Analyzing the propagation of influence and concept evolution in enterprise social networks through centrality and latent semantic analysis

Authors:
Weizhong Zhu;Chaomei Chen;Robert B. Allen
Affiliations:
College of Information Science and Technology, Drexel University, Philadelphia;College of Information Science and Technology, Drexel University, Philadelphia;College of Information Science and Technology, Drexel University, Philadelphia
Venue:
PAKDD'08 Proceedings of the 12th Pacific-Asia conference on Advances in knowledge discovery and data mining
Year:
2008

Citing 7
Cited 1

The anatomy of a large-scale hypertextual Web search engine

WWW7 Proceedings of the seventh international conference on World Wide Web 7
Authoritative sources in a hyperlinked environment

Journal of the ACM (JACM)
Extracting reputation in multi agent systems by means of social network topology

Proceedings of the first international joint conference on Autonomous agents and multiagent systems: part 1
Algorithms for estimating relative importance in networks

Proceedings of the ninth ACM SIGKDD international conference on Knowledge discovery and data mining
Graph-based anomaly detection

Proceedings of the ninth ACM SIGKDD international conference on Knowledge discovery and data mining
Fast discovery of connection subgraphs

Proceedings of the tenth ACM SIGKDD international conference on Knowledge discovery and data mining
Visual analytics: Storylines: Visual exploration and analysis in latent semantic spaces

Computers and Graphics

Discovering target groups in social networking sites: An effective method for maximizing joint influential power

Electronic Commerce Research and Applications

Quantified Score

Hi-index	0.00

Visualization

Abstract

Understanding the propagation of influence and the concept flow over a network in general has profound theoretical and practical implications. In this paper, we propose a novel approach to ranking individual members of a real-world communication network in terms of their roles in such propagation processes. We first improve the accuracy of the centrality measures by incorporating temporal attributes. Then, we integrate weighted PageRank and centrality scores to further improve the quality of these measures. We valid these ranking measures through a study of an email archive of a W3C working group against an independent list of experts. The results show that time-sensitive Degree, time-sensitive Betweenness and the integration of the weighted PageRank and these centrality measures yield the best ranking results. Our approach partially solves the rank sink problem of PageRank by adjusting flexible jumping probabilities with Betweenness centrality scores. Finally the text analysis based on Latent Semantic Indexing extracts key concepts distributed in different time frames and explores the evolution of the discussion topics in the social network. The overall study depicts an overview of the roles of the actors and conceptual evolution in the social network. These findings are important to understand the dynamics of the social networks.