Varanasi: Scientists from the Department of Computer Science and Engineering, Indian Institute of Technology (IIT) BHU have developed an efficient Artificial intelligence-based solutions blow abusive content exist social media platform.
This innovative research lays the foundation for the development of more precise and dedicated solutions. cyberbullying India’s large and diverse digital community.
This research has been published in the prestigious research journal “Language Resources and Assessment” in January 2024.
Ravindranath Chowdary C., associate professor in the Department of Computer Science and Engineering, said that India is home to the second largest digital population in the world, and activities on social media platforms (SMP) such as Facebook and X are very active.
“While these platforms have become an integral part of our daily lives, they also expose users to cyberbullying. A unique challenge in the Indian context is that user content is often multilingual, ranging from Hindi to English. language,” he said, adding that identifying abusive content is already a challenge in a single language, but if Mixed languages text.
Chaudhary said that despite significant efforts to address cyberbullying in high-resource languages such as English, mixed-language environments have received little attention.
Under the guidance of Chowdary, research scholar Paras Tiwari conducts an in-depth analysis of the complexity of Sanskrit-Roman hybrid texts and proposes a cost-effective method for collecting and annotating misused textual instances of code-mixing, relevance scores is 20.38%. The study produced a dataset that is eight times larger than existing state-of-the-art datasets. Furthermore, their work proposes effective solutions using traditional machine learning technology and advanced pre-trained large language models. Additionally, their work includes identifying hate speech and predict toxicity before it affects SMP users. Their research makes a significant contribution to improving the security of SMP for India’s diverse user base.
Professor Amit Patra, Director of the Institute, congratulated Ravindra Chowdary C. and his team for this pioneering research.
