SCIENCE@home

Overlapping Community Detection in Temporal Text Networks

Shuhan Yan,Yuting Jia,Xinbing Wang

Published 2021 in arXiv.org

ABSTRACT

Network is a powerful language to represent relational data. One way to understand network is to analyze groups of nodes which share same properties or functions. The task of discovering such groups is known as community detection. Generally, there are two types of information that can be utilized to fulfill this task, i.e., the link structures and the node attributes. The temporal text network is a special kind of network that contains both sources of information. Typical representatives include online blog networks, the World Wide Web (WWW) and academic citation networks. In this paper, we study the problem of overlapping community detection in temporal text network. We gather a large set of 32 temporal text networks with reliable groundtruth communities. By examining such networks, we find that a large proportion of edges connect two nodes which share no community in common. This scenario cannot be modeled by practically all existing community detection methods. Besides, we quantitatively analyze how node attributes help to improve the quality of detected communities and discover that nodes in the same community share similar textual contents. Motivated by these empirical observations, we propose MAGIC (Model Affiliation Graph with Interacting Communities), a generative model which captures community interactions and considers the information from both link structures and node attributes. Experimental results show that MAGIC achieves at least 40% relative improvements over 5 state-of-the-art methods in terms of 4 widelyused metrics. CCS Concepts •Data Mining → Graph Mining; •Networks → Network Clustering;

PUBLICATION RECORD

  • Publication year

    2021

  • Venue

    arXiv.org

  • Publication date

    2021-01-13

  • Fields of study

    Computer Science

  • Identifiers
    arXiv 2101.05137
  • External record

    Open on Semantic Scholar

  • Source metadata

    Semantic Scholar

CITATION MAP

EXTRACTION MAP

CLAIMS

  • No claims are published for this paper.

CONCEPTS

  • No concepts are published for this paper.

REFERENCES

Showing 1-37 of 37 references · Page 1 of 1

CITED BY