Towards Informative Open-ended Text Generation with Dynamic Knowledge Triples - ACL Anthology

Towards Informative Open-ended Text Generation with Dynamic Knowledge Triples

Zixuan Ren, Yang Zhao, Chengqing Zong


Abstract
Pretrained language models (PLMs), especially large language models (LLMs) demonstrate impressive capabilities in open-ended text generation. While our statistical results show that LLMs often suffer from over-concentrated information, where the generated texts overly focus on the given prompt and fail to provide sufficient background and detailed information as humans do. To address this issue, we propose a dynamic knowledge-guided informative open-ended text generation approach, that utilizes a knowledge graph to help the model generate more contextually related entities and detailed facts. Specifically, we first employ a local knowledge filter to extract relevant knowledge from the comprehensive knowledge graph for a given topic sentence. Then we introduce a dynamic knowledge selector to predict the entity to be mentioned in the subsequent sentence. Finally, we utilize a knowledge-enhanced text generator to produce a more informative output. To evaluate the effectiveness of our approach, we evaluate the proposed approach in two scenarios: fine-tuning for small PLMs and prompt tuning for LLMs. Experimental results show that our approach could generate more informative texts than baselines.
Anthology ID:
2023.findings-emnlp.210
Volume:
Findings of the Association for Computational Linguistics: EMNLP 2023
Month:
December
Year:
2023
Address:
Singapore
Editors:
Houda Bouamor, Juan Pino, Kalika Bali
Venue:
Findings
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
3189–3203
Language:
URL:
https://aclanthology.org/2023.findings-emnlp.210
DOI:
10.18653/v1/2023.findings-emnlp.210
Bibkey:
Cite (ACL):
Zixuan Ren, Yang Zhao, and Chengqing Zong. 2023. Towards Informative Open-ended Text Generation with Dynamic Knowledge Triples. In Findings of the Association for Computational Linguistics: EMNLP 2023, pages 3189–3203, Singapore. Association for Computational Linguistics.
Cite (Informal):
Towards Informative Open-ended Text Generation with Dynamic Knowledge Triples (Ren et al., Findings 2023)
Copy Citation:
PDF:
https://aclanthology.org/2023.findings-emnlp.210.pdf