These materials are designed to teach students about the topic of fairness in machine learning. The materials are described in more detail in the SIGCSE paper: Towards Machine Learning Fairness Education in a Natural Language Processing Course, with a discussion on what went well and what can be improved in the future.
Lecture: Bias in NLP. pptx file
Activity 1 Exploring Vectors Representations for Words and Texts and their Similarities. Details are in the ipynb file.
Activity 2 Measuring Bias in BERT Contextualized Embeddings. Details are in the ipynb file.
Activity 3 Measuring Disability Bias in BERT. Details are in the ipynb file
Final Project: Details are in the ipynb file.
Related work: Teaching Accessibility in AI Courses
Parts of this work are supported by the SIGCSE Special Projects Grant award 56004A.