Assessing the Effects of a Soft Cut-off in the Twitter Social Network Niloy Ganguly, Saptarshi Ghosh.

Slides:



Advertisements
Similar presentations
Resource Management §A resource can be a logical, such as a shared file, or physical, such as a CPU (a node of the distributed system). One of the functions.
Advertisements

Traffic-driven model of the World-Wide-Web Graph A. Barrat, LPT, Orsay, France M. Barthélemy, CEA, France A. Vespignani, LPT, Orsay, France.
Jure Leskovec, CMU Lars Backstrom, Cornell Ravi Kumar, Yahoo! Research Andrew Tomkins, Yahoo! Research.
Week 5 - Models of Complex Networks I Dr. Anthony Bonato Ryerson University AM8002 Fall 2014.
Comparison of Social Networks by Likhitha Ravi. Outline What is a social network? Elements of social network Previous studies What is missing in previous.
Community Detection Algorithm and Community Quality Metric Mingming Chen & Boleslaw K. Szymanski Department of Computer Science Rensselaer Polytechnic.
IEEE ICDCS, Toronto, Canada, June 2007 (LA-UR ) 1 Scale-Free Overlay Topologies with Hard Cutoffs for Unstructured Peer-to-Peer Networks Hasan Guclu.
Modeling and Analysis of Random Walk Search Algorithms in P2P Networks Nabhendra Bisnik, Alhussein Abouzeid ECSE, Rensselaer Polytechnic Institute.
Information Retrieval Lecture 8 Introduction to Information Retrieval (Manning et al. 2007) Chapter 19 For the MSc Computer Science Programme Dell Zhang.
Sogang University ICC Lab Using Game Theory to Analyze Wireless Ad Hoc networks.
CS 599: Social Media Analysis University of Southern California1 The Basics of Network Analysis Kristina Lerman University of Southern California.
Farnoush Banaei-Kashani and Cyrus Shahabi Criticality-based Analysis and Design of Unstructured P2P Networks as “ Complex Systems ” Mohammad Al-Rifai.
4. PREFERENTIAL ATTACHMENT The rich gets richer. Empirical evidences Many large networks are scale free The degree distribution has a power-law behavior.
The influence of search engines on preferential attachment Dan Li CS3150 Spring 2006.
1 Evolution of Networks Notes from Lectures of J.Mendes CNR, Pisa, Italy, December 2007 Eva Jaho Advanced Networking Research Group National and Kapodistrian.
TDTS21: Advanced Networking Lecture 8: Online Social Networks Based on slides from P. Gill Revised 2015 by N. Carlsson.
Directional triadic closure and edge deletion mechanism induce asymmetry in directed edge properties.
Masoud Valafar †, Reza Rejaie †, Walter Willinger ‡ † University of Oregon ‡ AT&T Labs-Research WOSN’09 Barcelona, Spain Beyond Friendship Graphs: A Study.
Observing Self-Organized Criticality - an approach to evolution dynamics By Won-Min Song.
CS 728 Lecture 4 It’s a Small World on the Web. Small World Networks It is a ‘small world’ after all –Billions of people on Earth, yet every pair separated.
Web as Graph – Empirical Studies The Structure and Dynamics of Networks.
CS Lecture 6 Generative Graph Models Part II.
Interconnect Implications of Growth-Based Structural Models for VLSI Circuits* Chung-Kuan Cheng, Andrew B. Kahng and Bao Liu UC San Diego CSE Dept.
Maximal Independent Set Distributed Algorithms for Multi-Agent Networks Instructor: K. Sinan YILDIRIM.
Computer Science 1 Web as a graph Anna Karpovsky.
Correctness of Gossip-Based Membership under Message Loss Maxim Gurevich, Idit Keidar Technion.
Tomo-gravity Yin ZhangMatthew Roughan Nick DuffieldAlbert Greenberg “A Northern NJ Research Lab” ACM.
Optimization Based Modeling of Social Network Yong-Yeol Ahn, Hawoong Jeong.
Analyzing the Vulnerability of Superpeer Networks Against Churn and Attack Niloy Ganguly Department of Computer Science & Engineering Indian Institute.
(Social) Networks Analysis III Prof. Dr. Daning Hu Department of Informatics University of Zurich Oct 16th, 2012.
Lecture 6 - Models of Complex Networks II Dr. Anthony Bonato Ryerson University AM8002 Fall 2014.
Modeling Information Diffusion in Networks with Unobserved Links Quang Duong Michael P. Wellman Satinder Singh Computer Science and Engineering University.
Exploiting Web Matrix Permutations to Speedup PageRank Computation Presented by: Aries Chan, Cody Lawson, and Michael Dwyer.
Developing Analytical Framework to Measure Robustness of Peer-to-Peer Networks Niloy Ganguly.
+ Offline Optimal Ads Allocation in SNS Advertising Hui Miao, Peixin Gao.
Author: M.E.J. Newman Presenter: Guoliang Liu Date:5/4/2012.
Niloy Ganguly Department of Computer Science & Engineering Indian Institute of Technology, Kharagpur Kharagpur Stability analysis of peer to peer.
Popularity versus Similarity in Growing Networks Fragiskos Papadopoulos Cyprus University of Technology M. Kitsak, M. Á. Serrano, M. Boguñá, and Dmitri.
Using Transactional Information to Predict Link Strength in Online Social Networks Indika Kahanda and Jennifer Neville Purdue University.
Percolation in self-similar networks Dmitri Krioukov CAIDA/UCSD M. Á. Serrano, M. Boguñá UNT, March 2011.
Theory of αBiNs: Alphabetic Bipartite Networks Animesh Mukherjee Dept. of Computer Science and Engineering Indian Institute of Technology, Kharagpur Collaborators:
Web Caching and Content Distribution: A View From the Interior Syam Gadde Jeff Chase Duke University Michael Rabinovich AT&T Labs - Research.
COLOR TEST COLOR TEST. Social Networks: Structure and Impact N ICOLE I MMORLICA, N ORTHWESTERN U.
Thang N. Dinh, Dung T. Nguyen, My T. Thai Dept. of Computer & Information Science & Engineering University of Florida, Gainesville, FL Hypertext-2012,
Part 1: Biological Networks 1.Protein-protein interaction networks 2.Regulatory networks 3.Expression networks 4.Metabolic networks 5.… more biological.
Project funded by the Future and Emerging Technologies arm of the IST Programme Analytical Insights into Immune Search Niloy Ganguly Center for High Performance.
Analyzing the Vulnerability of Superpeer Networks Against Attack Niloy Ganguly Department of Computer Science & Engineering Indian Institute of Technology,
Web Science & Technologies University of Koblenz ▪ Landau, Germany Micro-interactions and Macro-observations Deciding Between Competing Models Steffen.
Lecture 10: Network models CS 765: Complex Networks Slides are modified from Networks: Theory and Application by Lada Adamic.
Measuring Behavioral Trust in Social Networks
Most of contents are provided by the website Network Models TJTSD66: Advanced Topics in Social Media (Social.
SCALABLE INFORMATION-DRIVEN SENSOR QUERYING AND ROUTING FOR AD HOC HETEROGENEOUS SENSOR NETWORKS Paper By: Maurice Chu, Horst Haussecker, Feng Zhao Presented.
 Tree in Sensor Network Patrick Y.H. Cheung, and Nicholas F. Maxemchuk, Fellow, IEEE 3 rd New York Metro Area Networking Workshop (NYMAN 2003)
1 Epidemic Spreading Parameters: External Model based on population density and travel statistics.
Stefanos Antaris A Socio-Aware Decentralized Topology Construction Protocol Stefanos Antaris *, Despina Stasi *, Mikael Högqvist † George Pallis *, Marios.
1 An Arc-Path Model for OSPF Weight Setting Problem Dr.Jeffery Kennington Anusha Madhavan.
“MASHABLE” by Suman Kalyan Maity Dept. of CSE IIT Kharagpur CNeRG Retreat
Brief Announcement : Measuring Robustness of Superpeer Topologies Niloy Ganguly Department of Computer Science & Engineering Indian Institute of Technology,
1 Finding Spread Blockers in Dynamic Networks (SNAKDD08)Habiba, Yintao Yu, Tanya Y., Berger-Wolf, Jared Saia Speaker: Hsu, Yu-wen Advisor: Dr. Koh, Jia-Ling.
Project funded by the Future and Emerging Technologies arm of the IST Programme Search in Unstructured Networks Niloy Ganguly, Andreas Deutsch Center for.
Jure Leskovec, CMU Lars Backstrom, Cornell Ravi Kumar, Yahoo! Research Andrew Tomkins, Yahoo! Research.
Dynamic Network Analysis Case study of PageRank-based Rewiring Narjès Bellamine-BenSaoud Galen Wilkerson 2 nd Second Annual French Complex Systems Summer.
Lecture 23: Structure of Networks
A Study of Group-Tree Matching in Large Scale Group Communications
What Stops Social Epidemics?
On Growth of Limited Scale-free Overlay Network Topologies
Comparison of Social Networks by Likhitha Ravi
Lecture 23: Structure of Networks
Lecture 23: Structure of Networks
Modelling and Searching Networks Lecture 2 – Complex Networks
Presentation transcript:

Assessing the Effects of a Soft Cut-off in the Twitter Social Network Niloy Ganguly, Saptarshi Ghosh

Restrictions in OSNs Restrictions on the number of social links that a user can have  Hard cut-offs: 1000 in Orkut, 5000 in Facebook  Soft cut-off in Twitter Why restrictions?  Scalability issues: reduce strain on OSN infrastructure due to user-to-all-friends communication  Prevent indiscriminate linking by spammers

Need to study restrictions in OSNs Conjecture  Restrictions only affect spammers and very few hyper-active legitimate users Reality in today’s OSNs  Thousands of legitimate users are getting blocked  Restrictions being increasingly criticized by socially active and popular users Twitter imposed a soft cut-off that adapts to requirements of popular users

The soft cut-off in Twitter u → v: user u ‘follows’ user v Conjectured Twitter follow- limit (“10% rule” ): Restriction on out-degree based on in-degree Need at least 1820 followers to follow more than 2000 Soft cut-off: Can follow up to 110% of number of followers Details in WOSN 2010, Computer Communication 2012 …

Does the Twitter follow-limit really affect many users?

Empirical measurements on Twitter Several measurements before restriction was imposed (in August 2008) Publicly available crawl of entire Twitter network as in July 2009  41.7 million nodes  1.47 billion social links

Scatter plot of followers/following spread Reproduced from [Krishnamurthy, WOSN 2008] In Jan-Feb 2008, before restriction imposed (x, y) implies a user following x (out-degree) y followers (in-degree)

Scatter plot of followers/following spread Reproduced from [Krishnamurthy, WOSN 2008] In Jan-Feb 2008, before restriction imposed In Oct-Nov 2009, a year after restriction imposed

Degree Distributions In-degree distribution: power-law over a large range of in- degrees

Degree Distributions Out-degree distribution (right): sharp spike around out- degree 2000 due to blocked users

Objectives Develop an analytical model to predict effects of restrictions  Fraction of users likely to get blocked  Effects of varying linking dynamics Design restrictions balancing the two conflicting objectives  Desired reduction in system-load due to communication  Minimize dissatisfaction among blocked users

Directed Network growth model Model by [Krapivsky et. al., PRL 86(23), 2001] extended by incorporating restrictions Growth event 1 (with probability p)  new user u joins and ‘follows’ existing user v  v chosen preferentially on in-degree (popularity) Growth event 2 (with probability 1-p)  existing user u ‘follows’ another existing user v  u chosen preferentially on out-degree (social activity), v on in-degree

Growth model (contd.) N ij (t) : number of nodes having in-degree i, out-degree j at time t Change in N ij (t) due to change in in-degrees Change in N ij (t) due to change in out-degrees Details in Networking 2011…

Modeling restrictions β ij = 1 if users having in-degree i allowed to have out-degree j, 0 otherwise For a κ % Twitter follow-limit at out-degree s (κ = 10, s = 2000 in reality ) Model solved to derive closed-form expressions for degree distributions in presence of restrictions Details in Networking 2011 …

Predictions by the model Accurately matches degree distributions of Twitter OSN Explains decrease in power-law exponent of out-degree distribution in Twitter after imposing restriction

Predictions by the model (contd.) Fraction of users who are likely to get blocked  Varies inversely proportional to network density  Reduces rapidly as link-formation becomes more random (as opposed to preferential)  Power-law decrease with starting point of cut-off s  Parabolic increase with κ (κ % (1 + κ -1 ) rule in Twitter)

Objectives Develop an analytical model to predict effects of restrictions  Fraction of users likely to get blocked Design restrictions balancing the two conflicting objectives  Desired reduction in system-load due to communication  Minimize dissatisfaction among blocked users

Using model to design restrictions Utility function for restrictions  L : reduction in links (communication-overhead)  B : fraction of users blocked / dissatisfied  w u : importance of minimizing user-dissatisfaction (value decided by system engineers) Optimizing U helps fix values of parameters in the restriction function to balance both objectives U = L – w u B Details in ComCom 2012 …

Using model to design restrictions (contd.) What values of restriction parameters s, κ will maximize achieved utility U, for given w u ? Values for s,κ chosen by Twitter justified for w u = 50

Summary till now … First study of restrictions in OSNs  First attempt to theoretically model effects of soft cut-offs on network growth Soft cut-offs likely to be favored in OSNs over hard cut-offs  Can be applied in undirected OSNs (e.g. Facebook) by distinguishing initiator and acceptor of social links

Thank you

Contact: Complex Network Research Group (CNeRG) CSE, IIT Kharagpur, India Thank You