Milling wisdom

摘要

Simple text classification algorithms perform remarkably well when used for detecting famous quotes in literary or philosophical text, with f-scores approaching 95%. We compare the task to topic classification, polarity classification and authorship attribution.