Enhancing LLMs via High-Knowledge Data Selection
Feiyu Duan, Xuemiao Zhang, Sirui Wang, Haoran Que, Yuqi Liu, Wenge Rong, Xunliang Cai
Abstract
The performance of Large Language Models (LLMs) is intrinsically linked to the quality of its training data. Although several studies have proposed methods for high-quality data selection, they do not consider the importance of knowledge richness in text corpora. In this paper, we propose a novel and gradient-free High-Knowledge Scorer (HKS) to select high-quality data from the dimension of knowledge, to alleviate the problem of knowledge scarcity in the pre-trained corpus. We propose a comprehensive multi-domain knowledge element pool and introduce knowledge density and coverage as metrics to assess the knowledge content of the text. Based on this, we propose a comprehensive knowledge scorer to select data with intensive knowledge, which can also be utilized for domain-specific high-knowledge data selection by restricting knowledge elements to the specific domain. We train models on a high-knowledge bilingual dataset, and experimental results demonstrate that our scorer improves the model's performance in knowledge-intensive and general comprehension tasks, and is effective in enhancing both the generic and domain-specific capabilities of the model.
BibTeX
@article{Duan_Zhang_Wang_Que_Liu_Rong_Cai_2025, title={Enhancing LLMs via High-Knowledge Data Selection}, volume={39}, url={https://ojs.aaai.org/index.php/AAAI/article/view/34555}, DOI={10.1609/aaai.v39i22.34555}, abstractNote={The performance of Large Language Models (LLMs) is intrinsically linked to the quality of its training data. Although several studies have proposed methods for high-quality data selection, they do not consider the importance of knowledge richness in text corpora. In this paper, we propose a novel and gradient-free High-Knowledge Scorer (HKS) to select high-quality data from the dimension of knowledge, to alleviate the problem of knowledge scarcity in the pre-trained corpus. We propose a comprehensive multi-domain knowledge element pool and introduce knowledge density and coverage as metrics to assess the knowledge content of the text. Based on this, we propose a comprehensive knowledge scorer to select data with intensive knowledge, which can also be utilized for domain-specific high-knowledge data selection by restricting knowledge elements to the specific domain. We train models on a high-knowledge bilingual dataset, and experimental results demonstrate that our scorer improves the model’s performance in knowledge-intensive and general comprehension tasks, and is effective in enhancing both the generic and domain-specific capabilities of the model.}, number={22}, journal={Proceedings of the AAAI Conference on Artificial Intelligence}, author={Duan, Feiyu and Zhang, Xuemiao and Wang, Sirui and Que, Haoran and Liu, Yuqi and Rong, Wenge and Cai, Xunliang}, year={2025}, month={Apr.}, pages={23832-23840} }