May 12, 2025
We are offering a new course in the fall, focused on the use of natural language processing to education and education research. The course is designed to accommodate a broad range of student experience and expertise and to provide an opportunity to explore and apply NLP to unsupervised learning — with the goal of using NLP to understand what students say and how they learn.
If you have any questions about the course, please feel free to respond to this email.
Best,
Denise
400
598
EE598C/EE400A, Offered Autumn 2025
The Application of Natural Language Processing to Education
Tuesday, Thursday: 10:30-12:20
The Application of Natural Language Processing to Education
Tuesday, Thursday: 10:30-12:20
EE598C/EE400A introduces graduate and undergraduate students to the use of natural language processing (NLP), a subset of machine learning, to understand and classify text whose meaning and organization start out as “unknown” (i.e., unsupervised learning). Students will have the opportunity to learn how a human being analyzes such text using qualitative research methods such as thematic analysis and then compare the performance of various NLP-based approaches to traditional human-driven analysis. The course serves as an introduction to the application of NLP to the analysis of language and text in the context of education. How can NLP be used to assess student feedback? Student essays? Student exam questions? How can NLP be effectively used to support education and education research?
The focus of this course is on the appropriate selection and optimal application of NLP techniques rather than the underlying algorithms and mathematics, although a brief introduction to the latter will be provided for the unsupervised learning techniques covered in the course. These techniques include semantic similarity, probabilistic, matrix factorization, linear transformation, and deep learning approaches using large language models. Students will complete a project comparing natural language processing techniques of their choice to analyze a real (not simulated) dataset or text corpora. Project grading will consist of homework (five assignments, 30%), one midterm (30%), and an open-ended project to be completed individually or in groups (40%).
No pre-requisite knowledge of machine learning is required. Basic fluency and familiarity with Python or similar programming language is essential.
—
***************************************************
Denise Wilson (she/her), Professor
Electrical and Computer Engineering
University of Washington
Seattle, Washington 98195-2500
***************************************************