Marketing E-Commerce by Social media using Product Recommendations and user Embedding

MarketingE-CommercebySocial media is the best way to improve marketing and business widely.The major issues faced with E-commerce and Social media interfacing is cold-start cross-site problem. The cold-start problem occurs at a situation when user is not having the history of purchase records.For the user who does not have a history of purchase records, we have introduced a method of finding the users' interested product without knowing any of the demographic information of the user. The product is recommended on basesof visits i.e., the item which is most likely to be visited by the users occur in the hit list. This product is rated at the top position for the users to purchase. The e-commerce with social media sites uses the strategy of user embedding and product recommendations. The product recommendations are achieved by incorporating LatentDirichlet Allocation(LDA), Re Ranking and Collaborative Filtering algorithms. The proposed framework can enhance the recommendation system by embedding products and users. This shows the potential of solving cold-start cross-site problem across the e-commerce and social media sites and enhances the marketing strategy.