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

Research Article

Combinational Meta-paths Mining for Correlation Relationship Evaluation in Bibliographic Networks

Download
9 downloads
  • @INPROCEEDINGS{10.1007/978-3-030-00916-8_60,
        author={Qinchen Wu and Peng Wu and Li Pan},
        title={Combinational Meta-paths Mining for Correlation Relationship Evaluation in Bibliographic Networks},
        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={Heterogeneous information networks Meta-paths Correlation relationships Node-pairs proximity},
        doi={10.1007/978-3-030-00916-8_60}
    }
    
  • Qinchen Wu
    Peng Wu
    Li Pan
    Year: 2018
    Combinational Meta-paths Mining for Correlation Relationship Evaluation in Bibliographic Networks
    COLLABORATECOM
    Springer
    DOI: 10.1007/978-3-030-00916-8_60
Qinchen Wu1, Peng Wu1, Li Pan1,*
  • 1: Shanghai Jiao Tong University
*Contact email: panli@sjtu.edu.cn

Abstract

Correlation relationships between objects are pervasive in heterogeneous information networks such as bibliographic networks, which made it possible to evaluate proximity between nodes from different perspectives. To explain these semantically rich correlations, meta-paths formed by interconnected node types and edge types have been widely used. This means, using meta-paths and their combinations we can explicitly evaluate relationships between nodes, and thus made it possible to search for proximate nodes according to specific correlations they carried. In this paper, we propose a combinational meta-paths mining algorithm to evaluate correlation relationships between nodes in bibliographic networks. Experiments with bibliographic networks have proved its effectiveness with respect to prior knowledge based results.