README.md

metadata

annotations_creators:
  - crowdsourced
language_creators:
  - expert-generated
  - other
language:
  - en
license:
  - cc-by-nc-4.0
multilinguality:
  - monolingual
pretty_name: spolin
size_categories:
  - 100K<n<1M
  - 10K<n<100K
source_datasets:
  - original
task_categories:
  - text-classification
  - text-generation
task_ids:
  - text-scoring
  - dialogue-modeling

SPOLIN

Table of Contents

• Dataset Description
  • Dataset Summary
  • Available SPOLIN Versions
  • Relevant Links
• Dataset Structure
• Dataset Statistics
• Other Information
  • ACL Presentation
  • Licensing Information
  • Citation Information

Dataset Description

Dataset Summary

This is the repo for the paper "Grounding Conversations with Improvised Dialogues" (ACL2020). The Selected Pairs of Learnable ImprovisatioN (SPOLIN) corpus is a collection of more than 68,000 "Yes, and" type dialogue pairs extracted from the Spontaneanation podcast by Paul F. Tompkins, the Cornell Movie-Dialogs Corpus, and the SubTle corpus. For more information, refer to our paper or our project page.

Available SPOLIN Versions:

The core dataset that was used for the experiments in the paper only includes yes-ands and non-yes-ands from Spontaneanation and most of what is provided in those extracted from the Cornell Movie-Dialogs Corpus. After the submitting the paper, we continued our iterative data augmentation process, repeating another iteration with the Cornell Movie-Dialogs Corpus and extracting from the SubTle corpus. This expanded version is also included in this repository here. This latest version of SPOLIN was used to train the model used in our demo.

In the data folder, we provide two versions of the SPOLIN training set:

1. Version used for experiments in the ACL paper: data/spolin-train-acl.csv
2. Expanded version: data/spolin-train.csv

Relevant Links:

• Project page: https://justin-cho.com/spolin
• Github repo: https://github.com/wise-east/spolin
• SpolinBot Demo: https://spolin.isi.edu
• ACL2020 Paper: https://aclanthology.org/2020.acl-main.218/

Dataset Structure

Fields

• id: unique identifier
• prompt: first utterance in utterance pair
• response: second utterance in utterance pair
• label: yesand = 1, non-yesand = 0
• source: the source for the sample
• split: whether the sample belongs to the training set or the validation set

Dataset Statistics

spolin-train.csv:

	yesands	non-yesands
Spontaneanation	10,459	5,587*
Cornell	16,426	18,310
SubTle	40,303	19,512
Total	67,188	43,409

spolin-train-acl.csv:

	yesands	non-yesands
Spontaneanation	10,459	5,587*
Cornell	14,976	17,851
Total	25,435	23,438

spolin-valid.csv:

	yesands	non-yesands
Spontaneanation	500	500*
Cornell	500	500
Total	1,000	1,000

*Artificially collected by mix & matching positive Spontaneanation samples to balance dataset for training classifier

Other Information

ACL Presentation

Video recording

Citation Information

If you use our data for your work, please cite our ACL2020 paper:

@inproceedings{cho2020spolin,
    title={Grounding Conversations with Improvised Dialogues},
    author={Cho, Hyundong and May, Jonathan},
    booktitle ={Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics},
    publisher = {Association for Computational Linguistics}, 
    location =  {Seattle, Washington, USA},
    year={2020}
}  

Licensing Information

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.