Publications

Abstract : Social media has become an important part of human life; not only has it become a platform for people to interact with each other, it has become the news portal, a stage for people to express themselves, and even for heinous tasks such as cyber bullying, stalking etc. India being a country with the world's second largest population, not to mention the lingual diversity, the usage of social media in all its forms, is at its heights. This resulted in the evolution of code-mixed data, which is a combination of more than one language. The Bengali-English code mixed data used in this work is provided by the NLP Tool Contest, SAIL @ ICON 2017. The Convolutional Neural Network has been used to classify the data as positive, negative or neutral. Later to analyze performance of the system in the native script of Indian languages, the same procedure has been applied on Telugu dataset which is created manually from various source of online movie reviews and the results are compared.