Loading...

Modeling Information Cascade in Social Network with Positive and Negative

Shafaei, Mahsa | 2013

512 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 45527 (19)
  4. University: Sharif University of Technology
  5. Department: Computer Engineering
  6. Advisor(s): Jalili, Mahdi
  7. Abstract:
  8. Information cascade or affection in a broad social network is introduced as a dynamic epidemic phenomenon in the society. As people notify a new innovation, technology, or hobby, they try to share it with their friends, colleges or neighbors. Till now most of the cascade models are presented for unsigned network, in which all links have the same sign (such as friends and trusted networks). In these networks cascade is independent of the edge sign. But in reality signed networks are as common as simple networks. Thus, in this thesis, we study information cascade in networks with positive and negative edges. We link the cascade size to community structure of signed networks; communities are defined such that positive inter-community and negative intra-community links are minimized. The cascade is initialized from a number of nodes that are selected randomly. Finally, the number of nodes that have not participated in the cascade is considered as cascade depth; the less the number of such nodes, the more the depth of the cascade. We investigate influence of community structure (i.e., percentage of inter-community positive and intra-community negative links) on the cascade depth in both model networks and real signed networks. We find significant effect of community structure on cascade depth in both model and real networks. Our results show that the more the intra-community negative links (i.e., the worse the community structure), the more the cascade depth, i.e., the less the number of the nodes participating in the cascade. However, increasing the number of inter-community positive links have the opposite effect and decreases the cascade depth. On the other hand we investigate the effect of initial active node and payoff matrix on cascade depth
  9. Keywords:
  10. Information Diffusion ; Social Networks ; Positive and Negative Links ; Clusterd Network

 Digital Object List

 Bookmark

...see more