🪐 Experiments with 20th-Century Speculative Fiction in HathiTrust

In order to make books that are still under copyright available for computational text analysis, HathiTrust Research Center has devised a machine-readable data format called Extracted Features. This repository contains data and Jupyter Notebooks for analyzing over three thousand works of speculative fiction in HathiTrust, all published between 1900-1999.

The tutorials here assume a working familiarity with Python and Jupyter Notebooks. For those new to both, The Programming Historian's "Introduction to Jupyter Notebooks" by Quinn Dombrowski, Tassie Gniady, and David Kloster is a great starting point. We'll also be working with Pandas, a Python library for working with tabular data. Melanie Walsh's Intro to Cultural Analytics course includes a fantastic overview of Pandas (as well as the conceptual and ethical challenges inherent to data work in the humanities).

Contents

Code

• 👉 htrc_sf_experiments.ipynb is the main tutorial. Start there.
• instructions for running TF-IDF on a volume (TK)

Data

• 📁 /data/SF_Extracted_Features_Full contains Extracted Features files for each volume (aka book)
• 📁 /data/thompson-mimno-SF-final-matches.tsv list of all SF volumes in HTRC identified by David Mimno and Laure Thompson