arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2510.07167cs.CL
收藏

层级文本分类的推理:以专利为例

Reasoning for Hierarchical Text Classification: The Case of Patents

University of Cambridge(剑桥大学);National Science Library, Chinese Academy of Sciences(中国科学院国家科学图书馆);Department of Information Resources Management, School of Economics and Management, University of Chinese Academy of Sciences(中国科学院大学经济管理学院信息资源管理系)

Lekang Jiang, Wenjun Sun, Stephan Goetz

AI总结:

针对层级文本分类缺乏可解释性的问题,提出RHC框架,将分类重构为逐步推理任务,结合冷启动与强化学习训练LLM,在专利及其他基准上提升准确率约3%并增强可解释性。

摘要:

层级文本分类(HTC)将文档分配到预定义分类体系的多个层级。由于领域知识难度大且标签数量庞大,自动化专利主题分类是最具挑战性的HTC场景之一。以往的方法仅输出扁平标签集,对预测背后的原因提供的洞察很少。因此,我们提出层级分类推理(RHC),一种将HTC重构为逐步推理任务以顺序推导层级标签的新框架。RHC分两个阶段训练大语言模型(LLM):冷启动阶段使输出与思维链(CoT)推理格式对齐,以及强化学习(RL)阶段以增强多步推理能力。RHC在我们的实验中展示了四个优势。(1)有效性:RHC超越了以往的基线,并在准确率和宏F1上比监督微调对应方法高出约3%。(2)可解释性:RHC在预测前生成自然语言论证,以促进人工检查。(3)可扩展性:RHC随模型规模有利地扩展,相比标准微调获得更大增益。(4)适用性:除专利外,我们进一步证明RHC在其他广泛使用的HTC基准上达到了最先进的性能,这突显了其广泛的适用性。

英文摘要:

Hierarchical text classification (HTC) assigns documents to multiple levels of a pre-defined taxonomy. Automated patent subject classification represents one of the hardest HTC scenarios because of professional difficulties and extensive labels. Prior approaches only output a flat label set, which offers little insight into the reason behind predictions. Therefore, we propose Reasoning for Hierarchical Classification (RHC), a novel framework that reformulates HTC as a step-by-step reasoning task to sequentially deduce hierarchical labels. RHC trains large language models (LLMs) in two stages: a cold-start stage that aligns outputs with chain-of-thought (CoT) reasoning format and a reinforcement learning (RL) stage to enhance multi-step reasoning ability. RHC demonstrates four advantages in our experiments. (1) Effectiveness: RHC surpasses previous baselines and outperforms the supervised fine-tuning counterparts by approximately 3% in accuracy and macro F1. (2) Explainability: RHC produces natural-language justifications before prediction to facilitate human inspection. (3) Scalability: RHC scales favorably with model size with larger gains compared to standard fine-tuning. (4) Applicability: Beyond patents, we further demonstrate that RHC achieves state-of-the-art performance on other widely used HTC benchmarks, which highlights its broad applicability.

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。

↑