Skip to content
No description, website, or topics provided.
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Type Name Latest commit message Commit time
Failed to load latest commit information.

Code and data for "Learning Scalar Adjective Intensity from Paraphrases"

If you use the code or data in this repo, please cite the following paper:

  author = 	"Cocos, Anne
		and Wharton, Veronica
		and Pavlick, Ellie
		and Apidianaki, Marianna
		and Callison-Burch, Chris",
  title = 	"Learning Scalar Adjective Intensity from Paraphrases",
  booktitle = 	"Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing",
  year = 	"2018",
  publisher = 	"Association for Computational Linguistics",
  pages = 	"1752--1762",
  location = 	"Brussels, Belgium",
  url = 	""

Directory contents

  • The adjective graph that encodes adjectival paraphrases from PPDB of the form (RB JJ1 <--> JJ2) is under jjgraph/
  • Code for computing a pairwise paraphrase-based intensity score is in pairpredict/
  • Code for the downstream tasks of global ordering and indirect question answering are in globalorder/ and iqap/ respectively
  • The crowdsourced dataset of adjective scales is in globalorder/data/crowd/gold_rankings. Each file contains adjectives belonging to one scale, where they are ranked from least to most intense by line.
You can’t perform that action at this time.