180 likes | 358 Views
COP 5725 ADVANCED DATABASE SYSTEM. Minimum-sized Positive Influential Node Set Selection for Social Networks: Considering Both Positive and Negative Influences. By: Manik Saini (ms13ad). Motivation . What is a Social Network? . Motivation . What is a social Network?
E N D
COP5725ADVANCEDDATABASESYSTEM Minimum-sized Positive Influential Node Set Selection for Social Networks: Considering Both Positive and Negative Influences. By: Manik Saini (ms13ad)
Motivation • What is a Social Network?
Motivation • What is a social Network? • Set of Nodes(which can be individuals and organizations) that share a common interest or purpose. • Represented as graph of relationships and interactions among a group of individuals.
Background & Numbers[1] • 58 million tweets per day • Founded in 2004 • 1.31 billion users, 2014 [1] • May be the 2nd largest “country” • Around 5 billion pictures • 3000 pictures uploaded per minute
Word of Mouth&Spread of Influence • Medium as a spread of Influence • Information, opinion, ideas • Crowd Sourcing : Some 2.3 million people are taking part in a crowdsourcing appeal to find missing Malaysian Airlines flight MH370 [2] • Word of Mouth • Can be used as a source of advertising/marketing products , brands etc.
Example • Gambling insulators problem • Whom should I select? • Constraint : • Minimum nodes & Net effect positive • Find minimum-sized node (user) set in a social network that could positively influence on every node in the network
Applications • Social intervention program • Like Quit Smoking, Quit Gambling . . . • Promote new products • (give sample products to chosen ones) • Advertising (Viral Marketing) • Social recommendation • Expert finding
Comparison David • v/s Kevin George Shamik Bose
NetworkModel • A social network is represented as an undirected graph • Social influence represented by weights on the edges • Positive/Negative influence • Nodes start either active or inactive • An active node may trigger activation of neighboring nodes based on a pre-defined threshold θ
MPINSProblem • Given • a social network • a threshold θ • Goal • The initially selected active node set denoted by I could positively influence every other node in the network no less than threshold θ • Objective • Minimize the size of I • NP-Hard Problem
Terminology • Positive Influence • Negative Influence • Ultimate Influence • Contribution Function
Approach • Two Phase approach: • Maximal Independent Set (M) • Greedy algorithm
Performance • Simulation Settings • Generate random graph • The weighs on edges are randomly generated • Simulation Results • Very close to Optimal solution • No existing state of the art that’s finds MPINS • Comparison with PIDS (Positive Influential Dominating Set)
References • [1] http://www.statisticbrain.com/social-networking-statistics/ • [2] http://www.theguardian.com/world/2014/mar/14/tomnod-online-search-malaysian-airlines-flight-mh370 • [3] Maximizing the spread of influence through a social network, David Kempe, Jon Kleinberg, ÉvaTardos