Data Analytics to Examine Spams


Industry experts help Indian and Japanese students to tackle Twitter spams with applications of data analytics

With the theme of ‘The Power Data Science for the Future’, Bitgrit organised the World Data Science Forum for students of data science and academicians from India and Japan. The event deliberated the future of the data science industry and discussed technology to solve real-world problems.

Platform for dialogue

The forum welcomed Indian and Japanese academicians, students and veterans from the IT, automotive, telecommunications and healthcare industries to discuss how innovations within the IoT, AI and the blockchain might foster disruption for individuals, businesses and institutions in the future.

While the forum was a platform for academicians and industry experts to share their knowledge and insights with young data scientists, it also hosted students from IIT Delhi and the University of Tokyo in an assignment to test their analytical capabilities.

Students inspect spammers

The participating students were engaged in an assignment, which was an attempt to tackle Twitter spam. Students were assigned the task of distinguishing spam tweets from real ones using data analytics. A set of 10,000 anonymous Twitter users were ordered by 50 parameters.

Students delineated the task with the unlabelled data causing them to use unsupervised and semi-supervised learning algorithms to identify attributes used by spammers. Consequently, they categorised users into spammers and non-spammers.

Holistically, the forum allowed a network for data scientists and students to form workable approaches using combinations of different methods. This may cultivate into a platform that utilises optimal application of data science and AI with societal contexts.