A modified dataset consisting of English dialogs between a user and an assistant discussing movie preferences in natural language. In the foundational dataset, it initially had 502 records.
- Name: ccpemv2.jsonl
- Version: Version 0.2
- Modifications Included: Transformed to ### User:... ### Assistant:... format
- Language: English
- License: Creative Commons Attribution 4.0
- Link to Adapted Dataset: ccpemv2 Dataset
@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
}