[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"similar-roomylee--awesome-relation-extraction":3,"tool-roomylee--awesome-relation-extraction":61},[4,18,26,36,44,53],{"id":5,"name":6,"github_repo":7,"description_zh":8,"stars":9,"difficulty_score":10,"last_commit_at":11,"category_tags":12,"status":17},4358,"openclaw","openclaw\u002Fopenclaw","OpenClaw 是一款专为个人打造的本地化 AI 助手，旨在让你在自己的设备上拥有完全可控的智能伙伴。它打破了传统 AI 助手局限于特定网页或应用的束缚，能够直接接入你日常使用的各类通讯渠道，包括微信、WhatsApp、Telegram、Discord、iMessage 等数十种平台。无论你在哪个聊天软件中发送消息，OpenClaw 都能即时响应，甚至支持在 macOS、iOS 和 Android 设备上进行语音交互，并提供实时的画布渲染功能供你操控。\n\n这款工具主要解决了用户对数据隐私、响应速度以及“始终在线”体验的需求。通过将 AI 部署在本地，用户无需依赖云端服务即可享受快速、私密的智能辅助，真正实现了“你的数据，你做主”。其独特的技术亮点在于强大的网关架构，将控制平面与核心助手分离，确保跨平台通信的流畅性与扩展性。\n\nOpenClaw 非常适合希望构建个性化工作流的技术爱好者、开发者，以及注重隐私保护且不愿被单一生态绑定的普通用户。只要具备基础的终端操作能力（支持 macOS、Linux 及 Windows WSL2），即可通过简单的命令行引导完成部署。如果你渴望拥有一个懂你",349277,3,"2026-04-06T06:32:30",[13,14,15,16],"Agent","开发框架","图像","数据工具","ready",{"id":19,"name":20,"github_repo":21,"description_zh":22,"stars":23,"difficulty_score":10,"last_commit_at":24,"category_tags":25,"status":17},3808,"stable-diffusion-webui","AUTOMATIC1111\u002Fstable-diffusion-webui","stable-diffusion-webui 是一个基于 Gradio 构建的网页版操作界面，旨在让用户能够轻松地在本地运行和使用强大的 Stable Diffusion 图像生成模型。它解决了原始模型依赖命令行、操作门槛高且功能分散的痛点，将复杂的 AI 绘图流程整合进一个直观易用的图形化平台。\n\n无论是希望快速上手的普通创作者、需要精细控制画面细节的设计师，还是想要深入探索模型潜力的开发者与研究人员，都能从中获益。其核心亮点在于极高的功能丰富度：不仅支持文生图、图生图、局部重绘（Inpainting）和外绘（Outpainting）等基础模式，还独创了注意力机制调整、提示词矩阵、负向提示词以及“高清修复”等高级功能。此外，它内置了 GFPGAN 和 CodeFormer 等人脸修复工具，支持多种神经网络放大算法，并允许用户通过插件系统无限扩展能力。即使是显存有限的设备，stable-diffusion-webui 也提供了相应的优化选项，让高质量的 AI 艺术创作变得触手可及。",162132,"2026-04-05T11:01:52",[14,15,13],{"id":27,"name":28,"github_repo":29,"description_zh":30,"stars":31,"difficulty_score":32,"last_commit_at":33,"category_tags":34,"status":17},1381,"everything-claude-code","affaan-m\u002Feverything-claude-code","everything-claude-code 是一套专为 AI 编程助手（如 Claude Code、Codex、Cursor 等）打造的高性能优化系统。它不仅仅是一组配置文件，而是一个经过长期实战打磨的完整框架，旨在解决 AI 代理在实际开发中面临的效率低下、记忆丢失、安全隐患及缺乏持续学习能力等核心痛点。\n\n通过引入技能模块化、直觉增强、记忆持久化机制以及内置的安全扫描功能，everything-claude-code 能显著提升 AI 在复杂任务中的表现，帮助开发者构建更稳定、更智能的生产级 AI 代理。其独特的“研究优先”开发理念和针对 Token 消耗的优化策略，使得模型响应更快、成本更低，同时有效防御潜在的攻击向量。\n\n这套工具特别适合软件开发者、AI 研究人员以及希望深度定制 AI 工作流的技术团队使用。无论您是在构建大型代码库，还是需要 AI 协助进行安全审计与自动化测试，everything-claude-code 都能提供强大的底层支持。作为一个曾荣获 Anthropic 黑客大奖的开源项目，它融合了多语言支持与丰富的实战钩子（hooks），让 AI 真正成长为懂上",148568,2,"2026-04-09T23:34:24",[14,13,35],"语言模型",{"id":37,"name":38,"github_repo":39,"description_zh":40,"stars":41,"difficulty_score":32,"last_commit_at":42,"category_tags":43,"status":17},2271,"ComfyUI","Comfy-Org\u002FComfyUI","ComfyUI 是一款功能强大且高度模块化的视觉 AI 引擎，专为设计和执行复杂的 Stable Diffusion 图像生成流程而打造。它摒弃了传统的代码编写模式，采用直观的节点式流程图界面，让用户通过连接不同的功能模块即可构建个性化的生成管线。\n\n这一设计巧妙解决了高级 AI 绘图工作流配置复杂、灵活性不足的痛点。用户无需具备编程背景，也能自由组合模型、调整参数并实时预览效果，轻松实现从基础文生图到多步骤高清修复等各类复杂任务。ComfyUI 拥有极佳的兼容性，不仅支持 Windows、macOS 和 Linux 全平台，还广泛适配 NVIDIA、AMD、Intel 及苹果 Silicon 等多种硬件架构，并率先支持 SDXL、Flux、SD3 等前沿模型。\n\n无论是希望深入探索算法潜力的研究人员和开发者，还是追求极致创作自由度的设计师与资深 AI 绘画爱好者，ComfyUI 都能提供强大的支持。其独特的模块化架构允许社区不断扩展新功能，使其成为当前最灵活、生态最丰富的开源扩散模型工具之一，帮助用户将创意高效转化为现实。",108111,"2026-04-08T11:23:26",[14,15,13],{"id":45,"name":46,"github_repo":47,"description_zh":48,"stars":49,"difficulty_score":32,"last_commit_at":50,"category_tags":51,"status":17},6121,"gemini-cli","google-gemini\u002Fgemini-cli","gemini-cli 是一款由谷歌推出的开源 AI 命令行工具，它将强大的 Gemini 大模型能力直接集成到用户的终端环境中。对于习惯在命令行工作的开发者而言，它提供了一条从输入提示词到获取模型响应的最短路径，无需切换窗口即可享受智能辅助。\n\n这款工具主要解决了开发过程中频繁上下文切换的痛点，让用户能在熟悉的终端界面内直接完成代码理解、生成、调试以及自动化运维任务。无论是查询大型代码库、根据草图生成应用，还是执行复杂的 Git 操作，gemini-cli 都能通过自然语言指令高效处理。\n\n它特别适合广大软件工程师、DevOps 人员及技术研究人员使用。其核心亮点包括支持高达 100 万 token 的超长上下文窗口，具备出色的逻辑推理能力；内置 Google 搜索、文件操作及 Shell 命令执行等实用工具；更独特的是，它支持 MCP（模型上下文协议），允许用户灵活扩展自定义集成，连接如图像生成等外部能力。此外，个人谷歌账号即可享受免费的额度支持，且项目基于 Apache 2.0 协议完全开源，是提升终端工作效率的理想助手。",100752,"2026-04-10T01:20:03",[52,13,15,14],"插件",{"id":54,"name":55,"github_repo":56,"description_zh":57,"stars":58,"difficulty_score":32,"last_commit_at":59,"category_tags":60,"status":17},4721,"markitdown","microsoft\u002Fmarkitdown","MarkItDown 是一款由微软 AutoGen 团队打造的轻量级 Python 工具，专为将各类文件高效转换为 Markdown 格式而设计。它支持 PDF、Word、Excel、PPT、图片（含 OCR）、音频（含语音转录）、HTML 乃至 YouTube 链接等多种格式的解析，能够精准提取文档中的标题、列表、表格和链接等关键结构信息。\n\n在人工智能应用日益普及的今天，大语言模型（LLM）虽擅长处理文本，却难以直接读取复杂的二进制办公文档。MarkItDown 恰好解决了这一痛点，它将非结构化或半结构化的文件转化为模型“原生理解”且 Token 效率极高的 Markdown 格式，成为连接本地文件与 AI 分析 pipeline 的理想桥梁。此外，它还提供了 MCP（模型上下文协议）服务器，可无缝集成到 Claude Desktop 等 LLM 应用中。\n\n这款工具特别适合开发者、数据科学家及 AI 研究人员使用，尤其是那些需要构建文档检索增强生成（RAG）系统、进行批量文本分析或希望让 AI 助手直接“阅读”本地文件的用户。虽然生成的内容也具备一定可读性，但其核心优势在于为机器",93400,"2026-04-06T19:52:38",[52,14],{"id":62,"github_repo":63,"name":64,"description_en":65,"description_zh":66,"ai_summary_zh":67,"readme_en":68,"readme_zh":69,"quickstart_zh":70,"use_case_zh":71,"hero_image_url":72,"owner_login":73,"owner_name":74,"owner_avatar_url":75,"owner_bio":76,"owner_company":77,"owner_location":78,"owner_email":79,"owner_twitter":80,"owner_website":81,"owner_url":82,"languages":80,"stars":83,"forks":84,"last_commit_at":85,"license":80,"difficulty_score":86,"env_os":87,"env_gpu":88,"env_ram":88,"env_deps":89,"category_tags":92,"github_topics":94,"view_count":32,"oss_zip_url":80,"oss_zip_packed_at":80,"status":17,"created_at":113,"updated_at":114,"faqs":115,"releases":116},6056,"roomylee\u002Fawesome-relation-extraction","awesome-relation-extraction","📖 A curated list of awesome resources dedicated to Relation Extraction, one of the most important tasks in Natural Language Processing (NLP).","awesome-relation-extraction 是一个专为自然语言处理（NLP）领域打造的精选资源库，核心聚焦于“关系抽取”这一关键任务。简单来说，它的目标是帮助计算机从文本中自动识别并理解实体之间的关联（例如从“乔布斯创立了苹果”中提取出“创立”这一关系）。\n\n面对该领域论文繁多、技术迭代快、资源分散的痛点，awesome-relation-extraction 提供了一站式的解决方案。它系统性地整理了从经典综述、前沿学术论文到公开数据集、教学视频及代码框架的各类高质量资源。内容覆盖广泛，不仅包含基于 CNN 的监督学习方法，还深入涵盖了远程监督、图神经网络（GNN）、基于 Transformer 的语言模型（如 BERT 等编码器与解码器架构）、知识图谱融合以及少样本学习等最新技术方向。\n\n这份资源清单特别适合 NLP 研究人员、算法工程师及相关领域的开发者使用。对于希望快速把握学术动态的研究者，它提供了详尽的文献指引；对于需要复现模型或寻找基线的开发者，它直接链接了宝贵的代码实现与数据集。无论是刚入门的新手还是资深专家，都能通过 awesome-relation-extr","awesome-relation-extraction 是一个专为自然语言处理（NLP）领域打造的精选资源库，核心聚焦于“关系抽取”这一关键任务。简单来说，它的目标是帮助计算机从文本中自动识别并理解实体之间的关联（例如从“乔布斯创立了苹果”中提取出“创立”这一关系）。\n\n面对该领域论文繁多、技术迭代快、资源分散的痛点，awesome-relation-extraction 提供了一站式的解决方案。它系统性地整理了从经典综述、前沿学术论文到公开数据集、教学视频及代码框架的各类高质量资源。内容覆盖广泛，不仅包含基于 CNN 的监督学习方法，还深入涵盖了远程监督、图神经网络（GNN）、基于 Transformer 的语言模型（如 BERT 等编码器与解码器架构）、知识图谱融合以及少样本学习等最新技术方向。\n\n这份资源清单特别适合 NLP 研究人员、算法工程师及相关领域的开发者使用。对于希望快速把握学术动态的研究者，它提供了详尽的文献指引；对于需要复现模型或寻找基线的开发者，它直接链接了宝贵的代码实现与数据集。无论是刚入门的新手还是资深专家，都能通过 awesome-relation-extraction 高效地获取所需信息，避免在海量资料中盲目摸索，从而更专注于算法创新与应用落地。","# Awesome Relation Extraction [![Awesome](https:\u002F\u002Fawesome.re\u002Fbadge.svg)](https:\u002F\u002Fawesome.re)\n![awesome_re](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Froomylee_awesome-relation-extraction_readme_1ceb71138243.png)\n\nA curated list of awesome resources dedicated to Relation Extraction, inspired by [awesome-nlp](https:\u002F\u002Fgithub.com\u002Fkeon\u002Fawesome-nlp) and [awesome-deep-vision](https:\u002F\u002Fgithub.com\u002Fkjw0612\u002Fawesome-deep-vision).\n\n**Contributing**: Please feel free to make *[pull requests](https:\u002F\u002Fgithub.com\u002Froomylee\u002Fawesome-relation-extraction\u002Fpulls)*.\n\n## Contents\n* [Research Trends and Surveys](#research-trends-and-surveys)\n* [Papers](#papers)\n\t* [Supervised Approaches](#supervised-approaches)\n\t* [Distant Supervision Approaches](#distant-supervision-approaches)\n    * [GNN-based Models](#gnn-based-models)\n\t* [Language Models](#language-models)\n\t    * [Encoder Representation from Transformer](#encoder-representation-from-transformer)\n\t    * [Decoder Representation from Transformer](#decoder-representation-from-transformer)\n    * [Knowledge Graph Based Approaches](#knowledge-graph-based-approaches)\n    * [Few-Shot Learning Approaches](#few-shot-learning-approaches)\n* [Datasets](#datasets)\n* [Videos and Lectures](#videos-and-lectures)\n* [Systems](#systems)\n* [Frameworks](#frameworks)\n\n\n## Research Trends and Surveys\n* [NLP progress: Relationship Extraction](https:\u002F\u002Fnlpprogress.com\u002Fenglish\u002Frelationship_extraction.html)\n* [Named Entity Recognition and Relation Extraction:State-of-the-Art](https:\u002F\u002Fwww.researchgate.net\u002Fprofile\u002FSyed-Waqar-Jaffry\u002Fpublication\u002F345315661_Named_Entity_Recognition_and_Relation_Extraction_State_of_the_Art\u002Flinks\u002F603015aaa6fdcc37a83aafd5\u002FNamed-Entity-Recognition-and-Relation-Extraction-State-of-the-Art.pdf) (Nasar et al., 2021)\n* [A Survey of Deep Learning Methods for Relation Extraction](https:\u002F\u002Farxiv.org\u002Fabs\u002F1705.03645) (Kumar, 2017)\n* [A Survey on Relation Extraction](https:\u002F\u002Fwww.cs.cmu.edu\u002F~nbach\u002Fpapers\u002FA-survey-on-Relation-Extraction.pdf) (Bach and Badaskar, 2017)\n* [Relation Extraction: A Survey](https:\u002F\u002Farxiv.org\u002Fabs\u002F1712.05191) (Pawar et al., 2017)\n* [A Review on Entity Relation Extraction](https:\u002F\u002Fieeexplore.ieee.org\u002Fstamp\u002Fstamp.jsp?arnumber=8269916) (Zhang et al., 2017)\n* [Review of Relation Extraction Methods: What is New Out There?](https:\u002F\u002Fciteseerx.ist.psu.edu\u002Fviewdoc\u002Fdownload?doi=10.1.1.727.1005&rep=rep1&type=pdf) (Konstantinova et al., 2014)\n* [100 Best Github: Relation Extraction](http:\u002F\u002Fmeta-guide.com\u002Fsoftware-meta-guide\u002F100-best-github-relation-extraction)\n\n\n## Papers\n\n### Supervised Approaches\n#### CNN-based Models\n* Convolution Neural Network for Relation Extraction [[paper]](https:\u002F\u002Flink.springer.com\u002Fchapter\u002F10.1007\u002F978-3-642-53917-6_21) [[code]](https:\u002F\u002Fgithub.com\u002Froomylee\u002Fcnn-relation-extraction) [[review]](https:\u002F\u002Fgithub.com\u002Froomylee\u002Fpaper-review\u002Fblob\u002Fmaster\u002Frelation_extraction\u002FConvolution%20Neural%20Network%20for%20Relation%20Extraction\u002Freview.md)\n\t* ChunYang Liu, WenBo Sun, WenHan Chao and WanXiang Che\n\t* ADMA 2013\n* Relation Classification via Convolutional Deep Neural Network [[paper]](http:\u002F\u002Fwww.aclweb.org\u002Fanthology\u002FC14-1220) [[code]](https:\u002F\u002Fgithub.com\u002Froomylee\u002Fcnn-relation-extraction) [[review]](https:\u002F\u002Fgithub.com\u002Froomylee\u002Fpaper-review\u002Fblob\u002Fmaster\u002Frelation_extraction\u002FRelation_Classification_via_Convolutional_Deep_Neural_Network\u002Freview.md)\n\t* Daojian Zeng, Kang Liu, Siwei Lai, Guangyou Zhou and Jun Zhao\n\t* COLING 2014\n* Relation Extraction: Perspective from Convolutional Neural Networks [[paper]](http:\u002F\u002Fwww.cs.nyu.edu\u002F~thien\u002Fpubs\u002Fvector15.pdf) [[code]](https:\u002F\u002Fgithub.com\u002Froomylee\u002Fcnn-relation-extraction) [[review]](https:\u002F\u002Fgithub.com\u002Froomylee\u002Fpaper-review\u002Fblob\u002Fmaster\u002Frelation_extraction\u002FRelation_Extraction-Perspective_from_Convolutional_Neural_Networks\u002Freview.md)\n\t* Thien Huu Nguyen and Ralph Grishman\n\t* NAACL 2015\n* Classifying Relations by Ranking with Convolutional Neural Networks [[paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1504.06580) [[code]](https:\u002F\u002Fgithub.com\u002Fpratapbhanu\u002FCRCNN)\n\t* Cicero Nogueira dos Santos, Bing Xiang and Bowen Zhou\n\t* ACL 2015\n* Attention-Based Convolutional Neural Network for Semantic Relation Extraction [[paper]](http:\u002F\u002Fwww.aclweb.org\u002Fanthology\u002FC16-1238) [[code]](https:\u002F\u002Fgithub.com\u002Fnicolay-r\u002Fmlp-attention)\n\t* Yatian Shen and Xuanjing Huang\n\t* COLING 2016\n* Relation Classification via Multi-Level Attention CNNs [[paper]](http:\u002F\u002Faclweb.org\u002Fanthology\u002FP16-1123) [[code]](https:\u002F\u002Fgithub.com\u002FlawlietAi\u002Frelation-classification-via-attention-model)\n\t* Linlin Wang, Zhu Cao, Gerard de Melo and Zhiyuan Liu\n\t* ACL 2016\n* MIT at SemEval-2017 Task 10: Relation Extraction with Convolutional Neural Networks [[paper]](https:\u002F\u002Faclanthology.info\u002Fpdf\u002FS\u002FS17\u002FS17-2171.pdf)\n\t* Ji Young Lee, Franck Dernoncourt and Peter Szolovits\n\t* SemEval 2017\n\n#### RNN-based Models\n* Relation Classification via Recurrent Neural Network [[paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1508.01006)\n\t* Dongxu Zhang and Dong Wang\n\t* arXiv 2015\n* Bidirectional Long Short-Term Memory Networks for Relation Classification [[paper]](http:\u002F\u002Fwww.aclweb.org\u002Fanthology\u002FY15-1009)\n\t* Shu Zhang, Dequan Zheng, Xinchen Hu and Ming Yang\n\t* PACLIC 2015\n* End-to-End Relation Extraction using LSTMs on Sequences and Tree Structure [[paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1601.00770)\n\t* Makoto Miwa and Mohit Bansal\n\t* ACL 2016\n* Attention-Based Bidirectional Long Short-Term Memory Networks for Relation Classification [[paper]](http:\u002F\u002Fanthology.aclweb.org\u002FP16-2034) [[code]](https:\u002F\u002Fgithub.com\u002FSeoSangwoo\u002FAttention-Based-BiLSTM-relation-extraction)\n\t* Peng Zhou, Wei Shi, Jun Tian, Zhenyu Qi, Bingchen Li, Hongwei Hao and Bo Xu\n\t* ACL 2016\n* Semantic Relation Classification via Hierarchical Recurrent Neural Network with Attention [[paper]](http:\u002F\u002Fwww.aclweb.org\u002Fanthology\u002FC16-1119)\n\t* Minguang Xiao and Cong Liu\n\t* COLING 2016\n* Semantic Relation Classification via Bidirectional LSTM Networks with Entity-aware Attention using Latent Entity Typing [[paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1901.08163) [[code]](https:\u002F\u002Fgithub.com\u002Froomylee\u002Fentity-aware-relation-classification)\n\t* Joohong Lee, Sangwoo Seo and Yong Suk Choi\n\t* arXiv 2019\n\n#### Dependency-based Models\n* Semantic Compositionality through Recursive Matrix-Vector Spaces [[paper]](http:\u002F\u002Faclweb.org\u002Fanthology\u002FD12-1110) [[code]](https:\u002F\u002Fgithub.com\u002Fpratapbhanu\u002FMVRNN)\n\t* Richard Socher, Brody Huval, Christopher D. Manning and Andrew Y. Ng\n\t* EMNLP-CoNLL 2012\n* Factor-based Compositional Embedding Models [[paper]](https:\u002F\u002Fwww.cs.cmu.edu\u002F~mgormley\u002Fpapers\u002Fyu+gormley+dredze.nipsw.2014.pdf)\n\t* Mo Yu, Matthw R. Gormley and Mark Dredze\n\t* NIPS Workshop on Learning Semantics 2014\n* A Dependency-Based Neural Network for Relation Classification [[paper]](http:\u002F\u002Fwww.aclweb.org\u002Fanthology\u002FP15-2047)\n\t* Yang Liu, Furu Wei, Sujian Li, Heng Ji, Ming Zhou and Houfeng Wang\n\t* ACL 2015\n* Classifying Relations via Long Short Term Memory Networks along Shortest Dependency Path [[paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1508.03720) [[code]](https:\u002F\u002Fgithub.com\u002FSshanu\u002FRelation-Classification)\n\t* Xu Yan, Lili Mou, Ge Li, Yunchuan Chen, Hao Peng and Zhi Jin\n\t* EMNLP 2015\n* Semantic Relation Classification via Convolutional Neural Networks with Simple Negative Sampling [[paper]](https:\u002F\u002Fwww.aclweb.org\u002Fanthology\u002FD\u002FD15\u002FD15-1062.pdf)\n\t* Kun Xu, Yansong Feng, Songfang Huang and Dongyan Zhao\n\t* EMNLP 2015\n* Improved Relation Classification by Deep Recurrent Neural Networks with Data Augmentation [[paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1601.03651)\n\t* Yan Xu, Ran Jia, Lili Mou, Ge Li, Yunchuan Chen, Yangyang Lu and Zhi Jin\n\t* COLING 2016\n* Bidirectional Recurrent Convolutional Neural Network for Relation Classification [[paper]](http:\u002F\u002Fwww.aclweb.org\u002Fanthology\u002FP16-1072)\n\t* Rui Cai, Xiaodong Zhang and Houfeng Wang\n\t* ACL 2016\n* Neural Relation Extraction via Inner-Sentence Noise Reduction and Transfer Learning [[paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1808.06738)\n\t* Tianyi Liu, Xinsong Zhang, Wanhao Zhou, Weijia Jia\n\t* EMNLP 2018\n\t\n#### GNN-based Models\n* Matching the Blanks: Distributional Similarity for Relation Learning [[paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1906.03158)\n    * Livio Baldini Soares, Nicholas FitzGerald, Jeffrey Ling, Tom Kwiatkowski\n    * ACL 2019\n* Relation of the Relations: A New Paradigm of the Relation Extraction Problem [[paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2006.03719)\n\t* Zhijing Jin, Yongyi Yang, Xipeng Qiu, Zheng Zhang\n\t* EMNLP 2020\n* GDPNet: Refining Latent Multi-View Graph for Relation Extraction \n    [[paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2012.06780.pdf)\n    [[code]](https:\u002F\u002Fgithub.com\u002FXueFuzhao\u002FGDPNet)\n    * Fuzhao Xue, Aixin Sun, Hao Zhang, Eng Siong Chng\n    * AAAI 21\n* RECON: Relation Extraction using Knowledge Graph Context in a Graph Neural Network\n    [[parer]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2009.08694.pdf)\n    [[code]](https:\u002F\u002Fgithub.com\u002Fansonb\u002FRECON)\n    * Anson Bastos, Abhishek Nadgeri, Kuldeep Singh, Isaiah Onando Mulang', Saeedeh Shekarpour, Johannes Hoffart, Manohar Kaul\n    * WWW'21\n\t\n### Distant Supervision Approaches\n* Distant supervision for relation extraction without labeled data [[paper]](https:\u002F\u002Fweb.stanford.edu\u002F~jurafsky\u002Fmintz.pdf) [[review]](https:\u002F\u002Fgithub.com\u002Froomylee\u002Fpaper-review\u002Fblob\u002Fmaster\u002Frelation_extraction\u002FDistant_supervision_for_relation_extraction_without_labeled_data\u002Freview.md)\n\t* Mike Mintz, Steven Bills, Rion Snow and Dan Jurafsky\n\t* ACL 2009\n* Knowledge-Based Weak Supervision for Information Extraction of Overlapping Relations [[paper]](http:\u002F\u002Fwww.aclweb.org\u002Fanthology\u002FP11-1055) [[code]](http:\u002F\u002Faiweb.cs.washington.edu\u002Fai\u002Fraphaelh\u002Fmr\u002F)\n\t* Raphael Hoffmann, Congle Zhang, Xiao Ling, Luke Zettlemoyer and Daniel S. Weld\n\t* ACL 2011\n* Multi-instance Multi-label Learning for Relation Extraction [[paper]](http:\u002F\u002Fwww.aclweb.org\u002Fanthology\u002FD12-1042) [[code]](https:\u002F\u002Fnlp.stanford.edu\u002Fsoftware\u002Fmimlre.shtml)\n\t* Mihai Surdeanu, Julie Tibshirani, Ramesh Nallapati and Christopher D. Manning\n\t* EMNLP-CoNLL 2012\n* Distant Supervision for Relation Extraction via Piecewise Convolutional Neural Networks [[paper]](http:\u002F\u002Fwww.emnlp2015.org\u002Fproceedings\u002FEMNLP\u002Fpdf\u002FEMNLP203.pdf) [[review]](https:\u002F\u002Fgithub.com\u002Froomylee\u002Fpaper-review\u002Fblob\u002Fmaster\u002Frelation_extraction\u002FDistant_Supervision_for_Relation_Extraction_via_Piecewise_Convolutional_Neural_Networks\u002Freview.md) [[code]](https:\u002F\u002Fgithub.com\u002Fnicolay-r\u002Fsentiment-pcnn)\n\t* Daojian Zeng, Kang Liu, Yubo Chen and Jun Zhao\n\t* EMNLP 2015\n* Relation Extraction with Multi-instance Multi-label Convolutional Neural Networks [[paper]](https:\u002F\u002Fpdfs.semanticscholar.org\u002F8731\u002F369a707046f3f8dd463d1fd107de31d40a24.pdf) [[review]](https:\u002F\u002Fgithub.com\u002Froomylee\u002Fpaper-review\u002Fblob\u002Fmaster\u002Frelation_extraction\u002FRelation_Extraction_with_Multi-instance_Multi-label_Convolutional_Neural_Networks\u002Freview.md) [[code]](https:\u002F\u002Fgithub.com\u002Fmay-\u002Fcnn-re-tf)\n\t* Xiaotian Jiang, Quan Wang, Peng Li, Bin Wang\n\t* COLING 2016\n* Incorporating Relation Paths in Neural Relation Extraction [[paper]](http:\u002F\u002Faclweb.org\u002Fanthology\u002FD17-1186) [[review]](https:\u002F\u002Fgithub.com\u002Froomylee\u002Fpaper-review\u002Fblob\u002Fmaster\u002Frelation_extraction\u002FIncorporating_Relation_Paths_in_Neural_Relation_Extraction\u002Freview.md)\n\t* Wenyuan Zeng, Yankai Lin, Zhiyuan Liu and Maosong Sun\n\t* EMNLP 2017\n* Neural Relation Extraction with Selective Attention over Instances [[paper]](http:\u002F\u002Fwww.aclweb.org\u002Fanthology\u002FP16-1200) [[code]](https:\u002F\u002Fgithub.com\u002Fthunlp\u002FOpenNRE\u002F)\n\t* Yankai Lin, Shiqi Shen, Zhiyuan Liu, Huanbo Luan and Maosong Sun\n\t* ACL 2017\n* Learning local and global contexts using a convolutional recurrent network model for relation classification in biomedical text [[paper]](http:\u002F\u002Fwww.aclweb.org\u002Fanthology\u002FK17-1032) [[code]](https:\u002F\u002Fgithub.com\u002Fdesh2608\u002Fcrnn-relation-classification) [[code]](https:\u002F\u002Fgithub.com\u002Fkwonmha\u002FConvolutional-Recurrent-Neural-Networks-for-Relation-Extraction)\n\t* Desh Raj, Sunil Kumar Sahu and Ashish Anan\n\t* CoNLL 2017\n* Hierarchical Relation Extraction with Coarse-to-Fine Grained Attention[[paper]](https:\u002F\u002Faclweb.org\u002Fanthology\u002FD18-1247)[[code]](https:\u002F\u002Fgithub.com\u002Fthunlp\u002FHNRE)\n\t* Xu Han, Pengfei Yu∗, Zhiyuan Liu, Maosong Sun, Peng Li\n\t* EMNLP 2018\n* RESIDE: Improving Distantly-Supervised Neural Relation Extraction using Side Information [[paper]](http:\u002F\u002Fmalllabiisc.github.io\u002Fpublications\u002Fpapers\u002Freside_emnlp18.pdf) [[code]](https:\u002F\u002Fgithub.com\u002Fmalllabiisc\u002FRESIDE)\n\t* Shikhar Vashishth, Rishabh Joshi, Sai Suman Prayaga, Chiranjib Bhattacharyya and Partha Talukdar\n\t* EMNLP 2018\n* Distant Supervision Relation Extraction with Intra-Bag and Inter-Bag Attentions\n    [[paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1904.00143.pdf)\n    [[code]](https:\u002F\u002Fgithub.com\u002FZhixiuYe\u002FIntra-Bag-and-Inter-Bag-Attentions)\n    * Zhi-Xiu Ye, Zhen-Hua Ling\n    * NAACL 2019\n\t\n### Language Models\n\n#### Encoder Representation from Transformer\n* Enriching Pre-trained Language Model with Entity Information for Relation Classification [[paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1905.08284.pdf)\n    * Shanchan Wu, Yifan He\n    * arXiv 2019\n* LUKE: Deep Contextualized Entity Representations with Entity-aware Self-attention\n    [[paper]](https:\u002F\u002Fwww.aclweb.org\u002Fanthology\u002F2020.emnlp-main.523\u002F)\n    [[code]](https:\u002F\u002Fgithub.com\u002Fstudio-ousia\u002Fluke)\n    * Ikuya Yamada, Akari Asai, Hiroyuki Shindo, Hideaki Takeda, Yuji Matsumoto\n    * EMNLP 2020\n* SpanBERT: Improving pre-training by representing and predicting spans [[paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1907.10529.pdf) [[code]](https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002FSpanBERT)\n    * Mandar Joshi, Danqi Chen, Yinhan Liu, Daniel S. Weld, Luke Zettlemoyer and Omer Levy\n    * TACL 2020 (Transactions of the Association for Computational Linguistics)\n* Efficient long-distance relation extraction with DG-SpanBERT\n    [[paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2004.03636)\n    * Jun Chen, Robert Hoehndorf, Mohamed Elhoseiny, Xiangliang Zhang\n    \n#### Decoder Representation from Transformer\n* Improving Relation Extraction by Pretrained Language Representations\n    [[paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1906.03088)\n    [[review]](https:\u002F\u002Fopenreview.net\u002Fforum?id=BJgrxbqp67)\n    [[code]](https:\u002F\u002Fgithub.com\u002FDFKI-NLP\u002FTRE)\n    * Christoph Alt, Marc Hübner, Leonhard Hennig\n    * AKBC 19\n\n### Knowledge Graph Based Approaches\n* KGPool: Dynamic Knowledge Graph Context Selection for Relation Extraction\n  \t[[paper]](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2106.00459.pdf)\n  \t[[code]](https:\u002F\u002Fgithub.com\u002Fnadgeri14\u002FKGPool)\n\t* Abhishek Nadgeri, Anson Bastos, Kuldeep Singh, Isaiah Onando Mulang, Johannes Hoffart, Saeedeh Shekarpour, and Vijay Saraswat\n\t* ACL 2021 (findings)\n\n### Few-Shot Learning Approaches\n* FewRel: A Large-Scale Supervised Few-Shot Relation Classification Dataset with State-of-the-Art Evaluation [[paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1810.10147) [[website]](http:\u002F\u002Fzhuhao.me\u002Ffewrel) [[code]](https:\u002F\u002Fgithub.com\u002FProKil\u002FFewRel)\n\t* Xu Han, Hao Zhu, Pengfei Yu, Ziyun Wang, Yuan Yao, Zhiyuan Liu, Maosong Sun\n\t* EMNLP 2018\n\n### Miscellaneous\n* Jointly Extracting Relations with Class Ties via Effective Deep Ranking [[paper]](http:\u002F\u002Faclweb.org\u002Fanthology\u002FP17-1166)\n\t* Hai Ye, Wenhan Chao, Zhunchen Luo and Zhoujun Li\n\t* ACL 2017\n* End-to-End Neural Relation Extraction with Global Optimization [[paper]](http:\u002F\u002Faclweb.org\u002Fanthology\u002FD17-1182)\n\t* Meishan Zhang, Yue Zhang and Guohong Fu\n\t* EMNLP 2017\n* Adversarial Training for Relation Extraction [[paper]](https:\u002F\u002Fpeople.eecs.berkeley.edu\u002F~russell\u002Fpapers\u002Femnlp17-relation.pdf)\n\t* Yi Wu, David Bamman and Stuart Russell\n\t* EMNLP 2017\n* A neural joint model for entity and relation extraction from biomedical text[[paper]](https:\u002F\u002Fbmcbioinformatics.biomedcentral.com\u002Farticles\u002F10.1186\u002Fs12859-017-1609-9)\n\t* Fei Li, Meishan Zhang, Guohong Fu and Donghong Ji\n\t* BMC bioinformatics 2017\n* Joint Extraction of Entities and Relations Using Reinforcement Learning and Deep Learning [[paper]](https:\u002F\u002Fwww.hindawi.com\u002Fjournals\u002Fcin\u002F2017\u002F7643065\u002F)\n\t* Yuntian Feng, Hongjun Zhang, Wenning Hao, and Gang Chen\n\t* Journal of Computational Intelligence and Neuroscience 2017\n* TDEER: An Efficient Translating Decoding Schema for Joint Extraction of Entities and Relations [[paper]](https:\u002F\u002Faclanthology.org\u002F2021.emnlp-main.635\u002F) [[code]](https:\u002F\u002Fgithub.com\u002F4ai\u002Ftdeer)\n\t* Xianming Li, Xiaotian Luo, Chenghao Dong, Daichuan Yang, Beidi Luan and Zhen He\n\t* EMNLP 2021\n\n[Back to Top](#contents)\n\n\n## Datasets\n* SemEval-2010 Task 8 [[paper]](http:\u002F\u002Fwww.aclweb.org\u002Fanthology\u002FS10-1006) [[download]](https:\u002F\u002Fdocs.google.com\u002Fleaf?id=0B_jQiLugGTAkMDQ5ZjZiMTUtMzQ1Yy00YWNmLWJlZDYtOWY1ZDMwY2U4YjFk&sort=name&layout=list&num=50)\n\t* Multi-Way Classification of Semantic Relations Between Pairs of Nominals\n* New York Times (NYT) Corpus [[paper]](http:\u002F\u002Fwww.riedelcastro.org\u002F\u002Fpublications\u002Fpapers\u002Friedel10modeling.pdf) [[download]](https:\u002F\u002Fcatalog.ldc.upenn.edu\u002FLDC2008T19)\n\t* This dataset was generated by aligning *Freebase* relations with the NYT corpus, with sentences from the years 2005-2006 used as the training corpus and sentences from 2007 used as the testing corpus.\n* FewRel: Few-Shot Relation Classification Dataset [[paper]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1810.10147) [[Website]](http:\u002F\u002Fzhuhao.me\u002Ffewrel)\n\t* This dataset is a supervised few-shot relation classification dataset. The corpus is Wikipedia and the knowledge base used to annotate the corpus is Wikidata.\n* TACRED: The TAC Relation Extraction Dataset \n    [[paper]](https:\u002F\u002Fwww.aclweb.org\u002Fanthology\u002FD17-1004.pdf) \n    [[Website]](https:\u002F\u002Fnlp.stanford.edu\u002Fprojects\u002Ftacred\u002F) \n    [[download]](https:\u002F\u002Fcatalog.ldc.upenn.edu\u002FLDC2018T24)\n    * Is a large-scale relation extraction dataset with built over newswire and web text from the corpus used in the yearly TAC Knowledge Base Population (TAC KBP) challenges.\n* ACE05: \n    [[Website]](https:\u002F\u002Fcatalog.ldc.upenn.edu\u002FLDC2006T06) \n    [[download-info]](https:\u002F\u002Fwww.ldc.upenn.edu\u002Flanguage-resources\u002Fdata\u002Fobtaining)\n    * This dataset represent texts extracted from a variety of sources: broadcast conversation, broadcast news, newsgroups, weblogs. The\n    6 relation types between 7 types on entities: acility (FAC), Geo-PoliticalEntity (GPE), Location (LOC), Organization (ORG), \n    Person (PER), Vehicle (VEH), Weapon (WEA).\n* SemEval-2018 Task 7\n    [[paper]](https:\u002F\u002Fwww.aclweb.org\u002Fanthology\u002FS18-1111.pdf)\n    [[Website]](https:\u002F\u002Fcompetitions.codalab.org\u002Fcompetitions\u002F17422)\n    [[download]](https:\u002F\u002Flipn.univ-paris13.fr\u002F~gabor\u002Fsemeval2018task7\u002F)\n    * The corpus is collected from abstracts and introductions of scientific papers, and\n    there are six types of semantic relations in total.\n    There are three subtasks of it: Subtask\n    1.1 and Subtask 1.2 are relation classification on\n    clean and noisy data, respectively; Subtask 2 is\n    the standard relation extraction.\n\nFor state of the art results check out [nlpprogress.com on relation extraction](https:\u002F\u002Fnlpprogress.com\u002Fenglish\u002Frelationship_extraction.html)\n\n[Back to Top](#contents)\n\n\n## Videos and Lectures\n* [Stanford University: CS124](https:\u002F\u002Fweb.stanford.edu\u002Fclass\u002Fcs124\u002F), Dan Jurafsky\n\t* (Video) [Week 5: Relation Extraction and Question](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=5SUzf6252_0&list=PLaZQkZp6WhWyszpcteV4LFgJ8lQJ5WIxK&ab_channel=FromLanguagestoInformation)\n* [Washington University: CSE517](https:\u002F\u002Fcourses.cs.washington.edu\u002Fcourses\u002Fcse517\u002F), Luke Zettlemoyer\n\t* (Slide) [Relation Extraction 1](https:\u002F\u002Fcourses.cs.washington.edu\u002Fcourses\u002Fcse517\u002F13wi\u002Fslides\u002Fcse517wi13-RelationExtraction.pdf)\n\t* (Slide) [Relation Extraction 2](https:\u002F\u002Fcourses.cs.washington.edu\u002Fcourses\u002Fcse517\u002F13wi\u002Fslides\u002Fcse517wi13-RelationExtractionII.pdf)\n* [New York University: CSCI-GA.2590](https:\u002F\u002Fcs.nyu.edu\u002Fcourses\u002Fspring17\u002FCSCI-GA.2590-001\u002F), Ralph Grishman\n\t* (Slide) [Relation Extraction: Rule-based Approaches](https:\u002F\u002Fcs.nyu.edu\u002Fcourses\u002Fspring17\u002FCSCI-GA.2590-001\u002FDependencyPaths.pdf)\n* [Michigan University: Coursera](https:\u002F\u002Fai.umich.edu\u002Fportfolio\u002Fnatural-language-processing\u002F), Dragomir R. Radev\n\t* (Video) [Lecture 48: Relation Extraction](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=TbrlRei_0h8&ab_channel=ArtificialIntelligence-AllinOne)\n* [Virginia University: CS6501-NLP](http:\u002F\u002Fweb.cs.ucla.edu\u002F~kwchang\u002Fteaching\u002FNLP16\u002F), Kai-Wei Chang\n\t* (Slide) [Lecture 24: Relation Extraction](http:\u002F\u002Fweb.cs.ucla.edu\u002F~kwchang\u002Fteaching\u002FNLP16\u002Fslides\u002F24-relation.pdf)\n\n\n[Back to Top](#contents)\n\n\n## Systems\n* [DeepDive](http:\u002F\u002Fdeepdive.stanford.edu\u002F)\n* [Stanford Relation Extractor](https:\u002F\u002Fnlp.stanford.edu\u002Fsoftware\u002FrelationExtractor.html)\n\n[Back to Top](#contents)\n\n\n## Frameworks\n* **OpenNRE** [[github]](https:\u002F\u002Fgithub.com\u002Fthunlp\u002FOpenNRE) [[paper]](https:\u002F\u002Faclanthology.org\u002FD19-3029.pdf)\n    * Is an open-source and extensible toolkit that provides a unified framework to implement neural models for relation extraction (RE) between named entities. \n    It is designed for various scenarios for RE, including sentence-level RE, bag-level RE, document-level RE, and few-shot RE. \n    It provides various functional RE modules based on both TensorFlow and PyTorch to maintain sufficient modularity and extensibility, making it becomes easy to incorporate new models into the framework.\n* **AREkit** [[github]](https:\u002F\u002Fgithub.com\u002Fnicolay-r\u002FAREkit) [[research-applicable-paper]](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2006.13730.pdf)\n    * Is an open-source and extensible toolkit focused on data preparation for document-level relation extraction organization. \n    It complements the OpenNRE functionality, as in terms of the latter, *document-level RE setting is not widely explored* (2.4 [[paper]](https:\u002F\u002Faclanthology.org\u002FD19-3029.pdf)).\n    The core functionality includes \n    (1) API for document presentation with EL (Entity Linking, i.e. Object Synonymy) support \n    for sentence level relations preparation (dubbed as contexts)\n    (2) API for contexts extraction\n    (3) relations transferring from sentence-level onto document-level, etc.\n    It provides \n    [neural networks](https:\u002F\u002Fgithub.com\u002Fnicolay-r\u002FAREkit\u002Ftree\u002F0.21.0-rc\u002Fcontrib\u002Fnetworks) (like OpenNRE)\n    and\n    [BERT](https:\u002F\u002Fgithub.com\u002Fnicolay-r\u002FAREkit\u002Ftree\u002F0.21.0-rc\u002Fcontrib\u002Fbert) modules,\n    both applicable for sentiment attitude extraction task.\n* **DeRE** [[github]](https:\u002F\u002Fgithub.com\u002Fims-tcl\u002FDeRE) [[paper]](https:\u002F\u002Faclanthology.org\u002FD18-2008\u002F)\n\t* Is an open-source framework for **de**claritive **r**elation **e**xtraction, and therefore allows to declare your own task (using XML schemas) and apply manually implemented models towards it (using a provided API).\n\tThe task declaration builds on top of the *spans* and *relations between spans*. In terms of the latter, authors propose *frames*, where every frame yelds of: (1) *trigger* (span) and (2) *n*-slots, where every slot\n\tmay refer to *frame* or *span*. \n\tThe framework poses no theoretical restrictions to the window from which frames are extracted.\n\tThus, this concept may cover sentence-level, document-level and multi-document RE tasks.\n    \n    \n\n[Back to Top](#contents)\n\n\n## License\n[![license](https:\u002F\u002Fcamo.githubusercontent.com\u002F60561947585c982aee67ed3e3b25388184cc0aa3\u002F687474703a2f2f6d6972726f72732e6372656174697665636f6d6d6f6e732e6f72672f70726573736b69742f627574746f6e732f38387833312f7376672f63632d7a65726f2e737667)](https:\u002F\u002Fcreativecommons.org\u002Fpublicdomain\u002Fzero\u002F1.0\u002F)\n\nTo the extent possible under law, [Joohong Lee](https:\u002F\u002Froomylee.github.io\u002F) has waived all copyright and related or neighboring rights to this work.\n","# 令人惊叹的关系抽取 [![Awesome](https:\u002F\u002Fawesome.re\u002Fbadge.svg)](https:\u002F\u002Fawesome.re)\n![awesome_re](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Froomylee_awesome-relation-extraction_readme_1ceb71138243.png)\n\n一份精心整理的、专注于关系抽取的优质资源列表，灵感来源于 [awesome-nlp](https:\u002F\u002Fgithub.com\u002Fkeon\u002Fawesome-nlp) 和 [awesome-deep-vision](https:\u002F\u002Fgithub.com\u002Fkjw0612\u002Fawesome-deep-vision)。\n\n**贡献**: 欢迎随时提交 *[pull requests](https:\u002F\u002Fgithub.com\u002Froomylee\u002Fawesome-relation-extraction\u002Fpulls)*。\n\n## 目录\n* [研究趋势与综述](#research-trends-and-surveys)\n* [论文](#papers)\n\t* [监督学习方法](#supervised-approaches)\n\t* [远程监督方法](#distant-supervision-approaches)\n    * [基于图神经网络的模型](#gnn-based-models)\n\t* [语言模型](#language-models)\n\t    * [来自Transformer的编码器表示](#encoder-representation-from-transformer)\n\t    * [来自Transformer的解码器表示](#decoder-representation-from-transformer)\n    * [基于知识图谱的方法](#knowledge-graph-based-approaches)\n    * [小样本学习方法](#few-shot-learning-approaches)\n* [数据集](#datasets)\n* [视频与讲座](#videos-and-lectures)\n* [系统](#systems)\n* [框架](#frameworks)\n\n\n## 研究趋势与综述\n* [NLP进展：关系抽取](https:\u002F\u002Fnlpprogress.com\u002Fenglish\u002Frelationship_extraction.html)\n* [命名实体识别与关系抽取：最新进展](https:\u002F\u002Fwww.researchgate.net\u002Fprofile\u002FSyed-Waqar-Jaffry\u002Fpublication\u002F345315661_Named_Entity_Recognition_and_Relation_Extraction_State_of_the_Art\u002Flinks\u002F603015aaa6fdcc37a83aafd5\u002FNamed-Entity-Recognition-and-Relation-Extraction-State-of-the-Art.pdf)（Nasar等，2021年）\n* [深度学习在关系抽取中的应用综述](https:\u002F\u002Farxiv.org\u002Fabs\u002F1705.03645)（Kumar，2017年）\n* [关系抽取综述](https:\u002F\u002Fwww.cs.cmu.edu\u002F~nbach\u002Fpapers\u002FA-survey-on-Relation-Extraction.pdf)（Bach和Badaskar，2017年）\n* [关系抽取：综述](https:\u002F\u002Farxiv.org\u002Fabs\u002F1712.05191)（Pawar等，2017年）\n* [实体关系抽取综述](https:\u002F\u002Fieeexplore.ieee.org\u002Fstamp\u002Fstamp.jsp?arnumber=8269916)（Zhang等，2017年）\n* [关系抽取方法回顾：有哪些新进展？](https:\u002F\u002Fciteseerx.ist.psu.edu\u002Fviewdoc\u002Fdownload?doi=10.1.1.727.1005&rep=rep1&type=pdf)（Konstantinova等，2014年）\n* [GitHub最佳100：关系抽取](http:\u002F\u002Fmeta-guide.com\u002Fsoftware-meta-guide\u002F100-best-github-relation-extraction)\n\n\n## 论文\n\n### 监督学习方法\n#### 基于CNN的模型\n* 用于关系抽取的卷积神经网络 [[论文]](https:\u002F\u002Flink.springer.com\u002Fchapter\u002F10.1007\u002F978-3-642-53917-6_21) [[代码]](https:\u002F\u002Fgithub.com\u002Froomylee\u002Fcnn-relation-extraction) [[综述]](https:\u002F\u002Fgithub.com\u002Froomylee\u002Fpaper-review\u002Fblob\u002Fmaster\u002Frelation_extraction\u002FConvolution%20Neural%20Network%20for%20Relation%20Extraction\u002Freview.md)\n\t* ChunYang Liu, WenBo Sun, WenHan Chao 和 WanXiang Che\n\t* ADMA 2013\n* 基于卷积深度神经网络的关系分类 [[论文]](http:\u002F\u002Fwww.aclweb.org\u002Fanthology\u002FC14-1220) [[代码]](https:\u002F\u002Fgithub.com\u002Froomylee\u002Fcnn-relation-extraction) [[综述]](https:\u002F\u002Fgithub.com\u002Froomylee\u002Fpaper-review\u002Fblob\u002Fmaster\u002Frelation_extraction\u002FRelation_Classification_via_Convolutional_Deep_Neural_Network\u002Freview.md)\n\t* Daojian Zeng, Kang Liu, Siwei Lai, Guangyou Zhou 和 Jun Zhao\n\t* COLING 2014\n* 关系抽取：来自卷积神经网络的视角 [[论文]](http:\u002F\u002Fwww.cs.nyu.edu\u002F~thien\u002Fpubs\u002Fvector15.pdf) [[代码]](https:\u002F\u002Fgithub.com\u002Froomylee\u002Fcnn-relation-extraction) [[综述]](https:\u002F\u002Fgithub.com\u002Froomylee\u002Fpaper-review\u002Fblob\u002Fmaster\u002Frelation_extraction\u002FRelation_Extraction-Perspective_from_Convolutional_Neural_Networks\u002Freview.md)\n\t* Thien Huu Nguyen 和 Ralph Grishman\n\t* NAACL 2015\n* 使用卷积神经网络进行排序的关系分类 [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1504.06580) [[代码]](https:\u002F\u002Fgithub.com\u002Fpratapbhanu\u002FCRCNN)\n\t* Cicero Nogueira dos Santos, Bing Xiang 和 Bowen Zhou\n\t* ACL 2015\n* 基于注意力机制的卷积神经网络用于语义关系抽取 [[论文]](http:\u002F\u002Fwww.aclweb.org\u002Fanthology\u002FC16-1238) [[代码]](https:\u002F\u002Fgithub.com\u002Fnicolay-r\u002Fmlp-attention)\n\t* Yatian Shen 和 Xuanjing Huang\n\t* COLING 2016\n* 基于多级注意力CNN的关系分类 [[论文]](http:\u002F\u002Faclweb.org\u002Fanthology\u002FP16-1123) [[代码]](https:\u002F\u002Fgithub.com\u002FlawlietAi\u002Frelation-classification-via-attention-model)\n\t* Linlin Wang, Zhu Cao, Gerard de Melo 和 Zhiyuan Liu\n\t* ACL 2016\n* MIT在SemEval-2017任务10中的表现：使用卷积神经网络进行关系抽取 [[论文]](https:\u002F\u002Faclanthology.info\u002Fpdf\u002FS\u002FS17\u002FS17-2171.pdf)\n\t* Ji Young Lee, Franck Dernoncourt 和 Peter Szolovits\n\t* SemEval 2017\n\n#### 基于RNN的模型\n* 基于循环神经网络的关系分类 [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1508.01006)\n\t* Dongxu Zhang 和 Dong Wang\n\t* arXiv 2015\n* 用于关系分类的双向长短期记忆网络 [[论文]](http:\u002F\u002Fwww.aclweb.org\u002Fanthology\u002FY15-1009)\n\t* Shu Zhang, Dequan Zheng, Xinchen Hu 和 Ming Yang\n\t* PACLIC 2015\n* 使用序列和树结构上的LSTM进行端到端关系抽取 [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1601.00770)\n\t* Makoto Miwa 和 Mohit Bansal\n\t* ACL 2016\n* 基于注意力机制的双向长短期记忆网络用于关系分类 [[论文]](http:\u002F\u002Fanthology.aclweb.org\u002FP16-2034) [[代码]](https:\u002F\u002Fgithub.com\u002FSeoSangwoo\u002FAttention-Based-BiLSTM-relation-extraction)\n\t* Peng Zhou, Wei Shi, Jun Tian, Zhenyu Qi, Bingchen Li, Hongwei Hao 和 Bo Xu\n\t* ACL 2016\n* 基于带有注意力机制的层次化循环神经网络的语义关系分类 [[论文]](http:\u002F\u002Fwww.aclweb.org\u002Fanthology\u002FC16-1119)\n\t* Minguang Xiao 和 Cong Liu\n\t* COLING 2016\n* 基于实体感知注意力和潜在实体类型标注的双向LSTM网络进行语义关系分类 [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1901.08163) [[代码]](https:\u002F\u002Fgithub.com\u002Froomylee\u002Fentity-aware-relation-classification)\n\t* Joohong Lee, Sangwoo Seo 和 Yong Suk Choi\n\t* arXiv 2019\n\n#### 基于依存句法的模型\n* 通过递归矩阵-向量空间实现语义组合性 [[论文]](http:\u002F\u002Faclweb.org\u002Fanthology\u002FD12-1110) [[代码]](https:\u002F\u002Fgithub.com\u002Fpratapbhanu\u002FMVRNN)\n\t* Richard Socher, Brody Huval, Christopher D. Manning 和 Andrew Y. Ng\n\t* EMNLP-CoNLL 2012\n* 基于因子的组合式嵌入模型 [[论文]](https:\u002F\u002Fwww.cs.cmu.edu\u002F~mgormley\u002Fpapers\u002Fyu+gormley+dredze.nipsw.2014.pdf)\n\t* Mo Yu, Matthw R. Gormley 和 Mark Dredze\n\t* NIPS关于语义学习的研讨会 2014\n* 用于关系分类的基于依存句法的神经网络 [[论文]](http:\u002F\u002Fwww.aclweb.org\u002Fanthology\u002FP15-2047)\n\t* Yang Liu, Furu Wei, Sujian Li, Heng Ji, Ming Zhou 和 Houfeng Wang\n\t* ACL 2015\n* 沿最短依存路径使用长短期记忆网络进行关系分类 [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1508.03720) [[代码]](https:\u002F\u002Fgithub.com\u002FSshanu\u002FRelation-Classification)\n\t* Xu Yan, Lili Mou, Ge Li, Yunchuan Chen, Hao Peng 和 Zhi Jin\n\t* EMNLP 2015\n* 基于卷积神经网络和简单负采样的语义关系分类 [[论文]](https:\u002F\u002Fwww.aclweb.org\u002Fanthology\u002FD\u002FD15\u002FD15-1062.pdf)\n\t* Kun Xu, Yansong Feng, Songfang Huang 和 Dongyan Zhao\n\t* EMNLP 2015\n* 通过数据增强的深度循环神经网络改进关系分类 [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1601.03651)\n\t* Yan Xu, Ran Jia, Lili Mou, Ge Li, Yunchuan Chen, Yangyang Lu 和 Zhi Jin\n\t* COLING 2016\n* 用于关系分类的双向循环卷积神经网络 [[论文]](http:\u002F\u002Fwww.aclweb.org\u002Fanthology\u002FP16-1072)\n\t* Rui Cai, Xiaodong Zhang 和 Houfeng Wang\n\t* ACL 2016\n* 基于句子内部噪声减少和迁移学习的神经关系抽取 [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1808.06738)\n\t* Tianyi Liu, Xinsong Zhang, Wanhao Zhou 和 Weijia Jia\n\t* EMNLP 2018\n\n#### 基于GNN的模型\n* 匹配空缺：用于关系学习的分布相似性 [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1906.03158)\n    * Livio Baldini Soares, Nicholas FitzGerald, Jeffrey Ling, Tom Kwiatkowski\n    * ACL 2019\n* 关系之间的关系：关系抽取问题的新范式 [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2006.03719)\n\t* Zhijing Jin, Yongyi Yang, Xipeng Qiu, Zheng Zhang\n\t* EMNLP 2020\n* GDPNet：为关系抽取优化潜在多视图图\n    [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2012.06780.pdf)\n    [[代码]](https:\u002F\u002Fgithub.com\u002FXueFuzhao\u002FGDPNet)\n    * Fuzhao Xue, Aixin Sun, Hao Zhang, Eng Siong Chng\n    * AAAI 21\n* RECON：在图神经网络中利用知识图谱上下文进行关系抽取\n    [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2009.08694.pdf)\n    [[代码]](https:\u002F\u002Fgithub.com\u002Fansonb\u002FRECON)\n    * Anson Bastos, Abhishek Nadgeri, Kuldeep Singh, Isaiah Onando Mulang', Saeedeh Shekarpour, Johannes Hoffart, Manohar Kaul\n    * WWW'21\n\n### 远程监督方法\n* 无标注数据的关系抽取远程监督 [[论文]](https:\u002F\u002Fweb.stanford.edu\u002F~jurafsky\u002Fmintz.pdf) [[综述]](https:\u002F\u002Fgithub.com\u002Froomylee\u002Fpaper-review\u002Fblob\u002Fmaster\u002Frelation_extraction\u002FDistant_supervision_for_relation_extraction_without_labeled_data\u002Freview.md)\n\t* 迈克·明茨、史蒂文·比尔斯、里昂·斯诺和丹·朱拉夫斯基\n\t* ACL 2009\n* 基于知识的弱监督用于重叠关系的信息抽取 [[论文]](http:\u002F\u002Fwww.aclweb.org\u002Fanthology\u002FP11-1055) [[代码]](http:\u002F\u002Faiweb.cs.washington.edu\u002Fai\u002Fraphaelh\u002Fmr\u002F)\n\t* 拉斐尔·霍夫曼、丛格勒·张、萧凌、卢克·泽特勒莫耶和丹尼尔·S·韦尔德\n\t* ACL 2011\n* 多实例多标签学习用于关系抽取 [[论文]](http:\u002F\u002Fwww.aclweb.org\u002Fanthology\u002FD12-1042) [[代码]](https:\u002F\u002Fnlp.stanford.edu\u002Fsoftware\u002Fmimlre.shtml)\n\t* 米哈伊·苏尔迪安、朱莉·蒂布希拉尼、拉梅什·纳拉帕蒂和克里斯托弗·D·曼宁\n\t* EMNLP-CoNLL 2012\n* 基于分段卷积神经网络的关系抽取远程监督 [[论文]](http:\u002F\u002Fwww.emnlp2015.org\u002Fproceedings\u002FEMNLP\u002Fpdf\u002FEMNLP203.pdf) [[综述]](https:\u002F\u002Fgithub.com\u002Froomylee\u002Fpaper-review\u002Fblob\u002Fmaster\u002Frelation_extraction\u002FDistant_Supervision_for_Relation_Extraction_via_Piecewise_Convolutional_Neural_Networks\u002Freview.md) [[代码]](https:\u002F\u002Fgithub.com\u002Fnicolay-r\u002Fsentiment-pcnn)\n\t* 道建·曾、康刘、宇博·陈和俊赵\n\t* EMNLP 2015\n* 基于多实例多标签卷积神经网络的关系抽取 [[论文]](https:\u002F\u002Fpdfs.semanticscholar.org\u002F8731\u002F369a707046f3f8dd463d1fd107de31d40a24.pdf) [[综述]](https:\u002F\u002Fgithub.com\u002Froomylee\u002Fpaper-review\u002Fblob\u002Fmaster\u002Frelation_extraction\u002FRelation_Extraction_with_Multi-instance_Multi-label_Convolutional_Neural_Networks\u002Freview.md) [[代码]](https:\u002F\u002Fgithub.com\u002Fmay-\u002Fcnn-re-tf)\n\t* 蒋晓天、权旺、李鹏、王斌\n\t* COLING 2016\n* 在神经关系抽取中融入关系路径 [[论文]](http:\u002F\u002Faclweb.org\u002Fanthology\u002FD17-1186) [[综述]](https:\u002F\u002Fgithub.com\u002Froomylee\u002Fpaper-review\u002Fblob\u002Fmaster\u002Frelation_extraction\u002FIncorporating_Relation_Paths_in_Neural_Relation_Extraction\u002Freview.md)\n\t* 曾文渊、林彦凯、刘志远和孙茂松\n\t* EMNLP 2017\n* 基于实例选择性注意力的神经关系抽取 [[论文]](http:\u002F\u002Fwww.aclweb.org\u002Fanthology\u002FP16-1200) [[代码]](https:\u002F\u002Fgithub.com\u002Fthunlp\u002FOpenNRE\u002F)\n\t* 林彦凯、沈世奇、刘志远、栾焕波和孙茂松\n\t* ACL 2017\n* 使用卷积循环网络模型学习局部与全局上下文，用于生物医学文本中的关系分类 [[论文]](http:\u002F\u002Fwww.aclweb.org\u002Fanthology\u002FK17-1032) [[代码]](https:\u002F\u002Fgithub.com\u002Fdesh2608\u002Fcrnn-relation-classification) [[代码]](https:\u002F\u002Fgithub.com\u002Fkwonmha\u002FConvolutional-Recurrent-Neural-Networks-for-Relation-Extraction)\n\t* 德什·拉杰、苏尼尔·库马尔·萨胡和阿希什·阿南\n\t* CoNLL 2017\n* 基于粗细粒度注意力的层次化关系抽取[[论文]](https:\u002F\u002Faclweb.org\u002Fanthology\u002FD18-1247)[[代码]](https:\u002F\u002Fgithub.com\u002Fthunlp\u002FHNRE)\n\t* 韩旭、于鹏飞∗、刘志远、孙茂松、李鹏\n\t* EMNLP 2018\n* RESIDE：利用辅助信息改进远程监督的神经关系抽取 [[论文]](http:\u002F\u002Fmalllabiisc.github.io\u002Fpublications\u002Fpapers\u002Freside_emnlp18.pdf) [[代码]](https:\u002F\u002Fgithub.com\u002Fmalllabiisc\u002FRESIDE)\n\t* 希卡尔·瓦西斯特、里沙布·乔希、赛·苏曼·普拉亚加、奇兰吉布·巴塔查里亚和帕尔塔·塔鲁克达尔\n\t* EMNLP 2018\n* 基于袋内与袋间注意力的远程监督关系抽取\n    [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1904.00143.pdf)\n    [[代码]](https:\u002F\u002Fgithub.com\u002FZhixiuYe\u002FIntra-Bag-and-Inter-Bag-Attentions)\n    * 叶志秀、凌振华\n    * NAACL 2019\n\t\n### 语言模型\n\n#### 基于Transformer的编码器表示\n* 利用实体信息丰富预训练语言模型以进行关系分类 [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1905.08284.pdf)\n    * 吴善婵、何一凡\n    * arXiv 2019\n* LUKE：具有实体感知自注意力的深度上下文化实体表示\n    [[论文]](https:\u002F\u002Fwww.aclweb.org\u002Fanthology\u002F2020.emnlp-main.523\u002F)\n    [[代码]](https:\u002F\u002Fgithub.com\u002Fstudio-ousia\u002Fluke)\n    * 山田郁也、浅井明里、新藤弘幸、武田英昭、松本裕司\n    * EMNLP 2020\n* SpanBERT：通过表示和预测跨度来改进预训练 [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1907.10529.pdf) [[代码]](https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002FSpanBERT)\n    * 曼达尔·乔希、陈丹琪、尹汉·刘、丹尼尔·S·韦尔德、卢克·泽特勒莫耶和奥默·列维\n    * TACL 2020（计算语言学协会汇刊）\n* 使用DG-SpanBERT高效地进行长距离关系抽取\n    [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F2004.03636)\n    * 陈俊、罗伯特·霍恩多夫、穆罕默德·埃尔霍塞尼、张翔亮\n\n#### 基于Transformer的解码器表示\n* 利用预训练语言表示改进关系抽取\n    [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1906.03088)\n    [[综述]](https:\u002F\u002Fopenreview.net\u002Fforum?id=BJgrxbqp67)\n    [[代码]](https:\u002F\u002Fgithub.com\u002FDFKI-NLP\u002FTRE)\n    * 克里斯托夫·阿尔特、马克·许布纳、莱昂哈德·亨尼格\n    * AKBC 19\n\n### 基于知识图谱的方法\n* KGPool：用于关系抽取的动态知识图谱上下文选择\n  \t[[论文]](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2106.00459.pdf)\n  \t[[代码]](https:\u002F\u002Fgithub.com\u002Fnadgeri14\u002FKGPool)\n\t* 阿比舍克·纳德格里、安森·巴斯托斯、库尔迪普·辛格、以赛亚·奥南多·穆兰格、约翰内斯·霍夫特、萨伊德·谢卡普尔和维杰·萨拉斯瓦特\n\t* ACL 2021（研究成果）\n\n### 少样本学习方法\n* FewRel：一个大规模有监督的少样本关系分类数据集，并提供最先进的评估 [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1810.10147) [[网站]](http:\u002F\u002Fzhuhao.me\u002Ffewrel) [[代码]](https:\u002F\u002Fgithub.com\u002FProKil\u002FFewRel)\n\t* 韩旭、朱浩、于鹏飞、王子云、姚元、刘志远和孙茂松\n\t* EMNLP 2018\n\n### 杂项\n* 通过有效的深度排序联合提取具有类别关联的关系 [[论文]](http:\u002F\u002Faclweb.org\u002Fanthology\u002FP17-1166)\n\t* 海叶、曹文翰、罗俊臣和李周军\n\t* ACL 2017\n* 基于全局优化的端到端神经关系抽取 [[论文]](http:\u002F\u002Faclweb.org\u002Fanthology\u002FD17-1182)\n\t* 张美珊、张岳和傅国洪\n\t* EMNLP 2017\n* 面向关系抽取的对抗训练 [[论文]](https:\u002F\u002Fpeople.eecs.berkeley.edu\u002F~russell\u002Fpapers\u002Femnlp17-relation.pdf)\n\t* 吴毅、大卫·巴曼和斯图尔特·拉塞尔\n\t* EMNLP 2017\n* 一种用于从生物医学文本中联合抽取实体与关系的神经网络模型[[论文]](https:\u002F\u002Fbmcbioinformatics.biomedcentral.com\u002Farticles\u002F10.1186\u002Fs12859-017-1609-9)\n\t* 李飞、张美珊、傅国洪和季东红\n\t* BMC生物信息学 2017\n* 使用强化学习和深度学习联合抽取实体与关系 [[论文]](https:\u002F\u002Fwww.hindawi.com\u002Fjournals\u002Fcin\u002F2017\u002F7643065\u002F)\n\t* 冯云天、张洪军、郝文宁和陈刚\n\t* 计算智能与神经科学杂志 2017\n* TDEER：一种高效的翻译解码模式，用于联合抽取实体与关系 [[论文]](https:\u002F\u002Faclanthology.org\u002F2021.emnlp-main.635\u002F) [[代码]](https:\u002F\u002Fgithub.com\u002F4ai\u002Ftdeer)\n\t* 李献明、罗晓天、董成浩、杨大川、栾贝迪和何震\n\t* EMNLP 2021\n\n[返回顶部](#contents)\n\n\n## 数据集\n* SemEval-2010 任务8 [[论文]](http:\u002F\u002Fwww.aclweb.org\u002Fanthology\u002FS10-1006) [[下载]](https:\u002F\u002Fdocs.google.com\u002Fleaf?id=0B_jQiLugGTAkMDQ5ZjZiMTUtMzQ1Yy00YWNmLWJlZDYtOWY1ZDMwY2U4YjFk&sort=name&layout=list&num=50)\n\t* 名词对之间语义关系的多分类\n* 纽约时报（NYT）语料库 [[论文]](http:\u002F\u002Fwww.riedelcastro.org\u002F\u002Fpublications\u002Fpapers\u002Friedel10modeling.pdf) [[下载]](https:\u002F\u002Fcatalog.ldc.upenn.edu\u002FLDC2008T19)\n\t* 该数据集是通过将*Freebase*关系与纽约时报语料库进行对齐生成的，其中2005—2006年的句子用作训练语料，2007年的句子用作测试语料。\n* FewRel：少样本关系分类数据集 [[论文]](https:\u002F\u002Farxiv.org\u002Fabs\u002F1810.10147) [[官网]](http:\u002F\u002Fzhuhao.me\u002Ffewrel)\n\t* 该数据集是一个有监督的少样本关系分类数据集。语料来源于维基百科，标注语料的知识库为Wikidata。\n* TACRED：TAC关系抽取数据集\n    [[论文]](https:\u002F\u002Fwww.aclweb.org\u002Fanthology\u002FD17-1004.pdf)\n    [[官网]](https:\u002F\u002Fnlp.stanford.edu\u002Fprojects\u002Ftacred\u002F)\n    [[下载]](https:\u002F\u002Fcatalog.ldc.upenn.edu\u002FLDC2018T24)\n    * 是一个大规模的关系抽取数据集，基于每年TAC知识库构建（TAC KBP）挑战赛中使用的新闻和网络文本构建而成。\n* ACE05：\n    [[官网]](https:\u002F\u002Fcatalog.ldc.upenn.edu\u002FLDC2006T06)\n    [[下载信息]](https:\u002F\u002Fwww.ldc.upenn.edu\u002Flanguage-resources\u002Fdata\u002Fobtaining)\n    * 该数据集包含来自广播对话、广播新闻、新闻组和博客等多种来源的文本。涉及实体之间的7种关系类型中的6种：设施（FAC）、地缘政治实体（GPE）、地点（LOC）、组织（ORG）、人物（PER）、车辆（VEH）、武器（WEA）。\n* SemEval-2018 任务7\n    [[论文]](https:\u002F\u002Fwww.aclweb.org\u002Fanthology\u002FS18-1111.pdf)\n    [[官网]](https:\u002F\u002Fcompetitions.codalab.org\u002Fcompetitions\u002F17422)\n    [[下载]](https:\u002F\u002Flipn.univ-paris13.fr\u002F~gabor\u002Fsemeval2018task7\u002F)\n    * 该语料收集自科技论文的摘要和引言，共包含六种语义关系。任务分为三个子任务：子任务1.1和1.2分别针对干净和噪声数据的关系分类；子任务2则是标准的关系抽取。\n\n有关最新研究进展，请参阅[nlpprogress.com上的关系抽取页面](https:\u002F\u002Fnlpprogress.com\u002Fenglish\u002Frelationship_extraction.html)\n\n[返回顶部](#contents)\n\n\n## 视频与讲座\n* [斯坦福大学：CS124](https:\u002F\u002Fweb.stanford.edu\u002Fclass\u002Fcs124\u002F)，丹·朱拉夫斯基\n\t* （视频）[第5周：关系抽取与问题](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=5SUzf6252_0&list=PLaZQkZp6WhWyszpcteV4LFgJ8lQJ5WIxK&ab_channel=FromLanguagestoInformation)\n* [华盛顿大学：CSE517](https:\u002F\u002Fcourses.cs.washington.edu\u002Fcourses\u002Fcse517\u002F)，卢克·泽特莫耶尔\n\t* （幻灯片）[关系抽取1](https:\u002F\u002Fcourses.cs.washington.edu\u002Fcourses\u002Fcse517\u002F13wi\u002Fslides\u002Fcse517wi13-RelationExtraction.pdf)\n\t* （幻灯片）[关系抽取2](https:\u002F\u002Fcourses.cs.washington.edu\u002Fcourses\u002Fcse517\u002F13wi\u002Fslides\u002Fcse517wi13-RelationExtractionII.pdf)\n* [纽约大学：CSCI-GA.2590](https:\u002F\u002Fcs.nyu.edu\u002Fcourses\u002Fspring17\u002FCSCI-GA.2590-001\u002F)，拉尔夫·格里什曼\n\t* （幻灯片）[关系抽取：基于规则的方法](https:\u002F\u002Fcs.nyu.edu\u002Fcourses\u002Fspring17\u002FCSCI-GA.2590-001\u002FDependencyPaths.pdf)\n* [密歇根大学：Coursera](https:\u002F\u002Fai.umich.edu\u002Fportfolio\u002Fnatural-language-processing\u002F)，德拉戈米尔·R·拉德夫\n\t* （视频）[第48讲：关系抽取](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=TbrlRei_0h8&ab_channel=ArtificialIntelligence-AllinOne)\n* [弗吉尼亚大学：CS6501-NLP](http:\u002F\u002Fweb.cs.ucla.edu\u002F~kwchang\u002Fteaching\u002FNLP16\u002F)，凯-伟·张\n\t* （幻灯片）[第24讲：关系抽取](http:\u002F\u002Fweb.cs.ucla.edu\u002F~kwchang\u002Fteaching\u002FNLP16\u002Fslides\u002F24-relation.pdf)\n\n\n[返回顶部](#contents)\n\n\n## 系统\n* [DeepDive](http:\u002F\u002Fdeepdive.stanford.edu\u002F)\n* [斯坦福关系抽取器](https:\u002F\u002Fnlp.stanford.edu\u002Fsoftware\u002FrelationExtractor.html)\n\n[返回顶部](#contents)\n\n## 框架\n* **OpenNRE** [[github]](https:\u002F\u002Fgithub.com\u002Fthunlp\u002FOpenNRE) [[paper]](https:\u002F\u002Faclanthology.org\u002FD19-3029.pdf)\n    * 是一个开源且可扩展的工具包，提供了一个统一的框架来实现命名实体之间关系抽取（RE）的神经网络模型。\n    它适用于多种关系抽取场景，包括句子级、袋级、文档级以及少样本关系抽取。该工具包基于 TensorFlow 和 PyTorch 提供了多种功能模块，以保持足够的模块化和可扩展性，从而便于将新模型集成到框架中。\n* **AREkit** [[github]](https:\u002F\u002Fgithub.com\u002Fnicolay-r\u002FAREkit) [[research-applicable-paper]](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2006.13730.pdf)\n    * 是一个专注于文档级关系抽取数据准备的开源且可扩展的工具包。\n    它补充了 OpenNRE 的功能，因为后者在文档级关系抽取方面“尚未得到广泛探索”（见 2.4 节 [[paper]](https:\u002F\u002Faclanthology.org\u002FD19-3029.pdf)）。\n    其核心功能包括：\n    (1) 支持实体链接（即对象同义性）的文档表示 API，用于准备句子级别的关系上下文；\n    (2) 上下文提取 API；\n    (3) 将句子级关系迁移到文档级等功能。\n    它提供了与 OpenNRE 类似的[神经网络](https:\u002F\u002Fgithub.com\u002Fnicolay-r\u002FAREkit\u002Ftree\u002F0.21.0-rc\u002Fcontrib\u002Fnetworks)模块，以及[BERT](https:\u002F\u002Fgithub.com\u002Fnicolay-r\u002FAREkit\u002Ftree\u002F0.21.0-rc\u002Fcontrib\u002Fbert)模块，两者均可用于情感态度抽取任务。\n* **DeRE** [[github]](https:\u002F\u002Fgithub.com\u002Fims-tcl\u002FDeRE) [[paper]](https:\u002F\u002Faclanthology.org\u002FD18-2008\u002F)\n    * 是一个用于**声明式**关系抽取的开源框架，因此允许用户通过 XML 模式声明自己的任务，并使用提供的 API 应用手动实现的模型。\n    任务声明基于“跨度”及“跨度之间的关系”。对于后者，作者提出了“框架”的概念：每个框架包含 (1) “触发器”（一个跨度）和 (2) n 个槽位，其中每个槽位可以引用“框架”或“跨度”。\n    该框架对提取框架的窗口大小没有理论限制。\n    因此，这一概念可以覆盖句子级、文档级以及多文档的关系抽取任务。\n    \n\n[返回顶部](#contents)\n\n\n## 许可证\n[![license](https:\u002F\u002Fcamo.githubusercontent.com\u002F60561947585c982aee67ed3e3b25388184cc0aa3\u002F687474703a2f2f6d6972726f72732e6372656174697665636f6d6d6f6e732e6f72672f70726573736b69742f627574746f6e732f38387833312f7376672f63632d7a65726f2e737667)](https:\u002F\u002Fcreativecommons.org\u002Fpublicdomain\u002Fzero\u002F1.0\u002F)\n\n在法律允许的最大范围内，[Joohong Lee](https:\u002F\u002Froomylee.github.io\u002F) 已放弃本作品的所有版权及相关权利或邻接权。","# Awesome Relation Extraction 快速上手指南\n\n`awesome-relation-extraction` 并非一个可直接安装的单一软件包或框架，而是一个**精选资源列表**，汇集了关系抽取（Relation Extraction, RE）领域的论文、数据集、代码库和综述。本指南将帮助你利用该列表快速找到适合的工具并运行示例。\n\n## 环境准备\n\n由于列表中包含多种基于不同深度学习框架（如 PyTorch, TensorFlow）的实现，建议先准备通用的开发环境。\n\n*   **操作系统**: Linux (推荐 Ubuntu 18.04+), macOS, 或 Windows (WSL2)\n*   **Python**: 3.6 或更高版本 (推荐 3.8+)\n*   **包管理工具**: `pip` 或 `conda`\n*   **硬件要求**: 若需复现深度学习模型（如 CNN, RNN, GNN, BERT），建议使用配备 NVIDIA GPU 的环境并安装 CUDA。\n\n**前置依赖安装：**\n```bash\n# 创建虚拟环境 (推荐)\npython -m venv re-env\nsource re-env\u002Fbin\u002Factivate  # Windows: re-env\\Scripts\\activate\n\n# 安装基础科学计算库\npip install torch torchvision torchaudio --index-url https:\u002F\u002Fdownload.pytorch.org\u002Fwhl\u002Fcu118 # 根据CUDA版本调整\npip install tensorflow\npip install transformers datasets scikit-learn pandas numpy\n```\n\n> **国内加速提示**：建议使用清华或阿里镜像源加速 pip 安装：\n> `pip install \u003Cpackage_name> -i https:\u002F\u002Fpypi.tuna.tsinghua.edu.cn\u002Fsimple`\n\n## 安装步骤\n\n由于这是一个资源列表，你需要从中选择具体的项目（代码库）进行克隆和安装。以下以列表中热门的 **OpenNRE** (由清华团队开发，支持多种监督及远程监督模型) 为例演示如何从该列表中提取并安装具体工具。\n\n1.  **浏览列表选择项目**：\n    在 `awesome-relation-extraction` 的 [Papers](https:\u002F\u002Fgithub.com\u002Froomylee\u002Fawesome-relation-extraction#papers) 或 [Frameworks](https:\u002F\u002Fgithub.com\u002Froomylee\u002Fawesome-relation-extraction#frameworks) 部分找到感兴趣的项目链接（例如 `Neural Relation Extraction with Selective Attention over Instances` 对应的代码库 `thunlp\u002FOpenNRE`）。\n\n2.  **克隆代码库**：\n    ```bash\n    git clone https:\u002F\u002Fgithub.com\u002Fthunlp\u002FOpenNRE.git\n    cd OpenNRE\n    ```\n\n3.  **安装项目依赖**：\n    大多数项目会在根目录提供 `requirements.txt`。\n    ```bash\n    # 使用国内镜像源加速安装\n    pip install -r requirements.txt -i https:\u002F\u002Fpypi.tuna.tsinghua.edu.cn\u002Fsimple\n    ```\n\n    *注：如果是纯论文复现代码且无 setup 脚本，请参照该项目 README 中的具体依赖说明。*\n\n## 基本使用\n\n以下以安装好的 **OpenNRE** 为例，展示如何加载预训练模型并进行最简单的关系抽取预测。其他项目的用法类似，通常遵循“加载数据 -> 初始化模型 -> 训练\u002F预测”的流程。\n\n### 1. 快速预测示例 (Python)\n\n创建一个 `test_re.py` 文件，输入句子和实体位置，获取关系分类结果：\n\n```python\nimport opennre\n\n# 加载预训练模型 (例如：基于 BERT 的模型)\n# 模型会自动下载到 ~\u002F.opennre\u002Fcheckpoint\u002F\nmodel = opennre.get_model('bert_base_uncased')\n\n# 定义输入句子和实体位置\n# sentence: 输入文本\n# pos1: 第一个实体的起始和结束索引\n# pos2: 第二个实体的起始和结束索引\ntest_sentence = \"Mark Zuckerberg is the founder of Facebook.\"\ntest_data = {\n    'token': ['Mark', 'Zuckerberg', 'is', 'the', 'founder', 'of', 'Facebook', '.'],\n    'pos1': [0, 2],  # \"Mark Zuckerberg\"\n    'pos2': [6, 7]   # \"Facebook\"\n}\n\n# 进行预测\nresult = model.infer(test_data)\n\nprint(f\"预测关系：{result[0]}\")\nprint(f\"置信度：{result[1]}\")\n```\n\n### 2. 运行官方示例脚本\n\n大多数仓库都提供了 `example.py` 或 `train.py`。以 OpenNRE 为例，运行内置的 Few-Shot 学习示例：\n\n```bash\n# 进入 example 目录\ncd example\n\n# 运行 Few-Shot 关系抽取示例 (使用 ProtoNet 算法)\npython train_fewrel.py \\\n    --model proto \\\n    --encoder bert \\\n    --train_dataset fewrel_train \\\n    --val_dataset fewrel_val \\\n    --hidden_size 230 \\\n    --max_length 80\n```\n\n### 3. 探索其他资源\n\n回到 `awesome-relation-extraction` 列表，你可以：\n*   **查找数据集**：访问 [Datasets](https:\u002F\u002Fgithub.com\u002Froomylee\u002Fawesome-relation-extraction#datasets) 部分下载 SemEval, TAC-KBP, FewRel 等标准数据集。\n*   **复现经典论文**：在 [Supervised Approaches](https:\u002F\u002Fgithub.com\u002Froomylee\u002Fawesome-relation-extraction#supervised-approaches) 或 [Distant Supervision Approaches](https:\u002F\u002Fgithub.com\u002Froomylee\u002Fawesome-relation-extraction#distant-supervision-approaches) 中找到对应论文的 `[code]` 链接，克隆后按照其特定 README 运行。","某金融科技公司的情报分析团队需要从每日海量的财经新闻中，自动提取上市公司与其供应商、竞争对手或子公司之间的复杂关系，以构建动态知识图谱辅助投资决策。\n\n### 没有 awesome-relation-extraction 时\n- **资源检索低效**：团队成员需花费数天时间在 Google Scholar 和 GitHub 上盲目搜索，难以区分过时的 CNN 模型与最新的 Transformer 架构。\n- **技术选型盲目**：缺乏对“远程监督”或“少样本学习”等特定场景下主流方案的系统对比，导致初期尝试的模型在稀疏数据上准确率极低。\n- **复现成本高昂**：找不到经过验证的代码实现和配套数据集，工程师需从零编写数据处理管道，严重拖慢原型开发进度。\n- **前沿视野缺失**：无法及时获取最新的综述论文和行业趋势，导致技术方案停留在几年前的水平，难以处理复杂的长文本依赖。\n\n### 使用 awesome-relation-extraction 后\n- **一站式资源导航**：直接利用其分类清晰的目录，快速定位到基于语言模型（如 BERT）的最新 SOTA 方案，将调研时间从数天缩短至几小时。\n- **精准场景匹配**：针对财经数据标注少的痛点，直接采纳列表中推荐的“少样本学习”和“远程监督”专题论文与代码，显著提升模型冷启动效果。\n- **加速落地实施**：复用列表中提供的成熟框架和基准数据集，团队迅速搭建起可运行的基线系统，将开发周期压缩了 60%。\n- **持续技术迭代**：通过跟踪列表中的最新研究趋势和视频教程，团队能持续引入图神经网络（GNN）等新方法，不断优化知识图谱的推理能力。\n\nawesome-relation-extraction 通过聚合全球顶尖的关系抽取资源，将团队从繁琐的文献大海捞针中解放出来，使其能专注于核心业务逻辑的创新与落地。","https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Froomylee_awesome-relation-extraction_1ceb7113.png","roomylee","Joohong Lee","https:\u002F\u002Foss.gittoolsai.com\u002Favatars\u002Froomylee_f926551f.jpg","Machine Learning Researcher","@scatterlab","Seoul, South Korea","roomylee13@gmail.com",null,"https:\u002F\u002Froomylee.github.io","https:\u002F\u002Fgithub.com\u002Froomylee",1228,134,"2026-04-06T12:05:19",5,"","未说明",{"notes":90,"python":88,"dependencies":91},"该仓库是一个资源列表（Awesome List），汇集了关系抽取领域的论文、数据集、系统和框架链接，本身不是一个可直接运行的软件工具，因此 README 中未包含具体的操作系统、硬件配置或依赖库安装要求。用户需根据列表中引用的具体子项目（如 cnn-relation-extraction, OpenNRE 等）的独立文档来查询相应的运行环境需求。",[],[35,93,14],"其他",[95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112],"awesome","relation-extraction","relation-classification","distant-supervision","deep-learning","natural-language-processing","nlp","paper","machine-learning","semeval-2010","new-york-times","state-of-the-art","acl","emnlp","naacl","aaai","nips","trends","2026-03-27T02:49:30.150509","2026-04-10T10:34:32.203789",[],[]]