A multi-e-commerce cross-recommendation method based on clustering feature migration

Abstract

The invention discloses a multi-e-commerce cross-recommendation method based on clustering feature migration, which includes the following steps: 1) Scoring matrix construction stage: a. collecting data of each e-commerce; b. data cleaning and denoising; c. constructing a scoring matrix; d ending ;2) Auxiliary domain learning stage: a obtain rating matrix; b extract user / item feature matrix; c cluster user / item feature matrix; d calculate average score; e construct cluster feature matrix; f for each auxiliary e-commerce , repeat the above steps until the end; 3) Target domain learning stage: a) obtain the target e-commerce scoring matrix; b transfer clustering features, and complete matrix decomposition. c reconstructs the target e-commerce scoring matrix; d generates a recommendation list; e ends. The present invention uses transfer learning technology to provide a new solution to the data sparsity, cold start, diversity and accuracy dilemmas existing in the e-commerce recommendation system.

Description

technical field [0001] The invention relates to a multi-e-commerce cross-recommendation method, which solves the problem of low recommendation accuracy in the case of extremely sparse data and cold start of an e-commerce recommendation system. Background technique [0002] As the scale of e-commerce websites continues to expand, the problem of information overload is becoming more and more serious, and a very potential method to solve this problem is a personalized recommendation system. For example, the well-known e-commerce platform Amazon uses behavior records such as clicks, browsing, favorites, and shopping carts that reflect users' purchasing interests to recommend other products that may be of interest to users. According to each user's preference, "thousands of people and thousands of faces" intelligent content recommendation can effectively improve key indicators such as user activity, stay time, payment rate, retention rate, etc., and create huge value for society ...

AI Technical Summary

Problems solved by technology

Recommendations in a single field cannot effectively share Internet resources, resulting in relatively occluded information and easy formation of information islands

Images