A method and device for pushing products

Abstract

The present invention discloses a method and device for pushing products, and relates to the field of computer technology. A specific implementation of the method includes: collecting company information and product information; constructing a knowledge graph based on the company information and the product information; wherein the nodes of the knowledge graph at least include company, product, product type and crowd attribute; determining the target node based on the user portrait characteristics of the target user, filtering out the node of the target product connected to the target node from the knowledge graph; and pushing the information of the target product to the target user. This implementation can solve the technical problem of inaccurate recommendations caused by user cold start and product cold start.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a method and device for pushing products. Background Art

[0002] The cold start of a recommendation system means that the system does not have user and product data to use algorithms for recommendation analysis. Cold start is divided into user cold start and product cold start. User cold start refers to the first time a user visits a product page and has no previous visit information, so recommendations cannot be made based on previous data. Product cold start refers to when a new product is launched and how to recommend the new product to users who may be interested in it.

[0003] For user cold start, the existing technology mainly provides a bottom-up recommendation, that is, prepare a batch of products in advance, and randomly select some products from this batch to recommend to users; for product cold start, it mainly recommends new products to users by interspersing them with old products in a certain proportion.

[0004] In the process of implementing the present invention, the inventors found that there are at least the following problems in the prior art:

[0005] For user cold start, we generally choose products with higher sales volume to recommend to users, which easily forms a long-tail distribution and causes many products not to be recommended to users. For product cold start, the randomness is relatively strong. Some users may be interested in the new product while others may not, and it is impossible to make accurate recommendations to users. Summary of the invention

[0006] In view of this, an embodiment of the present invention provides a method and device for pushing products to solve the technical problem of inaccurate recommendations caused by user cold start and product cold start.

[0007] To achieve the above object, according to one aspect of an embodiment of the present invention, a method for pushing a product is provided, comprising:

[0008] Collect company information and product information;

[0009] Constructing a knowledge graph based on the company information and the product information; wherein the nodes of the knowledge graph include at least company, product, product type and population attributes;

[0010] Determine a target node according to the user portrait characteristics of the target user, and filter out nodes of the target product connected to the target node from the knowledge graph;

[0011] Push information about the target product to the target user.

[0012] Optionally, collect company information and product information, including:

[0013] Collect company information from the company information table, wherein the company information includes the company name;

[0014] Product information is collected from the product information table, wherein the product information includes product name, company, and basic attribute information, wherein the basic attribute information includes type and population attributes.

[0015] Optionally, the collection of company information and product information also includes:

[0016] Natural language processing is performed on the text information of the product, and keywords in the text information are extracted as characteristic attribute information of the product.

[0017] Optionally, constructing a knowledge graph based on the company information and the product information includes:

[0018] Build the relationship between product nodes and company nodes based on the relationship between the product and the company to which the product belongs;

[0019] According to the product type corresponding to the product, build the relationship between the product node and the product type node;

[0020] According to the crowd attributes corresponding to the product, the relationship between the product node and the crowd attribute node is constructed.

[0021] Optionally, the nodes of the knowledge graph further include feature attributes;

[0022] Constructing a knowledge graph according to the company information and the product information also includes:

[0023] According to the characteristic attribute information corresponding to the product, the relationship between the product node and the characteristic attribute node is constructed.

[0024] Optionally, the target node is at least one of a company node, a product type node, and a crowd attribute node.

[0025] Optionally, filtering out nodes of target products connected to the target node from the knowledge graph includes:

[0026] Determine whether there is a node of the first target product that is directly connected to the target node and has not been pushed in the knowledge graph;

[0027] If yes, then obtaining the information of the first target product;

[0028] If not, the nodes of the second target product that are indirectly connected to the target node and have not been pushed are filtered out from the knowledge graph, and the information of the second target product is obtained.

[0029] In addition, according to another aspect of an embodiment of the present invention, there is provided a device for pushing a product, comprising:

[0030] Collection module, used to collect company information and product information;

[0031] A construction module, used to construct a knowledge graph based on the company information and the product information; wherein the nodes of the knowledge graph include at least company, product, product type and crowd attribute;

[0032] A screening module, used to determine a target node according to the user portrait characteristics of a target user, and to screen out nodes of a target product connected to the target node from the knowledge graph;

[0033] The push module is used to push the information of the target product to the target user.

[0034] Optionally, the acquisition module is further used for:

[0035] Collect company information from the company information table, wherein the company information includes the company name;

[0036] Product information is collected from the product information table, wherein the product information includes product name, company, and basic attribute information, wherein the basic attribute information includes type and population attributes.

[0037] Optionally, the acquisition module is further used for:

[0038] Natural language processing is performed on the text information of the product, and keywords in the text information are extracted as characteristic attribute information of the product.

[0039] Optionally, the building block is further used to:

[0040] Build the relationship between product nodes and company nodes based on the relationship between the product and the company to which the product belongs;

[0041] According to the product type corresponding to the product, build the relationship between the product node and the product type node;

[0042] According to the crowd attributes corresponding to the product, the relationship between the product node and the crowd attribute node is constructed.

[0043] Optionally, the nodes of the knowledge graph further include feature attributes;

[0044] The building blocks are also used to:

[0045] According to the characteristic attribute information corresponding to the product, the relationship between the product node and the characteristic attribute node is constructed.

[0046] Optionally, the target node is at least one of a company node, a product type node, and a crowd attribute node.

[0047] Optionally, the screening module is further used for:

[0048] Determine whether there is a node of the first target product that is directly connected to the target node and has not been pushed in the knowledge graph;

[0049] If yes, then obtaining the information of the first target product;

[0050] If not, the nodes of the second target product that are indirectly connected to the target node and have not been pushed are filtered out from the knowledge graph, and the information of the second target product is obtained.

[0051] According to another aspect of an embodiment of the present invention, there is further provided an electronic device, including:

[0052] one or more processors;

[0053] a storage device for storing one or more programs,

[0054] When the one or more programs are executed by the one or more processors, the one or more processors implement the method described in any one of the above embodiments.

[0055] According to another aspect of an embodiment of the present invention, a computer-readable medium is provided, on which a computer program is stored. When the program is executed by a processor, the method described in any one of the above embodiments is implemented.

[0056] An embodiment of the above invention has the following advantages or beneficial effects: because a technical means is adopted to construct a knowledge graph based on company information and product information features, determine a target node based on the user portrait features of a target user, and filter out nodes of a target product connected to the target node from the knowledge graph, the technical problem of inaccurate recommendations caused by user cold start and product cold start in the prior art is overcome. The embodiment of the present invention solves the product cold start problem of the recommendation system through the knowledge graph, and solves the user cold start problem by utilizing the association relationship of the user portrait features of the user in the knowledge graph, thereby improving the accuracy and efficiency of recommendations.

[0057] The further effects of the above-mentioned non-conventional optional manner will be described below in conjunction with specific implementation examples. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] The accompanying drawings are used to better understand the present invention and do not constitute an improper limitation of the present invention.

[0059] Figure 1 is a schematic diagram of the main process of the method for pushing products according to an embodiment of the present invention;

[0060] Figure 2 It is a schematic diagram of the main process of a method for pushing products according to a reference embodiment of the present invention;

[0061] Figure 3 is a schematic diagram of main modules of an apparatus for pushing products according to an embodiment of the present invention;

[0062] Figure 4 is an exemplary system architecture diagram to which embodiments of the present invention may be applied;

[0063] Figure 5 It is a schematic diagram of the structure of a computer system of a terminal device or a server suitable for implementing an embodiment of the present invention. DETAILED DESCRIPTION

[0064] The following is a description of exemplary embodiments of the present invention in conjunction with the accompanying drawings, including various details of the embodiments of the present invention to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present invention. Similarly, for clarity and conciseness, the description of well-known functions and structures is omitted in the following description.

[0065] Figure 1 Schematic diagram of the main process of the method for pushing products according to an embodiment of the present invention. As an embodiment of the present invention, Figure 1 As shown, the method for pushing a product may include:

[0066] Step 101, collect company information and product information.

[0067] First, data collection is carried out, which is mainly the preparation work for the construction of the knowledge graph. The collected data sources are divided into structured data and unstructured data. Structured data is stored in database tables, including company information tables and product information tables, and unstructured data is stored in unstructured databases, including text information of each product, such as advertising text, promotional text, etc.

[0068] Optionally, step 101 may include: collecting company information from a company information table, the company information including the company name; collecting product information from a product information table, the product information including the product name, the company to which it belongs, basic attribute information, the basic attribute information including type and population attributes. The company name may include the full name and the abbreviation; the basic attribute information of the product may include the type (such as health insurance, accident insurance, serious illness insurance, auto insurance, etc.), premium, insured amount, waiting period, hesitation period, application age, suitable population (such as the elderly, children, adults), etc.

[0069] Optionally, step 101 may also include: performing natural language processing on the text information of the product, and extracting keywords from the text information as characteristic attribute information of the product. By performing natural language processing on the text, characteristic attribute information of the product can be extracted from it, such as "high price", "high cost performance", etc. Extracting characteristic attribute information from the text information of the product can more accurately describe the product and further enrich the information of the product.

[0070] Taking insurance products as an example, the product information is shown in the following table:

[0071]

[0072]

[0073] Step 102: construct a knowledge graph based on the company information and the product information.

[0074] The knowledge graph aims to describe various entities or concepts and their relationships in the real world, which constitutes a huge semantic network graph. In the knowledge graph, nodes are also called entities or concepts, and edges are the relationships between nodes.

[0075] In an embodiment of the present invention, the nodes of the knowledge graph include at least companies, products, product types, and crowd attributes. Optionally, step 102 may include: constructing a relationship between a product node and a company node based on the relationship between the product and the company to which the product belongs; constructing a relationship between a product node and a product type node based on the product type corresponding to the product; and constructing a relationship between a product node and a crowd attribute node based on the crowd attribute corresponding to the product. Therefore, the relationship between each node is constructed through product-company, product-product type, and product-crowd attributes, thereby obtaining a knowledge graph.

[0076] Optionally, the nodes of the knowledge graph may also include characteristic attributes; accordingly, step 102 may also include: constructing a relationship between a product node and a characteristic attribute node according to the characteristic attribute information corresponding to the product. In order to construct a complete knowledge graph, the characteristic attribute may also be used as one of the nodes to construct a product-characteristic attribute relationship.

[0077] Take insurance products as an example to build the relationship shown in the following table:

[0078]

[0079]

[0080] in:

[0081] COMPANY_SKU: The relationship between a company and a product. Based on the product and the company to which it belongs, an edge between the product node and the company node is constructed. This is mainly to find all products under the company.

[0082] SKU_TYPE: The relationship between products and their main types. Based on the main type of the product, such as life insurance, health insurance, accident insurance, and auto insurance, the edge between the product node and the main type node is constructed, mainly to prevent the push of products of the same type.

[0083] SKU_SUB_TYPE: The relationship between products and product subtypes. Based on the above main product types, more fine-grained divisions can be made. For example, health insurance can be divided into million-dollar medical insurance, critical illness insurance, cancer insurance, etc.

[0084] SKU_PRICE: The relationship between products and feature attributes (such as price), which constructs the edge between product nodes and feature attribute nodes. For example, for different product price ranges, insurance with annual premiums greater than 8,000 yuan is defined as high-priced insurance, which is mainly recommended for high-net-worth people.

[0085] SKU_TARGET_AGE: The relationship between products and crowd attributes, which builds the edge between product nodes and crowd attribute nodes. For example, children's insurance is matched based on whether there are children in the user profile features.

[0086] By importing the nodes with the above-mentioned relationships and the data corresponding to each node into a graph database (such as a neo4j database), a knowledge graph can be obtained.

[0087] When a new product is launched, the new product can be added to the knowledge graph through steps 101-102, thus solving the cold start problem of the product.

[0088] Step 103, determine the target node according to the user portrait characteristics of the target user, and filter out the nodes of the target product connected to the target node from the knowledge graph.

[0089] If the target user visits the product page for the first time and has no previous visit information, the user portrait features of the target user are first obtained, and then the target node is determined based on the user portrait features of the target user (such as gender, age, occupation, whether there is a car, whether there are children, etc.), and then the node of the target product connected to the target node is filtered out from the knowledge graph. Among them, the target node is at least one of a company node, a product type node, and a crowd attribute node. Optionally, after receiving the target user's access request, the user portrait features of the target user can be collected from the user portrait database. The user portrait features may include gender, age, occupation, whether there is a car, whether there are children, spending power, etc.

[0090] For example, if the user profile feature of the target user is that he has children, then the target node is determined to be "Children"; if the user profile feature of the target user is high net worth, then the target node is determined to be "High Price"; if the user profile feature of the target user is that he has a car, then the target node is determined to be "Auto Insurance".

[0091] It should be pointed out that the target user generally has multiple user profile features. In order to push products efficiently, the priority of user profile features can be configured, for example: having children is higher than having a car, having a car is higher than having a high net worth, etc. The target node is determined based on the user profile features with high priority.

[0092] Optionally, filtering out the nodes of the target product connected to the target node from the knowledge graph includes: determining whether there is a node of the first target product that is directly connected to the target node and not pushed in the knowledge graph; if so, obtaining the information of the first target product; if not, filtering out the node of the second target product that is indirectly connected to the target node and not pushed from the knowledge graph, and obtaining the information of the second target product. In an embodiment of the present invention, it is first determined whether there is a node of the first target product that is directly connected to the target node and not pushed in the knowledge graph (i.e., the first-degree node of the target node); if so, obtaining the information of the first target product; if not, determining whether the node type of the second-degree node of the target node (i.e., the node directly connected to the first node) is a product.

[0093] If yes, then obtain the product information of the second-degree node; if no, filter out the node of the second target product that is directly connected to the second-degree node and has not been pushed from the knowledge graph (i.e., the third-degree node of the target node), and obtain the information of the second target product. And so on, no further details will be given.

[0094] Step 104: Push the target product information to the target user.

[0095] The embodiment of the present invention can use the constructed knowledge graph to recommend products to new users, thereby solving the user cold start problem. Moreover, the embodiment of the present invention uses the knowledge graph for a real-time recommendation system, utilizes high-speed retrieval of the graph, and queries products that meet user characteristics, thereby achieving real-time recommendations and improving recommendation efficiency.

[0096] Use AB split testing:

[0097] By using the recommendation of knowledge graph and the recommendation model in the prior art to conduct group tests and observe the click-to-purchase behavior of new users, it was found that the user's click-through rate increased by about 30%-70% compared with the old model, which proves the advantage of the method provided in this application in cold start.

[0098] According to the various embodiments described above, it can be seen that the embodiments of the present invention solve the technical problem of inaccurate recommendations caused by user cold start and product cold start in the prior art by constructing a knowledge graph based on company information and product information, determining a target node based on the user portrait features of the target user, and filtering out the nodes of the target product connected to the target node from the knowledge graph. The embodiments of the present invention solve the product cold start problem of the recommendation system through the knowledge graph, and solve the user cold start problem by utilizing the association relationship between the user portrait features of the user in the knowledge graph, thereby improving the accuracy and efficiency of recommendations.

[0099] Figure 2 Schematic diagram of the main process of a method for pushing products according to a reference embodiment of the present invention. As another embodiment of the present invention, Figure 2 As shown, the method for pushing a product may include:

[0100] Step 201, collect company information and product information.

[0101] Optionally, step 201 may include: collecting company information from a company information table, wherein the company information includes the company name; collecting product information from a product information table, wherein the product information includes the product name, the company to which it belongs, and basic attribute information, wherein the basic attribute information includes type and population attributes; performing natural language processing on text information of the product, and extracting keywords from the text information as characteristic attribute information of the product.

[0102] Among them, the company name can include the full name and abbreviation; the basic attribute information of the product can include type (such as health insurance, accident insurance, critical illness insurance, car insurance, etc.), premium, insured amount, waiting period, hesitation period, application age, suitable population (such as the elderly, children, adults), etc.; characteristic attribute information can include "high price", "high cost performance", etc.

[0103] Step 202: construct a relationship between a product node and a company node according to the relationship between the product and the company to which the product belongs.

[0104] Step 203: construct a relationship between a product node and a product type node according to the product type corresponding to the product.

[0105] Step 204: construct a relationship between a product node and a crowd attribute node according to the crowd attribute corresponding to the product.

[0106] Step 205: construct the relationship between the product node and the feature attribute node according to the feature attribute information corresponding to the product.

[0107] It should be noted that steps 202 to 205 may be executed simultaneously or in sequence. The embodiment of the present invention does not limit the execution order of steps 202 to 205, and they may be executed in any order.

[0108] Step 206, import each node of the above-mentioned constructed relationship and the data corresponding to each node into the graph database to obtain a knowledge graph.

[0109] When a new product is launched, the new product can be added to the knowledge graph through steps 201-206, thus solving the cold start problem of the product.

[0110] Step 207, determine the target node according to the user portrait characteristics of the target user, and filter out the nodes of the target product connected to the target node from the knowledge graph.

[0111] If the target user visits the product page for the first time and has no previous visit information, the user profile features of the target user are first obtained, and then the target node is determined based on the user profile features of the target user (such as gender, age, occupation, whether there is a car, whether there are children, etc.), and then the node of the target product connected to the target node is filtered out from the knowledge graph. The target node is at least one of a company node, a product type node, a feature attribute node, and a crowd attribute node.

[0112] Specifically, determine whether there is a node of the first target product that is directly connected to the target node and has not been pushed in the knowledge graph; if so, obtain the information of the first target product; if not, filter out the node of the second target product that is indirectly connected to the target node and has not been pushed from the knowledge graph, and obtain the information of the second target product.

[0113] The target user usually has multiple user portrait features. In order to push products efficiently, you can configure the priority of the user portrait features and determine the target node based on the user portrait features with high priority.

[0114] Step 208: Push the target product information to the target user.

[0115] The embodiment of the present invention can use the constructed knowledge graph to recommend products to new users, thereby solving the user cold start problem. Moreover, the embodiment of the present invention uses the knowledge graph for a real-time recommendation system, utilizes high-speed retrieval of the graph, and queries products that meet user characteristics, thereby achieving real-time recommendations and improving recommendation efficiency.

[0116] In addition, the specific implementation content of the method for pushing products in a reference embodiment of the present invention has been described in detail in the method for pushing products described above, so the repeated content will not be described here.

[0117] Figure 3 is a schematic diagram of main modules of a device for pushing products according to an embodiment of the present invention, such as Figure 3 As shown, the device 300 for pushing products includes a collection module 301, a construction module 302, a screening module 303 and a push module 304; the collection module 301 is used to collect company information and product information; the construction module 302 is used to construct a knowledge graph based on the company information and the product information; wherein the nodes of the knowledge graph include at least company, product, product type and population attributes; the screening module 303 is used to determine the target node according to the user portrait characteristics of the target user, and filter out the nodes of the target product connected to the target node from the knowledge graph; the push module 304 is used to push the information of the target product to the target user.

[0118] Optionally, the acquisition module 301 is further used for:

[0119] Collect company information from the company information table, wherein the company information includes the company name;

[0120] Product information is collected from the product information table, wherein the product information includes product name, company, and basic attribute information, wherein the basic attribute information includes type and population attributes.

[0121] Optionally, the acquisition module 301 is further used for:

[0122] Natural language processing is performed on the text information of the product, and keywords in the text information are extracted as characteristic attribute information of the product.

[0123] Optionally, the construction module 302 is further used for:

[0124] Build the relationship between product nodes and company nodes based on the relationship between the product and the company to which the product belongs;

[0125] According to the product type corresponding to the product, build the relationship between the product node and the product type node;

[0126] According to the crowd attributes corresponding to the product, the relationship between the product node and the crowd attribute node is constructed.

[0127] Optionally, the nodes of the knowledge graph further include feature attributes;

[0128] The building block 302 is also used for:

[0129] According to the characteristic attribute information corresponding to the product, the relationship between the product node and the characteristic attribute node is constructed.

[0130] Optionally, the target node is at least one of a company node, a product type node, and a crowd attribute node.

[0131] Optionally, the screening module 303 is further used for:

[0132] Determine whether there is a node of the first target product that is directly connected to the target node and has not been pushed in the knowledge graph;

[0133] If yes, then obtaining the information of the first target product;

[0134] If not, the nodes of the second target product that are indirectly connected to the target node and have not been pushed are filtered out from the knowledge graph, and the information of the second target product is obtained.

[0135] According to the various embodiments described above, it can be seen that the embodiments of the present invention solve the technical problem of inaccurate recommendations caused by user cold start and product cold start in the prior art by constructing a knowledge graph based on company information and product information, determining a target node based on the user portrait features of the target user, and filtering out the nodes of the target product connected to the target node from the knowledge graph. The embodiments of the present invention solve the product cold start problem of the recommendation system through the knowledge graph, and solve the user cold start problem by utilizing the association relationship between the user portrait features of the user in the knowledge graph, thereby improving the accuracy and efficiency of recommendations.

[0136] It should be noted that the specific implementation content of the device for pushing products described in the present invention has been described in detail in the method for pushing products described above, so the repeated content will not be described again here.

[0137] Figure 4 An exemplary system architecture 400 is shown to which the method for pushing products or the apparatus for pushing products according to the embodiments of the present invention can be applied.

[0138] like Figure 4 As shown, system architecture 400 may include terminal devices 401, 402, 403, network 404 and server 405. Network 404 is used to provide a medium for communication links between terminal devices 401, 402, 403 and server 405. Network 404 may include various connection types, such as wired, wireless communication links or optical fiber cables, etc.

[0139] Users can use terminal devices 401, 402, 403 to interact with server 405 through network 404 to receive or send messages, etc. Various communication client applications can be installed on terminal devices 401, 402, 403, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, social platform software, etc. (only as examples).

[0140] The terminal devices 401 , 402 , and 403 may be various electronic devices having a display screen and supporting web browsing, including but not limited to smart phones, tablet computers, laptop computers, and desktop computers.

[0141] The server 405 may be a server that provides various services, such as a backend management server (only an example) that provides support for shopping websites browsed by users using the terminal devices 401, 402, and 403. The backend management server may analyze and process the received data such as product information query requests, and feed back the processing results (such as target push information, product information - only an example) to the terminal device.

[0142] It should be noted that the method for pushing products provided in the embodiment of the present invention is generally executed by the server 405 , and accordingly, the device for pushing products is generally disposed in the server 405 .

[0143] It should be understood that Figure 4 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to implementation requirements.

[0144] Reference below Figure 5 , which shows a schematic diagram of the structure of a computer system 500 of a terminal device suitable for implementing an embodiment of the present invention. Figure 5 The terminal device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0145] like Figure 5 As shown, the computer system 500 includes a central processing unit (CPU) 501, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 502 or a program loaded from a storage part 508 into a random access memory (RAM) 503. In the RAM 503, various programs and data required for the operation of the system 500 are also stored. The CPU 501, the ROM 502, and the RAM 503 are connected to each other via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0146] The following components are connected to the I / O interface 505: an input section 506 including a keyboard, a mouse, etc.; an output section 507 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 508 including a hard disk, etc.; and a communication section 509 including a network interface card such as a LAN card, a modem, etc. The communication section 509 performs communication processing via a network such as the Internet. A drive 510 is also connected to the I / O interface 505 as needed. A removable medium 511, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 510 as needed, so that a computer program read therefrom is installed into the storage section 508 as needed.

[0147] In particular, according to the embodiments disclosed in the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present invention include a computer program, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 509, and / or installed from the removable medium 511. When the computer program is executed by the central processing unit (CPU) 501, the above-mentioned functions defined in the system of the present invention are executed.

[0148] It should be noted that the computer-readable medium shown in the present invention may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device. In the present invention, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer-readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, optical cable, RF, etc., or any suitable combination of the above.

[0149] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, functions and operations of the system, method and computer program according to various embodiments of the present invention. In this regard, each box in the flow chart or block diagram can represent a module, a program segment, or a part of a code, and the above-mentioned module, program segment, or a part of a code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flow chart, and the combination of the boxes in the block diagram or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0150] The modules involved in the embodiments of the present invention may be implemented by software or hardware. The modules described may also be set in a processor, for example, it may be described as: a processor includes a collection module, a construction module, a screening module and a push module, wherein the names of these modules do not constitute a limitation on the modules themselves in some cases.

[0151] As another aspect, the present invention also provides a computer-readable medium, which may be included in the device described in the above embodiment; or it may exist independently without being assembled into the device. The above computer-readable medium carries one or more programs, and when the above one or more programs are executed by a device, the device implements the following method: collecting company information and product information; constructing a knowledge graph based on the company information and the product information; wherein the nodes of the knowledge graph include at least company, product, product type and population attributes; determining the target node based on the user portrait characteristics of the target user, and filtering out the nodes of the target product connected to the target node from the knowledge graph; and pushing the information of the target product to the target user.

[0152] According to the technical solution of the embodiment of the present invention, the technical problem of inaccurate recommendations caused by user cold start and product cold start in the prior art is overcome because the knowledge graph is constructed based on company information and product information, the target node is determined based on the user portrait characteristics of the target user, and the nodes of the target product connected to the target node are screened out from the knowledge graph. The embodiment of the present invention solves the product cold start problem of the recommendation system through the knowledge graph, and solves the user cold start problem by using the association relationship of the user portrait characteristics of the user in the knowledge graph, thereby improving the accuracy and efficiency of recommendations.

[0153] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions may occur depending on design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for pushing a product, characterized in that: include: Collect company information and product information; Constructing a knowledge graph based on the company information and the product information; wherein the nodes of the knowledge graph include at least company, product, product type and population attributes; Determine a target node according to the user portrait characteristics of the target user, and filter out nodes of target products connected to the target node from the knowledge graph; the target node is at least one of a company node, a product type node, and a crowd attribute node; Pushing information of the target product to the target user; According to the company information and the product information, a knowledge graph is constructed, including: Build the relationship between product nodes and company nodes based on the relationship between the product and the company to which the product belongs; According to the product type corresponding to the product, build the relationship between the product node and the product type node; According to the crowd attributes corresponding to the product, the relationship between the product node and the crowd attribute node is constructed; Filtering out nodes of target products connected to the target node from the knowledge graph includes: Determine whether there is a node of the first target product that is directly connected to the target node and has not been pushed in the knowledge graph; If yes, then obtaining the information of the first target product; If not, the nodes of the second target product that are indirectly connected to the target node and have not been pushed are filtered out from the knowledge graph, and the information of the second target product is obtained.

2. The method according to claim 1, characterized in that Collect company and product information, including: Collect company information from the company information table, wherein the company information includes the company name; Product information is collected from the product information table, wherein the product information includes product name, company, and basic attribute information, wherein the basic attribute information includes type and population attributes.

3. The method according to claim 2, characterized in that Collect company information and product information, including: Natural language processing is performed on the text information of the product, and keywords in the text information are extracted as characteristic attribute information of the product.

4. The method according to claim 1, characterized in that: The nodes of the knowledge graph also include feature attributes; Constructing a knowledge graph according to the company information and the product information also includes: According to the characteristic attribute information corresponding to the product, the relationship between the product node and the characteristic attribute node is constructed.

5. A device for pushing products, characterized in that: include: Collection module, used to collect company information and product information; A construction module, used to construct a knowledge graph based on the company information and the product information; wherein the nodes of the knowledge graph include at least company, product, product type and crowd attribute; A screening module, used to determine a target node according to the user portrait characteristics of a target user, and screen out nodes of target products connected to the target node from the knowledge graph; the target node is at least one of a company node, a product type node, and a crowd attribute node; A push module, used to push information of the target product to the target user; The building blocks are also used to: Build the relationship between product nodes and company nodes based on the relationship between the product and the company to which the product belongs; According to the product type corresponding to the product, build the relationship between the product node and the product type node; According to the crowd attributes corresponding to the product, the relationship between the product node and the crowd attribute node is constructed; The screening module is also used for: Determine whether there is a node of the first target product that is directly connected to the target node and has not been pushed in the knowledge graph; If yes, then obtaining the information of the first target product; If not, the nodes of the second target product that are indirectly connected to the target node and have not been pushed are filtered out from the knowledge graph, and the information of the second target product is obtained.

6. An electronic device, characterized in that: include: one or more processors; a storage device for storing one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 4.

7. A computer readable medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 4 is implemented.

Images