Recommendation Systems via Matrix Factorization

November 2020

This project was the second project for my machine learning class. We were given a dataset of several thousand single-sentence reviews collected from three domains: imdb.com, amazon.com, yelp.com. Each review consists of a sentence and a binary label indicating the sentence's emotional sentiment (1 for positive feelings; 0 for negative feelings). All the provided reviews in the training and test set were scraped from websites whose assumed audience is primarily English speakers. There are 2400 input, output pairs in the training set with 4510 unique words and 600 inputs in the test set with 1921 uniques words. Our main task is to develop a binary classifier that can correctly identify a new sentence's sentiment. Sentiment analysis is performed using three different models.

Project report