ccpemv2 / README.md
aloobun's picture
Update README.md
c6c3b39
|
raw
history blame
1.01 kB
metadata
license: cc-by-4.0
language:
  - en
size_categories:
  - 10K<n<100K
pretty_name: movies dialog dataset

A modified dataset consisting of 12,000 records of English dialogs between a user and an assistant discussing movie preferences in natural language. In the foundational dataset, it initially had 502 records. To enhance the dataset's scope, I expanded it to include a total of 12,000 records.

Dataset Information

  • Name: ccpemv2.jsonl
  • Version: Version 0.1
  • Modifications Included: Transformed to ### User:... ### Assistant:... format
  • Language: English
  • License: Creative Commons Attribution 4.0

Original Dataset Citation

@inproceedings{radlinski-etal-2019-ccpe,
title = {Coached Conversational Preference Elicitation: A Case Study in Understanding Movie Preferences},
author = {Filip Radlinski and Krisztian Balog and Bill Byrne and Karthik Krishnamoorthi},
booktitle = {Proceedings of the Annual Meeting of the Special Interest Group on Discourse and Dialogue ({SIGDIAL})},
year = 2019
}