GraphFlow+: Exploiting Conversation Flow in Conversational Machine Comprehension with Graph Neural Networks

doi:10.1007/s11633-023-1421-0

CASIA OpenIR > 学术期刊 > Machine Intelligence Research

	GraphFlow+: Exploiting Conversation Flow in Conversational Machine Comprehension with Graph Neural Networks
	Jing Hu 1; Lingfei Wu 2; Yu Chen 3; Po Hu 1; Mohammed J. Zaki 4
发表期刊	Machine Intelligence Research
ISSN	2731-538X
	2024
卷号	21 期号:2 页码:272-282
摘要	The conversation machine comprehension (MC) task aims to answer questions in the multi-turn conversation for a single passage. However, recent approaches don′t exploit information from historical conversations effectively, which results in some references and ellipsis in the current question cannot be recognized. In addition, these methods do not consider the rich semantic relationships between words when reasoning about the passage text. In this paper, we propose a novel model GraphFlow+, which constructs a context graph for each conversation turn and uses a unique recurrent graph neural network (GNN) to model the temporal dependencies between the context graphs of each turn. Specifically, we exploit three different ways to construct text graphs, including the dynamic graph, static graph, and hybrid graph that combines the two. Our experiments on CoQA, QuAC and DoQA show that the GraphFlow+ model can outperform the state-of-the-art approaches.
关键词	Conversational machine comprehension (MC), reading comprehension, question answering, graph neural networks (GNNs), natural language processing (NLP)
DOI	10.1007/s11633-023-1421-0
引用统计
文献类型	期刊论文
条目标识符	http://ir.ia.ac.cn/handle/173211/56038
专题	学术期刊_Machine Intelligence Research
作者单位	1.School of Computer Science, Central China Normal University, Wuhan 430079, China 2.Pinterest, San Francisco 94016, USA 3.Meta, Mountain View 94039, USA 4.Rensselaer Polytechnic Institute, Troy 12180, USA
推荐引用方式 GB/T 7714	Jing Hu,Lingfei Wu,Yu Chen,et al. GraphFlow+: Exploiting Conversation Flow in Conversational Machine Comprehension with Graph Neural Networks[J]. Machine Intelligence Research,2024,21(2):272-282.
APA	Jing Hu,Lingfei Wu,Yu Chen,Po Hu,&Mohammed J. Zaki.(2024).GraphFlow+: Exploiting Conversation Flow in Conversational Machine Comprehension with Graph Neural Networks.Machine Intelligence Research,21(2),272-282.
MLA	Jing Hu,et al."GraphFlow+: Exploiting Conversation Flow in Conversational Machine Comprehension with Graph Neural Networks".Machine Intelligence Research 21.2(2024):272-282.

条目包含的文件
文件名称/大小	文献类型	版本类型	开放类型	使用许可
MIR-2022-11-336.pdf（1612KB）	期刊论文	出版稿	开放获取	CC BY-NC-SA	浏览