Title: Virality Prediction in Social Networks leveraging Network characteristics

Speaker: Siddharth Bora

Abstract: Online Social networks face a plethora of topics that vary in their spread in the network. Among these there are certain topics that spread to a very large audience - the ones we term as "viral". A question that now arises is that are there certain states in the network that a viral topic passes through ? Can we identify such states or instances and be able to predict that a topic floating in an online network will go viral in the coming future ? In this talk, we would discuss how the early spreading pattern of a topic would influence its virality. Also, what role would communities in such a network play in spreading of the topic. Further what other information can be leveraged that could aid in identifying topics that exhibit such unusual behavior - this would encompasses discussing about geographies, community concentration, etc.