Abstract— Inrecent years, the bounds between e-commerce and social networking have turn outto be increasingly blurred. it is possible to access many E-commerce websitesby using users social network accounts like facebook, twitter etc. Users ofsocial networks can able to post their newly purchased products in the microblogs, and can give links to the E-commerce web pages from where they arepurchased. In this paper we have a tendency to propose a unique answer forcross-site cold-start product recommendation that aims to advocate product fromecommerce websites to users at social networking sites in “coldstart” things, ahaul that has seldom been explored before. A serious challenge is the way toleverage data extracted from social networking sites for cross-site cold-startproduct recommendation. This paper proposes, by usingneural networks extract user features or user embeddings and product feature orproduct embedding’s from the data collected from Ecommerce websites. Then byusing gradient boosting tree method on the social networking sites, collectuser features and combine this with the user embeddings. Then by using matrixfactorization method use these user embedding’s for the cold start productrecommendation.
Keywords:- e-commerce, product recommender, productdemographic, microblogs, recurrent neural networks