We analyse popular data set IMDB movies reviews 50K for sentiment analysis. We explore several methods and compare them by accuracy and f1_score
To re-run notebook it is required to download the dataset (IMDB Dataset.csv) from kaggle to the working directory. Glove vectors also needed to be downloaded from here: https://nlp.stanford.edu/data/glove.6B.zip and unzipped to the imdbdata folder.