A repository of experiments with building a sentiment analysis model for Amazon reviews using PyTorch.
Amazon dataset retrieved from [1].
GloVe Common Crawl Word Embeddings retrieved from [2].
This project follows a modified Conventional Commits format for its commit messages (https://www.conventionalcommits.org/).
[1] Ni, J., Li, J., McAuley J. (2019). Justifying recommendations using distantly-labeled reviews and fined-grained aspects. Empirical Methods in Natural Language Processing (EMNLP). Retrieved from https://nijianmo.github.io/amazon/index.html.
[2] Pennington, J., Socher, R., and Manning, C. D. (2014). GloVe: Global Vectors for Word Representation. Empirical Methods in Natural Language Processing (EMNLP). 1532-1543. Retrieved from https://nlp.stanford.edu/projects/glove/.