A method for acquiring and migrating website data

Abstract

A method for acquiring and migrating website data, used for vertical e-commerce websites, migrating historical data information of existing e-commerce websites to other e-commerce websites through transfer learning, and collecting basic user information of existing e-commerce websites through scripts As well as user behavior, the user feature database is obtained from the basic user information, the product feature database is obtained from the product information of the existing e-commerce website, and the correlation between user features and product features is analyzed according to user behavior to obtain a correlation matrix. Other e-commerce According to the basic information of the user who enters the website, combined with the correlation matrix, the website calculates the correlation between the user and the products of the e-commerce website, and recommends them. The present invention learns the relevant patterns of user characteristics and product characteristics in the original system with dense historical data, and migrates them to other similar e-commerce websites that cannot be effectively recommended due to insufficient historical data based on the transfer learning method.

Description

technical field [0001] The invention belongs to the technical field of network data application, relates to transfer learning technology, and is a method for acquiring and transferring website data. Background technique [0002] E-commerce, with electronics and information technology as the means and foundation, takes commerce as the core, and transfers the original traditional sales and purchase methods and channels to the Internet of Things. Breaking geographical restrictions enables enterprises to develop a broader market. E-commerce is vividly compared to the production and sales of mouse and cement. In recent years, it has achieved rapid growth because of its fast and convenient way. [0003] Personalized recommendation, due to the rapid expansion of Internet information, makes it difficult for users to quickly find products or information suitable for them. In this context, personalized recommendation technology has gradually been extensively researched and applied. ...

AI Technical Summary

Problems solved by technology

[0006] The technical problem to be solved by the present invention is: for e-commerce websites with scarce historical data, due to the lack of historical data, personalized recommendations cannot be made to users, and it is time-consuming and labor-intensive to directly copy the data of existing websites, requiring a new machine learning method method to solve the problem of website cold start with scarce historical data. Although the transfer learning method proposes a new machine learning theory, how to apply it to the cold start of the website remains to be solved.

Images