来自 | 知乎 作者 | 李家琦
链接｜/p/410984053
本文已获作者授权，未经许可禁止二次转载
Dialogue-based Machine Reading Comprehension任务是近两年比较新的机器阅读理解（MRC）任务，任务目标是让机器去理解人们之间的对话。本文简要整理了该任务现有数据集，并推荐几篇相关论文。
一、数据集
该任务现有的数据集主要有如下这些：
1.Ma et, al, , NAACL（数据集没有命名）
任务类型：完形填空
论文：Challenging reading comprehension on daily conversation: Passage completion on multiparty dialog.
数据集：GitHub - emorynlp/reading-comprehension: Reading comprehension on multiparty dialog.
2.DREAM,TACL
任务类型：单选题
论文：Dream: A challenge data set and models for dialogue-based reading comprehension.
数据集：A Challenge Dataset and Models for Dialogue-Based Reading Comprehension
3.FriendsQA, SIGDial
任务类型：Span-base
论文：FriendsQA: Open-Domain Question Answering on TV Show Transcripts
数据集：GitHub - emorynlp/FriendsQA: Question answering on multiparty dialogue
4.Molweni，COLING
任务类型：Span-based
论文：Molweni: A Challenge Multiparty Dialogue-based Machine Reading Comprehension Dataset with Discourse Structure
数据集：GitHub - HIT-SCIR/Molweni
5.QAConv,arXiv
任务类型：Span-based
论文：QAConv: Question Answering on Informative Conversations
数据集：GitHub - salesforce/QAConv: This repository maintains the QAConv dataset, a question-answering dataset on informative conversations including business emails, panel discussions, and work channels.
目前此任务上使用比较多的数据集主要是DREAM、FriendsQA和Molweni。在QAConv数据集论文中，作者还将现有的几个数据集进行了对比。
数据集对比，来自QAConv论文
二、模型
这部分主要推荐DREAM、FriendsQA和Molweni这3个数据集上比较有代表性的模型论文。
1.DREAM数据集相关模型论文推荐
a.DUMA: Reading Comprehension with Transposition Thinking. arXiv .
b.Multi-task Learning with Multi-head Attention for Multi-choice Reading Comprehension. arXiv .
2.FriendsQA数据集相关模型论文推荐
a.Transformers to Learn Hierarchical Contexts in Multiparty Dialogue for Span-based Question Answering. ACL .
b.Graph-based knowledge integration for question answering over dialogue. COLING .
3.Molweni数据集相关模型论文推荐
a.DADgraph: A Discourse-aware Dialogue Graph Neural Network for Multiparty Dialogue Machine Reading Comprehension. IJCNN .
b.Self-and Pseudo-self-supervised Prediction of Speaker and Key-utterance for Multi-party Dialogue Reading Comprehension. EMNLP Findings.
c.Enhanced Speaker-aware Multi-party Multi-turn Dialogue Comprehension. arXiv .
以上是我简单整理的Dialogue MRC任务数据集和推荐的几篇相关论文，欢迎补充！
往期精彩回顾适合初学者入门人工智能的路线及资料下载机器学习及深度学习笔记等资料打印机器学习在线手册深度学习笔记专辑《统计学习方法》的代码复现专辑AI基础下载黄海广老师《机器学习课程》视频课黄海广老师《机器学习课程》711页完整版课件
本站qq群554839127，加入微信群请扫码：
