Projet de recherche BL/IN/08 (Action de recherche BL)
Objective: Online social networks (OSNs) like Twitter are increasingly being used to gather real-time information on events happening 'now', including disasters, emergency situations, and so on. The aim of this project is to use the tools of network science to characterize and model the information cascades appearing in Twitter in response to specific emergency events. We propose to (a) focus on emergency situations in countries such as India, where Twitter provides a post-crisis combat instrument (since usage of OSN is low) (b) as well as to explore the role of Twitter in European countries for providing real time information propagation for situation awareness.
Challenges: The major challenges in this direction are to (a) Extract and summarize important situational updates and news about an ongoing event from the large amounts of generic comments being posted on social media. (b) In emergency scenario, a lot of rumours / misinformation are posted which may lead to undesirable circumstances. Hence, guarding against the spread of such rumors is another important requirement. Identification of time-varying evolving communities of users who tweet about similar events (or rumour) would be in the core of these challenges.
Expected result: Modeling the dynamics of critical information diffusion in social media will lead to the development of cost-effective mechanisms so that the information can be better used by the government or disaster management groups to deal with the situation.
Added value: The Large Graphs and Network group of UCL Belgium,and Complex Network Research Group (CNeRG) of IIT Kharagpur has a long history of collaboration in the field of Network Science. The agenda of the two groups are similar; nonetheless there are few subtle differences. In DYSCO project, UCL has made important contribution in understanding the modular nature of dynamical and networked systems from the point of the control theory and optimization. On the other hand, IIT Kharagpur has developed a unique expertise in the field of computational social network. The proposed project, leveraging on their complementary expertise, will significantly help both the groups to contribute in their respective running projects. For instance, (a) timely identification of emerging communities in the fast evolving (tweet) network, will help in (b) identification of critical events in social media (say in Twitter) and subsequently (c) can be leveraged in modeling the information spread. This will definitely add a unique flavor to the DYSCO project and will directly benefit the DiSARM project of IIT Kharagpur.
Coordinateur Inde : Niloy Ganguly, Indian Institute of Technology, Kharagpur
