The full dataset (39,000 rows) is sold at https://graunt.com/listings/04a848bf-8dd5-4dd5-87f2-3493591e123d for $49. This repo holds only a 100-row sample per table, under the Brick and Mortar data licence.

Classic English Novels, Computed: Sentiment Arcs, Genres and an Author Influence Graph (343 Novels)

Buy the full dataset ($49): https://graunt.com/listings/04a848bf-8dd5-4dd5-87f2-3493591e123d

Computed features for 343 public-domain English novels (1660-1928): a 100-point sentiment arc per book with its shape, eight genres recovered from opening prose alone, book-to-book arc similarity, and a 77-author graph of documented influences with stylistic and conceptual similarity for every author pair.

Seven years of computational-literature research, reduced to tables. The texts are Project Gutenberg's; every number is ours. Sentiment arcs: TextBlob polarity per 1% of each book, smoothed (9-part centred mean), classified into shapes (Rise, Fall, Rags to riches, ...). Genres: an unsupervised network on raw opening prose recovers eight genre communities (detective, science fiction, nautical adventure, ...). Influence: 205 documented author-to-author influences (Wikidata P737 plus model-attested edges, each marked with its source) and TF-IDF stylistic plus embedding conceptual similarity for all 2,926 author pairs. Useful for teaching, for digital-humanities baselines, and as labelled evaluation data for language models.

Rows in the full dataset: 39,000
Sources: Project Gutenberg full texts (public domain); Wikidata property P737 'influenced by' (CC0); computations by Brick and Mortar AI LLC.
Excluded: No book text. Genre labels from third-party reading lists, used only for validation, are left out.
This repo: a 100-row sample of each table (sample/) and the full schema (SCHEMA.md). The full files are not here.

Seller: Brick and Mortar AI LLC, Saint Paul MN, https://brickandmortar.dev, aidan@brickandmortar.dev.

Downloads last month: 1

Total file size:

37.5 kB