RLStop:一种用于 TAR 的强化学习停止方法
RLStop: A Reinforcement Learning Stopping Method for TAR
AI总结:
本文提出基于强化学习的 TAR 停止规则 RLStop,通过奖励函数学习最佳停止审查时机,在多个基准数据集上显著减少人工审查工作量并优于现有方法。
摘要:
我们提出 RLStop,这是一种基于强化学习的新型技术辅助审查(TAR)停止规则,旨在帮助在 TAR 应用中尽量减少需要人工审查的文档数量。RLStop 使用奖励函数在示例排序上进行训练,以识别停止检查文档的最佳时机。在多个基准数据集(CLEF e-Health、TREC Total Recall 和 Reuters RCV1)上、在一系列目标召回率水平下进行的实验表明,RLStop 显著减少了为相关性筛选文档集合所需的工作量。RLStop 优于多种替代方法,在某些情况下其性能接近该任务可能达到的最大值。
英文摘要:
We present RLStop, a novel Technology Assisted Review (TAR) stopping rule based on reinforcement learning that helps minimise the number of documents that need to be manually reviewed within TAR applications. RLStop is trained on example rankings using a reward function to identify the optimal point to stop examining documents. Experiments at a range of target recall levels on multiple benchmark datasets (CLEF e-Health, TREC Total Recall, and Reuters RCV1) demonstrated that RLStop substantially reduces the workload required to screen a document collection for relevance. RLStop outperforms a wide range of alternative approaches, achieving performance close to the maximum possible for the task under some circumstances.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。