Insert Commonsense Knowledge Through Semantics for Dialogue Generation

Siqi Hou, Dandan Song*, Zhijing Wu, Xiechao Guo, Ziyi Yang

*此作品的通讯作者

科研成果: 书/报告/会议事项章节会议稿件同行评审

摘要

Language models such as GPT and BART have been widely applied to various language generation tasks, including dialogue generation. While some works have improved the generation performance of these models by incorporating external knowledge, the effective extraction of valuable knowledge from diverse sources and its incorporation into generation models require further investigation. To tackle this problem, we propose a method called Knowledge-enhanced Multi-turn Dialogue Model (KMDM) to extract commonsense knowledge from external knowledge graphs and inject the knowledge into the encoding and decoding processes globally and locally. We first extract sub-graphs according to the semantic correlations with contexts. Then we add a hierarchical graph attention layer to the decoder in order to acquire local information from sub-graphs. Experiments conducted on Wizard of Wikipedia and DailyDialog show the effectiveness of the proposed method.

源语言英语
主期刊名Knowledge Science, Engineering and Management - 17th International Conference, KSEM 2024, Proceedings
编辑Cungeng Cao, Huajun Chen, Liang Zhao, Junaid Arshad, Yonghao Wang, Taufiq Asyhari
出版商Springer Science and Business Media Deutschland GmbH
305-317
页数13
ISBN(印刷版)9789819754946
DOI
出版状态已出版 - 2024
活动17th International Conference on Knowledge Science, Engineering and Management, KSEM 2024 - Birmingham, 英国
期限: 16 8月 202418 8月 2024

出版系列

姓名Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
14885 LNAI
ISSN(印刷版)0302-9743
ISSN(电子版)1611-3349

会议

会议17th International Conference on Knowledge Science, Engineering and Management, KSEM 2024
国家/地区英国
Birmingham
时期16/08/2418/08/24

指纹

探究 'Insert Commonsense Knowledge Through Semantics for Dialogue Generation' 的科研主题。它们共同构成独一无二的指纹。

引用此