Sarcasm Detection for Hindi English Code Mixed Twitter Data

• Published in: International journal for research in applied science and engineering technology Vol. 11; no. 3; pp. 159 - 164
• Main Authors: Tejasvi, Koti, Reddy, Borra Richa, Reddy, Sheri Sukhjeevan, Rishikesh, Sirisilla
• Format: Journal Article
• Language: English
• Published: 31.03.2023
• ISSN: 2321-9653; 2321-9653
• DOI: 10.22214/ijraset.2023.49378

Twitter, Instagram have grown to be the two most popular social media sites for users to voice their opinions on a variety of subjects. The generation of such huge amounts of user data has made NLP activities like sentiment analysis and opinion mining far more important. Sarcastic statements on social media,also referred to as memes, have recently become very popular. Sarcasm challenges many NLP tasks because it flips the meaning and polarity of what the language implies. A lot of resources were developed for the English language, but this does not hold true for Hinglish In this paper we include a tweet corpus for training unique word embeddings as well as a Hinglish dataset labelled forsarcasm detection. Although there have been various attempts to categorise a text’s sentiment, there aren’t many models that can do the same when given non-English data that contains sarcasm or irony . This study compares numerous sarcasm detection methods for Hinglish data in order to determine which approach performs the best on datasets of various sizes and types. We have presented a technique that will enhance the outcomes of sarcasm recognition for Hindi- English code-mixed tweets by examining and researching the prior work.