Sentiment Analysis on IMDb Movie Reviews Using Machine Learning - Logistic Regression

---

Abstract

Sentiment Analysis deals with handling large number of honest reviews given by the consumers and categorizing them into specific class labels. It helps the company or the brand to know if their product is useful to the public. This paper aims to use sentiment analysis for Movie Reviews. IMDb is one such popular and trustworthy platform for information on movies. It analyzes the reviews provided by the viewers which helps people to decide if the movie is worth watching. In this paper, Natural Language Processing (NLP), sklearn and Logistic regression tools are used to identify and examine the sentiments of the IMDb movie reviews to train the model. The model performance is studied in terms of accuracy score. It has achieved a good accuracy of 90.064%. Hence it is believed to have good potential in analyzing customer feedback, product reviews and survey responses. It is also expected to perform better if integrated with Deep Learning models.

---

Keywords

IMDb Movie Review; Long shortterm memory; Machine Learning; Natural Language Processing; Opinion mining; Sentiment Analysis; Term Frequency-Inverse document Frequency (TF-IDF)