现代情报2026，Vol.46Issue(2)：91-101,11.DOI:10.3969/j.issn.1008-0821.2026.02.008
科技文献文本知识抽取的提示框架研究
Research on the Prompt Framework for Knowledge Extraction From Scientific and Technological Literature Text
摘要
Abstract
[Purpose/Significance]At present,in the case of small samples,the knowledge extraction method for sci-entific literature based on large language models is widely used.The effectiveness of the model depends on the quality of the prompt word framework,so it is necessary to design a high-quality framework for scientific literature knowledge extrac-tion prompt words.[Method/Process]This study took the field of organic solar cells as an example and designs an automatic knowledge extraction system for the field of organic solar cells.This study designed an automatic knowledge extraction prompt word framework for scientific and technological literature,and conducted comparative experiments using different large models based on this framework.[Result/Conclusion]The experimental results demonstrate that,in the task of know-ledge extraction from journal literature in the field of organic solar cells,the method utilizing the prompt phrase framework proposed in this paper achieves superior performance compared to the conventional method of prompt phrase knowledge extraction.关键词
提示词框架/大模型/文本挖掘/有机太阳能电池/高质量数据基座建设Key words
prompt word framework/large model/text mining/organic solar cell/construction of high-quality data foundation分类
社会科学引用本文复制引用
陈昱成,韩涛,胡正银..科技文献文本知识抽取的提示框架研究[J].现代情报,2026,46(2):91-101,11.基金项目
国家社会科学基金项目"支撑AI4Science的科技图书馆知识服务内容研究"(项目编号:22BTQ019). （项目编号:22BTQ019）
