The raw data is hosted by Stanford University and is also available on Kaggle . IMDb Sentiment Analysis Using Naive Bayes - IJFMR
This paper introduced a dataset of , specifically balanced with 25,000 positive and 25,000 negative samples. It has since become the benchmark for testing various machine learning and deep learning models, including: jada-imdb
Andrew L. Maas, Raymond E. Daly, Peter T. Pham, Dan Huang, Andrew Y. Ng, and Christopher Potts. The raw data is hosted by Stanford University
Advanced models like CNNs , LSTMs, and Transformers are frequently tested on this dataset. specifically balanced with 25