1 / 29

Sparsification of Influence Networks

Sparsification of Influence Networks. Michael Mathioudakis 1 , Francesco Bonchi 2 , Carlos Castillo 2 , Aris Gionis 2 , Antti Ukkonen 2 1 University of Toronto, Canada 2 Yahoo! Research Barcelona, Spain. Introduction. users perform actions post messages, pictures, videos

trista
Download Presentation

Sparsification of Influence Networks

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Sparsification ofInfluence Networks Michael Mathioudakis1, Francesco Bonchi2, Carlos Castillo2, Aris Gionis2, Antti Ukkonen2 1University of Toronto, Canada 2Yahoo! Research Barcelona, Spain

  2. Introduction users perform actions post messages, pictures, videos connected with other users interact,influenceeach other actionspropagate online social networks facebook750m users twitter 100m+users nice read indeed! 09:00 09:30

  3. Problem sparsify network eliminate large number of connections keep important connections sparsification: a data reduction operation network visualization efficient graph analysis • which connections are mostimportant • for the propagation of actions?

  4. What We Do technical framework sparsify network according to observed activity keep connections that best explain propagations our approach socialnetwork& observed propagations learnindependent cascade model (ICM) select k connections most likely to have produced propagations

  5. Outline • introduction • setting • social network • propagation model • sparsification • optimal algorithm • greedy algorithm: spine • experiments

  6. Social Network B A • users –nodes • B follows A – arcA→B

  7. Propagation of Actions independent cascade model propagation of an action unfolds in timesteps • users perform actions • actionspropagate I liked this movie great movie influence probability p(A,B) A B t t+1

  8. Propagation of Actions icm generates propagations sequence of activations likelihood B D p(A,B) active action α C A E not active t-1 t t+1

  9. Estimating Influence Probabilities social network + set of propagations max likelihood p(A,B) EM – [Saito et.al.] B D p(A,B) active action α C A E not active t-1 t t+1

  10. Outline • introduction • setting • social network • propagation model • sparsification • optimal algorithm • greedy algorithm: spine • experiments

  11. Sparsification social network p(A,B) k arcs set of propagations most likely to explain all propagations B p(A,B) A

  12. Sparsification social network p(A,B) k arcs set of propagations most likely to explain all propagations B p(A,B) A

  13. Sparsification not the k arcs with largest probabilities NP-hard and inapproximable difficult to find solution with non-zero likelihood

  14. How to Solve? brute-force approach try all subsets of k arcs? no break down into smaller problems combine solutions

  15. Optimal Algorithm sparsify separately incoming arcs of individual nodes optimize corresponding likelihood A B C kA + + kC kB = k • dynamic programming • optimal solution • however…

  16. Spine sparsification of influence networks greedyalgorithm efficient, good results two phases phase 1 try to obtain a non-zero-likelihoodsolution k0 < k arcs phase 2 build on top of phase 1

  17. Spine – Phase 1 phase 1 obtain a non-zero-likelihood solution select greedily arcs that participate in mostpropagations until all propagations are explained B social network action α A C B D A C t t+1 D B action β A C D

  18. Spine – Phase 2 add one arc at a time, the one that offers largest increase in likelihood submodular # arcs k0 k approximation guarantee for phase 2

  19. Outline • introduction • setting • social network • propagation model • sparsification • optimal algorithm • greedy algorithm: spine • experiments

  20. Experiments datasets meme.yahoo.com actions: postings (photos), nodes: users, arcs: who follows whom data from 2010 memetracker.org actions: mentions of a phrase, nodes: blogs & news sources, arcs: who links to whom data from 2009

  21. Experiments sampled datasets of different sizes YMemememe.yahoo.com MTrackmemetracker.org

  22. Experiments algorithms optimal (very inefficient) spine (a few seconds to 3.5hrs) by arc probability random

  23. Experiments

  24. Model Selection using BIC BIC(k) = -2logL + klogN

  25. Application spine as a preprocessing step influence maximization select k nodes to maximize spread of action [Kempe, Kleinberg, Tardos, 03] NP-hard, greedy approximation perform on sparsified network instead large benefit in efficiency, little loss in quality

  26. Application

  27. Public Code and Data http://www.cs.toronto.edu/~mathiou/spine/

  28. The End Questions?

More Related