AI4S赋能信息资源管理:演进、方法与实践OACHSSCD

[Purpose/Significance]Artificial Intelligence for Science(AI4S)is driving the transformation of scientific research from the traditional hypothesis-driven paradigm toward a new paradigm jointly driven by data and models.Against this backdrop,it is necessary to systematically examine the evolutionary trajectory,method-ological system,practical scenarios,and governance challenges of AI4S-enabled information resource manage-ment.[Method/Process]AI4S-enabled information resource management has made significant progress at both the methodological and practical levels.Large language models are reshaping scientific productivity in a wide range of scenarios,including retrieval-augmented generation,multi-agent collaboration,multimodal semantic alignment,knowledge graph integration,scientific literature and intelligence acquisition,peer review,scholarly publishing,and cross-lingual scientific communication.However,issues such as model hallucination and knowl-edge-base contamination,algorithmic manipulation and systemic bias,copyright disputes over AIGC,data monop-olies,and information cocoons have also created new tensions in academic ethics and research governance.[Result/Conclusion]Accordingly,the discipline of information resource management should proactively shift from being a supporting discipline for literature services to becoming an infrastructure discipline for scientific innovation.By building open and trustworthy scientific data foundations,improving academic norms for human-AI collaboration,and strengthening algorithmic regulation and content provenance mechanisms,information resource management should assume new responsibilities in the AI era as a provider of high-quality data,a governor of the knowledge lifecycle,and a guardian of academic ethics.