ABSTRACT
Analysis of sentiments is an opinion mining technique that focuses on extraction of emotions of people towards a specific topic using structured or unstructured data. A lot of research work happens in text mining and natural language processing in recent times. It has an extensive range of applications since most of the activities happening today revolve around the opinions and feedback of people. Handling such immeasurable data manually is highly impossible and everyone finds the necessity of sentiment analysis using a deep learning algorithm. In a world where numerous movies of different genres are released daily, people cannot afford their time trying to figure out the best movie. A sentiment classification system built using deep learning would help in the movie recommender system and also assist people in classifying movies based on their reviews. To implement this, two deep learning algorithms namely CNN and RNN and its variant named LSTM are used, their performance is compared and the algorithm that gives better accuracy for the sentiment analysis is determined.
