Text classification depends on how language is represented and how a model learns from those representations. This course examines feature extraction for text and the evaluation of classification systems, including cases where some categories occur much more frequently than others. You will interpret accuracy, precision, and recall and consider why a single measure may give an incomplete picture of performance.
You will then explore neural approaches to text classification, including deep learning concepts, the structure of feedforward neural networks, and how neural networks learn. By the end of the course, you will be able to extract features for classification, interpret evaluation results on imbalanced data, and explain how a feedforward neural network supports text classification.