Can algorithms predict the success of a book?

BooksResearchers at Stony Brook University in New York may have found an algorithm to predict how popular a book will be. They predicted which works of fiction out of a selection of already published books were successful, based on some of their stylistic features, using a computer model. The researchers said: “Predicting the success of novels is a curious question among publishers, professional book reviewers, aspiring and even expert writers alike. Indeed, even some of the best sellers and award winners can go through several rejections before they are picked up by a publisher.”

For publishers, the application of such a tool could be used to predict the commercial success of a book, basing decisions on quantitative measurements rather than subjective decisions.

The researchers found it could effectively distinguish successful novels from not-so-popular works (these had still made the grade with publishers) with an 84 % success rate. It also had an 89 % success rate on film scripts. It works by looking at specific stylistic features, rather than overarching themes, plots, characters, or emotional tones

What do potential bestsellers need?

  • Lots of verbs which describe thought processes – tended to be more successful than using action verbs.
  • Not many over-emotional verbs in the dialogue.
  • Getting straight to the point – just use “say.”
  • Describing things using nouns and adjectives, not verbs and adverbs.

Download figures for each work (the books studied were from Project Gutenberg) were used as the main success indicator. For a few books, the researchers also took into account Amazon sales or prestigious awards like Pulitzer and Nobel prizes. For films, its average score on IMDB was used..

The researchers also used their model to see if the accepted wisdom of being easy to read is a desirable quality. “We made an unexpected observation on the connection between readability and the literary success—that they correlate into the opposite directions.”

For more on the project