Large Movie Review Dataset

This dataset, titled Large Movie Review Dataset v1.0, contains 50,000 movie reviews annotated with binary sentiment polarity labels, serving as a standard benchmark for sentiment classification tasks. The dataset is evenly divided into 25,000 training samples and 25,000 testing samples, with a balanced distribution of positive and negative labels (25,000 each). Reviews are collected in a way that limits the number of reviews per movie to prevent label correlation and overfitting, ensuring no more than 30 reviews per movie and disjoint movie sets between training and testing. Positive reviews have scores \u2265 7\/10, while negative reviews have scores \u2264 4\/10, excluding neutral sentiments for clearer classification boundaries. An additional 50,000 unlabeled reviews are included for unsupervised learning purposes. This dataset is suitable for benchmarking binary sentiment analysis models in natural language processing (NLP), supporting both supervised and unsupervised learning methods.