Unveiling Sarcasm in Hindi: Cutting-Edge Deep Learning Framework for Sarcasm Detection

Abstract

Opinion mining, also known as sentiment analysis (SA), is a pivotal computer-based approach for discerning and categorizing sentiments within textual data. With the burgeoning popularity of social media platforms and the proliferation of user-generated content in Hindi, there is an escalating need for proficient sentiment analysis techniques tailored specifically for this language. The internet's accessibility in regional languages has become imperative to accommodate a diverse user base irrespective of age or linguistic inclination. While a significant portion of SA research has been conducted in English, there exists a dearth of comprehensive studies focusing on Indian languages, notably Hindi. This paper addresses this gap by investigating sentiment analysis within the context of Hindi, delving into various sentiment categories including positive, negative, neutral, and the notably complex sentiment of sarcasm. Sarcasm, characterized by its subtle irony and ridicule, poses a formidable challenge in sentiment analysis, particularly in languages like Hindi, which boast intricate grammatical structures and nuanced contextual cues. This study aims to enhance sentiment analysis methodologies by unraveling the intricacies of sarcasm detection in Hindi textual data.