The Research of Big Data Correlation Based on Heterogeneous Network

Abstract

Nowadays, with the rapid development of network, the information of online big data becomes more and more complex. The heterogeneous and cyberization are the other names of the big data. Therefore, traditional methods of data correlation are not fit for to analyze and manage the data based on online user behaviors. However, there are many new questions with data correlation need to be concerned and solved. In this paper, we choose two kinds of methods of the data correlation to study. The correlation of keywords and the clustering rhythms are used to analyze the big data correlation based on the online user behaviors. We first analyzed the trends and networking characteristics of current information data development. With the analysis of current information data and the dimension of the relationship, we mined the network characteristic of information data. From the networking dimension of the data, based on the heterogeneous network and social network, we studied the proposal method of data correlation to solve the problems in clustering, recommendation, querying and prediction of the big data.