This was created for Ryerson Capstone Project.
The problem was given to classify the movies review phrases into 5 levels of sentiment (Negative, Somewhat negative, Neutral, Somewhat Positive and Positive) and determine the best models after training a collection of phrases with prelabelled sentiment scores.
The dataset was pulled from one of Kaggle competitions. The link is https://www.kaggle.com/c/movie-review-sentiment-analysis-kernels-only。
Feel free to use it for your own practices.