Social Aspects of Natural Language Processing
This interdisciplinary course provides an overview of current research at the intersection of natural language processing (NLP) and adjacent research areas that emphasize the social aspects of language, such as computational social science, cultural analytics, AI ethics, and HCI. Students will learn how NLP can help answer social scientific questions, and how social aspects of language are incorporated into NLP models. We’ll examine social issues and pitfalls within the field, and how we can use NLP methods to analyze and support communication and behavior.
This course is targeted towards graduate students and advanced undergraduates who have prior experience with NLP (e.g. word embeddings, topic modeling, large language models). Students will gain an in-depth understanding of on-going research questions, and conduct a hands-on project of their own design that intersects social phenomena with language data. Class time will be a mix of lectures, discussion, and collaborative project work.
Course Overview
- 1
- Introduction
- 2
- Data and model pitfalls
- 3
- “Good” research practices
- 4
- Computational social science
- 5
- NLP for conversations
- 6
- NLP for communities
- 7
- Computational sociolinguistics
- 8
- Misinformation, factuality, & toxicity
- 9
- NLP for literature & history
- 10
- NLP for education & political science
- 11
- Quantifying bias across domains
- 12
- Biases in the LLM pipeline
- 13
- HCI & NLP
- 14
- NLP & HCI