CN116228370A - Product recommendation method and device and electronic equipment - Google Patents

Abstract

The application discloses a product recommendation method and device and electronic equipment. Relates to the field of artificial intelligence, and the method comprises the following steps: screening a plurality of target users from a database through a preset credit evaluation model, and acquiring historical product browsing records of each target user; determining the evaluation values of each target user on different products based on the historical product browsing records to obtain an evaluation value list for recording the evaluation values of all the target users on different products; determining a user to be recommended from a plurality of target users, and determining a product to be recommended from different products; recommending the product to be recommended to the user to be recommended based on the first similarity between the product to be recommended and each evaluated product, the second similarity between the target user of each evaluated product to be recommended and the user to be recommended, and the evaluation value list. By the method and the device, the problem that in the related technology, products recommended to the clients to be recommended are not matched with the clients to be recommended, so that the success rate of product recommendation is low is solved.

Description

Product recommendation method and device and electronic equipment

Technical Field

The application relates to the field of artificial intelligence, in particular to a product recommendation method, a product recommendation device and electronic equipment.

Background

In the related art, the path of product recommendation development has evolved from product and ad driven recommendations to data and technology driven recommendations. Along with the continuous enrichment of data dimension and the continuous increase of application scenes, the continuous enrichment of position data and internet of things data caused by data movement is realized, data product recommendation is also rapidly evolving, and the introduction of an intelligent product recommendation era is already started. However, in the related art, when recommending products, a customer manager still searches a customer list everywhere, and through communication with customers to know information, the customers recommend products in a traditional mode of old, new, friend to introduce, company with community street, global information platform to acquire newly registered enterprises, and the like, so that the recommendation success rate is low, and the range of selecting recommended objects is contracted.

Aiming at the problem that the product recommended to the client to be recommended is not matched with the client to be recommended in the related technology, so that the success rate of product recommendation is low, no effective solution is proposed at present.

Disclosure of Invention

The main purpose of the application is to provide a product recommending method, a device and electronic equipment, so as to solve the problem that in the related technology, products recommended to a customer to be recommended are not matched with the customer to be recommended, and therefore the success rate of product recommendation is low.

In order to achieve the above object, according to one aspect of the present application, there is provided a product recommendation method. The method comprises the following steps: screening a plurality of target users from a database through a preset credit evaluation model, and acquiring historical product browsing records of each target user, wherein the preset credit evaluation model is obtained by training asset information of the users and is used for screening target users with credit evaluation values greater than or equal to an evaluation value threshold value from the database, and each historical product browsing record comprises a plurality of products; determining the evaluation values of each target user on different products based on the historical product browsing records to obtain an evaluation value list for recording the evaluation values of all the target users on different products; determining a user to be recommended from a plurality of target users, and determining a product to be recommended from different products; recommending the product to be recommended to the user to be recommended based on the first similarity between the product to be recommended and each evaluated product, the second similarity between the target user of each evaluated product to be recommended and the user to be recommended, and the evaluation value list, wherein the evaluated product is the product evaluated by the user to be recommended.

Optionally, recommending the product to be recommended to the user to be recommended based on the first similarity between the product to be recommended and each evaluated product, the second similarity between the target user who has evaluated the product to be recommended and the user to be recommended, and the evaluation value list includes: judging whether the number of the evaluated products in the evaluation value list is larger than or equal to a number threshold; calculating a first similarity between the product to be recommended and each evaluated product under the condition that the number of the evaluated products is greater than or equal to a number threshold value, and calculating a first recommended evaluation value of the product to be recommended based on the first similarity and the evaluation value list; calculating a second similarity of each target user who evaluates the products to be recommended and the users to be recommended under the condition that the number of the evaluated products is smaller than a number threshold value, and calculating a second recommendation evaluation value of the products to be recommended based on the second similarity and the evaluation value list; recommending the product to be recommended to the user to be recommended based on the first recommendation evaluation value or the second recommendation evaluation value.

Optionally, recommending the product to be recommended to the user to be recommended based on the first recommendation evaluation value or the second recommendation evaluation value includes: judging whether the first recommendation evaluation value or the second recommendation evaluation value is larger than or equal to a recommendation evaluation value threshold value; and recommending the product to be recommended to the user to be recommended under the condition that the first recommendation evaluation value or the second recommendation evaluation value is larger than or equal to the recommendation evaluation value threshold value.

Optionally, determining the evaluation value of each target user for the different products based on the historical product browsing records includes: determining browsing marks of a target user on each product in the historical product browsing records, wherein the browsing marks at least comprise: collection, purchase and long-time browsing; setting the evaluation value of the target user on the target product as a first preset value under the condition that the browsing mark of the target product is collection; setting the evaluation value of the target user on the target product to be a second preset value under the condition that the browsing mark of the target product is purchase; and setting the evaluation value of the target user on the target product to be a third preset value under the condition that the browsing mark of the target product is long-time browsing.

Optionally, calculating the first similarity of the product to be recommended to each of the evaluated products includes: determining other users for evaluating the products to be recommended in the evaluation value list; calculating the average value of the evaluation values of the products to be recommended of all target users to obtain the product evaluation average value of the products to be recommended; calculating the average value of the evaluation values of all the target users on each evaluated product to obtain the product evaluation average value of each evaluated product; and inputting the product evaluation average value of each evaluated product, the product evaluation average value of the product to be recommended, the evaluation value of each other user to be recommended and the evaluation value of each other user to the evaluated product into a Pearson correlation coefficient calculation formula to obtain the first similarity of each evaluated product.

Optionally, calculating the first recommendation rating value for the product to be recommended based on the first similarity and the rating value list comprises: calculating the difference value of the evaluation value of each evaluated product and the product evaluation mean value of the evaluated product of the user to be recommended to obtain a first difference value of each evaluated product; calculating the product of the first difference value of each evaluated product and the first similarity of the evaluated product to obtain a group of first product values, and calculating the sum of all first product values in the group of first product values to obtain a first target sum value; and calculating the sum of the first similarity of all the evaluated products to obtain a second target sum value, calculating a first ratio of the first target sum value to the second target sum value, and determining the sum of the first ratio and the product evaluation mean value of the products to be recommended as a first recommendation evaluation value of the products to be recommended.

Optionally, calculating the second similarity of each of the target users who evaluated the product to be recommended to the user to be recommended includes: determining other users for evaluating the products to be recommended in the evaluation value list; calculating the average value of the evaluation values of all the products of each other user to obtain the user evaluation average value of each other user; calculating the average value of the evaluation values of the users to be recommended on all the products to obtain the user evaluation average value of the users to be recommended; and inputting the user evaluation average value of each other user, the user evaluation average value of the user to be recommended, the evaluation value of the user to be recommended on each evaluated product and the evaluation value of the other user on each evaluated product into a Pearson correlation coefficient calculation formula to obtain the second similarity of each other user.

Optionally, calculating the second recommended evaluation value of the product to be recommended based on the second similarity and the evaluation value list includes: calculating the difference value between the evaluation value of the product to be recommended of each other user and the user evaluation mean value of the other users to obtain a second difference value of each other user; calculating the product of the second difference value of each other user and the second similarity of the other users to obtain a group of second product values, and calculating the sum of all second product values in the group of second product values to obtain a third target sum value; and calculating the sum of the second similarity of all other users to obtain a fourth target sum value, calculating a second ratio of the third target sum value to the fourth target sum value, and determining the sum of the second ratio and the user evaluation mean value of the user to be recommended as a second recommendation evaluation value of the product to be recommended.

In order to achieve the above object, according to another aspect of the present application, there is provided a product recommendation device. The device comprises: the screening unit is used for screening a plurality of target users from the database through a preset credit evaluation model, and acquiring historical product browsing records of each target user, wherein the preset credit evaluation model is obtained by training asset information of the users, and is used for screening target users with credit evaluation values greater than or equal to an evaluation value threshold value from the database, and each historical product browsing record comprises a plurality of products; the first determining unit is used for determining the evaluation values of each target user on different products based on the historical product browsing records, and obtaining an evaluation value list for recording the evaluation values of all the target users on different products; the second determining unit is used for determining a user to be recommended from a plurality of target users and determining a product to be recommended from different products; the recommendation unit is used for recommending the product to be recommended to the user to be recommended based on the first similarity between the product to be recommended and each evaluated product, the second similarity between the target user of each evaluated product to be recommended and the user to be recommended, and the evaluation value list, wherein the evaluated product is the product evaluated by the user to be recommended.

Through the application, the following steps are adopted: screening a plurality of target users from a database through a preset credit evaluation model, and acquiring historical product browsing records of each target user, wherein the preset credit evaluation model is obtained by training asset information of the users and is used for screening target users with credit evaluation values greater than or equal to an evaluation value threshold value from the database, and each historical product browsing record comprises a plurality of products; determining the evaluation values of each target user on different products based on the historical product browsing records to obtain an evaluation value list for recording the evaluation values of all the target users on different products; determining a user to be recommended from a plurality of target users, and determining a product to be recommended from different products; recommending the product to be recommended to the user to be recommended based on the first similarity between the product to be recommended and each evaluated product, the second similarity between the target user of each evaluated product to be recommended and the user to be recommended, and the evaluation value list, wherein the evaluated product is the product evaluated by the user to be recommended, and the problem that the product recommended to the user to be recommended and the user to be recommended are not matched in the related technology, so that the success rate of product recommendation is low is solved. The method comprises the steps of calculating the first similarity between the product to be recommended and the evaluated product, or calculating the second similarity between the user to be recommended and the target user, determining the product evaluation value of the product to be recommended of the user to be recommended based on the first similarity or the second similarity, and determining whether to recommend the product to be recommended to the user to be recommended or not according to the product evaluation value, so that the effect of accurately matching the product recommended to the user to be recommended with the user to be recommended is achieved, and the product recommendation success rate is improved.

Drawings

The accompanying drawings, which are included to provide a further understanding of the application, illustrate and explain the application and are not to be construed as limiting the application. In the drawings:

FIG. 1 is a flow chart of a product recommendation method provided according to an embodiment of the present application;

FIG. 2 is a flow chart of an assessment value determination method provided according to an embodiment of the present application;

FIG. 3 is a schematic diagram of a product recommendation device provided according to an embodiment of the present application;

fig. 4 is a schematic diagram of an electronic device provided according to an embodiment of the present application.

Detailed Description

It should be noted that, in the case of no conflict, the embodiments and features in the embodiments may be combined with each other. The present application will be described in detail below with reference to the accompanying drawings in conjunction with embodiments.

In order to make the present application solution better understood by those skilled in the art, the following description will be made in detail and with reference to the accompanying drawings in the embodiments of the present application, it is apparent that the described embodiments are only some embodiments of the present application, not all embodiments. All other embodiments, which can be made by one of ordinary skill in the art based on the embodiments herein without making any inventive effort, shall fall within the scope of the present application.

It should be noted that the terms "first," "second," and the like in the description and claims of the present application and the above figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate in order to describe the embodiments of the present application described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.

It should be noted that, the user information (including, but not limited to, user equipment information, user personal information, etc.) and the data (including, but not limited to, data for presentation, analyzed data, etc.) related to the present disclosure are information and data authorized by the user or sufficiently authorized by each party.

The invention will be described with reference to preferred implementation steps, and fig. 1 is a flowchart of a product recommendation method according to an embodiment of the present application, as shown in fig. 1, and the method includes the following steps:

Step S101, screening a plurality of target users from a database through a preset credit evaluation model, and acquiring historical product browsing records of each target user, wherein the preset credit evaluation model is obtained through asset information training of the users and is used for screening target users with credit evaluation values greater than or equal to an evaluation value threshold from the database, and each historical product browsing record contains a plurality of products.

Specifically, the preset credit evaluation model may be a model for screening a large number of clients with good quality, acquiring a large number of client data from a database, and confirming the specific credit condition of the client based on the asset information of the client, that is, determining the credit evaluation value of the client, and when the credit evaluation value of the client is greater than or equal to the threshold value of the evaluation value, indicating that the credit condition of the client is good, and recommending the client as a good quality target user. After screening out the target users, the historical product browsing records of each target user are required to be obtained so as to screen out the products to be recommended. And constructing client screening, credit measuring and calculating and risk monitoring models, namely a credit evaluation model, according to different business scenes based on multi-dimensional data such as transactions, assets, credits and the like of clients, namely asset information, and recommending products for clients meeting the requirements by using big data and an internet technology.

For example, the clients in the database may be client lists that are imported through channels such as customs single window, tax authority tax data, unionpay data, etc., interface importation of customs lending application, tax, etc., to enrich the client data population sample. The credit evaluation module can provide financial products for enterprises with good tax payment conditions according to the tax payment conditions of the enterprises, and simultaneously excavate the upstream and downstream of the core enterprises with good tax payment by establishing the tax payment model, so as to screen out target users; or obtaining client information based on a customs single window, and running out the high-quality client, namely the target user, through a credit evaluation model, and providing proper cross-border service for the high-quality client based on a collaborative filtering algorithm.

It should be noted that, when the credit evaluation model is constructed, the property risk situation of the customer needs to be considered, for example, a risk prevention and control model is established for trust of the core enterprise mainly occupied by the digital supply chain service, the operation states of the core enterprise include (credit rating of the core enterprise, dedicated trust use rate of the supply chain, four-color identification of the customer, whether a potential risk customer exists, whether admission monitoring exists, whether post-credit early warning exists, whether overdue exists, whether bad loan appears, etc.) real-time display to the customer manager through a business intelligent report, real-time reminding to the customer manager through a workstation waiting or mail for the customer with set index risk, and product recommendation and wind control capability of the financial institution are improved by combining the early-stage customer intelligent marketing with the intelligent wind control of the customer in the lifetime.

Step S102, based on the historical product browsing records, the evaluation values of each target user on different products are determined, and an evaluation value list for recording the evaluation values of all target users on different products is obtained.

Specifically, the history product browsing records have specific browsing conditions of each target user on the product, such as collection, purchase, long-time browsing and the like, and the interested degree of the user on the product can be represented based on different browsing conditions, so that the evaluation value list of the evaluation values of all target users on different products is obtained based on quantifying the different browsing conditions into evaluation values.

Step S103, determining a user to be recommended from a plurality of target users, and determining products to be recommended from different products.

Specifically, each target user can be used as a user to be recommended, one user is randomly selected from a plurality of target users to be used as the user to be recommended, and the product to be recommended can be a product which is not contained in the product browsing record of the user to be recommended.

Step S104, recommending the product to be recommended to the user to be recommended based on the first similarity between the product to be recommended and each evaluated product, the second similarity between the target user of each evaluated product to be recommended and the user to be recommended, and the evaluation value list, wherein the evaluated product is the product evaluated by the user to be recommended.

Specifically, the evaluated product is a product in a product browsing record of the user to be recommended, the product to be recommended is recommended to the user to be recommended according to a collaborative algorithm principle, and the collaborative algorithm principle refers to finding preference of the user based on mining of historical behavior data of the user and predicting the product possibly liked by the user to be recommended. I.e., the common "guess you like", and "people buying the commodity like" etc.

The collaborative algorithm comprises a collaborative algorithm based on a user side, namely, recommendation is carried out according to a target user which has common preference with a user to be recommended, so that the second similarity between the target user and the user to be recommended is calculated; the collaborative algorithm further comprises a collaborative algorithm based on item measurement, that is, similar items are recommended to the user to be recommended according to the items liked by the user to be recommended, so that the first similarity between the product to be recommended and each evaluated product needs to be calculated. And determining whether the product to be recommended is required to be recommended to the user to be recommended or not according to the first similarity or the second similarity and the evaluation value list.

According to the product recommendation method, a plurality of target users are screened from a database through a preset credit evaluation model, and historical product browsing records of each target user are obtained, wherein the preset credit evaluation model is obtained through asset information training of the users and is used for screening target users with credit evaluation values greater than or equal to an evaluation value threshold from the database, and each historical product browsing record comprises a plurality of products; determining the evaluation values of each target user on different products based on the historical product browsing records to obtain an evaluation value list for recording the evaluation values of all the target users on different products; determining a user to be recommended from a plurality of target users, and determining a product to be recommended from different products; recommending the product to be recommended to the user to be recommended based on the first similarity between the product to be recommended and each evaluated product, the second similarity between the target user of each evaluated product to be recommended and the user to be recommended, and the evaluation value list, wherein the evaluated product is the product evaluated by the user to be recommended, and the problem that the product recommended to the user to be recommended and the user to be recommended are not matched in the related technology, so that the success rate of product recommendation is low is solved. The method comprises the steps of calculating the first similarity between the product to be recommended and the evaluated product, or calculating the second similarity between the user to be recommended and the target user, determining the product evaluation value of the product to be recommended of the user to be recommended based on the first similarity or the second similarity, and determining whether to recommend the product to be recommended to the user to be recommended or not according to the product evaluation value, so that the effect of accurately matching the product recommended to the user to be recommended with the user to be recommended is achieved, and the product recommendation success rate is improved.

Optionally, in the product recommendation method provided in the embodiment of the present application, recommending the product to be recommended to the user to be recommended based on the first similarity between the product to be recommended and each evaluated product, the second similarity between the target user who evaluates the product to be recommended and the user to be recommended, and the evaluation value list include: judging whether the number of the evaluated products in the evaluation value list is larger than or equal to a number threshold; calculating a first similarity between the product to be recommended and each evaluated product under the condition that the number of the evaluated products is greater than or equal to a number threshold value, and calculating a first recommended evaluation value of the product to be recommended based on the first similarity and the evaluation value list; calculating a second similarity of each target user who evaluates the products to be recommended and the users to be recommended under the condition that the number of the evaluated products is smaller than a number threshold value, and calculating a second recommendation evaluation value of the products to be recommended based on the second similarity and the evaluation value list; recommending the product to be recommended to the user to be recommended based on the first recommendation evaluation value or the second recommendation evaluation value.

Specifically, the collaborative algorithm based on the user side is selected by whether the number of the evaluated products is equal to or greater than the number threshold, or the collaborative algorithm based on the item measurement is used for product recommendation, that is, when the number of the evaluated products is enough as a sample, that is, the number of the evaluated products is equal to or greater than the number threshold, the collaborative algorithm based on the item measurement may be selected to calculate the first similarity between the product to be recommended and each of the evaluated products, and when the number of the evaluated products is insufficient as a sample, that is, the number of the evaluated products is less than the number threshold, the collaborative algorithm based on the user measurement may be selected to calculate the second similarity between the target user and the user to be recommended. And calculating a first recommendation evaluation value or a second recommendation evaluation value through the first similarity or the second similarity, and determining whether to recommend the product to be recommended to the user to be recommended or not according to the first similarity or the second similarity.

It should be noted that, since the collaborative algorithm based on the user side is easy to have low overlapping property of the articles purchased by different users, the algorithm cannot find the data sparsity problem of the users with similar preference. In addition, the collaborative algorithm based on the user side needs to maintain a user similarity matrix so as to quickly find out similar users, the memory resource cost of the matrix is very high, and the memory space is increased along with the increase of the number of users. The collaborative filtering algorithm based on the object side can calculate the similarity between different objects in advance under the online condition because the similarity of the objects is relatively fixed, and the result is stored in a table, so that the problem of collaborative algorithm based on the user side can be solved by looking up the table when recommending. However, collaborative filtering algorithms based on item side require enough of the evaluated items as samples.

Therefore, collaborative filtering algorithms based on the user side (recommending you according to people who have a common preference with you) are more suitable for tasks with strong practicability, lean towards social recommendation, and easily push new things, while collaborative filtering algorithms based on the object side (recommending similar objects according to objects that you like) are more suitable for recommending products with fixed interests, and lean towards personalized recommendation. For example, aiming at the tax payment situation of the enterprises, the method provides financing for the enterprises with good tax payment situation, simultaneously, the tax payment model is established to excavate the upstream and downstream of the core enterprises with good tax payment, and the method is suitable for adopting a coordination filtering algorithm based on users for providing financing for the upstream and downstream; the customer information is acquired based on a customs single window, and the customer information is run out of the high-quality customer through the model, so that the method is more suitable for providing proper cross-border service for the high-quality customer by adopting an article-based collaborative filtering algorithm.

Optionally, in the product recommendation method provided in the embodiment of the present application, recommending the product to be recommended to the user to be recommended based on the first recommendation evaluation value or the second recommendation evaluation value includes: judging whether the first recommendation evaluation value or the second recommendation evaluation value is larger than or equal to a recommendation evaluation value threshold value; and recommending the product to be recommended to the user to be recommended under the condition that the first recommendation evaluation value or the second recommendation evaluation value is larger than or equal to the recommendation evaluation value threshold value.

Specifically, when the first recommendation evaluation value or the second recommendation evaluation value is greater than or equal to the evaluation value threshold, the degree of interest of the user to be recommended in the product to be recommended may be higher, the product to be recommended may be recommended to the user to be recommended, and when the first recommendation evaluation value or the second recommendation evaluation value is less than the evaluation value threshold, the user to be recommended is not interested in the product to be recommended, and the product to be recommended is refused to be recommended to the user to be recommended. And determining whether the to-be-recommended product is interested in the to-be-recommended product by setting an evaluation value threshold value, and further determining whether to recommend the to-be-recommended product to the to-be-recommended user.

Optionally, in the product recommendation method provided in the embodiment of the present application, fig. 2 is a flowchart of the method for determining an evaluation value provided in the embodiment of the present application, and as shown in fig. 2, determining, based on a historical product browsing record, an evaluation value of each target user for different products includes: step S201, determining a browsing mark of the target user for each product in the historical product browsing record, where the browsing mark at least includes: collection, purchase and long-time browsing; step S202, setting an evaluation value of a target user on a target product as a first preset value under the condition that the browsing mark of the target product is collection; step S203, setting the evaluation value of the target user on the target product as a second preset value under the condition that the browsing mark of the target product is purchase; in step S204, in the case that the browsing flag of the target product is long-time browsing, the evaluation value of the target product by the target user is set to the third preset value.

Specifically, feature vectors preferred by the target user are extracted and quantified by numbers, and quantified features can be counted according to behavior statistics of the target user, for example, the target user purchases a certain article, then the evaluation value of the target user on the purchased article can be quantified to be a first preset value (for example, 5 points), the evaluation value of the target user on the collected article can be quantified to be a second preset value (for example, 4 points), the evaluation value of the target user on the article browsed by the target user for a long time can be quantified to be a third preset value (for example, 3 points), other browsing conditions can be quantified to be a fourth preset value (for example, 2 points), for example, table 1 is an evaluation value list of evaluation values of the target user on different products, and the evaluation value is not filled to represent that the target user evaluates the products. And the evaluation value of the target user on the product is quantified through the product browsing condition, so that the behavior of each user on the article is characterized as numbers, and the calculation is convenient.

TABLE 1

Evaluation value	Product 1	Product 2	Product 3	Product 4	Average of
						Target user A	4	5	3	5	4
Target user B		5	4		4.5
						Target user C	5	4	2		3.67
Target user D	2	4		3	3
						Target user E	3	4	5		4
Average of	3.5	4.2	3.5	4

Optionally, in the product recommendation method provided in the embodiment of the present application, calculating the first similarity between the product to be recommended and each evaluated product includes: determining other users for evaluating the products to be recommended in the evaluation value list; calculating the average value of the evaluation values of the products to be recommended of all target users to obtain the product evaluation average value of the products to be recommended; calculating the average value of the evaluation values of all the target users on each evaluated product to obtain the product evaluation average value of each evaluated product; and inputting the product evaluation average value of each evaluated product, the product evaluation average value of the product to be recommended, the evaluation value of each other user to be recommended and the evaluation value of each other user to the evaluated product into a Pearson correlation coefficient calculation formula to obtain the first similarity of each evaluated product.

Specifically, different similarity calculation methods are required to be selected according to different data characteristics, and common similarity calculation methods include cosine similarity, pearson correlation coefficient, jacquard similarity coefficient, euclidean distance, manhattan distance and the like. For example, the first similarity may be calculated by the embodiments of the present application as pearson correlation coefficients. The calculation formula of the pearson correlation coefficient is as follows:

where u, v are two sets of elements that require similarity calculation, i is an element in the intersection of u, v, r _u，i Is the ith element in the u set, r _v，i Is the ith element in the v set, r _u Is the average value of the elements in the u-set,

is the average of the elements in the v set, s (u, v) is the pearson correlation coefficient of u, v. The closer the pearson correlation coefficient is to 1, u, v, the more similar the pearson correlation coefficient is to-1, u, vDissimilar.

For example, as shown in table 1, the product to be recommended is the product 4, the user to be recommended is the target user C, the other users are the target user a and the target user D, the evaluated product is the evaluation value of the product 1 and the product 2,u to be recommended may be the evaluation value of each other user to be recommended, and v may be the evaluation value of each other user to the evaluated product, and then the first similarity of the product 1 and the product to be recommended 4 is:

The first similarity between the product 2 and the product 4 to be recommended is:

optionally, in the product recommendation method provided in the embodiment of the present application, calculating the first recommendation evaluation value of the product to be recommended based on the first similarity and the evaluation value list includes: calculating the difference value of the evaluation value of each evaluated product and the product evaluation mean value of the evaluated product of the user to be recommended to obtain a first difference value of each evaluated product; calculating the product of the first difference value of each evaluated product and the first similarity of the evaluated product to obtain a group of first product values, and calculating the sum of all first product values in the group of first product values to obtain a first target sum value; and calculating the sum of the first similarity of all the evaluated products to obtain a second target sum value, calculating a first ratio of the first target sum value to the second target sum value, and determining the sum of the first ratio and the product evaluation mean value of the products to be recommended as a first recommendation evaluation value of the products to be recommended.

Specifically, the first recommendation evaluation value may be derived from the following calculation formula:

wherein P is _u,i Is the first recommended evaluation value of the recommendation,

is the product evaluation mean value of the product to be recommended, +. >

Is the product evaluation mean value of the evaluated product, r _v，i The evaluation value of each evaluated product of the user to be recommended is the first recommendation evaluation value of recommending the product to be recommended 4 to the user to be recommended C, and the first recommendation evaluation value is as follows:

P _u,i the value of 4.66 is greater than 4.5, and the recommendation of the product to be recommended 4 to the user to be recommended C may be decided based on the first recommendation evaluation value.

Optionally, in the product recommendation method provided in the embodiment of the present application, calculating the second similarity between the target user who evaluates the product to be recommended and the user to be recommended includes: determining other users for evaluating the products to be recommended in the evaluation value list; calculating the average value of the evaluation values of all the products of each other user to obtain the user evaluation average value of each other user; calculating the average value of the evaluation values of the users to be recommended on all the products to obtain the user evaluation average value of the users to be recommended; and inputting the user evaluation average value of each other user, the user evaluation average value of the user to be recommended, the evaluation value of the user to be recommended on each evaluated product and the evaluation value of the other user on each evaluated product into a Pearson correlation coefficient calculation formula to obtain the second similarity of each other user.

For example, as shown in table 1, the product to be recommended is product 4, the user to be recommended is target user C, the other users are target user a and target user D, the evaluated products are product 1 and product 2,u may be evaluation values of the user to be recommended for each evaluated product, and v may be evaluation values of the other users for each evaluated product, and the second similarity between the user to be recommended C and the target user a is:

The second similarity between the user C to be recommended and the target user D is as follows:

/>

optionally, in the product recommendation method provided in the embodiment of the present application, calculating the second recommendation evaluation value of the product to be recommended based on the second similarity and the evaluation value list includes: calculating the difference value between the evaluation value of the product to be recommended of each other user and the user evaluation mean value of the other users to obtain a second difference value of each other user; calculating the product of the second difference value of each other user and the second similarity of the other users to obtain a group of second product values, and calculating the sum of all second product values in the group of second product values to obtain a third target sum value; and calculating the sum of the second similarity of all other users to obtain a fourth target sum value, calculating a second ratio of the third target sum value to the fourth target sum value, and determining the sum of the second ratio and the user evaluation mean value of the user to be recommended as a second recommendation evaluation value of the product to be recommended.

Specifically, the second recommendation evaluation value may be derived from the following calculation formula:

wherein P is _u,i Is the second recommended evaluation value of the recommendation,

is the user evaluation mean value of the user to be recommended, +.>

Is the user evaluation mean value of other users, r _v，i Is that every other user treats the recommended product The evaluation value of the product, the second recommendation evaluation value of the product to be recommended 4 recommended to the user to be recommended C is:

the second recommendation evaluation value 4.270 is larger than 4.167, and based on the second recommendation evaluation value, it can be decided to recommend the product to be recommended 4 to the user to be recommended C.

It should be noted that the steps illustrated in the flowcharts of the figures may be performed in a computer system such as a set of computer executable instructions, and that although a logical order is illustrated in the flowcharts, in some cases the steps illustrated or described may be performed in an order other than that illustrated herein.

The embodiment of the application also provides a product recommendation device, and the product recommendation device can be used for executing the product recommendation method provided by the embodiment of the application. The following describes a product recommendation device provided in the embodiment of the present application.

Fig. 3 is a schematic diagram of a product recommendation device provided according to an embodiment of the present application. As shown in fig. 3, the apparatus includes:

a screening unit 10, configured to screen a plurality of target users from a database through a preset credit evaluation model, and obtain a historical product browsing record of each target user, where the preset credit evaluation model is obtained by training asset information of the user, and is used to screen target users with credit evaluation values greater than or equal to an evaluation value threshold from the database, and each historical product browsing record includes a plurality of products;

A first determining unit 20, configured to determine an evaluation value of each target user for different products based on the historical product browsing records, and obtain an evaluation value list for recording evaluation values of all target users for different products;

a second determining unit 30 for determining a user to be recommended from a plurality of target users, and determining a product to be recommended from different products;

the recommending unit 40 is configured to recommend the product to be recommended to the user to be recommended based on the first similarity between the product to be recommended and each evaluated product, the second similarity between the target user of each evaluated product to be recommended and the user to be recommended, and the evaluation value list, where the evaluated product is the product evaluated by the user to be recommended.

According to the product recommendation device provided by the embodiment of the application, a plurality of target users are screened from a database through a screening unit 10 through a preset credit evaluation model, and historical product browsing records of each target user are obtained, wherein the preset credit evaluation model is obtained through asset information training of the users and is used for screening target users with credit evaluation values greater than or equal to an evaluation value threshold value from the database, and each historical product browsing record comprises a plurality of products; a first determining unit 20 that determines an evaluation value of each target user for different products based on the history product browsing records, and obtains an evaluation value list that records evaluation values of all target users for different products; a second determining unit 30 that determines a user to be recommended from among a plurality of target users, and determines a product to be recommended from among different products; the recommending unit 40 recommends the product to be recommended to the user to be recommended based on the first similarity between the product to be recommended and each evaluated product, the second similarity between the target user of each evaluated product to be recommended and the user to be recommended, and the evaluation value list, wherein the evaluated product is the product evaluated by the user to be recommended, the problem that the product recommendation success rate is low due to mismatching between the product recommended to the user to be recommended and the user to be recommended in the related art is solved, the product recommendation success rate is improved by calculating the first similarity between the product to be recommended and the evaluated product or calculating the second similarity between the user to be recommended and the target user, and determining the product evaluation value of the product to be recommended by the user to be recommended based on the first similarity or the second similarity, and determining whether the product to be recommended to the user to be recommended is recommended to the user to be recommended by the product evaluation value, so that the product recommended to the user to be recommended is accurately matched with the user to be recommended is achieved, and the effect of improving the product recommendation success rate is achieved.

Optionally, in the product recommendation device provided in the embodiment of the present application, the recommendation unit 40 includes: the judging module is used for judging whether the number of the evaluated products in the evaluation value list is larger than or equal to a number threshold value; the first calculation module is used for calculating the first similarity between the product to be recommended and each evaluated product under the condition that the number of the evaluated products is greater than or equal to a number threshold value, and calculating a first recommended evaluation value of the product to be recommended based on the first similarity and an evaluation value list; the second calculation module is used for calculating second similarity between the target user of each estimated product to be recommended and the user to be recommended under the condition that the number of the estimated products is smaller than a number threshold value, and calculating a second recommendation estimated value of the product to be recommended based on the second similarity and an estimated value list; and the recommendation module is used for recommending the product to be recommended to the user to be recommended based on the first recommendation evaluation value or the second recommendation evaluation value.

Optionally, in the product recommendation device provided in the embodiment of the present application, the recommendation module includes: the judging sub-module is used for judging whether the first recommendation evaluation value or the second recommendation evaluation value is larger than or equal to a recommendation evaluation value threshold value; and the pushing sub-module is used for recommending the product to be recommended to the user to be recommended under the condition that the first recommendation evaluation value or the second recommendation evaluation value is larger than or equal to the recommendation evaluation value threshold value.

Optionally, in the product recommendation device provided in the embodiment of the present application, the first determining unit 20 includes: the determining module is used for determining browsing marks of the target user on each product in the historical product browsing records, wherein the browsing marks at least comprise: collection, purchase and long-time browsing; the first setting module is used for setting the evaluation value of the target product by the target user to be a first preset value under the condition that the browsing mark of the target product is collection; the second setting module is used for setting the evaluation value of the target product by the target user to a second preset value under the condition that the browsing mark of the target product is purchase; and the third setting module is used for setting the evaluation value of the target user on the target product to a third preset value under the condition that the browsing mark of the target product is long-time browsing.

Optionally, in the product recommendation device provided in the embodiment of the present application, the first calculation module includes: the first determining submodule is used for determining other users for evaluating the products to be recommended in the evaluation value list; the first computing sub-module is used for computing the average value of the evaluation values of the products to be recommended of all target users to obtain the product evaluation average value of the products to be recommended; the second computing sub-module is used for computing the average value of the evaluation values of all the target users on each evaluated product to obtain the product evaluation average value of each evaluated product; the first input sub-module is used for inputting the product evaluation average value of each evaluated product, the product evaluation average value of the product to be recommended, the evaluation value of each other user to be recommended and the evaluation value of each other user to the evaluated product into the pearson correlation coefficient calculation formula to obtain the first similarity of each evaluated product.

Optionally, in the product recommendation device provided in the embodiment of the present application, the first calculation module includes: the third calculation sub-module is used for calculating the difference value between the evaluation value of each evaluated product and the product evaluation mean value of the evaluated product by the user to be recommended to obtain a first difference value of each evaluated product; a fourth calculation sub-module, configured to calculate a product of the first difference value of each evaluated product and the first similarity of the evaluated product, to obtain a set of first product values, and calculate a sum of all first product values in the set of first product values, to obtain a first target sum value; and the fifth calculation sub-module is used for calculating the sum of the first similarity of all the evaluated products to obtain a second target sum value, calculating a first ratio of the first target sum value to the second target sum value, and determining the sum of the first ratio and the product evaluation mean value of the products to be recommended as a first recommendation evaluation value of the products to be recommended.

Optionally, in the product recommendation device provided in the embodiment of the present application, the second calculation module includes: the second determining submodule is used for determining other users for evaluating the products to be recommended in the evaluation value list; a sixth calculation sub-module, configured to calculate a mean value of the evaluation values of all the products by each other user, so as to obtain a user evaluation mean value of each other user; a seventh calculation sub-module, configured to calculate a mean value of the evaluation values of the to-be-recommended user on all the products, so as to obtain a user evaluation mean value of the to-be-recommended user; and the second input sub-module is used for inputting the user evaluation average value of each other user, the user evaluation average value of the user to be recommended, the evaluation value of the user to be recommended on each evaluated product and the evaluation value of the other user on each evaluated product into the pearson correlation coefficient calculation formula to obtain the second similarity of each other user.

Optionally, in the product recommendation device provided in the embodiment of the present application, the second calculation module includes: an eighth calculation sub-module, configured to calculate a difference between an evaluation value of the product to be recommended by each other user and a user evaluation mean value of the other users, to obtain a second difference of each other user; a ninth calculation sub-module, configured to calculate a product of a second difference value of each other user and a second similarity of the other users to obtain a set of second product values, and calculate a sum of all second product values in the set of second product values to obtain a third target sum value; and a tenth calculation sub-module, configured to calculate a sum of second similarities of all other users to obtain a fourth target sum value, calculate a second ratio of the third target sum value to the fourth target sum value, and determine a sum of the second ratio and a user evaluation mean of the user to be recommended as a second recommendation evaluation value of the product to be recommended.

The product recommendation device comprises a processor and a memory, wherein the screening unit 10, the first determining unit 20, the second determining unit 30, the recommendation unit 40 and the like are stored in the memory as program units, and the processor executes the program units stored in the memory to realize corresponding functions.

The processor includes a kernel, and the kernel fetches the corresponding program unit from the memory. The kernel can be provided with one or more than one, and products recommended to the clients to be recommended are accurately matched with the clients to be recommended by adjusting kernel parameters, so that the product recommendation success rate is improved.

The memory may include volatile memory, random Access Memory (RAM), and/or nonvolatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM), among other forms in computer readable media, the memory including at least one memory chip.

The embodiment of the invention provides a computer readable storage medium, on which a program is stored, which when executed by a processor, implements a product recommendation method.

The embodiment of the invention provides a processor, which is used for running a program, wherein the program runs to execute a product recommendation method.

Fig. 4 is a schematic diagram of an electronic device provided according to an embodiment of the present application. As shown in fig. 4, the device 401 includes a processor, a memory, and a program stored on the memory and executable on the processor, and the processor implements the following steps when executing the program: product recommendation methods. The device herein may be a server, PC, PAD, cell phone, etc.

The present application also provides a computer program product adapted to perform, when executed on a data processing device, a program initialized with the method steps of: product recommendation methods.

It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

In one typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.

The memory may include volatile memory in a computer-readable medium, random Access Memory (RAM) and/or nonvolatile memory, etc., such as Read Only Memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.

Computer readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of storage media for a computer include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium, which can be used to store information that can be accessed by a computing device. Computer-readable media, as defined herein, does not include transitory computer-readable media (transmission media), such as modulated data signals and carrier waves.

It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article or apparatus that comprises an element.

It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

The foregoing is merely exemplary of the present application and is not intended to limit the present application. Various modifications and changes may be made to the present application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc. which are within the spirit and principles of the present application are intended to be included within the scope of the claims of the present application.

Claims (10)

1. A method of product recommendation, comprising:

screening a plurality of target users from a database through a preset credit evaluation model, and acquiring historical product browsing records of each target user, wherein the preset credit evaluation model is obtained by training asset information of the users and is used for screening the target users with credit evaluation values greater than or equal to an evaluation value threshold from the database, and each historical product browsing record comprises a plurality of products;

Determining the evaluation values of each target user on different products based on the historical product browsing records to obtain an evaluation value list for recording the evaluation values of all target users on different products;

determining a user to be recommended from the target users, and determining a product to be recommended from the different products;

recommending the product to be recommended to the user to be recommended based on the first similarity of the product to be recommended and each evaluated product, the second similarity of the target user of each evaluated product to be recommended and the user to be recommended, and the evaluation value list, wherein the evaluated product is the product evaluated by the user to be recommended.

2. The method of claim 1, wherein recommending the product to be recommended to the user to be recommended based on the first similarity of the product to be recommended to each evaluated product, the second similarity of each evaluated target user of the product to be recommended to the user to be recommended, and the evaluation value list comprises:

judging whether the number of the evaluated products in the evaluation value list is larger than or equal to a number threshold value;

calculating a first similarity between the product to be recommended and each of the evaluated products when the number of the evaluated products is greater than or equal to the number threshold, and calculating a first recommendation evaluation value of the product to be recommended based on the first similarity and the evaluation value list;

Calculating a second similarity of each target user who evaluates the products to be recommended and the users to be recommended under the condition that the number of the evaluated products is smaller than the number threshold value, and calculating a second recommendation evaluation value of the products to be recommended based on the second similarity and the evaluation value list;

recommending the product to be recommended to the user to be recommended based on the first recommendation evaluation value or the second recommendation evaluation value.

3. The method of claim 2, wherein recommending the product to be recommended to the user to be recommended based on the first recommendation-assessment value or the second recommendation-assessment value comprises:

judging whether the first recommendation evaluation value or the second recommendation evaluation value is larger than or equal to a recommendation evaluation value threshold;

and recommending the product to be recommended to the user to be recommended under the condition that the first recommendation evaluation value or the second recommendation evaluation value is larger than or equal to the recommendation evaluation value threshold value.

4. The method of claim 1, wherein determining an evaluation value for each target user for a different product based on the historical product review record comprises:

determining browsing marks of the target user on each product in the historical product browsing records, wherein the browsing marks at least comprise: collection, purchase and long-time browsing;

Setting an evaluation value of the target user on the target product as a first preset value under the condition that the browsing mark of the target product is the collection;

setting an evaluation value of the target user on the target product to a second preset value under the condition that the browsing mark of the target product is the purchase;

and setting the evaluation value of the target user on the target product to be a third preset value under the condition that the browsing mark of the target product is the long-time browsing.

5. The method of claim 2, wherein calculating a first similarity of the product to be recommended to each of the evaluated products comprises:

determining other users evaluating the products to be recommended in the evaluation value list;

calculating the average value of the evaluation values of all target users on the products to be recommended to obtain the product evaluation average value of the products to be recommended;

calculating the average value of the evaluation values of all the target users on each evaluated product to obtain the product evaluation average value of each evaluated product;

and inputting a product evaluation average value of each evaluated product, the product evaluation average value of the product to be recommended, the evaluation value of each other user on the product to be recommended and the evaluation value of each other user on the evaluated product into a pearson correlation coefficient calculation formula to obtain a first similarity of each evaluated product.

6. The method of claim 2, wherein calculating a first recommendation-assessment value for the product-to-be-recommended based on the first similarity and the list of assessment values comprises:

calculating the difference value of the evaluation value of each evaluated product of the user to be recommended and the product evaluation mean value of the evaluated product to obtain a first difference value of each evaluated product;

calculating the product of the first difference value of each evaluated product and the first similarity of the evaluated product to obtain a group of first product values, and calculating the sum of all first product values in the group of first product values to obtain a first target sum value;

and calculating the sum of the first similarity of all the evaluated products to obtain a second target sum value, calculating a first ratio of the first target sum value to the second target sum value, and determining the sum of the first ratio and the product evaluation average value of the products to be recommended as a first recommendation evaluation value of the products to be recommended.

7. The method of claim 2, wherein calculating a second similarity of each of the target users who evaluated the product to be recommended to the user to be recommended comprises:

determining other users evaluating the products to be recommended in the evaluation value list;

Calculating the average value of the evaluation values of all the products of each other user to obtain the user evaluation average value of each other user;

calculating the average value of the evaluation values of the users to be recommended on all products to obtain the user evaluation average value of the users to be recommended;

and inputting the user evaluation average value of each other user, the user evaluation average value of the user to be recommended, the evaluation value of the user to be recommended for each evaluated product and the evaluation value of the other user for each evaluated product into a pearson correlation coefficient calculation formula to obtain the second similarity of each other user.

8. The method of claim 2, wherein calculating a second recommendation-assessment value for the product-to-be-recommended based on the second similarity and the list of assessment values comprises:

calculating the difference value of the evaluation value of each other user on the product to be recommended and the user evaluation mean value of the other users to obtain a second difference value of each other user;

calculating the product of the second difference value of each other user and the second similarity of the other users to obtain a group of second product values, and calculating the sum of all second product values in the group of second product values to obtain a third target sum value;

And calculating the sum of second similarity of all other users to obtain a fourth target sum value, calculating a second ratio of the third target sum value to the fourth target sum value, and determining the sum of the second ratio and a user evaluation mean value of the user to be recommended as a second recommendation evaluation value of the product to be recommended.

9. A product recommendation device, comprising:

the screening unit is used for screening a plurality of target users from a database through a preset credit evaluation model, and acquiring historical product browsing records of each target user, wherein the preset credit evaluation model is obtained by training asset information of the users, and is used for screening the target users with credit evaluation values greater than or equal to an evaluation value threshold from the database, and each historical product browsing record comprises a plurality of products;

the first determining unit is used for determining the evaluation values of each target user on different products based on the historical product browsing records to obtain an evaluation value list for recording the evaluation values of all the target users on different products;

the second determining unit is used for determining a user to be recommended from the target users and determining a product to be recommended from the different products;

And the recommending unit is used for recommending the product to be recommended to the user to be recommended based on the first similarity of the product to be recommended and each evaluated product, the second similarity of each target user evaluated by the product to be recommended and the user to be recommended, and the evaluation value list, wherein the evaluated product is the product evaluated by the user to be recommended.

10. An electronic device comprising one or more processors and a memory for storing one or more programs, wherein the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the product recommendation method of any of claims 1-8.

Images