SIGIR

LLM-Evaluation Tropes: Perspectives on the Validity of LLM-Evaluations

LLM评估的陈词滥调：对LLM评估有效性的视角

Laura Dietz, Oleg Zendel, Peter Bailey, Charles Clarke, Ellese Cotterill, Jeff Dalton, Faegheh Hasibi, Mark Sanderson, Nick Craswell

AI总结 本文探讨了LLM评估在信息检索系统中的有效性问题，指出其可能带来的偏见和自我强化风险，并提出测试和协作框架以规范LLM在评估中的使用。

Journal ref In Proceedings of the 2025 International ACM SIGIR Conference on Innovative Concepts and Theories in Information Retrieval (ICTIR). Association for Computing Machinery, New York, NY, USA, 218-229

Rules, Resources, and Restrictions: A Taxonomy of Task-Based Information Request Intents

规则、资源与限制：任务导向信息请求意图的分类

Melanie A. Kilian, David Elsweiler

AI总结 本文提出任务导向信息请求意图的分类，以弥补传统查询方法与AI任务支持需求之间的差距。

Comments 11 pages, 1 figure, to be published in: 2026 ACM SIGIR Conference on Human Information Interaction and Retrieval (CHIIR '26), March 22-26, 2026, Seattle, WA, USA. ACM, New York, NY, USA, 11 pages. https://doi.org/10.1145/3786304.3787863

Information Farming: From Berry Picking to Berry Growing

信息耕作：从摘Berry到种Berry

Leif Azzopardi, Adam Roegiest

AI总结 本文提出信息耕作概念，认为生成式AI使人们从碎片化信息采集转向主动信息培养，标志着人类与信息互动方式的根本转变。

Comments ACM CHIIR 2026

Journal ref ACM SIGIR Conference on Human Information Interaction and Retrieval 2026

A Transfer Framework for Enhancing Temporal Graph Learning in Data-Scarce Settings

一种用于数据稀缺场景下增强时序图学习的迁移框架

Sidharth Agarwal, Tanishq Dubey, Shubham Gupta, Srikanta Bedathur

机构 * Indian Institute of Technology, Delhi（印度理工学院德里） 纠错

AI总结 本文提出一种迁移框架，通过解耦节点表示与特征，提升时序图学习在数据稀缺场景下的性能。

Journal ref SIGIR 2025: Proceedings of the 48th International ACM SIGIR Conference on Research and Development in Information Retrieval