Natural Language Processing

3 credits · Graduate School

This course is colloquially titled “from tokens to transformers” and will focus on using computers to automatically analyze language data for linguistic features. The goal of the course is to provide students with the background and computing skills necessary to independently analyze and assess language in naturally occurring text. The course is broken into three parts with the first part focusing on classic natural language processing (NLP). At the lexical level, this will include assessments of words and phrases for features related to complexity, semantics, and sentiment. At the syntactic level, analyses will examine part-of-speech tagging and dependency parsing for annotating phrasal and clausal complexity. The second part of the course will focus on early word embedding approaches to capture meaning. This will include identifying key words and terms, topic modeling, and identifying semantic relationships between words using high dimension vectors. The final part of the class will introduce transformed-based approaches for language modeling including pre-trained transformers, fine-tuned transformers, sentence transformers, and prompt engineering. The course will be taught using the Python programming language and appropriate Python packages (e.g., Pandas, spaCy, PyTorch, Gensim, Transformers). The course will be practice-based, and the course assignments will require students to analyze language data appropriate to their field of interest, discuss the data from an analytical language perspective, and model and interpret outcomes. SPRING [3]