Preliminary Class Business
- Extending discussion of text analysis in our last class to film/video:
- Class 6 on Social Network Analysis:
- Visit by Mark Algee-Hewitt and Ryan Heuser, co-associate directors of the Stanford Literary Lab. Please read in advance France Moretii, "'Operationalizing': Or, The Function of Measurement in Modern Literary Theory" [PDF] (Stanford Lit Lab Pamphlet #6, 2013)
- Rough agenda for Class 6:
- 1. Introductions all around
2. Mark and Ryan talk informally "about our own work and backgrounds, maybe discuss a couple of projects under the rubric of how they operationalize various questions or concepts and then open the floor for some more general discussion."
3. Alan will note the connections between the Stanford Lit Lab work (and the "Operationalizing" essay) and the social network analysis topic that is the specific subject of this particular class.
4. Depending on the time remaining, we'll take a look at some of the student beginner practicums in using Gephi and social network visualization as a springboard to discussing larger issues related to network analysis. - Preview of readings and practicum for Class 6
- Using Gephi
- Note: the practicums for classes 5 and 6 are the most technically advanced in this course, which is designed for beginners.
- Not covered in this course are more advanced, customized, and specialized DH methods and workflows that require setting up a "development environment" on one's computer, usually operated at the command-line level. Such a development environment often includes a combination of the following:
- Python (a programming language/environment used in diverse research disciplines, with many extensible "packages" or "libraries" for a variety of text-processing, statistical, and other purposes).
- R (programming language and software environment for "statistical computing and graphics")
- Git and Github (version control and repository system for managing project revisions and sharing, collaborating on, or branching from projects).
1. Our Topic Modeling Practicums (and our questions)
- Practicum folder
- Student practicums.
- Some of Alan Liu's practicums:
- Lyrical Ballads sample (using In-Browser Topic Modeling Tool, javascript implementation of LDA)
- Input text file: lyrical-ballads.txt
- topic-model-lyrical-ballads.pdf
- topic-model-lyrical-ballads-words.pdf
- topic-model-lyrical-ballads-correlations.pdf
- 4Humanities WhatEvery1Says corpus sample (using Topic Modeling Tool, and MALLET)
2. Theory, Practice, & Applied Uses of Topic Modeling
- Theory:
- Ted Underwood, "Topic Modeling Made Just Simple Enough" (2012)
- David M. Blei, "Topic Modeling and Digital Humanities" (2012)
- Matt Burton, "The Joy of Topic Modeling" (2013)
- Ted Underwood, "What Kinds of "Topics" Does Topic Modeling Actually Produce?" (2012)
- Matthew L. Jockers, "The LDA Buffet is Now Open; or, Latent Dirichlet Allocation for English Majors" (2011)
- John Mohr and Petko Bogdanov, "Topic Models: What They Are and Why They Matter" (2013)
- Benjamin M. Schmidt, "Words Alone: Dismantling Topic Models in the Humanities" (2012)
- Additional resources for discussion:
- Practice: "Recipes" for Text Preparation and Modeling (plus research on next-step processes):
- Matthew L. Jockers, "'Secret' Recipe for Topic Modeling Themes"
- Stop lists ("Expanded Stopwords List")
- Parts of speech filtering
- "Chunking"
- Andrew Goldstone & Ted Underwood, "Code Used ... in Analyzing Topic Models of Literary-studies Journals"
- Jonathan Chang, et al., "Reading Tea Leaves: How Humans Interpret Topic Models"
- Daniel McFarland, et al., "Differentiating Language Usage Through Topic Models"
- Applied Uses
- Matthew L. Jockers, Macroanalysis: Digital Methods and Literary History (2013) -- Chap. 8: "Theme" (Alan's highlighted copy; course password required)
- Andrew Goldstone and Ted Underwood, "The Quiet Transformations of Literary Studies: What Thirteen Thousand Scholars Could Tell Us" (2014) (preprint version)
- See Andrew Goldstone's site for exploring and visualizing the topic models on which "The Quiet Transformations of Literary Studies" is based. (See also Andrew Goldstone's new site for exploring and visualizing topic models 40 years of the Signs journal)
- Lisa M. Rhody, "Topic Modeling and Figurative Language" (2012)
4. The Conceptual Challenge of Topic Modeling
- Phenomena that appear in distant reading:
(in escalating scale of possible disturbance to humanists)
- Pattern
- Scale phenomena / Cyclical phenomena
- Quantitative phenomena
- Probabilistic phenomena
- [Network phenomena]
Conceptual graph by Matt Burton
- "signal"
- "model"
- "field" (cf. Marshall McLuhan, excerpts from Understanding Media)
- the unknown