Recommendation of online products using microblogging information in social media

Abstract

(Please read carefully abstract of the template). With the usage of social network like Facebook or Twitter sign-in’s as e-commerce sign-in where user are enabled to post on micro blogs linking to the e-commerce webpages, this mechanism has weakened the restrictions between e-commerce and social networking. A process of referring products to users of social networking by E-commerce sites in Cold-Start situations which is very rare is proposed where the question for knowledge isolation becomes the issue. This problem can be overcome by linking social networking and E-commerce pages as a bridge using features signs and information from neural networks adopting an improved ascent helping Tree method to move social networking users as user embedding’s. This can be done by employing a matrix factorization and trial done on Chinese micro-blogging service SINA WEIBO and Chinese B2C E-commerce website JINGDONG show that the proposed method is efficient.