Sentiment Analysis of Code-Mixed Social Media Text (SA-CMSMT) in Indian-Languages

The arrival of web 2.0 platforms and increasing usage of social networking sites have proliferated social media content on the web. These platforms also provides multilingual interface to allow people to write freely in their native language. Over the past few decades, a new phenomenon called code-mixing has been observed in social media data which has attracted attention of researchers in sociolinguists and Natural Language Processing domains. However, due to informal nature of the text present in code-mixing phenomenon, there are a number of challenges ranging from data extraction to summarization. Sentiment Analysis of code-mixed data is a key research field which has emerged in the recent past. NLP researchers aim to provide natural language processing (NLP) tools that can collect, analyzed, evaluate, and summarize code-mixed data. The researchers had to deal with dataset construction, preprocessing, annotation, language identification, feature extraction, feature selection, and sentiment classification when it came to sentiment analysis of Code-mixed Social Media Text. This paper provides an overview of work carried out in sentiment analysis of code-mixed Indian languages textual data.