由于公式存在多种可能解释,解析与评估都变得复杂,因此分层且复杂的数学表达式识别(MER)颇具挑战性。本文提出 Hierarchical Detail-Focused Recognition dataset(HDR,分层细节聚焦识别数据集),这是首个专门为解决上述问题而设计的数据集。它包含大规模训练集 HDR-100M,以一亿个训练实例提供前所未有的规模和多样性;测试集 HDR-Test 则包含复杂分层公式的多种解释,用于全面评估模型性能。此外,复杂公式解析常出现细粒度细节错误。为此,本文提出 Hierarchical Detail-Focused Recognition Network(HDNet,分层细节聚焦识别网络),这一创新框架引入分层子公式模块,专注于精确处理公式细节,从而显著提升 MER 性能。实验结果表明,HDNet 在多个数据集上均优于现有 MER 模型。
英文摘要:
Hierarchical and complex Mathematical Expression Recognition (MER) is challenging due to multiple possible interpretations of a formula, complicating both parsing and evaluation. In this paper, we introduce the Hierarchical Detail-Focused Recognition dataset (HDR), the first dataset specifically designed to address these issues. It consists of a large-scale training set, HDR-100M, offering an unprecedented scale and diversity with one hundred million training instances. And the test set, HDR-Test, includes multiple interpretations of complex hierarchical formulas for comprehensive model performance evaluation. Additionally, the parsing of complex formulas often suffers from errors in fine-grained details. To address this, we propose the Hierarchical Detail-Focused Recognition Network (HDNet), an innovative framework that incorporates a hierarchical sub-formula module, focusing on the precise handling of formula details, thereby significantly enhancing MER performance. Experimental results demonstrate that HDNet outperforms existing MER models across various datasets.
Language Technologies Institute, Carnegie Mellon University(卡内基梅隆大学语言技术研究所)
事件接地的目的是链接文本语料库中的提及引用从知识库(KB)的事件。此任务以前的工作主要集中在链接到单个KB事件,从而忽略了事件的层次方面。文档中的事件通常以不同的时空粒度级别进行描述(Glavas et al. 2014)。这些层次关系被用于叙事理解和图式构建的下游任务。在这项工作中,我们提出了一个扩展的事件接地任务,需要解决层次事件结构的知识库。我们建议的任务涉及到链接一个提到引用的一组事件标签从一个子事件层次结构中的知识库。我们提出了一种检索方法,通过辅助分层损失来利用事件分层(Murty et al. 2018)。在维基百科和维基数据自动创建的多语言数据集上,我们的实验证明了分层损失对检索和重新排名基线的有效性(Wu et al. 2020; Pratapa,Gupta和Mitamura 2022)。此外,我们证明了系统的能力,以帮助层次发现之间看不见的事件。
Aggregation of Reasoning: A Hierarchical Framework for Enhancing Answer Selection in Large Language Models
School of Computer Science, Fudan University(复旦大学计算机科学技术学院);National University of Singapore(新加坡国立大学);Shanghai AI Laboratory(上海人工智能实验室);Midea AI Research Center(美的AI研究中心)
EEG-MedRAG: Enhancing EEG-based Clinical Decision-Making via Hierarchical Hypergraph Retrieval-Augmented Generation
College of Information Science and Engineering, Northeastern University(信息科学与工程学院,东北大学);Foshan Graduate School of Innovation, Northeastern University(创新佛山研究生院,东北大学);College of Computing and Data Science, Nanyang Technological University(计算与数据科学学院,南洋理工大学);Institute of Automation, Chinese Academy of Sciences(自动化研究所,中国科学院);Squirrel Ai Learning
Enhancing Test-Time Scaling of Large Language Models with Hierarchical Retrieval-Augmented MCTS
The Ohio State University(俄亥俄州立大学);Case Western Reserve University(凯斯西储大学);Duke University(杜克大学);University of Hong Kong(香港大学);City University of Hong Kong(香港城市大学);ByteDance(字节跳动)
Dynamic Language Group-Based MoE: Enhancing Code-Switching Speech Recognition with Hierarchical Routing
School of Informatics, Xiamen University(厦门大学信息学院);School of Electronic Science and Engineering, Xiamen University(厦门大学电子科学与技术学院);Kuaishou Technology Co., Ltd(快手科技有限公司)
Enhancing Knowledge Density via Hierarchical Context Pruning for Retrieval-Augmented Repository-Level Code Completion
MIIT Key Laboratory of Pattern Analysis and Machine Intelligence(信息科技模式分析与机器智能国家重点实验室);College of Computer Science and Technology(计算机科学与技术学院);Huaneng Information Technology Co Ltd(华能信息技术有限公司);Hunan Vanguard Group Corporation Co Ltd(湖南 Vanguard集团有限公司);Tsinghua University(清华大学);Zhejiang University(浙江大学);PowerChina Huadong Engineering Co Ltd(国家电网华东工程有限公司)
Hierarchical Softmax for End-to-End Low-resource Multilingual Speech Recognition
Kyoto University(京都大学);Xinjiang University(新疆大学);The University of Tokyo(东京大学);National Institute of Information and Communications Technology (NICT)(日本信息通信研究机构)