A survey and comparative study on negative sentiment analysis in social media data

With the rapid growth of internet usage, especially on social media, forums, review platforms, and blogs, an enormous amount of data is being generated. This data often contains users' opinions, emotions, and arguments on various topics. To make informed decisions or predictions, it's crucial to analyze and organize this unstructured data effectively. Sentiment analysis of social media data has become essential, aiming to identify different forms of sentiments like hate speech, profanity, sentiment, and targeted insults. However, in the field of natural language processing (NLP), a significant challenge in sentiment analysis is the scarcity of labeled data. Researchers have traditionally used methods like lexicon-based and traditional machine learning approaches to process this unstructured social media data. Recent studies indicate that deep learning techniques have proven effective in handling this task. This study aims to provide a comprehensive overview of various classical machine learning and deep learning techniques employed in sentiment analysis. We explore different sentiment analysis categories and compare their performance using various evaluation metrics.