At a time when social media is playing a dominant role in our lives, the menace of fake news is also rising. To curb the same, IIT-Kharagpur has come up with a solution that uses artificial intelligence to extract critical information from viral social media content that is manually not possible.
IIT Kharagpur's newly developed technology has helped in detecting the authenticity of the post and pass on data to relief and rescue organisations.
Saptarshi Ghosh of Department of Computer Science & Engineering at IIT Kharagpur, told Economic Times, that how the technology can detect fake news and alert users about disasters through deep machine learning algorithms.
“We are developing AI methods to extract specific types of critical and actionable information from social media content posted during disasters such as resource needs, resource availability and information about trapped people,” said Ghosh.
Ghosh heads a team which develops algorithms for this project.
However, the solution only identifies fake news or rumours related to a disaster. Gosh claims that their solution is 90 per cent accurate in detecting the difference between fake and real news. Besides, it helps to glean information that can help disaster victims.
It has been found that on an average 2 per cent tweets have relevant information on the disaster, the rest of the tweets contain sympathetic and mostly, gibberish conversation.
How does the solution work?
Gosh explains, as a region gets hit by a calamity, the news usually goes viral by a tweet. So the computer reads out the empathetic posts and sends the relevant information to relief operators nearby.
The genesis of the project:
During the infamous 2015 Nepal earthquake, the institute began this project and worked for the 2016 Chennai floods as well. The project was jointly funded by IIT Kharagpur’s Institute Scheme for Innovative Research and Development grant, Microsoft Research India and ITRA, Media Labs Asia and Department of Electronics and Information Technology.
IIT-K along with its collaborators at Qatar Computing Research Institute (QCRI) has submitted a proposal to Microsoft Research Indi and if the proposal gets an acceptance, Kharagpur will collaborate with Microsoft Research and QCRI to develop the systems for post-disaster relief operations.
The institute is now aiming to develop a web-based system that will be based on their own algorithm perform tasks like identifying and extracting actionable information. In addition to this, a person sitting in a control room will get details on the requirement and availability of the resources and coordinate relief operation accordingly. Or a relief worker can be updated through smartphone about trapped people.