SherlockBox-RuNeg

SherlockBox-RuNeg is a data set in Russian that features annotations of negation cues and negation scopes. It consists of 120 sentences from a story by Sir Arthur Conan Doyle, The Adventure of the Cardboard Box ("Картонная коробка"). The story itself and this Russian translation of the story are in the public domain. The Russian text was obtained from http://www.lib.ru/AKONANDOJL/sh_cbox.txt. Only sentences that contain negation were included in SherlockBox-RuNeg. Annotations were performed by one native Russian speaker with the Prodigy annotation tool. The data is presented in two files. One is a json file that stores each sentence as 3 vectors: list of tokens, list of cue labels (1 = cue, 3 = not a cue), list of scope labels (1 = in scope, 0 = not in scope). Another is a text file that gives a more visual representation of each sentence in 5 lines: tokens, cue labels, scope labels, cue tokens, scope tokens.