Collaborative Computing: Networking, Applications and Worksharing. 13th International Conference, CollaborateCom 2017, Edinburgh, UK, December 11–13, 2017, Proceedings

Research Article

Ranking the Influence of Micro-blog Users Based on Activation Forwarding Relationship

Download
135 downloads
  • @INPROCEEDINGS{10.1007/978-3-030-00916-8_36,
        author={Yiwei Yang and Wenbin Yao and Dongbin Wang},
        title={Ranking the Influence of Micro-blog Users Based on Activation Forwarding Relationship},
        proceedings={Collaborative Computing: Networking, Applications and Worksharing. 13th International Conference, CollaborateCom 2017, Edinburgh, UK, December 11--13, 2017, Proceedings},
        proceedings_a={COLLABORATECOM},
        year={2018},
        month={10},
        keywords={Activation forwarding Random forest Independent cascade model Micro-blog},
        doi={10.1007/978-3-030-00916-8_36}
    }
    
  • Yiwei Yang
    Wenbin Yao
    Dongbin Wang
    Year: 2018
    Ranking the Influence of Micro-blog Users Based on Activation Forwarding Relationship
    COLLABORATECOM
    Springer
    DOI: 10.1007/978-3-030-00916-8_36
Yiwei Yang,*, Wenbin Yao,*, Dongbin Wang1
  • 1: Beijing University of Posts and Telecommunications
*Contact email: yiwei_yang@163.com, yaowenbin_cdc@163.com

Abstract

How to predict the influence of users in micro-blog is a challenging task. Although numerous attempts have been made for this topic, few of them analyze the influence of users from the perspective filtration mechanism. In this paper, we propose a novel Activation Forwarding Relationship Independent Cascade algorithm for analyzing the influence of users. The algorithm mainly consists of two parts: forwarding prediction and activation process. We predict the forwarding relationship by Random Forest (RF) and improve the Independent Cascade algorithm to construct an activation network. The algorithm can filter non influence users during the construction of the activation network, thus reducing the amount of ranking time. By calculating the user’s activation capability, we rank user’s influence. The experimental results show that our algorithm can achieve 95% accuracy in predicting forwarding relationships. Besides, our algorithm not only saves computing time, but also shows that the Top-10 users in the ranking list have better ability to spread information than the existing ranking algorithms.