Sentiment Analysis Using VADER and Logistic Regression Techniques

A prominent subject named sentiment analysis in natural language processing aims to detect the sentiment or emotion contained in a text This paper aims to compare the efficacy of two approaches in order to analyze sentiment: VADER and logistic regression. VADER is a lexicon-based method that uses a predefined set of guidelines to analyze the sentiment of a text, while Logistic Regression is a machine learning-based approach that relies on training data to learn how to classify the sentiment of a text. To evaluate the performance of these approaches, they were applied to a dataset of tweets about airlines and compared the results in terms of accuracy. The results showed that Logistic Regression outperformed VADER on this dataset, achieving an accuracy of 79%. These results suggest that Logistic Regression may be a more effective approach for sentiment analysis in the context of airline tweets.