How to classify text using machine learning via Facebook's FastText library
In this Tech Talk, Michelle Scharfstein discusses how FastText, an open-source library created by Facebook AI Research, enables text classification through supervised learning. Michelle provides a high level overview of how text classification models are built and evaluated using FastText. She also takes a deeper look at some of the tricks used in FastText that improve the time complexity of model building compared to that of other prevailing techniques. Finally, Michelle shows how FastText can be used to identify spam in a sample inbox.
Project Members: Michelle Scharfstein