Description: This one-day workshop aims to help people from non-technical backgrounds to understand what happens when we apply artificial intelligence (AI) to language-related tasks. Short lectures and interactive activities will introduce participants to the general principles under the hood of AI algorithms. Then participants will apply what they learned by working with a simple computational language model, such as an automatic part-of-speech tagger for English. They will explore how the model performs when only a handful of examples are available to train it. Then they will try to improve it. The workshop assumes no background in computer science. It assumes participants have a basic understanding of statistics and some linguistic training, but it aims to be accessible to anyone interested in the topic.

Learning Objectives

• Differentiate: artificial intelligence, computational linguistics, natural language processing (NLP), machine learning, and deep learning or neural networks
• Explain simply modern AI’s key elements (annotation, data features, and testing) and formal foundation (how predictions are made)
• Know at least three language tasks that computers can perform
• Understand how these tasks are evaluated when done automatically
• Perform an error analysis of a simple model and see the impacts on the model’s performance

Speaker

• DIU alumna and former faculty member, Dr. Sarah Moeller
• Bio: sarahrmoeller.github.io