A retweet prediction method of micro-blog big data users based on Map/Reduce
Yuelong Zhao, Meng Fang · 2016
In all kinds of social network software, retweet is a common behavior and a important mechanism for information dissemination.Especially retweet prediction of microblog users is very important to deep research.However traditional method can't be effectively applied to big data.To solve this problem, in this paper, first study the most relevant features of retweet, such as proximity social network, retweet activity, etc.Second, use Map/Reduce programming framework to achieve the extraction of feature set and an improved random forests algorithm.Third, gives a distributed approach based on Hadoop platform and use this algorithm for parallel retweet prediction of user's concern edge.Finally, do some experiment by using real data sets of Sina's Micro-blog.Experiments show that this distributed approach based on hadoop platform is better than traditional design, it can effectively predict retweet of Micro-blog user in less time.