47
Lecture 4 Influence Maximization Ding-Zhu Du University of Texas at Dallas

Lecture 4 Influence Maximization Ding-Zhu Du University of Texas at Dallas

Embed Size (px)

Citation preview

  • Slide 1
  • Lecture 4 Influence Maximization Ding-Zhu Du University of Texas at Dallas
  • Slide 2
  • Outline Kate Middleton effect Submodular Function Max Independent Cascade 2
  • Slide 3
  • The trend effect that Kate, Duchess of Cambridge has on others, from cosmetic surgery for brides, to sales of coral-colored jeans. Kate Middleton effect Kate Middleton effect 3
  • Slide 4
  • According to Newsweek, "The Kate Effect may be worth 1 billion to the UK fashion industry." Tony DiMasso, L. K. Bennetts US president, stated in 2012, "...when she does wear something, it always seems to go on a waiting list." Hike in Sales of Special Products 4
  • Slide 5
  • Influential persons often have many friends. Kate is one of the persons that have many friends in this social network. For more Kates, its not as easy as you might think! How to Find Kate? 5
  • Slide 6
  • Given a digraph and k>0, Find k seeds (Kates) to maximize the number of influenced persons. Influence Maximization 6
  • Slide 7
  • 7 Theorem Proof
  • Slide 8
  • Outline Kate Middleton effect Submodular Function Max Independent Cascade 8
  • Slide 9
  • What is a submodular function? Consider a function f on all subsets of a set E. f is submodular if
  • Slide 10
  • Max Coverage Given a collection C of subsets of a set E, find a subcollection C of C, with |C|63%) of the number of nodes that any size-k set could activate.
  • Slide 32
  • Proof of Submodularity 32
  • Slide 33
  • 33
  • Slide 34
  • Decision Version of InfMax in IC 34 Theorem Corollary Is it in NP?
  • Slide 35
  • 35 Theorem (Chen et al., 2010) Proof
  • Slide 36
  • 36
  • Slide 37
  • Disadvantage Lack of efficiency. Computing m (S) is # P-hard under both IC and LT models. Selecting a new vertex u that provides the largest marginal gain m (S+u) - m (S), which can only be approximated by Monte-Carlo simulations (10,000 trials). Assume a weighted social graph as input. How to learn influence probabilities from history?
  • Slide 38
  • Monte-Carlo Method 38 Buffon's needle
  • Slide 39
  • Research done by our group in UTD 39
  • Slide 40
  • Zaixin Lu, Wei Zhang, Weili Wu, Bin Fu, Ding- Zhu Du: Approximation and Inapproximation for the Influence Maximization Problem in Social Networks under Deterministic Linear Threshold Model. ICDCS Workshops 2011: 160-165ICDCS Workshops 2011 40
  • Slide 41
  • Zaixin Lu, Lidan Fan, Weili Wu, Bhavani Thuraisingham and Kai Yang, Efficient influence spread estimation for influence maximization under the linear threshold model, Computational Social Networks, 1 (2014) 41
  • Slide 42
  • Wen Xu, Zaixin Lu, Weili Wu, Zhiming Chen: A novel approach to online social influence maximization. Social Netw. Analys. Mining 4(1) (2014) 42
  • Slide 43
  • Editor-in-Chief: Ding-Zhu Du My T. Thai Computational Social Networks 43 A New Springer Journal Welcome to Submit Papers
  • Slide 44
  • Yuqing Zhu, Zaixin Lu, Yuanjun Bi, Weili Wu, Yiwei Jiang, Deying Li: Influence and Profit: Two Sides of the Coin. ICDM 2013: 1301-1306 44
  • Slide 45
  • Lidan Fan, Zaixin Lu, Weili Wu, Yuanjun Bi, Ailian Wang: A New Model for Product Adoption over Social Networks. COCOON 2013: 737-746COCOON 2013 45
  • Slide 46
  • Songsong Li, Yuqing Zhu, Deying Li, Donghyun Kim, Huan Ma, Hejiao Huang: Influence maximization in social networks with user attitude modification. ICC 2014: 3913-3918ICC 2014 46
  • Slide 47
  • THANK YOU!