Lecture: Topic Modeling English 147
How would a machine read our course texts?
Topic modeling assumes that every word in each of our texts is taken from an existing topic in the world. It sorts our texts into groups of words, which often overlap. What topics appear to shape the readings for this course?
Exploration of the visualized data reveals where these topics begin blending together, overlapping in complex and compelling ways. In order to draw meaningful connections across the modeled texts, we must move from topics to themes.
Themes revealed by the model
This topic model allows us to ask the following questions about our course readings:
- Is there a meaningful difference between the representation of male and female characters across these works?
- Are men always associated with public action and women with the home? Do any works challenge this claim?
- What is the relationship between the individual and the group? What role does kinship play in determining individual identity?
- How do characters experience their environment? Is there a connection between setting (where a scene/novel is set) and identity (who a character thinks him or herself to be)?
- Why is the body tied to looking, motion, and time? If there is a connection between the body and one’s place in the world, then how does that connection work?
Further Reading
This topic model was constructed based on existing work by Aaron Mauro and Jentery Sayers. The D3 visualization is created with mobstock’s D3 examples.
BACK