Adaptive Influence Maximization

Overview

Information diffusion and social influence are more and more present in today's Web ecosystem. Having algorithms that optimize the presence and message diffusion on social media is indeed crucial to all actors (media companies, political parties, corporations, etc.) who advertise on the Web. Motivated by the need for effective viral marketing strategies, influence estimation and influence maximization have therefore become important research problems, leading to a plethora of methods. However, the majority of these methods are non-adaptive, and therefore not appropriate for scenarios in which influence campaigns may be ran and observed over multiple rounds, nor for scenarios which cannot assume full knowledge over the diffusion networks and the ways information spreads in them.

In this tutorial we intend to present the recent research on adaptive influence maximization, which aims to address these limitations. This can be seen as a particular case of the influence maximization problem (where seeds in a social graph are selected to maximize information spread), one in which the decisions are taken as the influence campaign unfolds, over multiple rounds, and where knowledge about the graph topology and the influence process may be partial or even entirely missing. This setting, depending on the underlying assumptions, leads to variate and original approaches and algorithmic techniques, as we have witnessed in recent literature. We will review the most relevant research in this area, by organizing it along several key dimensions, and by discussing the methods' advantages and shortcomings, along with open research questions and the practical aspects of their implementation.

Instructors

Bogdan Cautis: LRI, Université Paris-Sud, France (bogdan.cautis@u-psud.fr)

Bogdan Cautis is a Professor at the Department of Computer Science of University of Paris-Sud, France, since September 2013. Before that, he was an Associate Professor at Telecom ParisTech, Paris (2007-2013). He received his PhD in 2007, from INRIA and University of Paris-Sud. His current research interests lie in the broad area of data management and data mining, with a particular focus on social networks and information diffusion.

Silviu Maniu: LRI, Université Paris-Sud, France (silviu.maniu@lri.fr)

Silviu Maniu is an Associate Professor in the Department of Computer Science of the University of Paris-Sud, France, since September 2015. Before that, he was an Postdoctoral Fellow at the University of Hong Kong (2012-2014) and a Researcher at Huawei Noah's Ark Lab (2014-2015). He received his PhD in 2012, from Telecom ParisTech. His research interests lie in the general area of graph data mining, with a focus on models and algorithms dealing with uncertain

Nikolaos Tziortziotis: Tradelab and LRI, Université Paris-Sud, France (ntziorzi@gmail.com)

Nikolaos Tziortziotis is a Data Scientist R&D at Tradelab Programmatic platform, Paris, France. Right before that, he was a postdoctoral researcher in the Department of Computer Science of the University of Paris-Sud, Orsay, France (Nov-Dec 2018). He was also a postdoctoral researcher at the Computer Science Laboratory (LIX), École Polytechnique, France (2015–2018). He received his PhD from the Department of Computer Science & Engineering of the University of Ioannina, Greece. His research interests span the broad areas of machine learning and data mining, with focus on reinforcement learning, Bayesian learning, influence maximization and real-time bidding.

Slides

[preliminary version]