Advanced Hybrid Information Processing. First International Conference, ADHIP 2017, Harbin, China, July 17–18, 2017, Proceedings

Research Article

Topic-Aware Influence Maximization in Large Recommendation Social Networks

Download
195 downloads
  • @INPROCEEDINGS{10.1007/978-3-319-73317-3_24,
        author={Jinghua Zhu and Qian Ming and Nan Wang},
        title={Topic-Aware Influence Maximization in Large Recommendation Social Networks},
        proceedings={Advanced Hybrid Information Processing. First International Conference, ADHIP 2017, Harbin, China, July 17--18, 2017, Proceedings},
        proceedings_a={ADHIP},
        year={2018},
        month={2},
        keywords={Influence maximization Topic-aware Recommendation social network},
        doi={10.1007/978-3-319-73317-3_24}
    }
    
  • Jinghua Zhu
    Qian Ming
    Nan Wang
    Year: 2018
    Topic-Aware Influence Maximization in Large Recommendation Social Networks
    ADHIP
    Springer
    DOI: 10.1007/978-3-319-73317-3_24
Jinghua Zhu1,*, Qian Ming1, Nan Wang1
  • 1: Heilongjiang University
*Contact email: zhujinghua@hlju.edu.cn

Abstract

Influence maximization () is a problem of finding several influential individuals in a social network so that their influence spread is maximized under certain propagation model. In recommendation social network such as Douban, information diffuses with multiple origins: internal and external influence. Furthermore, pairs of individuals usually have different influence strength on different topics, information, ideas and rumors etc. In this paper, we focus on the topic-aware problem for large recommendation social networks. We propose a novel TSID propagation model to formulate the multiple topics diffusion in recommendation social networks. We propose TIP algorithm to solve the influence maximization problem under TSID propagation model. Our experiment results show that TSID model can well depict the mix information propagation process in recommendation social network, the TIP algorithm has competitive response time and influence spread.