CN114022246B

CN114022246B - Product information pushing method and device, terminal equipment and storage medium

Info

Publication number: CN114022246B
Application number: CN202111308093.0A
Authority: CN
Inventors: 张莉; 张茜; 余雯; 姜敏华
Original assignee: Ping An Technology Shenzhen Co Ltd
Current assignee: Ping An Technology Shenzhen Co Ltd
Priority date: 2021-11-05
Filing date: 2021-11-05
Publication date: 2023-08-11
Anticipated expiration: 2041-11-05
Also published as: CN114022246A

Abstract

The application is suitable for the technical field of data processing, and provides a pushing method and device of product information, terminal equipment and a storage medium. The method comprises the following steps: determining an original product associated with a target customer and a new product associated with the original product according to a pre-constructed product relationship map; calculating the probability of purchasing the original product by the target client according to the product relation graph; determining attribute similarity of the new product and the original product according to the product relation map; according to the probability that the target client purchases the original product and the attribute similarity, calculating to obtain the probability that the target client purchases the new product; if the probability of the target client purchasing the new product is greater than a set threshold, pushing information of the new product to the target client. By adopting the method, the probability of purchasing a new product by the client can be predicted, so that the relevant information of the new product can be accurately pushed to the client.

Description

Product information pushing method and device, terminal equipment and storage medium

Technical Field

The application relates to the technical field of data processing, and provides a pushing method and device of product information, terminal equipment and a storage medium.

Background

At present, a plurality of product transaction platforms integrate corresponding product pushing algorithms, and the basic thought of the algorithms is to predict target products possibly purchased by a customer based on user portrait information and purchase history of the products by adopting a designed algorithm model so as to push related information of the target products for the customer. However, the new product on-line on the platform has no corresponding purchase history, and the existing algorithm model is difficult to predict the probability of the new product being purchased by the client, so that the relevant information of the new product cannot be accurately pushed to the client.

Disclosure of Invention

In view of the above, the present application provides a method, apparatus, terminal device and storage medium for pushing product information, which can predict the probability of purchasing a new product by a customer, so as to accurately push relevant information of the new product for the customer.

In a first aspect, an embodiment of the present application provides a method for pushing product information, including:

determining an original product associated with a target customer and a new product associated with the original product according to a pre-constructed product relationship graph, wherein the product relationship graph records the interrelationship between the customer and the customer, between the customer and the product and between the product and the product;

Calculating the probability of purchasing the original product by the target client according to the product relation graph;

determining attribute similarity of the new product and the original product according to the product relation map;

according to the probability that the target client purchases the original product and the attribute similarity, calculating to obtain the probability that the target client purchases the new product;

if the probability of the target client purchasing the new product is greater than a set threshold, pushing information of the new product to the target client.

The embodiment of the application constructs a relation graph for recording the interrelationship among customers, the products and the products in advance, and can calculate the probability of purchasing the original products and the attribute similarity between the new products and the original products through the relation graph; then, the probability of purchasing a new product by the client can be calculated according to the probability of purchasing the original product by the client and the attribute similarity, so that the probability of purchasing the new product by the client is predicted; finally, the information of the new product is pushed to the client who has a probability of purchasing the new product greater than the set threshold value, thereby accurately pushing the relevant information of the new product to the client.

In one implementation manner of the embodiment of the present application, the product relationship map may be constructed by the following steps:

acquiring customer attribute data, product purchase data and product attribute data, wherein the customer attribute data comprises a customer name and a customer attribute, the product purchase data comprises a record of a product purchased by a customer, and the product attribute data comprises a product name and a product attribute;

constructing a first sub-relationship graph according to the client attribute data, wherein nodes of the first sub-relationship graph are client names and client attributes, and edges of the first sub-relationship graph represent similar relationships of the client attributes;

constructing a second sub-relationship graph according to the product purchase data, wherein nodes of the second sub-relationship graph are customer names and product names, and edges of the second sub-relationship graph represent product purchase relationships;

constructing a third sub-relationship graph according to the product attribute data, wherein nodes of the third sub-relationship graph are product names and product attributes, and edges of the third sub-relationship graph represent product attribute similarity relationships, product attribute complementary relationships and product attribute upgrading relationships;

and the first sub-relationship map, the second sub-relationship map and the third sub-relationship map are mutually connected according to corresponding nodes to obtain the product relationship map, wherein the nodes of the product relationship map are a customer name, a product name, a customer attribute and a product attribute, and the edges of the product relationship map represent a customer attribute similarity relationship, a product purchase relationship, a product attribute similarity relationship, a product attribute complementary relationship and a product attribute upgrading relationship.

Further, the calculating the probability that the target customer purchases the original product according to the product relationship graph may include:

calculating product attribute conditional probabilities of the original products and customer attribute conditional probabilities of the original products according to the product relation graph, wherein the product attribute conditional probabilities of the original products represent the probability that the original products are purchased by the target customers based on the conditions of all product attributes, and the customer attribute conditional probabilities of the original products represent the probability that the original products are purchased by the target customers based on the conditions of all customer attributes;

the calculating the product attribute conditional probability of the original product and the customer attribute conditional probability of the original product according to the product relation graph may include:

counting the number of clients of the product with the product attribute and the total number of clients of the product relationship graph aiming at each product attribute of the original product, and dividing the number of clients of the product with the product attribute by the total number of clients to obtain product attribute conditional probability corresponding to the product attribute;

And counting the number of clients which have the client attribute and have purchased the original product in the product relationship graph and the total number of clients which have the product relationship graph aiming at each client attribute of the target client, and dividing the number of clients which have the client attribute and have purchased the original product by the total number of clients to obtain the client attribute conditional probability corresponding to the client attribute.

In an implementation manner of the embodiment of the present application, the calculating, according to the probability that the target client purchases the original product and the attribute similarity, the probability that the target client purchases the new product may include:

calculating to obtain the product attribute conditional probability of the new product according to the product attribute conditional probability of the original product and the attribute similarity, wherein the product attribute conditional probability of the new product represents the probability that the new product is purchased by the target client based on the conditions of all product attributes;

calculating to obtain the client attribute conditional probability of the new product according to the client attribute conditional probability of the original product and the attribute similarity, wherein the client attribute conditional probability of the new product represents the probability that the new product is purchased by the target client based on the conditions of all client attributes;

And calculating the probability of the target client purchasing the new product according to the product attribute conditional probability of the new product and the client attribute conditional probability of the new product.

Further, the number of the original products is plural, and the calculating to obtain the product attribute conditional probability of the new product according to the product attribute conditional probability of the original products and the attribute similarity may include:

respectively carrying out normalization processing on the attribute similarity corresponding to each original product to obtain the conditional probability weight of each original product;

determining a first original product of the plurality of original products for each product attribute of the new product, the first original product being an original product connected with the new product by the product attribute in the product relationship graph;

performing weighted summation operation on the product attribute conditional probability of the product attribute of the first original product according to the conditional probability weight of the first original product to obtain the product attribute conditional probability of the product attribute of the new product;

the calculating the client attribute conditional probability of the new product according to the client attribute conditional probability of the original product and the attribute similarity may include:

Determining, for each customer attribute of the target customer, a second original product of the plurality of original products, the second original product being an original product connected to the new product in the product relationship graph by the customer attribute;

and carrying out weighted summation operation on the client attribute conditional probability of the client attribute of the second original product according to the conditional probability weight of the second original product to obtain the client attribute conditional probability of the client attribute of the new product.

Further, after normalization processing is performed on the attribute similarity corresponding to each original product, to obtain a conditional probability weight of each original product, the method may further include:

for each original product, comment information of the target client on the original product is obtained;

carrying out semantic recognition on the comment information to obtain a semantic recognition result;

and adjusting the conditional probability weight of the original product according to the semantic recognition result.

In an implementation manner of the embodiment of the present application, the determining, according to the product relationship graph, the attribute similarity between the new product and the original product may include:

Counting types and quantity of attribute connection relations between the new product and the original product in the product relation map, wherein the types of the attribute connection relations comprise product attribute similarity relations, product attribute complementary relations and product attribute upgrading relations;

and determining the attribute similarity of the new product and the original product according to the type and the number of the attribute connection relations.

In a second aspect, an embodiment of the present application provides a pushing device for product information, including:

the product determining module is used for determining an original product associated with a target customer and a new product associated with the original product according to a pre-constructed product relationship graph, wherein the product relationship graph records the interrelationship among the customers, the customers and the products;

the original product purchase probability calculation module is used for calculating the probability of purchasing the original product by the target client according to the product relationship graph;

the attribute similarity calculation module is used for determining the attribute similarity of the new product and the original product according to the product relation graph;

the new product purchase probability calculation module is used for calculating the probability of the target client purchasing the new product according to the probability of the target client purchasing the original product and the attribute similarity;

And the product information pushing module is used for pushing the information of the new product to the target client if the probability of purchasing the new product by the target client is larger than a set threshold value.

In a third aspect, an embodiment of the present application provides a terminal device, including a memory, a processor, and a computer program stored in the memory and capable of running on the processor, where the processor implements a method for pushing product information according to the first aspect of the embodiment of the present application when the processor executes the computer program.

In a fourth aspect, an embodiment of the present application provides a computer readable storage medium, where a computer program is stored, where the computer program is executed by a processor to implement a method for pushing product information according to the first aspect of the present application.

In a fifth aspect, an embodiment of the present application provides a computer program product, which when run on a terminal device, causes the terminal device to perform a pushing method of product information as set forth in the first aspect of the embodiment of the present application.

The advantages achieved by the second to fifth aspects described above may be referred to in the description of the first aspect described above.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are needed in the embodiments or the description of the prior art will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present application, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

FIG. 1 is a flowchart of an embodiment of a method for pushing product information according to an embodiment of the present application;

FIG. 2 is a block diagram of one embodiment of a pushing device for product information according to an embodiment of the present application;

fig. 3 is a schematic diagram of a terminal device according to an embodiment of the present application.

Detailed Description

In the following description, for purposes of explanation and not limitation, specific details are set forth such as the particular system architecture, techniques, etc., in order to provide a thorough understanding of the embodiments of the present application. It will be apparent, however, to one skilled in the art that the present application may be practiced in other embodiments that depart from these specific details. In other instances, detailed descriptions of well-known systems, devices, circuits, and methods are omitted so as not to obscure the description of the present application with unnecessary detail. Furthermore, the terms "first," "second," "third," and the like in the description of the present specification and in the appended claims, are used for distinguishing between descriptions and not necessarily for indicating or implying a relative importance.

The embodiment of the application provides a pushing method, a pushing device, terminal equipment and a storage medium of product information, which can predict the probability of a new product purchased by a customer, so that relevant information of the new product can be accurately pushed to the customer.

It should be understood that the execution body of the product information pushing method provided by the embodiment of the present application may be a mobile phone, a tablet computer, a wearable device, a vehicle-mounted device, an augmented reality (augmented reality, AR)/Virtual Reality (VR) device, a notebook computer, an ultra-mobile personal computer (UMPC), a netbook, a personal digital assistant (personal digital assistant, PDA), a large-screen television, or a server, and the specific types of the terminal device and the server are not limited in the embodiment of the present application. The server may be an independent server, or may be a cloud server that provides cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (ContentDelivery Network, CDN), and basic cloud computing services such as big data and artificial intelligence platforms.

Referring to fig. 1, an embodiment of a method for pushing product information according to an embodiment of the present application includes:

101. determining an original product associated with a target customer and a new product associated with the original product according to a pre-constructed product relationship graph, wherein the product relationship graph records the interrelationship between the customer and the customer, between the customer and the product and between the product and the product;

the execution body of the embodiment of the application can be various types of servers or terminal equipment, for example, can be a system server of a shopping platform. Before step 101 is performed, a relationship graph of the relationships between the clients and the products and between the products in the record platform needs to be constructed, where the relationship graph is similar to the knowledge graph, and includes a plurality of nodes and edges connected between the nodes, where the nodes may include entities such as client names, client attributes, product names, product attributes, and the like, and the edges may include connection relationships such as client attribute similarity, product purchase, product attribute similarity, product attribute complementation, and product attribute upgrading, and the like.

(1) Acquiring customer attribute data, product purchase data and product attribute data, wherein the customer attribute data comprises a customer name and a customer attribute, the product purchase data comprises a record of a product purchased by a customer, and the product attribute data comprises a product name and a product attribute;

(2) Constructing a first sub-relationship graph according to the client attribute data, wherein nodes of the first sub-relationship graph are client names and client attributes, and edges of the first sub-relationship graph represent similar relationships of the client attributes;

(3) Constructing a second sub-relationship graph according to the product purchase data, wherein nodes of the second sub-relationship graph are customer names and product names, and edges of the second sub-relationship graph represent product purchase relationships;

(4) Constructing a third sub-relationship graph according to the product attribute data, wherein nodes of the third sub-relationship graph are product names and product attributes, and edges of the third sub-relationship graph represent product attribute similarity relationships, product attribute complementary relationships and product attribute upgrading relationships;

(5) And the first sub-relationship map, the second sub-relationship map and the third sub-relationship map are mutually connected according to corresponding nodes to obtain the product relationship map, wherein the nodes of the product relationship map are a customer name, a product name, a customer attribute and a product attribute, and the edges of the product relationship map represent a customer attribute similarity relationship, a product purchase relationship, a product attribute similarity relationship, a product attribute complementary relationship and a product attribute upgrading relationship.

For the step (1), based on the entity extraction and relationship connection concept of the knowledge graph, the entity connection relationship in the data can be obtained by mining the customer attribute data representing the customer relationship dimension, the product attribute data representing the product relationship dimension and the product purchase data representing the relationship dimension between the customer and the product from the enterprise data of the shopping platform.

For the step (2), the first sub-relationship graph represents a relationship dimension between clients, and the relationship label between the clients can be established specifically through the similarity degree of the client attributes, that is, the clients with similar client attributes are connected with each other in the first sub-relationship graph. The nodes of the first sub-relationship graph are customer names and customer attributes, wherein the customer attributes may include age segmentation, marital status, life stage, occupation, academic, financial and hobbies, and the like. The edges of the first sub-relationship graph are similar relationships of client attributes, the similarity degree between certain client attributes of two clients can be calculated, and if the similarity degree exceeds a set threshold value, graph nodes where the two clients are located are connected through the similar relationships of the client attributes. For example, if both client X and client Y are married, then client X and client Y may be connected by a "married" client attribute similarity relationship.

For step (3) above, the second sub-relationship graph represents a relationship dimension between the customer and the product, and a "customer-purchase-product" relationship network may be constructed according to the history of the customer's purchase of the product. The nodes of the second sub-relationship graph are customer names and product names, and edges represent product purchase relationships. For example, if customer X had purchased product Z, customer X and product Z may be connected by a product purchase relationship.

For the step (4), the third sub-relationship graph represents the relationship dimension between the products, specifically, the relationship labels, such as the similar product label, the complementary product label, and the upgrade product label, between the products may be established through the update record between the products and the attribute similarity/complementary degree/upgrade relationship between the products, so that the products having the similar product attributes, the complementary product attributes, and the upgrade relationship between the product attributes may be connected to each other in the third sub-relationship graph. The nodes of the third sub-relationship graph are product names and product attributes, wherein the product attributes can include production date, product category, manufacturer information, product specification, product price, and the like. The edges of the third sub-relationship graph are product attribute similarity relationship, product attribute complementary relationship and product attribute upgrading relationship, the similarity degree between certain product attributes of two products can be calculated, if the similarity degree exceeds a set threshold value, graph nodes where the two products are located are connected through the product attribute similarity relationship, in addition, the two products can be connected through the product attribute complementary relationship and the product attribute upgrading relationship, the correlations can be set manually, and the correlations can be automatically generated by identifying product information (such as product version numbers, product attribute complementary comparison tables and the like).

And (5) connecting the first sub-relationship map, the second sub-relationship map and the third sub-relationship map with each other according to the corresponding nodes to obtain a final relationship map. For example, a first sub-relationship graph and a second sub-relationship graph may be connected by a node "customer name", while a second sub-relationship graph and a third sub-relationship graph may be connected by a node "product name", thereby connecting the 3 sub-relationship graphs into one overall product relationship graph. Specifically, the nodes of the product relationship map are customer names, product names, customer attributes and product attributes, and edges represent customer attribute similarity relationships, product purchase relationships, product attribute similarity relationships, product attribute complementary relationships and product attribute upgrading relationships.

Because the product relationship map records the interconnection relationship between each customer and each product, for example, for a certain customer X, it can be determined through the product relationship map which products the customer X purchases, which other customers have an association relationship with the customer X, which other products the products purchased by the customer X have an association relationship with, based on the information, an original product associated with the customer X (i.e., a product that has been put on the platform for a certain period of time) can be determined, and a new product associated with the original product (i.e., a product that has been put on the platform for sale or is about to be put on the platform for sale) can be determined. It should be noted that there may be a plurality of the number of the original products.

In the embodiment of the application, the target client can be any client registered on the shopping platform, and the embodiment of the application determines whether to push the relevant information of a new product for the target client by calculating the probability of the target client purchasing the new product. From the product relationship graph, an original product associated with the target customer (e.g., a product that is interconnected with the target customer in the product relationship graph) may be determined, as well as a new product associated with the original product (e.g., a new product that is interconnected with the original product in the product relationship graph).

102. Calculating the probability of purchasing the original product by the target client according to the product relation graph;

according to the attribute connection relation between the target customer and the original product recorded by the product relation map, the probability of purchasing the original product by the target customer can be calculated. For example, the probability that the target customer purchases the original product may be determined according to the relationship between the target customer and the original product in the product relationship map, the customer attribute of the target customer, the product attribute of the original product, and other factors.

In an implementation manner of the embodiment of the present application, the calculating, according to the product relationship graph, a probability that the target customer purchases the original product may include:

And calculating the product attribute conditional probability of the original product and the customer attribute conditional probability of the original product according to the product relation graph, wherein the product attribute conditional probability of the original product represents the probability that the original product is purchased by the target customer based on the condition of each product attribute, and the customer attribute conditional probability of the original product represents the probability that the original product is purchased by the target customer based on the condition of each customer attribute.

The conditional probability refers to the occurrence probability of the event a under the condition that another event B has occurred. The conditional probability is expressed as: p (A|B), read as "probability of A occurring under the condition that B occurs". For a certain original product Z, which has N product attributes, each product attribute can be calculated to obtain a corresponding product attribute conditional probability, i.e. the original product Z has N corresponding product attribute conditional probabilities. For example, for the product attribute "medical care" of the original product Z, the probability that the condition of the original product Z based on the product attribute of "medical care" is purchased by the target customer can be calculated through the product relationship map, that is, the product attribute conditional probability of "medical care" of the original product Z is obtained. On the other hand, assuming that the target client has M client attributes, each client attribute can be calculated to obtain a corresponding client attribute conditional probability, that is, the original product Z has M corresponding client attribute conditional probabilities. For example, for the "married" customer attribute of the target customer, the probability that the original product Z was purchased by the target customer based on the condition of the "married" customer attribute, i.e., the "married" customer attribute conditional probability of the original product Z, can be calculated by the relationship map.

Specifically, the calculating the product attribute conditional probability of the original product and the customer attribute conditional probability of the original product according to the product relationship map may include:

(1) Counting the number of clients of the product with the product attribute and the total number of clients of the product relationship graph aiming at each product attribute of the original product, and dividing the number of clients of the product with the product attribute by the total number of clients to obtain product attribute conditional probability corresponding to the product attribute;

(2) And counting the number of clients which have the client attribute and have purchased the original product in the product relationship graph and the total number of clients which have the product relationship graph aiming at each client attribute of the target client, and dividing the number of clients which have the client attribute and have purchased the original product by the total number of clients to obtain the client attribute conditional probability corresponding to the client attribute.

For example, for a certain product attribute "healthcare" of the original product Z, the number S of customers who have purchased a product with the "healthcare" product attribute can be counted by the product relationship map ₁ And counting the total number S of clients of the product relationship graph ₂ The "medical care" product attribute conditional probability of the original product Z can be employed (S ₁ /S ₂ ) 100% calculation, and the like, the conditional probabilities of the N product attributes of the original product Z (corresponding to the N product attributes of the original product Z respectively) can be calculated. For the customer attribute "married" of the target customer, the number S of customers who have been married and purchased the original product can be counted by the product relationship map ₃ And counting the total number S of clients of the product relationship graph ₂ Then the conditional probability of the "married" customer property of the original product Z can be employed (S ₃ /S ₂ ) 100% calculation, and so on, the M product attribute conditional probabilities (corresponding to the M customer attributes of the target customer respectively) of the original product Z can be calculated.

Here, the product attribute conditional probability of the original product and the customer attribute conditional probability may be used to characterize the probability size of the target customer purchasing the original product.

103. Determining attribute similarity of the new product and the original product according to the product relation map;

in order to calculate the probability that a target customer purchases a new product based on the probability that the target customer purchases the original product, it has previously been necessary to determine the attribute similarity of the new product to the original product from the product relationship graph. In general, the product attribute of a new product may be compared with the product attribute of an original product to obtain an attribute similarity, and since in the product relationship map, the new product is related to the original product, that is, the new product and the original product are connected through more than one attribute relationship (edge), the attribute similarity of the new product and the original product may also be determined according to the number and the category (for example, the product attribute similarity relationship, the product attribute complementary relationship, and the product attribute upgrading relationship) of the relationship (edge) between the new product and the original product.

(1) Counting types and quantity of attribute connection relations between the new product and the original product in the product relation map, wherein the types of the attribute connection relations comprise product attribute similarity relations, product attribute complementary relations and product attribute upgrading relations;

(2) And determining the attribute similarity of the new product and the original product according to the type and the number of the attribute connection relations.

For example, if a certain original product a has 10 product attributes and new product B has 10 product attributes, and the original product a and new product B are connected by 5 product attributes in the relationship graph, it may be determined that the attribute similarity of the original product a and new product B is 5/10×100% =50%. For another example, in the relationship graph, the original product a and the new product B are connected by 5 product attributes, wherein 3 product attribute connection relationships are product attribute similarity relationships, 1 product attribute connection relationship is a product attribute complementary relationship, and 1 product attribute connection relationship is a product attribute upgrading relationship, then a corresponding scaling factor may also be set according to the type of the attribute connection relationship, for example, the scaling factor of the product attribute similarity relationship may be set to be 100%, the scaling factor of the product attribute complementary relationship may be set to be 80%, and the scaling factor of the product attribute complementary relationship may be set to be 200%, and then it may be determined that the attribute similarity of the original product a and the new product B is 3/10×100% +1/10×80% +1/10×200% =58%.

It will be appreciated that there may be a plurality of original products associated with a new product, so that a similarity of attributes with the new product can be calculated for each original product in step 103. For example, if there are N original products associated with a new product B, the attribute similarity between each of the N original products and the new product B may be calculated.

104. According to the probability that the target client purchases the original product and the attribute similarity, calculating to obtain the probability that the target client purchases the new product;

next, the probability that the target customer purchases the new product may be calculated according to the probability that the target customer purchases the original product and the attribute similarity. It can be understood that, if the probability that the target customer purchases the original product is higher, the similarity between the new product and the attribute of the original product is higher, the probability that the target customer purchases the new product is also determined to be higher, wherein the specific probability calculation mode is not limited by the present application.

(1) Calculating to obtain the product attribute conditional probability of the new product according to the product attribute conditional probability of the original product and the attribute similarity, wherein the product attribute conditional probability of the new product represents the probability that the new product is purchased by the target client based on the conditions of all product attributes;

(2) Calculating to obtain the client attribute conditional probability of the new product according to the client attribute conditional probability of the original product and the attribute similarity, wherein the client attribute conditional probability of the new product represents the probability that the new product is purchased by the target client based on the conditions of all client attributes;

(3) And calculating the probability of the target client purchasing the new product according to the product attribute conditional probability of the new product and the client attribute conditional probability of the new product.

It should be appreciated that if there are multiple original products, each original product has a corresponding product attribute conditional probability and customer attribute conditional probability, and each original product and the new product have a corresponding attribute similarity. In addition, as described above, each original product has a corresponding plurality of product attribute conditional probabilities and a corresponding plurality of customer attribute conditional probabilities, and the same new product also has a corresponding plurality of product attribute conditional probabilities and a corresponding plurality of customer attribute conditional probabilities.

(1) Respectively carrying out normalization processing on the attribute similarity corresponding to each original product to obtain the conditional probability weight of each original product;

(2) Determining a first original product of the plurality of original products for each product attribute of the new product, the first original product being an original product connected with the new product by the product attribute in the product relationship graph;

(3) And carrying out weighted summation operation on the product attribute conditional probability of the product attribute of the first original product according to the conditional probability weight of the first original product to obtain the product attribute conditional probability of the product attribute of the new product.

Firstly, carrying out normalization processing on attribute similarity corresponding to each original product to obtain conditional probability weight of each original product. Assuming that 3 original products are provided, the attribute similarity between the original products and the new products is 50%,50% and 100%, normalization processing can be performed on the three attribute similarity, namely, the sum of the three attribute similarity is 1, so that 0.25,0.25 and 0.5 are obtained and are respectively the conditional probability weights of each original product. Then, for each product attribute of the new product, the product attribute conditional probability of that product attribute may be calculated in the same manner as follows. For example, for a product attribute "healthcare" of a new product, a first original product connected to the new product by a "healthcare" product attribute may be determined from the product relationship graph, assuming that there are 10 original products associated with the new product, 3 of which are connected to the new product by a "healthcare" product attribute, The 3 original products are the first original products. And then, carrying out weighted summation operation by combining the product attribute conditional probability of the 'medical care' product attribute of the first original product and the conditional probability weight of the first original product, so as to obtain the product attribute conditional probability of the 'medical care' product attribute of the new product. Assume that the conditional probability weights of the 3 first original products are w respectively ₁ ，w ₂ And w ₃ The product attribute conditional probabilities of the respective "medical care" product attributes are p ₁ ，p ₂ And p ₃ The product attribute conditional probability of the "medical care" product attribute that the new product has can be expressed as p _x ＝p ₁ *w ₁ +p ₂ *w ₂ +p ₃ *w ₃ 。

(1) For each original product, comment information of the target client on the original product is obtained;

(2) Carrying out semantic recognition on the comment information to obtain a semantic recognition result;

(3) And adjusting the conditional probability weight of the original product according to the semantic recognition result.

For example, for a certain original product Z, after normalization processing is performed according to the attribute similarity to obtain the conditional probability weight of the original product Z, comment information of the target client for the original product Z may also be obtained. And carrying out NLP natural semantic recognition on the part of comment information to obtain semantic recognition results, such as good evaluation, bad evaluation, medium evaluation and the like. Then, the conditional probability weight of the original product Z can be adjusted according to the semantic recognition result, for example, if the semantic recognition result is good, the conditional probability weight of the original product Z can be improved according to a certain proportion, and if the semantic recognition result is bad, the conditional probability weight of the original product Z can be reduced according to a certain proportion.

Further, the calculating the client attribute conditional probability of the new product according to the client attribute conditional probability of the original product and the attribute similarity may include:

(1) Determining, for each customer attribute of the target customer, a second original product of the plurality of original products, the second original product being an original product connected to the new product in the product relationship graph by the customer attribute;

(2) And carrying out weighted summation operation on the client attribute conditional probability of the client attribute of the second original product according to the conditional probability weight of the second original product to obtain the client attribute conditional probability of the client attribute of the new product.

Similar to the method of calculating the product attribute conditional probability of a new product, in calculating the customer attribute conditional probability of a new product, the customer attribute conditional probability of the customer attribute possessed by the new product can be calculated for each customer attribute of the target customer in the same manner as follows. For example, for a customer attribute "married" of a target customer, a second original product connected to the new product by the "married" customer attribute may be determined according to the product relationship map, assuming that there are 10 original products associated with the new product, wherein 2 original products are connected to the new product by the "married" customer attribute, and the 2 original products are the second original products. Then, a weighted summation operation is performed by combining the client attribute conditional probability of the 'married' client attribute of the second original product and the conditional probability weights of the second original product, so as to obtain the client attribute conditional probability of the 'married' client attribute of the new product, and the conditional probability weights of the 2 second original products are assumed to be w respectively ₁ And w ₂ The conditional probabilities of the client attributes of the 'married' client attributes are p respectively ₁ And p ₂ The customer attribute conditional probability of the "married" customer attribute that the new product has can be expressed as p _y ＝p ₁ *w ₁ +p ₂ *w ₂ 。

And then, calculating the probability of the target customer purchasing the new product according to the product attribute conditional probability of the new product and the customer attribute conditional probability of the new product. Specific calculation modes can include:

(1) Calculating to obtain the total product attribute probability of the new product according to the product attribute conditional probabilities of the new product;

(2) Calculating to obtain the total customer attribute probability of the new product according to the customer attribute conditional probabilities of the new product;

(3) And determining the probability of the target client purchasing the new product according to the total product attribute probability of the new product and the total client attribute probability of the new product.

When the total product attribute probability of a new product is calculated, the average value of the conditional probabilities of all the product attributes of the new product can be directly calculated to be used as the total product attribute probability, and the total product attribute probability can be calculated by using a total probability formula or a Bayes formula and other formulas. Similarly, when calculating the total customer attribute probability of a new product, the average value of the conditional probabilities of the customer attributes of the new product can be directly calculated as the total customer attribute probability, or the total customer attribute probability can be calculated by using a full probability formula or a Bayes formula and other formulas. Finally, the average or maximum of the total product attribute probability and the total customer attribute probability may be taken as the probability that the target customer purchases the new product.

105. If the probability of the target client purchasing the new product is greater than a set threshold, pushing information of the new product to the target client.

Next, if the probability that the target customer purchases the new product is greater than some set threshold (e.g., 50%), the target customer may be considered a potential customer for the new product, and relevant information for the new product may be pushed to the target customer. If the probability of the target customer purchasing the new product is less than the set threshold, the target customer may be considered to be not a potential customer of the new product, and the relevant information of the new product may not be pushed to the target customer. Obviously, each customer aiming at the shopping platform can adopt the same processing mode as the target customer, thereby realizing the accurate product information pushing effect. In actual operation, the system server of the shopping platform can send the related information of the new product to the client of the target user in a short message or shopping platform APP message mode, so that the pushing of the product information is completed.

It should be understood that the sequence number of each step in the foregoing embodiment does not mean that the execution sequence of each process should be determined by the function and the internal logic, and should not limit the implementation process of the embodiment of the present application.

Corresponding to the pushing method of the product information described in the foregoing embodiments, fig. 2 shows a block diagram of a pushing device of the product information provided in the embodiment of the present application, and for convenience of explanation, only the portion relevant to the embodiment of the present application is shown.

Referring to fig. 2, the apparatus includes:

a product determining module 201, configured to determine an original product associated with a target customer and a new product associated with the original product according to a pre-constructed product relationship map, where the relationship map records correlations between customers and each other, and between products and each other;

an original product purchase probability calculation module 202, configured to calculate a probability of the target customer purchasing the original product according to the product relationship graph;

the attribute similarity calculation module 203 is configured to determine attribute similarity between the new product and the original product according to the product relationship graph;

a new product purchase probability calculation module 204, configured to calculate, according to the probability that the target customer purchases the original product and the attribute similarity, a probability that the target customer purchases the new product;

And the product information pushing module 205 is configured to push the information of the new product to the target client if the probability that the target client purchases the new product is greater than a set threshold.

In an implementation manner of the embodiment of the present application, the pushing device for product information may further include:

the data acquisition module is used for acquiring customer attribute data, product purchase data and product attribute data, wherein the customer attribute data comprises a customer name and a customer attribute, the product purchase data comprises a record of a product purchased by a customer, and the product attribute data comprises a product name and a product attribute;

the first sub-relationship graph construction module is used for constructing a first sub-relationship graph according to the client attribute data, wherein nodes of the first sub-relationship graph are client names and client attributes, and edges of the first sub-relationship graph represent similar relationships of the client attributes;

the second sub-relationship graph construction module is used for constructing a second sub-relationship graph according to the product purchase data, wherein nodes of the second sub-relationship graph are customer names and product names, and edges of the second sub-relationship graph represent product purchase relationships;

the third sub-relationship graph construction module is used for constructing a third sub-relationship graph according to the product attribute data, wherein nodes of the third sub-relationship graph are product names and product attributes, and edges of the third sub-relationship graph represent product attribute similarity relationships, product attribute complementary relationships and product attribute upgrading relationships;

The product relation map construction module is used for connecting the first sub-relation map, the second sub-relation map and the third sub-relation map with each other according to corresponding nodes to obtain the product relation map, wherein the nodes of the product relation map are a customer name, a product name, a customer attribute and a product attribute, and the edges of the product relation map represent a customer attribute similarity relation, a product purchase relation, a product attribute similarity relation, a product attribute complementary relation and a product attribute upgrading relation.

Further, the original product purchase probability calculation module may include:

an original product conditional probability calculation unit, configured to calculate a product attribute conditional probability of the original product and a customer attribute conditional probability of the original product according to the product relationship graph, where the product attribute conditional probability of the original product represents a probability that the original product is purchased by the target customer based on conditions of each product attribute, and the customer attribute conditional probability of the original product represents a probability that the original product is purchased by the target customer based on conditions of each customer attribute;

wherein the original product conditional probability calculation unit may include:

A product attribute conditional probability calculation subunit of an original product, configured to count, for each product attribute of the original product, the number of customers who have purchased a product with the product attribute in the product relationship graph and the total number of customers that the product relationship graph has, and divide the number of customers who have purchased a product with the product attribute by the total number of customers, so as to obtain a product attribute conditional probability corresponding to the product attribute;

the customer attribute conditional probability calculation subunit is configured to count, for each customer attribute of the target customer, the number of customers who have the customer attribute and have purchased the original product in the product relationship graph and the total number of customers who have the product relationship graph, and divide the number of customers who have the customer attribute and have purchased the original product by the total number of customers, so as to obtain a customer attribute conditional probability corresponding to the customer attribute.

In one implementation manner of the embodiment of the present application, the new product purchase probability calculation module may include:

the product attribute conditional probability calculation unit of the new product is used for calculating the product attribute conditional probability of the new product according to the product attribute conditional probability of the original product and the attribute similarity, and the product attribute conditional probability of the new product represents the probability that the new product is purchased by the target client based on the conditions of all product attributes;

A new product customer attribute conditional probability calculation unit, configured to calculate, according to the original product customer attribute conditional probability and the attribute similarity, a new product customer attribute conditional probability, where the new product customer attribute conditional probability represents a probability that the new product is purchased by the target customer based on conditions of each customer attribute;

and the new product purchase probability calculation unit is used for calculating the probability of the target client purchasing the new product according to the product attribute conditional probability of the new product and the client attribute conditional probability of the new product.

Further, the number of the original products is plural, and the product attribute conditional probability calculation unit of the new product may include:

the attribute similarity normalization subunit is used for respectively carrying out normalization processing on the attribute similarity corresponding to each original product to obtain the conditional probability weight of each original product;

a first original product determining subunit configured to determine, for each product attribute of the new product, a first original product of the plurality of original products, the first original product being an original product connected to the new product by the product attribute in the product relationship map;

The first weighted summation subunit is used for executing weighted summation operation on the product attribute conditional probability of the product attribute of the first original product according to the conditional probability weight of the first original product to obtain the product attribute conditional probability of the product attribute of the new product;

the customer attribute conditional probability calculation unit of the new product may include:

a second original product determination subunit configured to determine, for each customer attribute of the target customer, a second original product of the plurality of original products, the second original product being an original product connected to the new product by the customer attribute in the product relationship map;

and the second weighted summation subunit is used for executing weighted summation operation on the client attribute conditional probability of the client attribute of the second original product according to the conditional probability weight of the second original product to obtain the client attribute conditional probability of the client attribute of the new product.

Further, the product attribute conditional probability calculation unit of the new product may further include:

the comment information acquisition subunit is used for acquiring comment information of the target client on each original product;

The semantic recognition subunit is used for carrying out semantic recognition on the comment information to obtain a semantic recognition result;

and the conditional probability weight adjustment subunit is used for adjusting the conditional probability weight of the original product according to the semantic recognition result.

In one implementation manner of the embodiment of the present application, the attribute similarity calculation module may include:

the attribute connection relation statistics unit is used for counting types and quantity of attribute connection relations between the new product and the original product in the product relation map, wherein the types of the attribute connection relations comprise product attribute similarity relations, product attribute complementary relations and product attribute upgrading relations;

and the attribute similarity calculation unit is used for determining the attribute similarity of the new product and the original product according to the type and the number of the attribute connection relations.

The embodiment of the application also provides a computer readable storage medium, wherein the computer readable storage medium stores computer readable instructions, and the computer readable instructions realize any product information pushing method shown in fig. 1 when being executed by a processor.

The embodiment of the application also provides a computer program product, which when run on a server, causes the server to execute a pushing method for implementing any one of the product information shown in fig. 1.

Fig. 3 is a schematic diagram of a terminal device according to an embodiment of the present application. As shown in fig. 3, the terminal device 3 of this embodiment includes: a processor 30, a memory 31, and computer readable instructions 32 stored in the memory 31 and executable on the processor 30. The processor 30, when executing the computer readable instructions 32, implements the steps of the push method embodiment of the respective product information described above, such as steps 101 to 105 shown in fig. 1. Alternatively, the processor 30, when executing the computer readable instructions 32, performs the functions of the modules/units of the apparatus embodiments described above, such as the functions of the modules 201-205 shown in fig. 2.

Illustratively, the computer readable instructions 32 may be partitioned into one or more modules/units that are stored in the memory 31 and executed by the processor 30 to complete the present application. The one or more modules/units may be a series of computer readable instruction segments capable of performing specific functions describing the execution of the computer readable instructions 32 in the terminal device 3.

The terminal device 3 may be a computing device such as a smart phone, a notebook, a palm computer, a cloud terminal device, and the like. The terminal device 3 may include, but is not limited to, a processor 30, a memory 31. It will be appreciated by those skilled in the art that fig. 3 is merely an example of the terminal device 3 and does not constitute a limitation of the terminal device 3, and may comprise more or less components than shown, or may combine certain components, or different components, e.g. the terminal device 3 may further comprise input and output devices, network access devices, buses, etc.

The processor 30 may be a central processing unit (Central Processing Unit, CPU), but may also be other general purpose processors, digital signal processors (Digital Signal Processor, DSP), application specific integrated circuits (AppLication Specific Integrated Circuit, ASIC), off-the-shelf programmable gate arrays (fierld-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or the like. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.

The memory 31 may be an internal storage unit of the terminal device 3, such as a hard disk or a memory of the terminal device 3. The memory 31 may be an external storage device of the terminal device 3, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card) or the like, which are provided on the terminal device 3. Further, the memory 31 may also include both an internal storage unit and an external storage device of the terminal device 3. The memory 31 is used for storing the computer readable instructions and other programs and data required by the terminal device. The memory 31 may also be used for temporarily storing data that has been output or is to be output.

It should be noted that, because the content of information interaction and execution process between the above devices/units is based on the same concept as the method embodiment of the present application, specific functions and technical effects thereof may be referred to in the method embodiment section, and will not be described herein.

It will be apparent to those skilled in the art that, for convenience and brevity of description, only the above-described division of the functional units and modules is illustrated, and in practical application, the above-described functional distribution may be performed by different functional units and modules according to needs, i.e. the internal structure of the apparatus is divided into different functional units or modules to perform all or part of the above-described functions. The functional units and modules in the embodiment may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit, where the integrated units may be implemented in a form of hardware or a form of a software functional unit. In addition, the specific names of the functional units and modules are only for distinguishing from each other, and are not used for limiting the protection scope of the present application. The specific working process of the units and modules in the above system may refer to the corresponding process in the foregoing method embodiment, which is not described herein again.

The integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the present application may implement all or part of the flow of the method of the above embodiments, and may be implemented by a computer program to instruct related hardware, where the computer program may be stored in a computer readable storage medium, and when the computer program is executed by a processor, the computer program may implement the steps of each of the method embodiments described above. Wherein the computer program comprises computer program code which may be in source code form, object code form, executable file or some intermediate form etc. The computer readable medium may include at least: any entity or device capable of carrying computer program code to a photographing device/terminal apparatus, recording medium, computer Memory, read-Only Memory (ROM), random access Memory (RAM, random Access Memory), electrical carrier signals, telecommunications signals, and software distribution media. Such as a U-disk, removable hard disk, magnetic or optical disk, etc.

In the foregoing embodiments, the descriptions of the embodiments are emphasized, and in part, not described or illustrated in any particular embodiment, reference is made to the related descriptions of other embodiments.

The above embodiments are only for illustrating the technical solution of the present application, and not for limiting the same; although the application has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present application, and are intended to be included in the scope of the present application.

Claims

1. The pushing method of the product information is characterized by comprising the following steps:

if the probability of the target client purchasing the new product is greater than a set threshold value, pushing information of the new product to the target client;

wherein, the product relationship map is constructed by the following steps:

the first sub-relationship map, the second sub-relationship map and the third sub-relationship map are connected with each other according to corresponding nodes to obtain the product relationship map, wherein the nodes of the product relationship map are a customer name, a product name, a customer attribute and a product attribute, and the edges of the product relationship map represent a customer attribute similarity relationship, a product purchase relationship, a product attribute similarity relationship, a product attribute complementary relationship and a product attribute upgrading relationship;

the calculating the probability of purchasing the original product by the target client according to the product relation graph comprises the following steps:

The calculating product attribute conditional probability of the original product and customer attribute conditional probability of the original product according to the product relation graph comprises the following steps:

counting the number of clients which have the client attribute and have purchased the original product in the product relationship graph and the total number of clients which have the product relationship graph aiming at each client attribute of the target client, and dividing the number of clients which have the client attribute and have purchased the original product by the total number of clients to obtain client attribute conditional probability corresponding to the client attribute;

the calculating, according to the probability that the target customer purchases the original product and the attribute similarity, the probability that the target customer purchases the new product includes:

calculating the probability of purchasing the new product by the target client according to the product attribute conditional probability of the new product and the client attribute conditional probability of the new product;

the number of the original products is multiple, and the product attribute conditional probability of the new product is calculated according to the product attribute conditional probability of the original products and the attribute similarity, including:

the calculating to obtain the customer attribute conditional probability of the new product according to the customer attribute conditional probability of the original product and the attribute similarity comprises the following steps:

2. The method of claim 1, further comprising, after normalizing the attribute similarity corresponding to each of the original products to obtain a conditional probability weight for each of the original products:

3. The method of claim 1 or 2, wherein said determining the attribute similarity of the new product to the original product from the product relationship graph comprises:

4. A pushing device for product information, comprising:

the product information pushing module is used for pushing the information of the new product to the target client if the probability of purchasing the new product by the target client is larger than a set threshold value;

the product relation map construction module is used for connecting the first sub-relation map, the second sub-relation map and the third sub-relation map with each other according to corresponding nodes to obtain the product relation map, wherein the nodes of the product relation map are a customer name, a product name, a customer attribute and a product attribute, and the edges of the product relation map represent a customer attribute similarity relation, a product purchase relation, a product attribute similarity relation, a product attribute complementary relation and a product attribute upgrading relation;

The original product purchase probability calculation module includes:

wherein the original product conditional probability calculation unit includes:

the product attribute conditional probability calculation subunit of the original product is used for counting the number of clients of the product with the product attribute and the total number of clients of the product relationship graph aiming at each product attribute of the original product, and dividing the number of clients of the product with the product attribute by the total number of clients to obtain product attribute conditional probability corresponding to the product attribute;

the customer attribute conditional probability calculation subunit of the original product is used for counting the number of customers which have the customer attribute and have purchased the original product and the total number of customers which have the product relationship graph in the product relationship graph aiming at each customer attribute of the target customer, and dividing the number of customers which have the customer attribute and have purchased the original product by the total number of customers to obtain customer attribute conditional probability corresponding to the customer attribute;

The new product purchase probability calculation module includes:

the new product purchase probability calculation unit is used for calculating the probability of the target customer purchasing the new product according to the product attribute conditional probability of the new product and the customer attribute conditional probability of the new product;

the number of the original products is plural, and the product attribute conditional probability calculation unit of the new product includes:

the customer attribute conditional probability calculation unit of the new product includes:

5. A terminal device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that the processor implements the pushing method of product information according to any of claims 1 to 3 when executing the computer program.

6. A computer readable storage medium storing a computer program, characterized in that the computer program, when executed by a processor, implements the pushing method of product information according to any of claims 1 to 3.