Text analysis: classification and topic modeling

Overview

• Introduce supervised text classification
• Implement a tidymodels workflow using text features
• Define topic modeling
• Explain Latent Dirichlet allocation and how this process works
• Demonstrate how to use LDA to recover topic structure from an unknown set of topics
• Identify methods for selecting the appropriate parameter for $k$

Before class

• Read Supervised classification with text data
  • For greater depth of coverage, read Classification in Supervised Machine Learning for Text Analysis in R
• Read chapter 6 in Tidy Text Mining with R
  • Topic modeling from the lecture notes demonstrates how to implement this in a (semi)-tidy workflow

Class materials

• Predicting song artist from lyrics
• Text analysis: topic modeling

What you need to do after class

• Complete the text analysis homework