IIT-KGP developing mechanism to extract info from social-media during disasters
(Sourced from PTI)
A research team of IIT Kharagpur has been developing a mechanism to filter critical situational information from social media during disasters from tonnes of other data. During a disaster, people such as victims and on-site empathizers flood social media platforms with real-time information, but the huge volume of posts makes it difficult for one to locate relevant messages from the spot, IIT-KGP said.
While the critical information is "less than 2%", the problem is compounded as the posts often contain informal-language without any grammar, arbitrary shortening of words and images. The research team from the Department of Computer Science and Engineering at the institute is developing the algorithms that would be able to search various social-media handles and give situational information from English and Hindi posts.
The algorithm can identify and extract critical, actionable information in real-time and in summarized form, lead researcher, Prof Saptarshi Ghosh, said. He added that they studied group posts exchanged among medical practitioners of an NGO on an internal messaging mobile application, during a three-month period following the 2015 Nepal earthquake. From these posts, they extracted detailed information on medical-requirements after an earthquake.
At present, the team is developing web-based systems and mobile apps for aiding post-disaster relief operations. For instance, a person sitting in the control room can get live updates about what resources are needed where, what resources are available and coordinate relief operations accordingly. A relief worker can also get updates on his or her smartphone about people being trapped in the vicinity.
The project has been jointly funded by IIT Kharagpur's Institute Scheme for Innovative Research and Development grant, Microsoft Research India and ITRA, Media Labs Asia and the Department of Electronics and Information Technology, Government of India, the IIT KGP statement said.