Information Technology Reference
In-Depth Information
In this paper we take the micro-blog for example, put forward the method of
evaluating node's influence of online social networks. This paper is organized as
follows: Section 2 we give an overview of related work. We introduce the basic idea of
our method in Section 3. Section 4 we introduce the method of this paper in a great
detail. The analysis of experiment in Section 5. We draw conclusions of this paper in
Section 6.
2
Related Work
In earlier research, the people according to the method of system science use indexes
like node degree, betweenness, closeness, information, eigenvector, the network
diameter and so on to measure the influence. In some recent studies of influence people
gradually combine the methods of social network analysis and methods of Internet
search [4].
There are a lot of study of micro-blog mainly focus on the Twitter. Efforts have
been made to evaluating influence of online social network [5-19]. Leila [5]
investigates the power of retweet mechanism and findings suggest that relations of
“friendship” at Twitter are important but not enough. Sun [6] proposes a graph model
to represent the relationships between online posts of one topic, in order to identify the
influential users.Jianshu Weng [7] proposed TwitterRank which measures the
influence taking both the topical similarity between users and the link structure into
account.Meeyoung analyzed Propagation characteristics of Twitter, micro-blog
forwarding and uses three parameters, by the study of a large number of Twitter data
[8]. So they found effects of the user in the topic in the process of communication. Pal
performed an extensive study about Twitter follower-following topology analysis [18].
Wu [10] utilize power multiplication iterative to calculate Markov matrix, by
optimizing and improving the PageRank algorithm. Yang [11] Starting from the two
angles of active users and blog quality, constructed the evaluation index of the blogger
influence, introduced the blogger communication ability factor, using the idea of
PageRank algorithm to design a new influence ranking algorithm to evaluate the
blogger influence.Guo [12] proposed the quantitative definition of user information
dissemination scope, and gives the method for calculating the influence.
3
The Basic Ideas of Algorithm
As we know, every micro-blog users in the network corresponding to individual or unit
of reality. User can enhance his own prestige by publishing micro-blog, forwarding
and commenting of others, concern for others. The attribute in the micro-blog included
two parts e.g. user attributes and micro-blog properties. The user itself includes the
user ID, user type, attention number, number of fans , number of micro-blog, number
of mentions, and micro-blog attributes including number of micro-blog, publish time,
the forwarding numbers, numbers of comments. Micro-blog network and online
community network, the user can according to their own preferences selectively use
"forward", "collection", "comment" on a piece of information or micro-blog do
corresponding operations.
Node's attribute is the basic characteristic of node. If the user measure the node's
influence only by itself or micro-blog attribute to measure the node's influence is
