Yang, H., Zhu, C., Lin, L., & Yuan, P. (2017, July). Towards Truss-Based Temporal Community Search. In Conference'17 (pp. 1-8).
This paper addresses the limitations of existing community search methods in temporal networks that primarily rely on lower-order connectivity and often overlook higher-order temporal connectivity. The authors aim to develop an efficient and effective method for identifying higher-order temporal communities, specifically focusing on the truss model.
The authors propose a novel temporal community model called maximal-𝛿-truss (MDT) that emphasizes maximal temporal support, ensuring all edges are connected by a sequence of triangles with specific temporal properties. To efficiently search for the MDT containing a user-specified query node, they develop a two-pronged approach:
The proposed MDT model and associated algorithms provide an efficient and effective solution for identifying meaningful higher-order temporal communities in dynamic networks. The methods demonstrate superior performance compared to existing approaches, particularly in handling large-scale temporal graphs.
This research significantly contributes to the field of temporal network analysis by introducing a novel community model and efficient algorithms for community search. The proposed methods have broad applications in various domains, including social network analysis, recommendation systems, and anomaly detection.
While the truss model effectively captures higher-order structural cohesiveness, exploring more complex temporal motifs could potentially reveal even more meaningful community structures. Future research could investigate incorporating such motifs while maintaining computational efficiency. Additionally, extending the proposed methods to handle dynamic updates in evolving temporal networks is a promising direction for future work.
To Another Language
from source content
arxiv.org
Principais Insights Extraídos De
by Huihui Yang,... às arxiv.org 10-22-2024
https://arxiv.org/pdf/2410.15046.pdfPerguntas Mais Profundas