SENTIMENT ANALYSIS OF YOUTUBE COMMENTS USING PYTHON NLP TOOLS

Abstract

With the rapid growth of video platforms such as YouTube, the comments left by viewers have turned into a valuable source of insights into public opinion and emotional expression. Understanding these sentiments is important for researchers, businesses, and content creators who want to gauge audience reactions and improve engagement. This paper provides an indepth review of different methods and tools used to analyze YouTube comments through Pythonbased Natural Language Processing (NLP) libraries. The focus is on categorizing user opinions into positive, negative, or neutral sentiments using tools such as TextBlob, VADER (Valence Aware Dictionary for Sentiment Reasoning), and spaCy. We also focus on the preprocessing methods required to handle noisy comment data and evaluate both the performance and accuracy of different techniques on real YouTube datasets. The analysis compares the accuracy and execution time of several Python-based NLP tools used for sentiment classification tasks. In conclusion, the study presents observations on key challenges, possible improvements, and future opportunities for advancements in multilingual sentiment analysis and context-aware systems