Methods Fellow Workshop: Tracing meaning through time with digital methods: how did Darwin’s language evolve? New
The aim of this course is to support students, researchers, and professionals interested in exploring the changing nature of the English vocabulary in historical texts at scale, and to reflect critically on the limitations of these computational analyses. We will focus on computational methods for representing word meaning and word meaning change from large-scale historical text corpora. The corpus used will consist of Darwin’s letters from the (Darwin Project https://www.darwinproject.ac.uk/) at Cambridge University Library. All code will be in online Python notebooks.
If you are interested in attending this course, please fill in the application form
Postgraduate students and staff at the University of Cambridge, particularly in the arts, humanities, social sciences, libraries, archives, and history of science.
Prior knowledge of Python (and specifically text processing with Python) is desirable.
If you are unsure whether your Python experience is sufficient, please apply anyway and we can talk about it together.
Number of sessions: 4
# | Date | Time | Venue | Trainer |
---|---|---|---|---|
1 | Tue 8 Jun 2021 10:00 - 11:00 | 10:00 - 11:00 | Cambridge Digital Humanities Online | Barbara McGillivray |
2 | Tue 8 Jun 2021 15:00 - 16:00 | 15:00 - 16:00 | Cambridge Digital Humanities Online | Barbara McGillivray |
3 | Tue 15 Jun 2021 10:00 - 11:00 | 10:00 - 11:00 | Cambridge Digital Humanities Online | Barbara McGillivray |
4 | Tue 15 Jun 2021 15:00 - 16:00 | 15:00 - 16:00 | Cambridge Digital Humanities Online | Barbara McGillivray |
We will cover:
- From words to vectors: how can geometry help us understand words’ meaning? introducing word embeddings technologies and their use (and limitations) in representing word meaning, with examples from the Darwin corpus.
- Finding word meaning change over time: can we automate it? exploring current computational methods for detection of words’ semantic change, testing them on the Darwin corpus.
By the end of this course you should be able to:
- Understand how word meaning representations can be created from texts and their limitations;
- Generate your own word representations from historical texts;
- Apply existing methods for tracking word meaning change over time and interpret their results.
The course will consist of four one-hour remote video sessions, spread over two days in two successive weeks. The first session of each day will introduce the topics and methods, while the second session of each day will consist of practical exercises and group work. A chat forum will be used on Moodle for asking/answering questions between the first and the second day.
If you do not have experience with Python and are not interested in the coding aspect, you are welcome to attend only the first session of each day.
You will need a laptop/desktop to join the online sessions and follow the self-paced materials. As we will be using Google Colab notebooks, you do not need to have Python installed. You should have a Google account, which you can set up with your work address.
Booking / availability