E-Commerce for Cold Start Product Suggestion using Micro Blogging Data Through Connecting Social Media

Abstract

In recent years, the boundaries between e-commerce and social networking became progressively blurred. Several e-commerce websites support the mechanism of social login wherever users will sign in the websites victimization their social network identities like their Facebook or Twitter accounts. Users also can post their fresh purchased product on microblogs with links to the e-commerce product websites. During this paper we have a tendency to propose a unique answer for cross-site cold-start product recommendation, that aims to advocate product from e-commerce websites to users at social networking sites in “coldstart” things, a haul that has seldom been explored before. A serious challenge is the way to leverage data extracted from social networking sites for cross-site cold-start product recommendation.We propose to use the coupled users across social networking sites and e-commerce websites (users United Nations agency have social networking accounts and have created purchases on e-commerce websites) as a bridge to map users’ social networking options to a different feature illustration for product recommendation. In specific, we have a tendency to propose learning each users’ and merchandises’ feature representations (called user embeddings and product embeddings, respectively) from information collected from e-commerce websites victimization continual neural networks so apply a changed gradient boosting trees methodology to remodel users’ social networking options into user embeddings. We have a tendency to then develop a feature-based matrix factorisation approach which might leverage the learnt user embeddings for cold-start product recommendation. Experimental results on an oversized dataset made from the biggest Chinese microblogging service SINA WEIBO and also the largest Chinese B2C e-commerce web site JINGDONG have shown the effectiveness of our planned framework.