E-commerce platform potential customer recommendation method and system based on comprehensive similarity

Abstract

The invention relates to an e-commerce platform potential customer recommendation method and system based on comprehensive similarity. The method comprises the following steps: selecting a motif, finding out all instances of the motif, constructing a high-order adjacency matrix of customers, then obtaining a customer sequence, obtaining representation vectors of the customers, and further calculating the contact closeness between the customers; calculating the similarity between the features carried by the clients; weighting the connection closeness and the characteristic similarity to obtain the comprehensive similarity of the clients, and further obtaining a KNN graph; selecting the customers with the highest core from the customers which are not distributed with the groups as core customers; taking the selected core customer as an initial group, and sequentially adding the customer which enables the group fitness function increment to be maximum in group neighbor customers into the group; and core customer selection and group expansion are repeated until all customers in the customer network of the e-commerce platform belong to the groups. According to the invention, the group existing in the e-commerce platform client network can be effectively mined, and the commodity which is purchased most frequently is recommended for the group.

Description

technical field

[0001] The invention relates to a method and system for recommending potential customers of an e-commerce platform based on comprehensive similarity. Background technique

[0002] Due to the interaction between social members in work, study, life, entertainment and other activities, a certain stable relationship is gradually formed, and then a social network is formed. With the rapid development of Internet technology, people introduced the concept of early social network into the Internet, and created an online social network oriented to social network services. Representative products of online social networks include domestic WeChat, Weibo, Taobao, and foreign Facebook and Twitter. The vigorous development of online social networks has greatly changed people's lifestyles. For example, online shopping has become a mainstream shopping method, and more than 80% of Internet customers often use online shopping. Internet users can use social networks to make f...

Examples

Embodiment Construction

[0022] The technical solution of the present invention will be specifically described below in conjunction with the accompanying drawings.

[0023] Such as figure 1 As shown, this example provides a method for recommending potential customers of an e-commerce platform based on comprehensive similarity, which is characterized in that it includes the following steps:

[0024] Step S1, first select a motif, find out all instances of the motif in the social network, and construct the customer's high-order adjacency matrix; then, perform random walks on the customer's low-order adjacency matrix and high-order adjacency matrix respectively , to obtain the customer sequence; finally, use the Skip-gram model to train the customer sequence to obtain the customer's representation vector, and calculate the relationship between customers according to the customer's representation vector;

[0025] Step S2, calculating the degree of similarity between the features carried by the customers;...

technical field

[0001] The invention relates to a method and system for recommending potential customers of an e-commerce platform based on comprehensive similarity. Background technique

[0002] Due to the interaction between social members in work, study, life, entertainment and other activities, a certain stable relationship is gradually formed, and then a social network is formed. With the rapid development of Internet technology, people introduced the concept of early social network into the Internet, and created an online social network oriented to social network services. Representative products of online social networks include domestic WeChat, Weibo, Taobao, and foreign Facebook and Twitter. The vigorous development of online social networks has greatly changed people's lifestyles. For example, online shopping has become a mainstream shopping method, and more than 80% of Internet customers often use online shopping. Internet users can use social networks to make f...

Examples

Embodiment Construction

[0022] The technical solution of the present invention will be specifically described below in conjunction with the accompanying drawings.

[0023] Such as figure 1 As shown, this example provides a method for recommending potential customers of an e-commerce platform based on comprehensive similarity, which is characterized in that it includes the following steps:

[0024] Step S1, first select a motif, find out all instances of the motif in the social network, and construct the customer's high-order adjacency matrix; then, perform random walks on the customer's low-order adjacency matrix and high-order adjacency matrix respectively , to obtain the customer sequence; finally, use the Skip-gram model to train the customer sequence to obtain the customer's representation vector, and calculate the relationship between customers according to the customer's representation vector;

[0025] Step S2, calculating the degree of similarity between the features carried by the customers;...

Images