A Distributed Algorithm for Overlapped Community Detection in Large-Scale Networks
Dibakar Saha, Partha Sarathi Mandal

TL;DR
This paper introduces DOCD, a distributed algorithm for overlapped community detection in large-scale networks, addressing privacy and scalability issues by enabling in-network computation and communication.
Contribution
The paper presents a novel decentralized algorithm, DOCD, for detecting overlapped communities in large networks, matching centralized methods in accuracy while improving scalability.
Findings
DOCD achieves comparable modularity to centralized algorithms.
It efficiently identifies overlapped nodes and communities.
Requires few communication rounds for accurate detection.
Abstract
Overlapped community detection in social networks has become an important research area with the increasing popularity and complexity of the networks. Most of the existing solutions are either centralized or parallel algorithms, which are computationally intensive - require complete knowledge of the entire networks. But it isn't easy to collect entire network data because the size of the actual networks may be prohibitively large. This may be a result of either privacy concerns or technological impediments. Performing in-network computation solves both problems utilizing the computational capability of the individual nodes of the network. Simultaneously, nodes communicate and share data with their neighbors via message passing, which may go a long way toward mitigating individual nodes' privacy concerns in the network. All the aforementioned concerns motivated us to design a…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
