Analysis and Classification of Tweets for Sentiment Analysis on Twitter Dataset

Sentiment analysis on Twitter helps us to take public opinion on diverse topics and present-day situations. This work is used to compare conventional methods, which gain knowledge of Methods with advanced Techniques for analysing sentiment in tweets. Traditional and Popular methods for analysis are probabilistic class, logistic regression and then we will also study about the advanced models like transformer-based strategies. These advanced models use special techniques to extract significant functions from textual content, assisting to investigate feelings, evaluations, emotions, and more successful stories. In this work, We will discuss about the challenges of sentiment analysis on Twitter, which includes informal language, slang, and the noisy nature of social media records. To compare those methods, we use datasets containing extraordinary types of tweets, which include reviews, news, and product comments. Our proposed methodology is a hybrid approach which uses RNN, LSTM and Bi-LSTM.our results provide an in-depth evaluation of hybrid approach with the existing approaches and results shows that this provides a best accuracy and processing performance in real-world scenarios.