Analysis And Research Of User Online And Offline Social Networks For E-commerce Websites

With the rapid development of mobile Internet,more and more people choose to use mobile phones to browse information online,which has led to the explosive growth of network data.By the of rapid rise of artificial intelligence,using data mining and machine learning to mine information from these data can help to provide better serve to users.Social network is a theoretical construct to study relationships between individuals.It can improve performance of recommender system by mining the social relationship between people,thus achieving more accuracy of recommendation,and bringing users a better online experience.Firstly,this thesis introduces the background,research significance and organizational structure,and then introduces the general framework of the analysis,the data processing platform Spark,the data sets and data preprocessing.The latter part mainly carries on detailed construction and analysis to the social network from two angles,i.e.,online and offline.First,we use the user data of three e-commerce websites,namely Jingdong,Tmall and Taobao,to construct online social network.Relevant indicators of network are calculated and compared with those of the traditional network.From the perspective of products,multidimensional analysis has been carried out to enrich characteristics of products.Second,based on user's data of Jingdong e-commerce website,the offline social network has been constructed,to analyze mobile characteristics of users,and to explore the overlap between online and offline social networks.Finally,we further explore the extensibility of online and offline social networks,by using the characteristics of offline user and social network to predict whether users are friends online.Based on real users' browsing traffic data of e-commerce websites,the analysis and research of online and offline social networks are carried out,which plays an important role in describing the characteristics of users and studying the relationship between online and offline behaviors of users.