Insta
IIT Kharagpur’s Neural Network, Deep Learning Algorithm To Help Manage Disasters Via Social Media Message Sorting
Swarajya Staff
Nov 21, 2018, 02:49 PM | Updated 02:49 PM IST
Save & read from anywhere!
Bookmark stories for easy access on any device or the Swarajya app.
A research team from the Department of Computer Science and Engineering, Indian Institute of Technology, Kharagpur (IIT-K) developed an algorithm for effective management of disaster time social-media posts in a bid to effectively manage the disaster, reports India Today.
The recent earthquakes in Nepal and floods in Kerala have proved the inevitability of social media platforms as they have become essential sources of real-time information regarding disasters. The victims, on-site volunteers and empathisers all of whom act as ‘social sensors’ are the ones that provide information.
However, the vast volume of flow and arbitrary nature of the content makes it challenging to locate relevant information. The social media posts often contain informal language without any grammar, arbitrary shortening of words and images.
The algorithm, therefore, attempts to sort the variegated, overwhelming number of social media posts during disasters. This can help real-time disaster-related news to reach the right places.
The algorithm developed by IITK can search for various social media platforms, especially micro-blogging sites, with broad access, and report on situational information.
The research team has employed ‘Neural Network and Deep Learning’ models to understand social media text which can often contain shortened words, images or informal information. It can also filter rumours and hate posts. The team is also developing web-based systems and mobile apps for aiding post-disaster relief operations.
Save & read from anywhere!
Bookmark stories for easy access on any device or the Swarajya app.
Support Swarajya's 50 Ground Reports Project & Sponsor A Story
Every general election Swarajya does a 50 ground reports project.
Aimed only at serious readers and those who appreciate the nuances of political undercurrents, the project provides a sense of India's electoral landscape. As you know, these reports are produced after considerable investment of travel, time and effort on the ground.
This time too we've kicked off the project in style and have covered over 30 constituencies already. If you're someone who appreciates such work and have enjoyed our coverage please consider sponsoring a ground report for just Rs 2999 to Rs 19,999 - it goes a long way in helping us produce more quality reportage.
You can also back this project by becoming a subscriber for as little as Rs 999 - so do click on this links and choose a plan that suits you and back us.
Click below to contribute.