Title:	DSIN : DYNAMIC SELECTION OF INFLUENTIAL NODES IN SOCIAL NETWORKS USING NETWORK ATTRIBUTES
Author(s):	Myriam Jaouadi and Lotfi Ben Romdhane
ISBN:	978-989-8533-82-1
Editors:	Pedro Isaías and Hans Weghorn
Year:	2018
Edition:	Single
Keywords:	Dynamic Networks, Social Actions
Type:	Full Paper
First Page:	261
Last Page:	268
Language:	English
Cover:
Full Contents:	click to dowload
Paper Abstract:	Social networks have attracted many researchers due to the importance of data raised from them. The explosive growth of such data has taken a central place in social networks analysis. Talking about social network analysis, the influence maximization problem represents a fundamental issue that contributes to identify most influential users in on-line social networks. The measurement of influence refers to find users that make the most of information adoption. A considerable number of approaches have focused on how to maximize influence in static social networks. However, real social networks evolve, both structure and information change over time. Even some works have considered dynamic networks, they have focused on topological structure evolution and ignored information dynamics. In this paper, our challenge is how to mine and adapt the influencers in a dynamic environment where information changes over time that we call ‘Dynamic Influence Maximization’. In this work, information dynamics can be modeled by the set of social actions (e.g., 'like','share', 'comment', 'retweet') performed by users in a certain time window. In fact, we propose a parameterless model called DSIN (Dynamic Selection of Influential Nodes) that takes into account users' social actions to detect influential nodes. The main idea of our proposal is to consider the network’s attributes change in the influence maximization process. Experimental results on synthetic social networks demonstrate that DSIN is efficient for identifying influential nodes, compared with two newly known proposals.
	Go Back