CN114998042A

CN114998042A - Product recommendation processing method, device and equipment and computer-readable storage medium

Info

Publication number: CN114998042A
Application number: CN202210744778.8A
Authority: CN
Inventors: 李雨洁
Original assignee: Ping An Property and Casualty Insurance Company of China Ltd
Current assignee: Ping An Property and Casualty Insurance Company of China Ltd
Priority date: 2022-06-28
Filing date: 2022-06-28
Publication date: 2022-09-02

Abstract

The application belongs to the technical field of data processing, and provides a processing method, a device, computer equipment and a computer readable storage medium for product recommendation, aiming at solving the problem of low processing efficiency of product recommendation, a preset product recommendation multilayer graph model is obtained according to a user identifier, the preset product recommendation multilayer graph model comprises at least three graph layers, each graph layer comprises a plurality of nodes, the adjacent graph layers have business incidence relation, the adjacent graph layers comprise a user layer, the user layer only comprises one node, then the recommendation probability of each node is obtained according to all the nodes and based on a recommendation algorithm of a Personalrank, then the product recommendation is carried out according to the recommendation probability and based on a preset probability use mode by using the recommendation probability corresponding to the nodes of different types, the search efficiency can be improved, and the node based on the plurality of the graph layers realizes the recommendation of multilayer entities, the recommendation efficiency, diversity and accuracy of product recommendation are improved.

Description

Product recommendation processing method, device and equipment and computer-readable storage medium

Technical Field

The present application relates to the field of data processing technologies, and in particular, to a method and an apparatus for processing product recommendation, a computer device, and a computer-readable storage medium.

Background

With the development of information technology and internet, new opportunities and challenges are met by the insurance industry, and the marketing channels of insurance products are diversified, and the digital transformation of the marketing of the insurance products also needs to follow the era. In the conventional technology, when the insurance industry carries out insurance product marketing by using a digital channel, a machine learning technology is generally adopted to support the recommendation of the insurance product, but the recommendation mode of the insurance product can only realize the recommendation of one entity of the insurance product, for example, for the recommendation of various entities of the insurance product, a sales channel or an insurance category, and the like, if the recommendation of various entities of the insurance product is carried out, models corresponding to various entities are required to be adopted for operation, so that the recommendation efficiency of the insurance entity is reduced.

Disclosure of Invention

The application provides a processing method and device for product recommendation, computer equipment and a computer readable storage medium, which can solve the technical problem of low processing efficiency of product recommendation in the prior art.

In a first aspect, the present application provides a processing method for product recommendation, including: the method comprises the steps of obtaining a user identification of a target user, obtaining a preset product recommendation multilayer graph model corresponding to the user identification according to the user identification, wherein the preset product recommendation multilayer graph model comprises at least three graph layers, each graph layer comprises a plurality of nodes, business incidence relation exists between adjacent graph layers, a user layer is included, the user layer only comprises one node as a user node, and the user node is used for describing the user identification; acquiring the recommendation probability of each node according to all the nodes based on a recommendation algorithm of the PersonalRank; and recommending the product by using the recommendation probabilities corresponding to the nodes of different types respectively according to the recommendation probabilities and based on a preset probability using mode.

In a second aspect, the present application further provides a product recommendation processing apparatus, including: the system comprises a first obtaining unit and a second obtaining unit, wherein the first obtaining unit is used for obtaining a user identifier of a target user and obtaining a preset product recommendation multilayer graph model corresponding to the user identifier according to the user identifier, the preset product recommendation multilayer graph model comprises at least three graph layers, each graph layer comprises a plurality of nodes, business association relation exists between adjacent graph layers, the adjacent graph layers comprise a user layer, the user layer only comprises one node as a user node, and the user node is used for describing the user identifier; the second obtaining unit is used for obtaining the recommendation probability of each node according to all the nodes and based on the recommendation algorithm of the PersonalRank; and the probability using unit is used for recommending the product by using the recommendation probability corresponding to each of the nodes of different types according to the recommendation probability and based on a preset probability using mode.

In a third aspect, the present application further provides a computer device, which includes a memory and a processor, where the memory stores a computer program, and the processor implements the steps of the processing method for product recommendation when executing the computer program.

In a fourth aspect, the present application further provides a computer-readable storage medium storing a computer program which, when executed by a processor, causes the processor to perform the steps of the method of processing the product recommendation.

The application provides a processing method and device for product recommendation, computer equipment and a computer readable storage medium. The processing method comprises the steps of obtaining a user identification of a target user, obtaining a preset product recommendation multilayer graph model corresponding to the user identification according to the user identification, wherein the preset product recommendation multilayer graph model comprises at least three graph layers, each graph layer comprises a plurality of nodes, business incidence relation exists between adjacent graph layers, the adjacent graph layers comprise a user layer, the user layer only comprises one node, obtaining recommendation probability of each node according to all the nodes and based on a recommendation algorithm of a Personalrank, and then carrying out product recommendation by using recommendation probabilities corresponding to different types of nodes respectively according to the recommendation probabilities and based on a preset probability use mode, wherein the preset product recommendation multilayer graph model occupies a small storage space due to the reduction of the structure of the Personalrank model, so that the search efficiency can be improved, and the recommendation of a multilayer entity can be realized based on the nodes of the plurality of the graph layers, and then the diversity, accuracy and recommendation efficiency of product recommendation are promoted.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present application, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.

Fig. 1 is a schematic flowchart of a processing method for product recommendation provided in an embodiment of the present application;

FIG. 2 is a model structure example of a preset product recommendation multi-layer graph model involved in applying a processing method of product recommendation in a non-vehicle insurance product according to an embodiment of the present application;

FIG. 3 is a schematic view of a first sub-flow of a processing method for product recommendation provided in an embodiment of the present application;

FIG. 4 is a second sub-flowchart of a processing method for product recommendation provided by an embodiment of the present application;

FIG. 5 is a third sub-flowchart of a processing method for product recommendation provided in an embodiment of the present application;

fig. 6 is a fourth sub-flowchart of a processing method for product recommendation provided in an embodiment of the present application;

FIG. 7 is a schematic block diagram of a processing device for product recommendation provided by an embodiment of the present application;

fig. 8 is a schematic block diagram of a computer device provided in an embodiment of the present application.

Detailed Description

The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are some, but not all, embodiments of the present application. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.

It will be understood that the terms "comprises" and/or "comprising," when used in this specification and the appended claims, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.

The embodiment of the application provides a processing method for product recommendation, which can be applied to computer equipment such as a server and the like and is used for recommending products with multi-layer service nodes in service scenes such as non-vehicle insurance products, online movie tickets or e-commerce shopping and the like in the insurance industry. Referring to fig. 1, fig. 1 is a schematic flowchart of a processing method for product recommendation according to an embodiment of the present application. As shown in FIG. 1, the method includes the following steps S11-S13:

s11, obtaining a user identification of a target user, and obtaining a preset product recommendation multilayer graph model corresponding to the user identification according to the user identification, wherein the preset product recommendation multilayer graph model comprises at least three graph layers, each graph layer comprises a plurality of nodes, business incidence relation exists between adjacent graph layers, a user layer is included, the user layer only comprises one node as a user node, and the user node is used for describing the user identification.

Specifically, in an application scenario of a non-vehicle insurance product (referred to as a non-vehicle product for short), online movie ticket sale or e-commerce shopping in the insurance industry, for example, if it is detected that an old user logs in an online website or APP, the old user may be considered as a target user, or when an insurance agent clicks a preset evaluation button to know a product recommendation mode of a certain target insurance user, the target insurance user may be used as the target user, and meanwhile, an instruction for starting product recommendation is considered to be started, and a user identifier of the target user is obtained in response to a start instruction for product recommendation, where the user identifier may be a content of a unique identification user such as a login account number, a user name, an identity number or a mobile phone number of the user.

Recommending a product with multiple layers of service nodes, constructing a preset product recommendation multilayer Graph Model suitable for product recommendation in advance according to service requirements, wherein the preset product recommendation multilayer Graph Model can be a Graph (algorithm) -Based Model (Graph-Based Model in English), the preset product recommendation multilayer Graph Model comprises at least three Graph layers, each Graph layer is used for describing the same type of entity, each Graph layer comprises a plurality of nodes, each node describes an entity, different nodes of each Graph layer describe different entities belonging to the same type, the entities are used for describing discrete objects existing in the service, for example, in a non-vehicle insurance product, the entities can be objects such as user names, insurance product Graph layers, insurance category names, sales channel names and the like, service incidence relations exist between adjacent entities, and the incidence relations are correlations between the entities corresponding to the nodes, for example, a purchase behavior may be generated for a user on a certain product, or a certain product may be related to a business such as a sale based on a certain sale channel, where the preset product recommendation multilayer graph model includes a user layer, and the user layer includes only one node as a user node, and the user node is used for describing a user identifier. Especially for the application scenario of recommending products based on historical transaction data, the preset product recommendation multilayer graph model may further include a historical transaction data layer, nodes of the historical transaction data layer are used for describing historical transaction data, the historical transaction data may include an identifier corresponding to a product or a service that a target user has purchased, the product may be a tangible commodity such as a book, a smart phone, and the service may be an intangible commodity such as insurance service, financial service, and communication service, for example, in the insurance service, related entities such as a product name, a product type, an APP, or an online channel or an offline channel such as a website of an insurance product may be involved, statistics of product recommendation performed on the user based on the historical transaction data such as the product or the service that the user has purchased may more accurately predict the preference of the user, thereby improving the accuracy of product recommendation, and the effectiveness of marketing resources is improved. For example, referring to fig. 2, fig. 2 is a model structure example of a preset product recommendation multilayer graph model involved in applying a processing method for product recommendation in a non-vehicle insurance product provided in an embodiment of the present application, as shown in fig. 2, in the non-vehicle insurance product example, the built preset product recommendation multilayer graph model may include 4 graph layers, and there is an association relationship in service between adjacent graph layers, where the association relationship may be an association relationship in a front-back order of service nodes, or a correlation of product elements involved in service, and respectively:

1) a first layer (namely a starting layer of a graph structure) is generally a user layer, nodes of the user layer are used for describing recommended users (namely target users), the user layer only comprises one node as a user node, and the user node adopts user identification to describe the target users of recommended products;

2) the second layer is an insurance product layer purchased and used for describing insurance products purchased by the recommended user, each node describes an insurance product purchased by the recommended user, the node corresponding to the insurance product 1 purchased by the recommended user describes the insurance product 1 purchased by the recommended user, and the node corresponding to the insurance product 2 purchased by the recommended user describes the insurance product 2 purchased by the recommended user;

3) the third layer is a sales channel layer and is used for describing sales channels of purchased insurance products in the second layer, and therefore, the second layer and the third layer are used as business association relations between adjacent layers;

4) and the fourth layer is an insurance category layer and is used for describing insurance category sets sold by each sales channel in the third layer, so that the third layer and the fourth layer are used as business association relations between adjacent layers.

Wherein, the third layer and the fourth layer are used for recommending products, the third layer is used for screening out a sales channel for recommending insurance products to the user 1, the fourth layer is used for screening out an insurance category for recommending insurance products to the user 1, and further, according to the insurance category screened out by the fourth layer, a specific insurance product contained in the insurance category is screened out as a recommended insurance product, in the example shown in fig. 2, if the development channel is most important in all sales channels related to the user, the user can be considered to be more favorable to the development channel, and subsequently, the marketing and the recommendation of the insurance products can be carried out in the development channel, so that the accuracy and the effectiveness of the sales channel recommendation are improved, and further, the accuracy and the adaptability of the insurance product to the target user recommendation are improved, to improve the marketing effectiveness of insurance products. Meanwhile, as shown in fig. 2, after the fourth layer, a fifth layer may be further constructed, where the fifth layer may be an insurance product layer and is used to describe a set of insurance products included in each insurance category, and a specific insurance product included in the insurance category may be screened out as a recommended insurance product according to the fifth layer, where edge connections between nodes in fig. 2 may be undirected connections or directed connections, and directed connections are adopted to describe directed correlations between nodes.

Based on the above-mentioned multi-layer graph model for recommending preset products, for each target user, user target data may be combined, where the user target data may include user history data, the user history data may include data of different behaviors such as historical transaction data and historical search data, and may also include data generated in different processes such as original history data and incremental data, where the historical search data is behavior data searched by the user in the past time, for example, search data in which the user searches for a certain product in the e-commerce APP without purchasing, and the historical transaction data may include an identifier and a sales channel corresponding to a product or service purchased by the target user, and historical transaction time, so that the user target data may be combined, for example, with the historical transaction data, to fill the user target data when needed, or fill the user target data in time to the multi-layer graph model for recommending preset products when the data is generated And each node contained in the model is constructed, so that the preset product recommendation multilayer graph model of the target user is constructed, and the preset product recommendation multilayer graph model of the specific user can be obtained. Continuing to refer to FIG. 2, as shown in FIG. 2, in this example, for user 1, the historical data describing the purchase of products by user 1 can be described by the structure of the graph of FIG. 2 as follows:

1) insurance products which are purchased by a user comprise a purchased insurance product 1 and a purchased insurance product 2;

2) the bought insurance product 1 can be sold through a sale channel 1, and the bought insurance product 2 can be sold through the sale channel 1 and a sale channel 2;

3) the insurance categories sold by the sale channel 1 comprise an insurance category 1 and an insurance category 3, and the insurance categories sold by the sale channel 2 comprise an insurance category 2 and an insurance category 3.

Meanwhile, optionally, the set of insurance products included in the insurance category 1 may be an insurance product set 1, the insurance product set 1 belongs to a plurality of insurance products of the insurance category 1, the set of insurance products included in the insurance category 2 may be an insurance product set 2, the insurance product set 2 belongs to a plurality of insurance products of the insurance category 2, the set of insurance products included in the insurance category 3 may be an insurance product set 3, and the insurance product set 3 includes a plurality of insurance products belonging to the insurance category 3.

After a user identification of a target user is obtained based on the constructed preset product recommendation multilayer graph model, a preset product recommendation multilayer graph model corresponding to the user identification is obtained according to the user identification, nodes contained in the preset product recommendation multilayer graph model are based on, the preset product recommendation multilayer graph model describes related entities and incidence relations contained in user target data, and the entities relate to multilayer service nodes and/or service factors, subsequently, the recommendation probability of each entity in the preset product recommendation multilayer graph model is counted according to the preset product recommendation multilayer graph model, and product recommendation is carried out on the target user based on the recommendation probability of each entity.

S12, obtaining the recommendation probability of each node according to all the nodes and based on the recommendation algorithm of the PersonalRank.

Specifically, for the preset product recommendation multilayer graph model, based on a recommendation algorithm of the PersonalRank, that is, according to the nodes of the preset product recommendation multilayer graph model, if there is a correlation between different nodes, the associated nodes are connected, and if a user has made a purchase action for a certain commodity (that is, there is a purchase correlation between the user and the commodity), the respective nodes of the user and the commodity are connected. For example, continuing to refer to fig. 2, as shown in fig. 2, for the user 1, according to the purchased insurance product node generated by purchasing insurance products, based on the business logic relationship existing between the adjacent image layers, the purchased insurance product node may associate a sales channel node and an insurance category node, and connect a plurality of nodes having business association relationship, so as to obtain a path: user 1-bought insurance product 1-sales channel 1-insurance category 1; user 1-bought insurance product 1-sales channel 1-insurance category 3; user 2-bought insurance product 2-sales channel 1-insurance category 3; the method comprises the steps that a user 1 purchases insurance products 2, a sales channel 2 insurance category 2 and other paths, then according to the connected paths, the user nodes on the first layer can move to the nodes on the last layer from the farthest to the farthest, the importance degree of all the nodes to the user can converge to a certain value after multiple iterations, the value is used for describing the importance degree of each node, and then the node set required by product recommendation can be obtained by sequencing according to the importance degree of each node. The relevance of the vertices (i.e., nodes) in the graph mainly depends on the following factors: 1) the number of paths between two vertices; 2) the path length between two vertices; 3) the vertex through which the path between two vertices passes. The vertex with high correlation generally has the following characteristics: 1) two vertexes are connected by a plurality of paths; 2) the path length connecting the two vertices is relatively short; 3) the path connecting the two vertices does not pass through the vertex with the greater degree of departure.

Referring to fig. 2, as shown in fig. 2, for personalized recommendation to the user 1, when the user 1 starts to walk from the node corresponding to the user in the figure and walks to a node, it is first determined whether to continue walking according to a preset probability, or to stop the walking and start to walk again from the node corresponding to the user 1. If the decision is made to continue the wandering, a node is randomly selected from the nodes pointed by the current node as the next-time node according to uniform distribution, so that after many wandering, the probability that each node is visited converges to a number, and finally the weight of each node in the recommendation list is the visit probability of the node, namely the recommendation probability of each node, so that the correlation between the node of the user 1 and all other nodes is obtained, and then particularly, products connected with the node of the user 1 without a direct edge can be taken, and the recommendation list is generated according to the correlation. The method and the device construct the preset product recommendation multilayer graph model which belongs to each user and only one node is reserved on a user layer, so that each search and iteration during product recommendation are only performed among necessary nodes, compared with the traditional PersonalRank algorithm structure which comprises a plurality of user nodes and operates among all the user nodes, the method and the device can reduce the time complexity and the memory utilization rate of the whole operation of the model, improve the search efficiency, expand the graph layers contained in the preset product recommendation multilayer graph model into at least three graph layers, determine the next layer of nodes through the incidence relation between the adjacent graph layers, and analogize the nodes until the preset product recommendation multilayer graph model is constructed and searched, can utilize the incidence relation among the multiple graph layers, can process the incidence analysis among complex and diverse entities, and realize the multilayer recommendation effect, compared with the traditional PersonalRank algorithm structure which only comprises two layers and can only realize recommendation of one entity, the product recommendation processing method provided by the embodiment of the application can realize recommendation of a multilayer entity, and is more suitable for a service scene recommended by the multilayer entity, so that the preset product recommendation multilayer graph model is smaller in structure, the storage space is smaller, the search efficiency of product recommendation can be improved, the recommendation of the multilayer entity is realized, and the diversity and the accuracy of product recommendation can be improved.

And S13, recommending products by using the recommendation probabilities corresponding to the nodes of different types according to the recommendation probabilities and based on a preset probability using mode.

Specifically, according to the recommendation probability, a plurality of recommendation probabilities meeting a preset probability condition can be obtained as target probabilities, the target probabilities are respectively corresponding to different types of nodes, and the target probabilities are used for product recommendation based on a preset probability use mode.

Further, the recommending the product by using the recommendation probabilities corresponding to the nodes of different types based on the preset probability using manner includes:

based on a preset probability screening condition, screening respective recommended probabilities of different types of nodes as target probabilities, taking the nodes corresponding to the target probabilities as target nodes, and displaying recommended products on a preset terminal according to the target nodes;

or screening the recommendation probabilities of different types of nodes as target probabilities based on preset probability screening conditions, and displaying the target probabilities to a preset terminal so that relevant personnel can recommend products according to the recommendation probabilities.

Specifically, after the recommendation probability of each node is obtained, based on a preset probability screening condition, screening the recommendation probability of each node of different types as a target probability, for example, for the node of each type of layer, the recommendation probability meeting the preset probability condition may be obtained as the target probability, for example, according to the correlation from high to low of the recommendation probability, the recommendation probability with high correlation is determined as the target probability, so that the recommendation probability of each node of different types is screened as the target probability, thereby implementing recommendation of multiple entities, and in product recommendation, the maximum recommendation probability in each layer is generally used as the target probability. For example, in the insurance field, when marketing non-vehicle insurance products, for an application scenario in which the recommendation probability is used as an insurance agent to perform offline marketing, a plurality of entities related to the non-vehicle insurance products and corresponding target probabilities can be displayed to a terminal, so that the insurance agent can adopt entities of each node according to the target probabilities to determine an optimal marketing mode of the non-vehicle insurance products to perform the offline marketing of the non-vehicle insurance products. For example, referring to fig. 2, for the user 1, the specific contents and recommendation probabilities thereof described by a plurality of sales channels and each insurance category may be displayed, so that the insurance agent selects the products in the sales channels and the insurance categories suitable for the user 1 as the marketing means of the non-vehicle insurance products according to the recommendation probabilities. And selling application scenes such as e-commerce or movie tickets and the like on line, screening recommendation probabilities of different types of nodes as target probabilities based on preset probability screening conditions, taking the nodes corresponding to the target probabilities as the target nodes, particularly, generating a recommended product list according to the relevance (namely, the recommendation probability is reduced from large to small) of products which are not purchased by a user in a product map layer (namely, products which are not directly connected with the user in the graph structure), generating a recommended product list which is the same as or similar to the products purchased by the user, and displaying the recommended products on a preset terminal, thereby recommending the recommended products to the user, and also realizing the recommendation of a plurality of entities.

In the embodiment of the application, only one node is reserved for a user layer describing the user identifier, and the graph layer is expanded to include at least three graph layers, so that a corresponding preset product recommendation multilayer graph model is constructed for each user, compared with a traditional PersonalRank algorithm structure which includes a plurality of user nodes, behaviors generated by all users are stored by using only one complex graph structure, each iteration needs to iterate the whole graph structure, the time consumption is long, and the memory utilization rate is high, the graph structure of the preset product recommendation multilayer graph model provided by the embodiment of the application reduces the complexity and enables the storage space occupied by the model to be small, each search and iteration during product recommendation are only performed in the range of related nodes, the time complexity and the memory utilization rate of the whole operation of the model are reduced, the search efficiency is improved, and simultaneously, the graph layers included by the preset product recommendation multilayer graph model are expanded to include at least three graph layers, and an incidence relation exists among the layers, so that the recommendation probability of each node can be counted based on a recommendation algorithm of the PersonalRank, and a multi-layer recommendation effect is achieved.

Referring to fig. 3, fig. 3 is a schematic view of a first sub-flow of a processing method for product recommendation provided in an embodiment of the present application, and as shown in fig. 3, in this embodiment, before acquiring a user identifier of a target user, the method further includes:

s111, responding to a starting instruction for recommending the product, and acquiring an initial user object recommended by the product;

s112, judging whether historical transaction data of the initial user object exist or not;

s113, if the historical transaction data exist, taking the initial user object as a target user;

and S114, if the historical transaction data does not exist, the initial user object is not taken as a target user.

Specifically, if it is monitored that a user logs in an online website or APP, or if an insurance agent is received to know a product recommendation mode for a certain target insurance user by clicking a preset evaluation button, the logged-in user or the target insurance user can be used as an initial user object for recommending a product, the user logs in the online website or APP in response, or the insurance agent clicks a preset evaluation button to obtain the initial user object for recommending the product, and according to the initial user object, whether historical transaction data exists in the initial user object can be inquired in a database, if the historical transaction data exists, the initial user object is determined to be an old user, the initial user object is used as a target user, the user identifier of the target user is obtained, and if the historical transaction data does not exist, the initial user object is not used as the target user, the initial user object may be determined as a new user, and for an application scenario in which product recommendation is performed on the basis of historical transaction data, particularly on a purchased product based on the historical transaction data, if the historical transaction data does not exist, product recommendation cannot be performed according to the processing method for product recommendation provided in the embodiment of the present application, and for such a scenario, a handling manner needs to be additionally set, for example, random recommendation or recommendation of a product with a high sales volume which is popular is performed.

Referring to fig. 4, fig. 4 is a schematic view of a second sub-flow of a processing method for product recommendation provided in an embodiment of the present application, and as shown in fig. 4, in this embodiment, the obtaining a preset product recommendation multi-layer graph model corresponding to the user identifier according to the user identifier includes:

s114, obtaining user target data corresponding to the user identification according to the user identification, and obtaining an initial preset product recommendation multilayer graph model;

s115, filling the user target data into the initial preset product recommendation multilayer graph model to obtain a preset product recommendation multilayer graph model corresponding to the user identification.

Specifically, when a product is recommended for a certain user, a preset product recommendation multilayer graph model corresponding to the user is constructed based on data corresponding to the user, that is, an initial preset product recommendation multilayer graph model based on a graph is constructed in advance, the initial preset product recommendation multilayer graph model can be defined as an empty model or a model frame structure endowed with an initial value, the initial preset product recommendation multilayer graph model can only define how many layers are included, and which incidence relations on services exist between adjacent layers are determined according to specific services, so that which data of the user are stored in each layer is determined, user target data corresponding to the user identification is obtained, and after the initial preset product recommendation multilayer graph model based on the graph is obtained, the user target data is filled into the initial preset product recommendation multilayer graph model, filling specific entities (specific data) contained in the user target data into corresponding nodes of each layer contained in the initial preset product recommended multilayer graph model, so as to obtain a preset product recommended multilayer graph model corresponding to the user identifier, for example, for a preset product recommended multilayer graph model corresponding to zhangsan, if zhangsan is 3, 3 nodes are constructed for storing 3 entities in the layers of purchased insurance products contained in the preset product recommended multilayer graph model, for a preset product recommended multilayer graph model corresponding to liqing, if liqing is 5, 5 nodes are constructed for storing 5 entities in the layers of purchased insurance products contained in the preset product recommended multilayer graph model, because the preset product recommended multilayer graph model corresponding to the user identifier is constructed only when product recommendation is determined for the user, the product recommendation method and the product recommendation system can be built as required, and compared with the preset product recommendation multilayer graph model which is built in advance by each user or adopts a fixed structure, the use space of the preset product recommendation multilayer graph model for storage resources can be reduced, and the utilization rate of the storage resources is saved.

In an embodiment, the step 1 of the preset product recommendation multilayer graph model is a user layer, the step 2 is a product identifier corresponding to a product already purchased by a user and described by the user layer, and the step 2 to the step n are product element layers having a business association relationship between adjacent layers, where n is greater than or equal to 3, please refer to fig. 5, and fig. 5 is a third sub-process schematic diagram of the processing method for product recommendation provided in the embodiment of the present application, as shown in fig. 5, in this embodiment, the obtaining, according to all the nodes, a recommendation probability of each node based on a recommendation algorithm of a PersonalRank includes:

s51, performing edge connection between the user node included in the user layer and each node included in the 2 nd layer, and performing edge connection between nodes having business association in the 2 nd layer to the nth layer, to obtain a plurality of paths between nodes in the 1 st layer to the nth layer;

s52, according to each path, starting to walk from one end of the path until the other end of the path is walked, counting the visited probability of each node, and after multiple rounds of walking, until the probability convergence of each node visited tends to be stable, and obtaining the recommended probability of each node relative to the user node.

Specifically, when product recommendation is performed on a user based on a purchased product of the user, setting a 1 st layer of the preset product recommendation multilayer graph model as a user layer, a 2 nd layer is a product identifier corresponding to a product already purchased by the user and described by the user layer, and the 2 nd layer to the nth layer are product element layers with business correlation between adjacent layers, where n is greater than or equal to 3, for example, please refer to fig. 2 continuously, in fig. 2, a node of the user layer of the first layer is used for describing the user 1, the 2 nd layer is used for describing insurance products already purchased by the user 1 and includes a purchased insurance product 1 and a purchased insurance product 2, the 3 rd layer is used for describing that the purchased insurance product 1 can be sold through a sales channel 1, the purchased insurance product 2 can be sold through a sales channel 1 and a sales channel 2, the 4 th layer is used for describing that the sales channel 1 can sell insurance products of a major type 1 and a major type 3, therefore, the 2 nd image layer to the 4 th image layer are product element layers with business correlation between adjacent image layers.

When a multi-layer entity product recommendation is performed based on a purchased product, because the 2 nd layer is a product purchased by a user, that is, a user node and each node of the 2 nd layer generate an overarching action, and each node of the 2 nd layer has a correlation with the user node, first performing edge connection on the user node included in the user layer and each node included in the 2 nd layer, where the edge connection may be an undirected edge, for example, performing edge connection on the user 1 and the purchased insurance product 1 in fig. 2, performing edge connection on the user 1 and the purchased insurance product 2 in fig. 2, and performing edge connection on nodes having a business association relationship in the 2 nd layer to the nth layer to obtain a plurality of paths between the respective nodes of the 1 st layer to the nth layer, for example, performing edge connection on the purchased insurance product 1 in fig. 2 and a sales channel 1, connecting bought insurance products 2 and a sales channel 1 in fig. 2 at the same time, connecting bought insurance products 2 and sales channels 2 in fig. 2 at the same time, connecting sales channels 1 and insurance categories 1 and 3 in fig. 2 at the same time respectively, and connecting sales channels 2 and insurance categories 2 and 3 at the same time respectively to form a path: user 1-bought insurance product 1-sales channel 1-insurance category 1; user 1-bought insurance product 1-sales channel 1-insurance category 3; user 2-bought insurance product 2-sales channel 1-insurance category 3; the method comprises the steps that a user 1 buys insurance products 2, sells channels 2, insures category 2 and other paths, then walks from one end of the paths according to the connected paths and each path, for example, the paths can walk from user nodes of the paths until the nodes of the 4 th layer at the other end of the paths are walked, the visited probability of each node is counted, and through multiple rounds of walking, the probability convergence of the visited probability of each node tends to be stable, the importance degree of the node to the user can converge to a certain value, the value is used for describing the importance degree of each node, the recommendation probability of each node relative to the user nodes is obtained, the search efficiency can be improved, and the multi-layer entity recommendation can be realized based on the nodes of multiple layers, so that the product recommendation diversity is improved, Accuracy and recommendation efficiency.

Referring to fig. 6, fig. 6 is a fourth sub-flow diagram of a processing method for product recommendation provided in the embodiment of the present application, and as shown in fig. 6, in this embodiment, the screening recommendation probabilities of different types of nodes as target probabilities based on a preset probability screening condition includes:

s61, according to the sequence incidence relation of nodes corresponding to different image layers in a service, taking the image layer corresponding to the initial node on the service as an initial image layer, sequencing recommendation probabilities contained in the initial image layer from large to small to obtain an initial image layer probability sequence, and acquiring a plurality of previous recommendation probabilities in the initial image layer probability sequence as initial target probabilities;

s62, according to the initial target probability, determining a plurality of highest recommendation probabilities of paths existing in each layer and nodes of the initial target probability as corresponding associated target probabilities, and taking the initial target probability and the associated target probabilities as respective target probabilities of different types of screened nodes.

Specifically, because different nodes have precedence association relationships such as a determined relationship, a logical relationship in a front-back order, and the like in a service, for example, please continue to refer to fig. 2, a sales channel 1 can only sell insurance category 1 and insurance category 3, a sales channel 2 can only sell insurance category 2 and insurance category 3, and when the sales channel is used as a preamble node, a subsequent corresponding insurance category node and the like can be determined, so that according to the precedence association relationships of nodes corresponding to different layers in the service, a layer corresponding to a starting node on the service is used as a starting layer, recommendation probabilities included in the starting layer are sorted in a descending order to obtain a starting layer probability sequence, and a plurality of previous recommendation probabilities in the starting layer probability sequence are obtained as starting target probabilities, for example, a highest recommendation probability in the starting layer probability sequence can be obtained as a starting target probability, the first three highest recommendation probabilities in the initial layer probability sequence can also be obtained as initial target probabilities, three products are recommended, then at least one highest recommendation probability of paths existing between nodes of other layers and the initial target probabilities is determined as corresponding associated target probabilities according to the initial target probabilities, and the initial target probabilities and the associated target probabilities are used as respective target probabilities of different types of screened nodes. For example, with continuing reference to fig. 2, when recommending a sales channel and an insurance category corresponding to the sales channel, if the sales channel 1 has the highest recommendation probability, the recommendation probability of the sales channel 1 can be obtained as the initial target probability, the recommendation probability of the insurance category 1 having a path with the sales channel 1 can be obtained as the associated target probability, the recommendation probability of the sales channel 1 can be used as the target probability of the sales channel node, the recommendation probability of the insurance category 1 can be used as the target probability of the insurance category node, or the recommendation probabilities of the insurance categories 1 and 3 having a path with the sales channel 1 can be obtained as the associated target probability, the recommendation probability of the sales channel 1 can be used as the target probability of the sales channel node, and the recommendation probabilities of the insurance categories 1 and 3 can be used as the target probability of the insurance category node, therefore, recommendation of the associated multi-entity nodes is realized, and diversity, accuracy and recommendation efficiency of product recommendation can be realized.

It should be noted that, in the processing method for product recommendation described in each of the above embodiments, the technical features included in different embodiments may be recombined as needed to obtain a combined implementation, but all of them are within the protection scope claimed in the present application.

Referring to fig. 7, fig. 7 is a schematic block diagram of a processing device for product recommendation according to an embodiment of the present application. Corresponding to the product recommendation processing method, the embodiment of the application further provides a product recommendation processing device. As shown in fig. 7, the product recommendation processing device includes a unit for executing the product recommendation processing method, and the product recommendation processing device may be configured in a computer device. Specifically, referring to fig. 7, the processing device 70 for product recommendation includes a first obtaining unit 71, a second obtaining unit 72 and a probability using unit 73.

The first obtaining unit 71 is configured to obtain a user identifier of a target user, and obtain a preset product recommendation multilayer graph model corresponding to the user identifier according to the user identifier, where the preset product recommendation multilayer graph model includes at least three graph layers, each graph layer includes a plurality of nodes, and an association relationship in service exists between adjacent graph layers, where the association relationship includes a user layer, and the user layer includes only one node as a user node, where the user node is used to describe the user identifier;

a second obtaining unit 72, configured to obtain, according to all the nodes, a recommendation probability of each node based on a recommendation algorithm of the PersonalRank;

and the probability using unit 73 is configured to recommend a product by using the recommendation probabilities corresponding to the different types of nodes according to the recommendation probability and based on a preset probability using mode.

In one embodiment, the processing device 70 for product recommendation further comprises:

the response unit is used for responding to a starting instruction for recommending the product and acquiring an initial user object for recommending the product;

a judging unit, configured to judge whether there is historical transaction data of the initial user object;

and the determining unit is used for taking the initial user object as a target user if the historical transaction data exists.

In one embodiment, the first obtaining unit 71 includes:

the first obtaining subunit is used for obtaining user target data corresponding to the user identification according to the user identification, and obtaining an initial preset product recommendation multilayer graph model based on a graph;

and the filling subunit is used for filling the user target data into the initial preset product recommendation multilayer graph model to obtain a preset product recommendation multilayer graph model corresponding to the user identifier.

In an embodiment, the 1 st layer of the preset product recommendation multilayer graph model is a user layer, the 2 nd layer is a product identifier corresponding to a product already purchased by a user and described by the user layer, and the 2 nd layer to the nth layer are product element layers having a business association relationship between adjacent layers, where n is greater than or equal to 3, where the second obtaining unit 72 includes:

a first connection subunit, configured to perform edge connection on a user node included in the user layer and each node included in a 2 nd layer, and perform edge connection on nodes having business association relationships in the 2 nd layer to an nth layer, so as to obtain a plurality of paths between respective nodes of the 1 st layer to the nth layer;

and the second connecting subunit is used for starting to walk from one end of the path according to each path until the other end of the path is walked, counting the visited probability of each node, and performing multiple rounds of walking until the probability convergence of each node visited tends to be stable, so as to obtain the recommended probability of each node relative to the user node.

In one embodiment, the probability using unit 73 includes:

the first screening subunit is used for screening the respective recommendation probabilities of different types of nodes as target probabilities based on a preset probability screening condition;

and the first display subunit is used for taking the node corresponding to the target probability as a target node and displaying the recommended product on a preset terminal according to the target node.

In one embodiment, the probability using unit 73 includes:

and the second display subunit is used for displaying the target probability to a preset terminal so that related personnel can recommend the product according to the recommendation probability.

In one embodiment, the first screening subunit comprises:

the second obtaining subunit is configured to, according to the precedence association relationship of nodes corresponding to different layers in a service, use a layer corresponding to an initial node on the service as an initial layer, sort recommendation probabilities included in the initial layer in a descending order to obtain an initial layer probability sequence, and obtain a plurality of previous recommendation probabilities in the initial layer probability sequence as initial target probabilities;

and the determining subunit is used for determining a plurality of highest recommended probabilities of paths existing in the nodes of the initial target probability in each other layer as corresponding associated target probabilities according to the initial target probability, and taking the initial target probability and the associated target probability as respective target probabilities of the screened different types of nodes.

It should be noted that, as can be clearly understood by those skilled in the art, the specific implementation processes of the processing apparatus and each unit for product recommendation described above may refer to the corresponding descriptions in the foregoing method embodiments, and for convenience and conciseness of description, no further description is provided herein.

Meanwhile, the division and connection manner of each unit in the processing device for recommending products are only used for illustration, in other embodiments, the processing device for recommending products may be divided into different units as needed, and each unit in the processing device for recommending products may also adopt different connection orders and manners to complete all or part of the functions of the processing device for recommending products.

The processing means of the above product recommendation may be implemented in the form of a computer program which is executable on a computer device as shown in fig. 8.

Referring to fig. 8, fig. 8 is a schematic block diagram of a computer device according to an embodiment of the present application. The computer device 500 may be a computer device such as a desktop computer or a server, or may be a component or part of another device.

Referring to fig. 8, the computer device 500 comprises a processor 502, a memory, and a network interface 505 connected by a system bus 501, wherein the memory may comprise a non-volatile storage medium 503 and an internal memory 504, and the memory may be a volatile storage medium.

The non-volatile storage medium 503 may store an operating system 5031 and a computer program 5032. The computer program 5032, when executed, causes the processor 502 to perform a method of processing a product recommendation as described above.

The processor 502 is used to provide computing and control capabilities to support the operation of the overall computer device 500.

The internal memory 504 provides an environment for the operation of the computer program 5032 in the non-volatile storage medium 503, and when the computer program 5032 is executed by the processor 502, the processor 502 can be enabled to execute a processing method of the product recommendation.

The network interface 505 is used for network communication with other devices. Those skilled in the art will appreciate that the configuration shown in fig. 8 is a block diagram of only a portion of the configuration relevant to the present teachings and does not constitute a limitation on the computer device 500 to which the present teachings may be applied, and that a particular computer device 500 may include more or less components than those shown, or combine certain components, or have a different arrangement of components. For example, in some embodiments, the computer device may only include a memory and a processor, and in such embodiments, the structures and functions of the memory and the processor are consistent with those of the embodiment shown in fig. 8, and are not described herein again.

Wherein the processor 502 is configured to run the computer program 5032 stored in the memory to perform the steps of: the method comprises the steps of obtaining a user identification of a target user, obtaining a preset product recommendation multilayer graph model corresponding to the user identification according to the user identification, wherein the preset product recommendation multilayer graph model comprises at least three graph layers, each graph layer comprises a plurality of nodes, business incidence relation exists between adjacent graph layers, a user layer is included, the user layer only comprises one node as a user node, and the user node is used for describing the user identification; acquiring the recommendation probability of each node according to all the nodes based on a recommendation algorithm of a PersonalRank; and recommending the product by using the recommendation probabilities corresponding to the nodes of different types respectively according to the recommendation probabilities and based on a preset probability using mode.

In an embodiment, before implementing the obtaining of the user identifier of the target user, the processor 502 further implements the following steps:

responding to a starting instruction for recommending a product, and acquiring an initial user object for recommending the product;

judging whether historical transaction data of the initial user object exist or not;

and if the historical transaction data exist, taking the initial user object as a target user.

In an embodiment, when the processor 502 implements obtaining, according to the user identifier, the preset product recommendation multilayer graph model corresponding to the user identifier, the following steps are specifically implemented:

according to the user identification, user target data corresponding to the user identification are obtained, and an initial preset product recommendation multilayer graph model is obtained;

and filling the user target data into the initial preset product recommendation multilayer graph model to obtain a preset product recommendation multilayer graph model corresponding to the user identification.

In an embodiment, the 1 st layer of the preset product recommendation multilayer graph model is a user layer, the 2 nd layer is a product identifier corresponding to a product already purchased by a user and described by the user layer, and the 2 nd layer to the nth layer are product element layers having a business association relationship between adjacent layers, where n is greater than or equal to 3, and when the processor 502 obtains the recommendation probability of each node according to all the nodes based on a recommendation algorithm of a PersonalRank, the following steps are specifically implemented:

performing edge connection on a user node contained in the user layer and each node contained in the 2 nd layer, and performing edge connection on nodes with business association relation in the 2 nd layer to the nth layer to obtain a plurality of paths between the respective nodes of the 1 st layer to the nth layer;

and according to each path, starting to walk from one end of the path until the other end of the path is walked, counting the visited probability of each node, and performing multi-round walking until the probability convergence of the visited probability of each node tends to be stable, so as to obtain the recommended probability of each node relative to the user node.

In an embodiment, when implementing the product recommendation by using the recommendation probabilities corresponding to different types of nodes based on the preset probability using manner, the processor 502 specifically implements the following steps:

based on a preset probability screening condition, screening respective recommendation probabilities of different types of nodes as target probabilities;

and taking the node corresponding to the target probability as a target node, and displaying the recommended product on a preset terminal according to the target node.

In an embodiment, when the processor 502 implements the product recommendation by using the recommendation probabilities corresponding to the different types of nodes based on the preset probability using manner, the following steps are specifically implemented:

and displaying the target probability to a preset terminal so that relevant personnel recommend products according to the recommendation probability.

In an embodiment, when the processor 502 performs the screening of the respective recommended probabilities of the different types of nodes as the target probabilities based on the preset probability screening condition, the following steps are specifically performed:

according to the sequence incidence relation of nodes corresponding to different image layers in a service, the image layer corresponding to an initial node on the service is used as an initial image layer, the recommendation probabilities contained in the initial image layer are sequenced from large to small to obtain an initial image layer probability sequence, and a plurality of previous recommendation probabilities in the initial image layer probability sequence are obtained and used as initial target probabilities;

and determining a plurality of highest recommendation probabilities of paths existing between the nodes of each other layer and the initial target probability as corresponding associated target probabilities according to the initial target probability, and taking the initial target probability and the associated target probability as respective target probabilities of different types of screened nodes.

It should be understood that in the embodiment of the present Application, the Processor 502 may be a Central Processing Unit (CPU), and the Processor 502 may also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field Programmable Gate Arrays (FPGAs) or other Programmable logic devices, discrete Gate or transistor logic devices, discrete hardware components, and the like. Wherein a general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.

It will be understood by those skilled in the art that all or part of the processes in the method for implementing the above embodiments may be implemented by a computer program, and the computer program may be stored in a computer readable storage medium. The computer program is executed by at least one processor in the computer system to implement the flow steps of the embodiments of the method described above.

Accordingly, the present application also provides a computer-readable storage medium. The computer-readable storage medium may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, the computer-readable storage medium storing a computer program that, when executed by a processor, causes the processor to perform the steps of:

a computer program product which, when run on a computer, causes the computer to perform the steps of the method of processing of product recommendations described in the embodiments above.

The computer readable storage medium may be an internal storage unit of the aforementioned device, such as a hard disk or a memory of the device. The computer readable storage medium may also be an external storage device of the device, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), etc. provided on the device. Further, the computer-readable storage medium may also include both an internal storage unit and an external storage device of the apparatus.

It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described apparatuses, devices and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.

The storage medium is an entity and non-transitory storage medium, and may be various entity storage media capable of storing computer programs, such as a usb disk, a removable hard disk, a Read-Only Memory (ROM), a magnetic disk, or an optical disk.

Those of ordinary skill in the art will appreciate that the elements and algorithm steps of the examples described in connection with the embodiments disclosed herein may be embodied in electronic hardware, computer software, or combinations of both, and that the components and steps of the examples have been described in a functional general in the foregoing description for the purpose of illustrating clearly the interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.

In the several embodiments provided in the present application, it should be understood that the disclosed apparatus and method may be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative. For example, the division of each unit is only one logic function division, and there may be another division manner in actual implementation. For example, various elements or components may be combined or may be integrated into another system, or some features may be omitted, or not implemented.

The steps in the method of the embodiment of the application can be sequentially adjusted, combined and deleted according to actual needs. The units in the device of the embodiment of the application can be combined, divided and deleted according to actual needs. In addition, functional units in the embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit.

The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a storage medium. Based on such understanding, the technical solution of the present application may be substantially implemented or contributed to by the prior art, or all or part of the technical solution may be embodied in a software product, which is stored in a storage medium and includes instructions for causing an electronic device (which may be a personal computer, a terminal, or a network device) to perform all or part of the steps of the method according to the embodiments of the present application.

The above description is only for the specific embodiments of the present application, but the scope of the present application is not limited thereto, and any person skilled in the art can easily think of various equivalent modifications or substitutions within the technical scope of the present application, and these modifications or substitutions should be covered within the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

Claims

1. A method for processing product recommendations, comprising:

the method comprises the steps of obtaining a user identification of a target user, obtaining a preset product recommendation multilayer graph model corresponding to the user identification according to the user identification, wherein the preset product recommendation multilayer graph model comprises at least three graph layers, each graph layer comprises a plurality of nodes, business incidence relation exists between adjacent graph layers, a user layer is included, the user layer only comprises one node as a user node, and the user node is used for describing the user identification;

acquiring the recommendation probability of each node according to all the nodes based on a recommendation algorithm of a PersonalRank;

and recommending the product by using the recommendation probabilities corresponding to the nodes of different types respectively according to the recommendation probabilities and based on a preset probability using mode.

2. The processing method of the product recommendation according to claim 1, further comprising, before the obtaining the user identifier of the target user:

3. The product recommendation processing method according to claim 1, wherein the obtaining a preset product recommendation multilayer graph model corresponding to the user identifier according to the user identifier comprises:

4. The processing method for product recommendation according to claim 1, wherein a 1 st layer of the preset product recommendation multilayer graph model is a user layer, a 2 nd layer is a product identifier corresponding to a product already purchased by a user and described by the user layer, and the 2 nd to nth layers are product element layers having a business association relationship between adjacent layers, where n is greater than or equal to 3, and obtaining, according to all the nodes, a recommendation algorithm based on a PersonalRank, the recommendation probability of each node includes:

5. The method for processing the product recommendation according to claim 1, wherein the recommending the product using the recommendation probabilities corresponding to the different types of nodes based on the preset probability using manner comprises:

6. The method for processing the product recommendation according to claim 1, wherein the recommending the product using the recommendation probabilities corresponding to the different types of nodes based on the preset probability using manner comprises:

and displaying the target probability to a preset terminal so that related personnel recommend the product according to the recommendation probability.

7. The processing method for product recommendation according to claim 5 or 6, wherein the screening, based on the preset probability screening condition, the respective recommendation probabilities of the different types of nodes as the target probabilities includes:

according to the sequence incidence relation of nodes corresponding to different image layers in a service, the image layer corresponding to an initial node on the service is used as an initial image layer, recommendation probabilities included in the initial image layer are sequenced from large to small to obtain an initial image layer probability sequence, and a plurality of front recommendation probabilities in the initial image layer probability sequence are obtained and used as initial target probabilities;

8. A product recommendation processing device, comprising:

the system comprises a first obtaining unit, a second obtaining unit and a third obtaining unit, wherein the first obtaining unit is used for obtaining a user identifier of a target user and obtaining a preset product recommendation multilayer graph model corresponding to the user identifier according to the user identifier, the preset product recommendation multilayer graph model comprises at least three graph layers, each graph layer comprises a plurality of nodes, business association relation exists between adjacent graph layers, the adjacent graph layers comprise a user layer, the user layer only comprises one node as a user node, and the user node is used for describing the user identifier;

the second obtaining unit is used for obtaining the recommendation probability of each node according to all the nodes and based on the recommendation algorithm of the PersonalRank;

and the probability using unit is used for recommending the product by using the recommendation probability corresponding to each of the nodes of different types according to the recommendation probability and based on a preset probability using mode.

9. A computer device, comprising a memory and a processor coupled to the memory; the memory is used for storing a computer program; the processor is adapted to run the computer program to perform the steps of the method according to any of claims 1-7.

10. A computer-readable storage medium, characterized in that the storage medium stores a computer program which, when being executed by a processor, realizes the steps of the method according to any one of claims 1 to 7.