HiRAG: A Historical Information-Driven Retrieval-Augmented Generation Framework for Background Summarization

Dong Zhou, Binli Zeng, Nankai Lin, Yongmei Zhou, Aimin Yang
Proceedings of the 16th Asian Conference on Machine Learning, PMLR 260:1016-1031, 2025.

Abstract

In an era overwhelmed by a deluge of global information, it is often challenging for people to grasp the relationships that an event develops over time. The background summarization (BS) task facilitates a profound understanding of the relationships between the current background of an event at any given time and its historical backgrounds. To enhance comprehension and help news readers and professionals to quickly understand the evolution of events, we introduce a Historical information-driven Retrieval-Augmented Generation framework (HiRAG). This framework is designed to extract the most relevant information from historical backgrounds and supplement it to generate precise background summarization. HiRAG employs state-of-the-art retrieval-augmented generation technologies to produce relevant background summarization. We implement a multi-strategy similarity calculation and introduce a sliding window mechanism to optimize retrieval construction. Our framework has been rigorously tested through a series of experiments and extensive analyses of the latest datasets. The promising results affirm the effectiveness of our proposed HiRAG framework and its retrieval capabilities.

Cite this Paper


BibTeX
@InProceedings{pmlr-v260-zhou25a, title = {{HiRAG}: {A} Historical Information-Driven Retrieval-Augmented Generation Framework for Background Summarization}, author = {Zhou, Dong and Zeng, Binli and Lin, Nankai and Zhou, Yongmei and Yang, Aimin}, booktitle = {Proceedings of the 16th Asian Conference on Machine Learning}, pages = {1016--1031}, year = {2025}, editor = {Nguyen, Vu and Lin, Hsuan-Tien}, volume = {260}, series = {Proceedings of Machine Learning Research}, month = {05--08 Dec}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v260/main/assets/zhou25a/zhou25a.pdf}, url = {https://proceedings.mlr.press/v260/zhou25a.html}, abstract = {In an era overwhelmed by a deluge of global information, it is often challenging for people to grasp the relationships that an event develops over time. The background summarization (BS) task facilitates a profound understanding of the relationships between the current background of an event at any given time and its historical backgrounds. To enhance comprehension and help news readers and professionals to quickly understand the evolution of events, we introduce a Historical information-driven Retrieval-Augmented Generation framework (HiRAG). This framework is designed to extract the most relevant information from historical backgrounds and supplement it to generate precise background summarization. HiRAG employs state-of-the-art retrieval-augmented generation technologies to produce relevant background summarization. We implement a multi-strategy similarity calculation and introduce a sliding window mechanism to optimize retrieval construction. Our framework has been rigorously tested through a series of experiments and extensive analyses of the latest datasets. The promising results affirm the effectiveness of our proposed HiRAG framework and its retrieval capabilities.} }
Endnote
%0 Conference Paper %T HiRAG: A Historical Information-Driven Retrieval-Augmented Generation Framework for Background Summarization %A Dong Zhou %A Binli Zeng %A Nankai Lin %A Yongmei Zhou %A Aimin Yang %B Proceedings of the 16th Asian Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2025 %E Vu Nguyen %E Hsuan-Tien Lin %F pmlr-v260-zhou25a %I PMLR %P 1016--1031 %U https://proceedings.mlr.press/v260/zhou25a.html %V 260 %X In an era overwhelmed by a deluge of global information, it is often challenging for people to grasp the relationships that an event develops over time. The background summarization (BS) task facilitates a profound understanding of the relationships between the current background of an event at any given time and its historical backgrounds. To enhance comprehension and help news readers and professionals to quickly understand the evolution of events, we introduce a Historical information-driven Retrieval-Augmented Generation framework (HiRAG). This framework is designed to extract the most relevant information from historical backgrounds and supplement it to generate precise background summarization. HiRAG employs state-of-the-art retrieval-augmented generation technologies to produce relevant background summarization. We implement a multi-strategy similarity calculation and introduce a sliding window mechanism to optimize retrieval construction. Our framework has been rigorously tested through a series of experiments and extensive analyses of the latest datasets. The promising results affirm the effectiveness of our proposed HiRAG framework and its retrieval capabilities.
APA
Zhou, D., Zeng, B., Lin, N., Zhou, Y. & Yang, A.. (2025). HiRAG: A Historical Information-Driven Retrieval-Augmented Generation Framework for Background Summarization. Proceedings of the 16th Asian Conference on Machine Learning, in Proceedings of Machine Learning Research 260:1016-1031 Available from https://proceedings.mlr.press/v260/zhou25a.html.

Related Material