These teaching materials are used by the COST network OPINION.

We will use these materials during the COST Action Training School in Salamanca, Spain.

In this short one-hour session, we will mainly discuss the what machine learning is, how supervised and unsupervised methods differ, and how they can be relevant for opinion research. You can find the slides here.

In this this hands-on session, we will build our own classifiers. For this, we are essentially following Chapter 8 and parts of Chapter 11 of the book Computational Analysis of Communication by Wouter van Atteveldt, Damian Trilling, and Carlos Arcila Calderón. Materials:

• slides
• Full Chapter text with all code examples: Chapter 8, Chapter 11
• For R users:
• For Python users: