WhatsFarzi, a new app aims at detecting fake news at the end users. In order to fight fake news, a group of students from Delhi have launched a new app which allows users to analyse authenticity of a message in text or image form.
Three BTech students of Indraprastha Institute of Information Technology, Delhi (IIIT-Delhi or IIIT-D) Madhur Tandon, Suryatej Reddy Vyalla and Dhruv Kuchhal have developed this app, which will help end users to check the authenticity of news spreading on Facebook and WhatsApp.
The newly developed app named ‘WhatsFarzi‘ was made under the supervision of Ponnurangam Kumaraguru, associate professor, department of computer science, which enables people to check the authenticity of the content, message or an image on social media.
All you need to do is download ‘WhatsFarzi‘ from Google playstore. To check the authenticity of a viral text or image getting, one needs to copy and paste it in the designated space in the app. The app will then rate the content on the scale of fake to authentic through a coloured bar.
The team combined the knowledge of graphics, natural language processing, machine learning / AI and Graphics Processing Unit (GPU) to analyse the authenticity of news content and images. The app, however, is not equipped to check the authenticity of video messages. “It will help the end users, as it will equip them to check the authenticity of content circulated on WhatsApp,” said Ponnurangam Kumaraguru, the principal investigator of the project.
Before WhatsFarzi, a similar initiative to analyse the authenticity of news on Twitter and Facebook was tested by the PhD students of the college.
“We analysed Hurricane Sandy and Boston blasts in detail; using this understanding, we developed TweetCred, a Chrome browser extension and a REST API. Building upon the success of TweetCred, a PhD student who graduated in 2017 extended TweetCred for Facebook,” said Ponnurangam while talking to Education Times.
“Collecting enormous data in the form of fake content, image analysis were among the major challenges faced by the team while developing the app,” informed IIIT-D professor. “We also analysed Paris attacks in 2015 based on the visual theme, embedded text, and the sentiment of images. Using machine learning models, we built a Chrome and Firefox browser plugin called Facebook Inspector (FbI). With these understanding and seeing the lynching news and fake content spread on WhatsApp, we started working on WhtasFarzi app from August 2018 and now we launched WhatsFarzi,” added Ponnurangam.
“We have received a good response after launch of the app on February 11. We hope this app will help the users to differentiate between the genuine and fake information floating on the internet,” added Ponnurangam.
Courtesy: TOI