data and product data in the database as data acquired from the database in the next period service, thereby overcoming the problem of weak timeliness of product recommendation results caused by large change and strong time sequence of financial product information.
An embodiment of the present application further provides a product recommendation device, please refer to fig. 3, where fig. 3 is a schematic block diagram of the product recommendation device provided in the embodiment of the present application, and the device includes: an acquisition unit 201, a feature unit 202, a similarity calculation unit 202, and a processing unit 203, wherein,
an acquisition unit 201, configured to acquire user data of at least one user and product data of a plurality of products from a database;
a feature unit 202, configured to perform feature engineering on the user data of the at least one user and the product data of the multiple products to obtain feature information, where the feature information includes the user feature data of the at least one user and the product feature data of the multiple products;
a similarity calculation unit 203, configured to obtain similarity information of the multiple products based on the product feature data of the multiple products;
a processing unit 204, configured to obtain a product recommendation result of a target user of the at least one user by using a deep learning model based on the feature information, the similarity information of the multiple products, and a current popular product list, where the current popular product list includes at least one product of the multiple products.
The apparatus further comprises a storage unit 204 for storing the user data as well as the product data.
The processing unit 202 is further configured to take a product of the first products, the correlation of which is greater than a preset threshold, as the target recommended product, and push the target recommended product to a user terminal.
In some embodiments, the obtaining unit 201 is further configured to obtain the product requirement of the user;
in a possible implementation manner, the screening unit is specifically configured to:
determining feature information for characterizing product needs of the target user based on the user feature data of the target user, the product needs including at least one of: the type of a product which the target user desires to purchase, the risk acceptance information of the target user, the fund information of the target user and the expected product income information of the target user;
determining feature information for characterizing product attributes of the plurality of products based on product feature data of the plurality of products, the product attributes including at least one of: the product type, the product risk level, the open purchase time of the product, the minimum purchase fund of the product and the income information of the product;
and screening the products according to the characteristic information for representing the product requirements of the target user and the characteristic information for representing the product attributes of the products to obtain a candidate product list of the target user.
In a possible implementation manner, the similarity corresponding to the at least one candidate product includes a similarity between the at least one candidate product and at least one current popular product included in the current popular product list;
the processing unit is specifically configured to:
updating the candidate product list of the target user based on the similarity between the at least one candidate product and at least one current popular product contained in the current popular product list to obtain an updated candidate product list;
and obtaining a product recommendation result of the target user by utilizing a deep learning model based on the user characteristic data of the target user and the product characteristic data of at least one candidate product contained in the updated candidate product list.
In a possible implementation manner, the processing unit is specifically configured to:
determining whether a target candidate product with the similarity between the target candidate product and a target popular product in the current popular product list being greater than a first preset threshold exists in the candidate product list of the target user;
and responding to the target candidate product in the candidate product list of the target user, and adding the target popular product to the candidate product list of the target user to obtain an updated candidate product list.
In some embodiments, the adding the target trending product to the list of candidate products for the target user comprises: replacing the target candidate product in the candidate product list of the target user with the target popular product.
In some embodiments, the processing unit is specifically configured to:
based on the feature information, the similarity information of the products and the current popular product list, obtaining a product recommendation list of a target user by using a deep learning model;
and in response to the first product and the second product with the similarity reaching a second preset threshold exist in the product recommendation list and the first product included in the current popular product list is ranked behind the second product in the product recommendation list, interchanging the positions of the first product and the second product in the product recommendation list to obtain a product recommendation result.
In some embodiments, the user characteristic data of the target user includes user static information and user dynamic information,
the user static information includes at least one of: the user identification, the occupation of the user, the age of the user and the area where the user is located;
the user dynamic information comprises at least one of the following: the method comprises the steps of obtaining information of products purchased by a user, information of the number of the products purchased by the user, the buried point data of the products browsed by the user, the path of the products purchased by the user, the profit demand of the user, the fund state of the user and the risk threshold set by the user.
In some embodiments, the characteristic information further includes interaction characteristics between the target user and the plurality of products.
In some embodiments, the product characteristic data of the plurality of products comprises at least one of: product identification, product category, product open purchase time, product risk level, whether the product is a popular product, revenue information of the product, maximum withdrawal of the product, and minimum purchase funds of the product.
In some embodiments, the interactive features include the number of times the target user purchased different products, one or more ways the target user purchased different products, the number of times the target user purchased different products through a first way, the target user's attention to different products, the time and frequency of the target user purchasing different products, and the web page ranking pagerank of the products relative to the target user.
In some embodiments, the plurality of products includes a plurality of existing products and at least one added product, and the similarity information includes a similarity between each two products in the plurality of products.
It can be understood that the functions of the apparatus provided in the embodiment of the present application or the units included in the apparatus may be used to execute the method described in the above method embodiment, and for specific implementation, reference may be made to the description of the above method embodiment, and for brevity, detailed description is omitted here. It is to be understood that, when the method described in the foregoing method embodiment is executed, since the amount of data that may need to be processed is huge, the method may be executed by one apparatus or may be executed by two or more apparatus clusters, and the embodiment of the present application is not limited in particular.
Referring to fig. 4, fig. 4 is a schematic block diagram of an electronic device according to an embodiment of the present disclosure. The electronic device in the present embodiment as shown in fig. 4 may include: one or more processors 301, a communication interface 302, and a memory 303. The processor 301, the communication interface 302, and the memory 303 are connected by a bus 304. The memory 303 is adapted to store a computer program comprising program instructions, and the processor 301 is adapted to execute the program instructions stored by the memory 303 to perform the functions performed by the processing unit 204 or the method steps described in the above method embodiments.
It should be understood that, in the embodiment of the present Application, the Processor 301 may be a Central Processing Unit (CPU), and the Processor may also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field Programmable Gate Arrays (FPGAs) or other Programmable logic devices, discrete Gate or transistor logic devices, discrete hardware components, and the like. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
In a specific implementation, the processor 301, the communication interface 302, and the memory 303 described in this embodiment of the present application may execute the implementation described in the product recommendation method provided in this embodiment of the present application, and may also execute the implementation of the apparatus described in this embodiment of the present application, which is not described herein again.
In a specific implementation, the electronic Device in the embodiment of the present application is a terminal Device that can be used to execute the method described in the embodiment of the method provided in the present application, and may also be a terminal Device that includes the apparatus described in the present application, and the electronic Device may be various terminal devices such as a Mobile phone, a driving recorder, a tablet computer, a Mobile Internet Device (MID), a server, and the like that can perform image acquisition unit and image processing, which is not limited in the embodiment of the present application.
Those of ordinary skill in the art will appreciate that the elements and method steps of the examples described in connection with the embodiments disclosed herein may be embodied in electronic hardware, computer software, or combinations of both, and that the components and steps of the examples have been described in a functional general in the foregoing description for the purpose of clearly illustrating the interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.
It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described apparatuses, electronic devices and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.
In the several embodiments provided in the present application, it should be understood that the disclosed apparatus, electronic device and method may be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, the division of the units is only one logical division, and other divisions may be realized in practice, for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may also be an electric, mechanical or other form of connection.
In addition, functional units in the embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.
The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present application may be substantially or partially contributed by the prior art, or all or part of the technical solution may be embodied in a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
While the invention has been described with reference to specific embodiments, the scope of the invention is not limited thereto, and those skilled in the art can easily conceive various equivalent modifications or substitutions within the technical scope of the invention. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.
Claims (10)
1. A method for recommending products, the method comprising:
obtaining user data of at least one user and product data of a plurality of products from a database;
performing feature engineering on the user data of the at least one user and the product data of the plurality of products to obtain feature information, wherein the feature information comprises the user feature data of the at least one user and the product feature data of the plurality of products;
obtaining similarity information of the products based on the product characteristic data of the products;
and obtaining a product recommendation result of a target user in the at least one user by using a deep learning model based on the feature information, the similarity information of the plurality of products and a current popular product list, wherein the current popular product list comprises at least one product in the plurality of products.
2. The method of claim 1, wherein obtaining product recommendations of a target user of the at least one user using a deep learning model based on the feature information, the similarity information of the plurality of products, and a current popular product list comprises:
screening the products according to the user characteristic data of the target user and the product characteristic data of the products to obtain a candidate product list of the target user, wherein the candidate product list comprises at least one candidate product in the products;
and obtaining a product recommendation result of the target user by utilizing a deep learning model based on the feature data of the target user, the similarity corresponding to the at least one candidate product contained in the similarity information of the products and the current popular product list.
3. The method of claim 2, wherein the corresponding similarity of the at least one candidate product comprises a similarity between the at least one candidate product and at least one current popular product included in the list of current popular products;
the obtaining, by using a deep learning model, a product recommendation result of the target user based on the feature information, the similarity corresponding to the at least one candidate product included in the similarity information of the plurality of products, and the current popular product list, includes:
updating the candidate product list of the target user based on the similarity between the at least one candidate product and at least one current popular product contained in the current popular product list to obtain an updated candidate product list;
and obtaining a product recommendation result of the target user by utilizing a deep learning model based on the user characteristic data of the target user and the product characteristic data of at least one candidate product contained in the updated candidate product list.
4. The method according to any one of claims 1-3, wherein the obtaining of the product recommendation result of the target user by using a deep learning model based on the feature information, the similarity information of the plurality of products and the current popular product list comprises:
based on the feature information, the similarity information of the products and the current popular product list, obtaining a product recommendation list of a target user by using a deep learning model;
and in response to the first product and the second product with the similarity reaching a second preset threshold exist in the product recommendation list and the first product included in the current popular product list is ranked behind the second product in the product recommendation list, interchanging the positions of the first product and the second product in the product recommendation list to obtain a product recommendation result.
5. A product recommendation device, the device comprising:
an acquisition unit for acquiring user data of at least one user and product data of a plurality of products from a database;
a feature unit, configured to perform feature engineering on the user data of the at least one user and the product data of the multiple products to obtain feature information, where the feature information includes the user feature data of the at least one user and the product feature data of the multiple products;
the similarity calculation unit is used for obtaining similarity information of the products based on the product characteristic data of the products;
and the processing unit is used for obtaining a product recommendation result of a target user in the at least one user by utilizing a deep learning model based on the feature information, the similarity information of the plurality of products and a current popular product list, wherein the current popular product list comprises at least one product in the plurality of products.
6. The apparatus of claim 5, further comprising:
the screening unit is used for screening the products according to the user characteristic data of the target user and the product characteristic data of the products to obtain a candidate product list of the target user, wherein the candidate product list comprises at least one candidate product in the products;
the processing unit is specifically configured to obtain a product recommendation result of the target user by using a deep learning model based on the feature data of the target user, the similarity corresponding to the at least one candidate product included in the similarity information of the plurality of products, and the current popular product list.
7. The apparatus of claim 6, wherein the corresponding similarity of the at least one candidate product comprises a similarity between the at least one candidate product and at least one current popular product included in the list of current popular products;
the processing unit is specifically configured to:
updating the candidate product list of the target user based on the similarity between the at least one candidate product and at least one current popular product contained in the current popular product list to obtain an updated candidate product list;
and obtaining a product recommendation result of the target user by utilizing a deep learning model based on the user characteristic data of the target user and the product characteristic data of at least one candidate product contained in the updated candidate product list.
8. The apparatus according to any one of claims 5 to 7, wherein the processing unit is specifically configured to:
based on the feature information, the similarity information of the products and the current popular product list, obtaining a product recommendation list of a target user by using a deep learning model;
and in response to the first product and the second product with the similarity reaching a second preset threshold exist in the product recommendation list and the first product included in the current popular product list is ranked behind the second product in the product recommendation list, interchanging the positions of the first product and the second product in the product recommendation list to obtain a product recommendation result.
9. An electronic device comprising a processor, a communication interface, and a memory; the memory is configured to store instructions, the processor is configured to execute the instructions, and the communication interface is configured to receive or transmit data; wherein the processor, when executing the instructions, performs the method of any one of claims 1-4.
10. A computer readable storage medium having computer program instructions stored thereon, wherein the computer program instructions, when executed by a processor, implement the method of any of claims 1-4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910461563.3A CN112015970A (en) | 2019-05-29 | 2019-05-29 | Product recommendation method, related equipment and computer storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910461563.3A CN112015970A (en) | 2019-05-29 | 2019-05-29 | Product recommendation method, related equipment and computer storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112015970A true CN112015970A (en) | 2020-12-01 |
Family
ID=73501865
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910461563.3A Pending CN112015970A (en) | 2019-05-29 | 2019-05-29 | Product recommendation method, related equipment and computer storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112015970A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113837843A (en) * | 2021-09-29 | 2021-12-24 | 平安科技（深圳）有限公司 | Product recommendation method, device, medium and electronic equipment |
| CN114202380A (en) * | 2021-12-07 | 2022-03-18 | 中国建设银行股份有限公司 | Recommendation method, device and equipment for financial products |
-
2019
- 2019-05-29 CN CN201910461563.3A patent/CN112015970A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113837843A (en) * | 2021-09-29 | 2021-12-24 | 平安科技（深圳）有限公司 | Product recommendation method, device, medium and electronic equipment |
| CN113837843B (en) * | 2021-09-29 | 2023-11-24 | 平安科技（深圳）有限公司 | Product recommendation method and device, medium and electronic equipment |
| CN114202380A (en) * | 2021-12-07 | 2022-03-18 | 中国建设银行股份有限公司 | Recommendation method, device and equipment for financial products |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20180108029A1 (en) | Detecting differing categorical features when comparing segments | |
| CN108805598B (en) | Similarity information determination method, server and computer-readable storage medium | |
| CN110008973B (en) | Model training method, method and device for determining target user based on model | |
| CN109299356B (en) | Activity recommendation method and device based on big data, electronic equipment and storage medium | |
| US20200234218A1 (en) | Systems and methods for entity performance and risk scoring | |
| CN106062743A (en) | Systems and methods for keyword suggestion | |
| CN113744016B (en) | Object recommendation method and device, equipment and storage medium | |
| CN111061979B (en) | User tag pushing method and device, electronic equipment and medium | |
| US11200593B2 (en) | Predictive recommendation system using tiered feature data | |
| US20220180379A1 (en) | Transaction-based information processing system, method, and article | |
| CN109961351B (en) | Information recommendation method and device, storage medium and computer equipment | |
| CN112365283B (en) | Coupon issuing method and device, terminal equipment and storage medium | |
| US11810151B2 (en) | Predictive recommendation system using tiered feature data | |
| CN110689402A (en) | Method and device for recommending merchants, electronic equipment and readable storage medium | |
| CN112598472A (en) | Product recommendation method, device, system, medium and program product | |
| CN113516496A (en) | Advertisement conversion rate pre-estimation model construction method, device, equipment and medium thereof | |
| CN113744020A (en) | Commodity file processing method and device, electronic equipment and storage medium | |
| CN111861605A (en) | Business object recommendation method | |
| CN112015970A (en) | Product recommendation method, related equipment and computer storage medium | |
| CN111680213A (en) | Information recommendation method, data processing method and device | |
| KR20190056075A (en) | Server and method for service evaluation | |
| CN111767457A (en) | Recommendation method and device | |
| WO2020150597A1 (en) | Systems and methods for entity performance and risk scoring | |
| CN115809889A (en) | Intelligent passenger group screening method, system, medium and equipment based on marketing effect | |
| CN114119078A (en) | Target resource determination method, device, electronic equipment and medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |