This repository contains data and code for the paper below:

Learning to Ask Good Questions: Ranking Clarification Questions using Neural Expected Value of Perfect Information
Sudha Rao (raosudha@cs.umd.edu) and Hal Daumé III (hal@umiacs.umd.edu)
To appear in the proceedings of The 2018 Association of Computational Lingusitics (ACL 2018)

• Download the clarification questions dataset from google drive here: https://go.umd.edu/clarification_questions_dataset

• cp clarification_questions_dataset/data ranking_clarification_questions/data

• Download word embeddings trained on stackexchange datadump here: https://go.umd.edu/stackexchange_embeddings

• cp stackexchange_embeddings/embeddings ranking_clarification_questions/embeddings

The above dataset contains clarification questions for these three sites of stackexchange:

1. askubuntu.com
2. unix.stackexchange.com
3. superuser.com

To run models on a combination of the three sites above, check ranking_clarification_questions/src/models/README

To generate clarification questions for a different site of stackexchange, check ranking_clarification_questions/src/data_generation/README

To retrain word embeddings on a newer version of stackexchange datadump, check ranking_clarification_questions/src/embedding_generation/README

Please contact Sudha Rao (raosudha@cs.umd.edu) if you have any questions or any feedback.