PaSa

TextsIntroduced 2021-11-06

PaSa is a dataset to train Machine Learning algorithms to automate the highlighting of patent paragraphs with semantic annotations. It consists of 150k samples obtained by traversing USPTO patents over a decade