摘要
Exploration of influential people is really a hot issue for effective viral marketing these days. Prior studies overlooked to consider the influence of interaction between users and the authority of users during information dissemination. In this article, we proposed an innovative concept by employing communication strength and authority in order to explore the influential people. First, we calculated communication strength by three kinds of actions including retweets, comments and mentions. Second, we deliberated authority metric by employing total times getting re-tweeted, number of tweets and followers of a user. Then, we propose a novel algorithm called Influential People Label Propagation Algorithm(IPLPA) that considers the proposed metric to explore the influential people in micro-blog networks. We performed extensive experiments to measure the influence and rank of each person within micro-blog network. The comparative study presents that IPLPA depicted effective people while baseline algorithms retrieved low influenced people at top rank. Additionally, influence dissemination also measured for obtained influential people in order to validate proposed concept. The findings of this study would be useful for viral marketing and advertisement campaigns.
Exploration of influential people is really a hot issue for effective viral marketing these days. Prior studies overlooked to consider the influence of interaction between users and the authority of users during information dissemination. In this article, we proposed an innovative concept by employing communication strength and authority in order to explore the influential people. First, we calculated communication strength by three kinds of actions including retweets, comments and mentions. Second, we deliberated authority metric by employing total times getting re-tweeted, number of tweets and followers of a user. Then, we propose a novel algorithm called Influential People Label Propagation Algorithm（IPLPA） that considers the proposed metric to explore the influential people in micro-blog networks. We performed extensive experiments to measure the influence and rank of each person within micro-blog network. The comparative study presents that IPLPA depicted effective people while baseline algorithms retrieved low influenced people at top rank. Additionally, influence dissemination also measured for obtained influential people in order to validate proposed concept. The findings of this study would be useful for viral marketing and advertisement campaigns.
基金
supported in part by the following funding agencies of China:National Natural Science Foundation under Grant 61170274,61602050 and U1534201
