A System for Identifying and Visualizing Influential Communities

TL;DR

This paper presents a system that identifies and visualizes the most influential communities in a co-author network based on impact metrics like citations, enabling users to explore research communities interactively.

Contribution

It introduces a novel algorithm for finding top influential communities and develops a visualization tool for exploring these communities in social networks.

Findings

Effective identification of top influential communities

Interactive visualization of community details

Ability to search by research fields

Abstract

In this paper, we introduce the concept of influential communities in a co-author network. We term a community as the most influential if the community has the highest influence among all other communities in the entire network. Influence of a community depends on the impact of the contents (e.g., citations of papers) generated by the members of that community. We propose an algorithm to identify the top K influential communities of an online social network. As a working prototype, we develop a visualization system that allows a user to find the top K influential communities from a co-author network. A user can search top K influential communities of particular research fields and our system provides him/her with a visualization of these communities. A user can explore the details of a community, such as authors, citations, and collaborations with other communities.