[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"similar-imsheridan--DeepRec":3,"tool-imsheridan--DeepRec":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",150037,2,"2026-04-10T23:33:47",[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。",108322,"2026-04-10T11:39:34",[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":66,"readme_en":67,"readme_zh":68,"quickstart_zh":69,"use_case_zh":70,"hero_image_url":71,"owner_login":72,"owner_name":73,"owner_avatar_url":74,"owner_bio":75,"owner_company":75,"owner_location":76,"owner_email":75,"owner_twitter":75,"owner_website":77,"owner_url":78,"languages":75,"stars":79,"forks":80,"last_commit_at":81,"license":82,"difficulty_score":83,"env_os":84,"env_gpu":85,"env_ram":85,"env_deps":86,"category_tags":89,"github_topics":90,"view_count":32,"oss_zip_url":75,"oss_zip_packed_at":75,"status":17,"created_at":97,"updated_at":98,"faqs":99,"releases":100},4968,"imsheridan\u002FDeepRec","DeepRec","推荐、广告工业界经典以及最前沿的论文、资料集合\u002F Must-read Papers on Recommendation System and CTR Prediction","DeepRec 是一个专注于推荐系统与广告点击率（CTR）预估领域的开源知识宝库。它系统性地整理了工业界经典与学术界最前沿的论文及技术分享，旨在解决从业者在海量文献中难以快速定位核心资料、理清技术演进脉络的痛点。\n\n无论是刚入门的开发者、深耕算法的研究人员，还是希望了解行业动态的技术决策者，都能从中获益。DeepRec 不仅收录了来自阿里巴巴、谷歌、微软、华为等顶尖科技公司的实战成果（如 DIN、DIEN、DCN、xDeepFM 等里程碑式模型），还通过清晰的“论文脉络图”帮助用户直观把握从基础因子分解机到复杂注意力机制的技术发展路径。\n\n其独特亮点在于“动态更新”与“工业界视角”，确保内容紧贴实际应用场景，而非单纯的理论堆砌。此外，项目还关联了王喆老师等专家的经典资源列表，并提供了交流社群入口，促进了知识共享与技术探讨。对于希望提升推荐算法效果或追踪 CTR 预估最新趋势的用户而言，DeepRec 是一份不可或缺的高效指南。","# 深度推荐系统与CTR预估工业界相关论文、业界分享\n动态更新推荐、广告工业界经典以及最前沿的论文、业界分享集合。所有资料均整理来自于互联网，如有侵权，请联系小助手deepdeliver。同时欢迎对推荐、广告方面工业界感兴趣的小伙伴添加小助手，将自动拉入交流群：\n\u003Cdiv align=left>\n\u003Cimg width=\"200\" src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fimsheridan_DeepRec_readme_8536b366d82d.jpg\" alt=\"交流群\"\u002F>\n\u003C\u002Fdiv>\n\n### 其他相关资源\n* [王喆的推荐系统paper列表](https:\u002F\u002Fgithub.com\u002Fwzhe06\u002FReco-papers)\n* [王喆的广告paper列表](https:\u002F\u002Fgithub.com\u002Fwzhe06\u002FAd-papers)\n\n### 论文脉络\n\n\u003Cdiv align=center>\n\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fimsheridan_DeepRec_readme_ed85f0b99d5d.jpg\" alt=\"脉络图\"\u002F>\n\u003C\u002Fdiv>\n\n# 目录\n\n## CTR\n* [[FiBiNET][RecSys 19][Weibo] FiBiNET_Combining Feature Importance and Bilinear feature Interaction for Click-Through Rate Prediction](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FCTR\u002F%5BFiBiNET%5D%5BRecSys%2019%5D%5BWeibo%5D%20FiBiNET_Combining%20Feature%20Importance%20and%20Bilinear%20feature%20Interaction%20for%20Click-Through%20Rate%20Prediction.pdf)\n* [[DSIN][IJCAI 19][Alibaba] Deep Session Interest Network for Click-Through Rate Prediction](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FCTR\u002F%5BDSIN%5D%5BIJCAI%2019%5D%5BAlibaba%5D%20Deep%20Session%20Interest%20Network%20for%20Click-Through%20Rate%20Prediction.pdf)\n* [[FGCNN][WWW 19][Huawei] Feature Generation by Convolutional Neural Network for Click-Through Rate Prediction](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FCTR\u002F%5BFGCNN%5D%5BWWW%2019%5D%5BHuawei%5D%20Feature%20Generation%20by%20Convolutional%20Neural%20Network%20for%20Click-Through%20Rate%20Prediction.pdf)\n* [[AutoInt][CIKM 19] AutoInt_Automatic Feature Interaction Learning via Self-Attentive Neural Networks](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FCTR\u002F%5BAutoInt%5D%5BCIKM%2019%5D%20AutoInt_Automatic%20Feature%20Interaction%20Learning%20via%20Self-Attentive%20Neural%20Networks.pdf)\n* [[DIEN][AAAI 19][Alibaba] Deep Interest Evolution Network for Click-Through Rate Prediction](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FCTR\u002F%5BDIEN%5D%5BAAAI%2019%5D%5BAlibaba%5D%20Deep%20Interest%20Evolution%20Network%20for%20Click-Through%20Rate%20Prediction.pdf)\n* [[PNN][TOIS 18] Product-based Neural Networks for User Response Prediction](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FCTR\u002F%5BPNN%5D%5BTOIS%2018%5D%20Product-based%20Neural%20Networks%20for%20User%20Response%20Prediction.pdf)\n* [[xDeepFM][KDD 18][Microsoft] xDeepFM_Combining Explicit and Implicit Feature Interactions for Recommender Systems](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FCTR\u002F%5BxDeepFM%5D%5BKDD%2018%5D%5BMicrosoft%5D%20xDeepFM_Combining%20Explicit%20and%20Implicit%20Feature%20Interactions%20for%20Recommender%20Systems.pdf)\n* [[DCN][KDD 17][Google] Deep & Cross Network for Ad Click Predictions](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FCTR\u002F%5BDCN%5D%5BKDD%2017%5D%5BGoogle%5D%20Deep%20%26%20Cross%20Network%20for%20Ad%20Click%20Predictions.pdf)\n* [[DIN][KDD 18][Alibaba] Deep Interest Network for Click-Through Rate Prediction](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FCTR\u002F%5BDIN%5D%5BKDD%2018%5D%5BAlibaba%5D%20Deep%20Interest%20Network%20for%20Click-Through%20Rate%20Prediction.pdf)\n* [[FNN][ECIR 16] Deep Learning over Multi-field Categorical Data_A Case Study on User Response Prediction](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FCTR\u002F%5BFNN%5D%5BECIR%2016%5D%20Deep%20Learning%20over%20Multi-field%20Categorical%20Data_A%20Case%20Study%20on%20User%20Response%20Prediction.pdf)\n* [[AFM][IJCAI 17] Attentional Factorization Machines - Learning the Weight of Feature Interactions via Attention Networks](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FCTR\u002F%5BAFM%5D%5BIJCAI%2017%5D%20Attentional%20Factorization%20Machines%20-%20Learning%20the%20Weight%20of%20Feature%20Interactions%20via%20Attention%20Networks.pdf)\n* [[DeepFM][IJCAI 17][Huawei] DeepFM_A Factorization-Machine based Neural Network for CTR Prediction](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FCTR\u002F%5BDeepFM%5D%5BIJCAI%2017%5D%5BHuawei%5D%20DeepFM_A%20Factorization-Machine%20based%20Neural%20Network%20for%20CTR%20Prediction.pdf)\n* [[NFM][SIGIR 17] Neural Factorization Machines for Sparse Predictive Analytics](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FCTR\u002F%5BNFM%5D%5BSIGIR%2017%5D%20Neural%20Factorization%20Machines%20for%20Sparse%20Predictive%20Analytics.pdf)\n* [[WDL][DLRS 16][Google] Wide & Deep Learning for Recommender Systems](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FCTR\u002F%5BWDL%5D%5BDLRS%2016%5D%5BGoogle%5D%20Wide%20%26%20Deep%20Learning%20for%20Recommender%20Systems.pdf)\n\n## Match\n* [[JTM][NIPS 19] Joint Optimization of Tree-based Index and Deep Model for Recommender Systems](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FMatch\u002F%5BJTM%5D%5BNIPS%2019%5D%20Joint%20Optimization%20of%20Tree-based%20Index%20and%20Deep%20Model%20for%20Recommender%20Systems.pdf)\n* [[MIND][arxiv 19][Alibaba] Multi-Interest Network with Dynamic Routing for Recommendation at Tmall](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FMatch\u002F%5BMIND%5D%5Barxiv%2019%5D%5BAlibaba%5D%20Multi-Interest%20Network%20with%20Dynamic%20Routing%20for%20Recommendation%20at%20Tmall.pdf)\n* [[SDM][CIKM 19][Alibaba] Sequential Deep Matching Model for Online Large-scale Recommender System](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FMatch\u002F%5BSDM%5D%5BCIKM%2019%5D%5BAlibaba%5D%20Sequential%20Deep%20Matching%20Model%20for%20Online%20Large-scale%20Recommender%20System.pdf)\n* [[TDM][KDD 18][Alibaba] Learning Tree-based Deep Model for Recommender Systems](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FMatch\u002F%5BTDM%5D%5BKDD%2018%5D%5BAlibaba%5D%20Learning%20Tree-based%20Deep%20Model%20for%20Recommender%20Systems.pdf)\n* [[NCF][WWW 17] Neural Collaborative Filtering](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FMatch\u002F%5BNCF%5D%5BWWW%2017%5D%20Neural%20Collaborative%20Filtering.pdf)\n* [[YoutubeDNN][RecSys 16][Google] Deep Neural Networks for YouTube Recommendations](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FMatch\u002F%5BYoutubeDNN%5D%5BRecSys%2016%5D%5BGoogle%5D%20Deep%20Neural%20Networks%20for%20YouTube%20Recommendations.pdf)\n* [[DSSM][CIKM 13][Microsoft] Learning Deep Structured Semantic Models for Web Search using Clickthrough Data](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FMatch\u002F%5BDSSM%5D%5BCIKM%2013%5D%5BMicrosoft%5D%20Learning%20Deep%20Structured%20Semantic%20Models%20for%20Web%20Search%20using%20Clickthrough%20Data.pdf)\n\n## Ranking\n* [[PRM][RecSys 19][Alibaba] Personalized Re-ranking for Recommendation](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FRanking\u002F%5BPRM%5D%5BRecSys%2019%5D%5BAlibaba%5D%20Personalized%20Re-ranking%20for%20Recommendation.pdf)\n* [[BERT4Rec][CIKM 19][Alibaba] BERT4Rec_Sequential Recommendation with Bidirectional Encoder Representations from Transformer](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FRanking\u002F%5BBERT4Rec%5D%5BCIKM%2019%5D%5BAlibaba%5D%20BERT4Rec_Sequential%20Recommendation%20with%20Bidirectional%20Encoder%20Representations%20from%20Transformer.pdf)\n* [[BST][DLP-KDD 19][Alibaba] Behavior Sequence Transformer for E-commerce Recommendation in Alibaba](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FRanking\u002F%5BBST%5D%5BDLP-KDD%2019%5D%5BAlibaba%5D%20Behavior%20Sequence%20Transformer%20for%20E-commerce%20Recommendation%20in%20Alibaba.pdf)\n\n## Embedding\n* [[Airbnb Embedding][KDD 18][Airbnb] Real-time Personalization using Embeddings for Search Ranking at Airbnb](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FEmbedding\u002F%5BAirbnb%20Embedding%5D%5BKDD%2018%5D%5BAirbnb%5D%20Real-time%20Personalization%20using%20Embeddings%20for%20Search%20Ranking%20at%20Airbnb.pdf)\n* [[Alibaba Embedding][KDD 18][Alibaba] Billion-scale Commodity Embedding for E-commerce Recommendation in Alibaba](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FEmbedding\u002F%5BAlibaba%20Embedding%5D%5BKDD%2018%5D%5BAlibaba%5D%20Billion-scale%20Commodity%20Embedding%20for%20E-commerce%20Recommendation%20in%20Alibaba.pdf)\n* [[DeepWalk][KDD 14] DeepWalk- Online Learning of Social Representations](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FEmbedding\u002F%5BDeepWalk%5D%5BKDD%2014%5D%20DeepWalk-%20Online%20Learning%20of%20Social%20Representations.pdf)\n* [[LINE][WWW 15][Microsoft] LINE_Large-scale Information Network Embedding](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FEmbedding\u002F%5BLINE%5D%5BWWW%2015%5D%5BMicrosoft%5D%20LINE_Large-scale%20Information%20Network%20Embedding.pdf)\n* [[Node2vec][KDD 16] Node2vec_Scalable Feature Learning for Networks](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FEmbedding\u002F%5BNode2vec%5D%5BKDD%2016%5D%20Node2vec_Scalable%20Feature%20Learning%20for%20Networks.pdf)\n* [[SDNE][KDD 16] Structural Deep Network Embedding](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FEmbedding\u002F%5BSDNE%5D%5BKDD%2016%5D%20Structural%20Deep%20Network%20Embedding.pdf)\n* [[Struc2Vec][KDD 17]struc2vec_Learning Node Representations from Structural Identity](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FEmbedding\u002F%5BStruc2Vec%5D%5BKDD%2017%5Dstruc2vec_Learning%20Node%20Representations%20from%20Structural%20Identity.pdf)\n* [[GraphSAGE][NIPS 17] Inductive Representation Learning on Large Graphs](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FEmbedding\u002F%5BGraphSAGE%5D%5BNIPS%2017%5D%20Inductive%20Representation%20Learning%20on%20Large%20Graphs.pdf)\n* [[GCN][ICLR 17] Semi-supervised Classification with Graph Convolutional Networks](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FEmbedding\u002F%5BGCN%5D%5BICLR%2017%5D%20Semi-supervised%20Classification%20with%20Graph%20Convolutional%20Networks.pdf)\n\n## MTL\n* [[RecSys 19][Alibaba] A Pareto-Efficient Algorithm for Multiple Objective Optimization in E-Commerce Recommendation](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FMTL\u002F%5BRecSys%2019%5D%5BAlibaba%5D%20A%20Pareto-Efficient%20Algorithm%20for%20Multiple%20Objective%20Optimization%20in%20E-Commerce%20Recommendation.pdf)\n* [[MMoE][KDD 18][Google] Modeling Task Relationships in Multi-task Learning with Multi-gate Mixture-of-Experts](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FMTL\u002F%5BMMoE%5D%5BKDD%2018%5D%5BGoogle%5D%20Modeling%20Task%20Relationships%20in%20Multi-task%20Learning%20with%20Multi-gate%20Mixture-of-Experts.pdf)\n* [[ESMM][SIGIR 18][Alibaba] Entire Space Multi-Task Model_An Effective Approach for Estimating Post-Click Conversion Rate](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FMTL\u002F%5BESMM%5D%5BSIGIR%2018%5D%5BAlibaba%5D%20Entire%20Space%20Multi-Task%20Model_An%20Effective%20Approach%20for%20Estimating%20Post-Click%20Conversion%20Rate.pdf)\n\n## Diversity\n* [[CIKM 18][Google] Practical Diversified Recommendations on YouTube with Determinantal Point Processes](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FDiversity\u002F%5BCIKM%2018%5D%5BGoogle%5D%20Practical%20Diversified%20Recommendations%20on%20YouTube%20with%20Determinantal%20Point%20Processes.pdf)\n* [[NeurIPS 18][Hulu] Fast Greedy MAP Inference for Determinantal Point Process to Improve Recommendation Diversity](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FDiversity\u002F%5BNeurIPS%2018%5D%5BHulu%5D%20Fast%20Greedy%20MAP%20Inference%20for%20Determinantal%20Point%20Process%20to%20Improve%20Recommendation%20Diversity.pdf)\n\n## EE\n* [[LinUCB][WWW 10][Yahoo] A Contextual-Bandit Approach to Personalized News Article Recommendation](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FEE\u002F%5BLinUCB%5D%5BWWW%2010%5D%5BYahoo%5D%20A%20Contextual-Bandit%20Approach%20to%20Personalized%20News%20Article%20Recommendation.pdf)\n\n## RL\n* [[IJCAI 19][Google] Reinforcement Learning for Slate-based Recommender Systems_A Tractable Decomposition and Practical Methodology](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FRL\u002F%5BIJCAI%2019%5D%5BGoogle%5D%20Reinforcement%20Learning%20for%20Slate-based%20Recommender%20Systems_A%20Tractable%20Decomposition%20and%20Practical%20Methodology.pdf)\n* [[WSDM 19][Google] Top-K Off-Policy Correction for a REINFORCE Recommender System](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FRL\u002F%5BWSDM%2019%5D%5BGoogle%5D%20Top-K%20Off-Policy%20Correction%20for%20a%20REINFORCE%20Recommender%20System.pdf)\n* [[DRN][WWW 18][Microsoft] DRN_A Deep Reinforcement Learning Framework for News Recommendation](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FRL\u002F%5BDRN%5D%5BWWW%2018%5D%5BMicrosoft%5D%20DRN_A%20Deep%20Reinforcement%20Learning%20Framework%20for%20News%20Recommendation.pdf)","# 深度推荐系统与CTR预估工业界相关论文、业界分享\n动态更新推荐、广告工业界经典以及最前沿的论文、业界分享集合。所有资料均整理来自于互联网，如有侵权，请联系小助手deepdeliver。同时欢迎对推荐、广告方面工业界感兴趣的小伙伴添加小助手，将自动拉入交流群：\n\u003Cdiv align=left>\n\u003Cimg width=\"200\" src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fimsheridan_DeepRec_readme_8536b366d82d.jpg\" alt=\"交流群\"\u002F>\n\u003C\u002Fdiv>\n\n### 其他相关资源\n* [王喆的推荐系统paper列表](https:\u002F\u002Fgithub.com\u002Fwzhe06\u002FReco-papers)\n* [王喆的广告paper列表](https:\u002F\u002Fgithub.com\u002Fwzhe06\u002FAd-papers)\n\n### 论文脉络\n\n\u003Cdiv align=center>\n\u003Cimg src=\"https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fimsheridan_DeepRec_readme_ed85f0b99d5d.jpg\" alt=\"脉络图\"\u002F>\n\u003C\u002Fdiv>\n\n# 目录\n\n## CTR\n* [[FiBiNET][RecSys 19][Weibo] FiBiNET_结合特征重要性和双线性特征交互的点击率预估](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FCTR\u002F%5BFiBiNET%5D%5BRecSys%2019%5D%5BWeibo%5D%20FiBiNET_结合特征重要性和双线性特征交互的点击率预估.pdf)\n* [[DSIN][IJCAI 19][Alibaba] 针对点击率预估的深度会话兴趣网络](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FCTR\u002F%5BDSIN%5D%5BIJCAI%2019%5D%5BAlibaba%5D%20针对点击率预估的深度会话兴趣网络.pdf)\n* [[FGCNN][WWW 19][Huawei] 基于卷积神经网络的特征生成用于点击率预估](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FCTR\u002F%5BFGCNN%5D%5BWWW%2019%5D%5BHuawei%5D%20基于卷积神经网络的特征生成用于点击率预估.pdf)\n* [[AutoInt][CIKM 19] AutoInt_通过自注意力神经网络实现自动特征交互学习](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FCTR\u002F%5BAutoInt%5D%5BCIKM%2019%5D%20AutoInt_通过自注意力神经网络实现自动特征交互学习.pdf)\n* [[DIEN][AAAI 19][Alibaba] 针对点击率预估的深度兴趣演化网络](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FCTR\u002F%5BDIEN%5D%5BAAAI%2019%5D%5BAlibaba%5D%20针对点击率预估的深度兴趣演化网络.pdf)\n* [[PNN][TOIS 18] 基于产品的神经网络用于用户响应预测](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FCTR\u002F%5BPNN%5D%5BTOIS%2018%5D%20基于产品的神经网络用于用户响应预测.pdf)\n* [[xDeepFM][KDD 18][Microsoft] xDeepFM_结合显式与隐式特征交互的推荐系统](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FCTR\u002F%5BxDeepFM%5D%5BKDD%2018%5D%5BMicrosoft%5D%20xDeepFM_结合显式与隐式特征交互的推荐系统.pdf)\n* [[DCN][KDD 17][Google] 广告点击预测的深度与交叉网络](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FCTR\u002F%5BDCN%5D%5BKDD%2017%5D%5BGoogle%5D%20广告点击预测的深度与交叉网络.pdf)\n* [[DIN][KDD 18][Alibaba] 针对点击率预估的深度兴趣网络](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FCTR\u002F%5BDIN%5D%5BKDD%2018%5D%5BAlibaba%5D%20针对点击率预估的深度兴趣网络.pdf)\n* [[FNN][ECIR 16] 多字段分类数据上的深度学习——以用户响应预测为例](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FCTR\u002F%5BFNN%5D%5BECIR%2016%5D%20多字段分类数据上的深度学习——以用户响应预测为例.pdf)\n* [[AFM][IJCAI 17] 注意力因子分解机——通过注意力网络学习特征交互权重](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FCTR\u002F%5BAFM%5D%5BIJCAI%2017%5D%20注意力因子分解机——通过注意力网络学习特征交互权重.pdf)\n* [[DeepFM][IJCAI 17][Huawei] DeepFM_基于因子分解机的神经网络用于CTR预测](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FCTR\u002F%5BDeepFM%5D%5BIJCAI%2017%5D%5BHuawei%5D%20DeepFM_基于因子分解机的神经网络用于CTR预测.pdf)\n* [[NFM][SIGIR 17] 稀疏预测分析中的神经因子分解机](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FCTR\u002F%5BNFM%5D%5BSIGIR%2017%5D%20稀疏预测分析中的神经因子分解机.pdf)\n* [[WDL][DLRS 16][Google] 推荐系统的宽&深学习](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FCTR\u002F%5BWDL%5D%5BDLRS%2016%5D%5BGoogle%5D%20推荐系统的宽&深学习.pdf)\n\n## Match\n* [[JTM][NIPS 19] 推荐系统中树状索引与深度模型的联合优化](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FMatch\u002F%5BJTM%5D%5BNIPS%2019%5D%20推荐系统中树状索引与深度模型的联合优化.pdf)\n* [[MIND][arxiv 19][Alibaba] 具有动态路由的多兴趣网络用于天猫推荐](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FMatch\u002F%5BMIND%5D%5Barxiv%2019%5D%5BAlibaba%5D%20具有动态路由的多兴趣网络用于天猫推荐.pdf)\n* [[SDM][CIKM 19][Alibaba] 面向在线大规模推荐系统的顺序深度匹配模型](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FMatch\u002F%5BSDM%5D%5BCIKM%2019%5D%5BAlibaba%5D%20面向在线大规模推荐系统的顺序深度匹配模型.pdf)\n* [[TDM][KDD 18][Alibaba] 学习用于推荐系统的树状深度模型](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FMatch\u002F%5BTDM%5D%5BKDD%2018%5D%5BAlibaba%5D%20学习用于推荐系统的树状深度模型.pdf)\n* [[NCF][WWW 17] 神经协同过滤](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FMatch\u002F%5BNCF%5D%5BWWW%2017%5D%20神经协同过滤.pdf)\n* [[YoutubeDNN][RecSys 16][Google] 用于YouTube推荐的深度神经网络](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FMatch\u002F%5BYoutubeDNN%5D%5BRecSys%2016%5D%5BGoogle%5D%20用于YouTube推荐的深度神经网络.pdf)\n* [[DSSM][CIKM 13][Microsoft] 利用点击数据学习用于网页搜索的深度结构化语义模型](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FMatch\u002F%5BDSSM%5D%5BCIKM%2013%5D%5BMicrosoft%5D%20利用点击数据学习用于网页搜索的深度结构化语义模型.pdf)\n\n## 排序\n* [[PRM][RecSys 19][阿里巴巴] 面向推荐的个性化重排序](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FRanking\u002F%5BPRM%5D%5BRecSys%2019%5D%5BAlibaba%5D%20Personalized%20Re-ranking%20for%20Recommendation.pdf)\n* [[BERT4Rec][CIKM 19][阿里巴巴] BERT4Rec_基于Transformer双向编码器表示的序列推荐](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FRanking\u002F%5BBERT4Rec%5D%5BCIKM%2019%5D%5BAlibaba%5D%20BERT4Rec_Sequential%20Recommendation%20with%20Bidirectional%20Encoder%20Representations%20from%20Transformer.pdf)\n* [[BST][DLP-KDD 19][阿里巴巴] 阿里巴巴电商推荐中的行为序列Transformer](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FRanking\u002F%5BBST%5D%5BDLP-KDD%2019%5D%5BAlibaba%5D%20Behavior%20Sequence%20Transformer%20for%20E-commerce%20Recommendation%20in%20Alibaba.pdf)\n\n## 嵌入\n* [[Airbnb嵌入][KDD 18][爱彼迎] 爱彼迎搜索排名中基于嵌入的实时个性化](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FEmbedding\u002F%5BAirbnb%20Embedding%5D%5BKDD%2018%5D%5BAirbnb%5D%20Real-time%20Personalization%20using%20Embeddings%20for%20Search%20Ranking%20at%20Airbnb.pdf)\n* [[阿里巴巴嵌入][KDD 18][阿里巴巴] 阿里巴巴电商推荐中的十亿级商品嵌入](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FEmbedding\u002F%5BAlibaba%20Embedding%5D%5BKDD%2018%5D%5BAlibaba%5D%20Billion-scale%20Commodity%20Embedding%20for%20E-commerce%20Recommendation%20in%20Alibaba.pdf)\n* [[DeepWalk][KDD 14] DeepWalk- 社交表征的在线学习](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FEmbedding\u002F%5BDeepWalk%5D%5BKDD%2014%5D%20DeepWalk-%20Online%20Learning%20of%20Social%20Representations.pdf)\n* [[LINE][WWW 15][微软] LINE_大规模信息网络嵌入](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FEmbedding\u002F%5BLINE%5D%5BWWW%2015%5D%5BMicrosoft%5D%20LINE_Large-scale%20Information%20Network%20Embedding.pdf)\n* [[Node2vec][KDD 16] Node2vec_可扩展的网络特征学习](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FEmbedding\u002F%5BNode2vec%5D%5BKDD%2016%5D%20Node2vec_Scalable%20Feature%20Learning%20for%20Networks.pdf)\n* [[SDNE][KDD 16] 结构化深度网络嵌入](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FEmbedding\u002F%5BSDNE%5D%5BKDD%2016%5D%20Structural%20Deep%20Network%20Embedding.pdf)\n* [[Struc2Vec][KDD 17] struc2vec_从结构身份中学习节点表示](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FEmbedding\u002F%5BStruc2Vec%5D%5BKDD%2017%5Dstruc2vec_Learning%20Node%20Representations%20from%20Structural%20Identity.pdf)\n* [[GraphSAGE][NIPS 17] 大型图上的归纳式表示学习](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FEmbedding\u002F%5BGraphSAGE%5D%5BNIPS%2017%5D%20Inductive%20Representation%20Learning%20on%20Large%20Graphs.pdf)\n* [[GCN][ICLR 17] 基于图卷积网络的半监督分类](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FEmbedding\u002F%5BGCN%5D%5BICLR%2017%5D%20Semi-supervised%20Classification%20with%20Graph%20Convolutional%20Networks.pdf)\n\n## 多任务学习\n* [[RecSys 19][阿里巴巴] 电子商务推荐中多目标优化的帕累托高效算法](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FMTL\u002F%5BRecSys%2019%5D%5BAlibaba%5D%20A%20Pareto-Efficient%20Algorithm%20for%20Multiple%20Objective%20Optimization%20in%20E-Commerce%20Recommendation.pdf)\n* [[MMoE][KDD 18][谷歌] 多任务学习中基于多门混合专家模型的任务关系建模](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FMTL\u002F%5BMMoE%5D%5BKDD%2018%5D%5BGoogle%5D%20Modeling%20Task%20Relationships%20in%20Multi-task%20Learning%20with%20Multi-gate%20Mixture-of-Experts.pdf)\n* [[ESMM][SIGIR 18][阿里巴巴] 全空间多任务模型_一种有效估算点击后转化率的方法](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FMTL\u002F%5BESMM%5D%5BSIGIR%2018%5D%5BAlibaba%5D%20Entire%20Space%20Multi-Task%20Model_An%20Effective%20Approach%20for%20Estimating%20Post-Click%20Conversion%20Rate.pdf)\n\n## 多样性\n* [[CIKM 18][谷歌] 使用行列式点过程在YouTube上实现实用的多样化推荐](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FDiversity\u002F%5BCIKM%2018%5D%5BGoogle%5D%20Practical%20Diversified%20Recommendations%20on%20YouTube%20with%20Determinantal%20Point%20Processes.pdf)\n* [[NeurIPS 18][Hulu] 快速贪婪MAP推理用于行列式点过程以提升推荐多样性](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FDiversity\u002F%5BNeurIPS%2018%5D%5BHulu%5D%20Fast%20Greedy%20MAP%20Inference%20for%20Determinantal%20Point%20Process%20to%20Improve%20Recommendation%20Diversity.pdf)\n\n## 探索与利用\n* [[LinUCB][WWW 10][雅虎] 一种基于上下文Bandit的个性化新闻文章推荐方法](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FEE\u002F%5BLinUCB%5D%5BWWW%2010%5D%5BYahoo%5D%20A%20Contextual-Bandit%20Approach%20to%20Personalized%20News%20Article%20Recommendation.pdf)\n\n## 强化学习\n* [[IJCAI 19][谷歌] 面板式推荐系统中的强化学习_一种可行的分解与实用方法](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FRL\u002F%5BIJCAI%2019%5D%5BGoogle%5D%20Reinforcement%20Learning%20for%20Slate-based%20Recommender%20Systems_A%20Tractable%20Decomposition%20and%20Practical%20Methodology.pdf)\n* [[WSDM 19][谷歌] REINFORCE推荐系统中的Top-K离策略校正](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FRL\u002F%5BWSDM%2019%5D%5BGoogle%5D%20Top-K%20Off-Policy%20Correction%20for%20a%20REINFORCE%20Recommender%20System.pdf)\n* [[DRN][WWW 18][微软] DRN_用于新闻推荐的深度强化学习框架](https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec\u002Fblob\u002Fmaster\u002FRL\u002F%5BDRN%5D%5BWWW%2018%5D%5BMicrosoft%5D%20DRN_A%20Deep%20Reinforcement%20Learning%20Framework%20for%20News%20Recommendation.pdf)","# DeepRec 快速上手指南\n\n## 项目简介\nDeepRec 是一个专注于**深度推荐系统与 CTR（点击率）预估**的工业界论文与资源集合。本项目并非一个可直接运行的软件框架，而是一个动态更新的**知识库**，汇集了来自阿里巴巴、谷歌、微软、华为等顶尖公司的经典及前沿论文、技术分享。它涵盖了从特征嵌入（Embedding）、匹配（Match）、排序（Ranking）到多任务学习（MTL）等推荐系统全链路的最新研究成果。\n\n> **注意**：本指南旨在帮助开发者快速获取和学习该项目整理的核心资源。如需复现论文算法，请根据下方提供的论文链接查阅具体代码实现（通常链接指向论文 PDF，部分论文对应开源代码可在其原始仓库找到）。\n\n---\n\n## 环境准备\n\n由于 DeepRec 主要是文献资源库，无需复杂的运行时环境。但为了高效阅读和研究其中的深度学习论文，建议开发者准备以下基础环境：\n\n### 系统要求\n- **操作系统**：Windows \u002F macOS \u002F Linux (推荐 Ubuntu 20.04+)\n- **浏览器**：支持 PDF 预览的现代浏览器（Chrome, Edge, Firefox）\n- **网络环境**：由于部分资源托管在 GitHub，国内用户建议配置网络加速或使用镜像。\n\n### 前置依赖（用于复现论文算法）\n若你计划根据论文复现模型，通常需要以下深度学习栈：\n- **Python**: 3.7+\n- **深度学习框架**: TensorFlow 2.x 或 PyTorch 1.8+ (视具体论文而定)\n- **包管理工具**: `pip` 或 `conda`\n\n---\n\n## 安装与获取步骤\n\nDeepRec 本身是一个代码仓库，获取方式主要为克隆仓库或直接浏览在线目录。\n\n### 方法一：克隆仓库（推荐）\n使用 Git 将完整资源下载到本地，方便离线阅读和检索。\n\n```bash\n# 推荐使用国内镜像源加速克隆（如 Gitee 镜像，若可用）\n# 若无特定镜像，直接使用 GitHub 地址\ngit clone https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec.git\n\n# 进入目录\ncd DeepRec\n```\n\n*注：若 GitHub 连接缓慢，可尝试使用代理或搜索相关的 Gitee 镜像仓库进行克隆。*\n\n### 方法二：在线浏览\n直接访问 GitHub 仓库页面，按目录结构浏览各类论文 PDF 文件：\n- 访问地址：`https:\u002F\u002Fgithub.com\u002Fimsheridan\u002FDeepRec`\n\n---\n\n## 基本使用\n\nDeepRec 的使用核心在于**按需检索**和**研读论文**。项目已按推荐系统的关键模块进行了分类整理。\n\n### 1. 目录结构说明\n下载后，你可以在本地看到以下核心分类目录：\n\n- **CTR\u002F**: 点击率预估经典模型（如 DeepFM, DIN, DCN, xDeepFM 等）。\n- **Match\u002F**: 召回与匹配阶段模型（如 YoutubeDNN, DSSM, MIND, TDM 等）。\n- **Ranking\u002F**: 精排序阶段模型（如 BST, BERT4Rec, PRM 等）。\n- **Embedding\u002F**: 图嵌入与序列嵌入技术（如 GraphSAGE, Node2vec, Airbnb Embedding 等）。\n- **MTL\u002F**: 多任务学习模型（如 MMoE, ESMM 等）。\n- **Diversity\u002F**: 推荐多样性相关研究。\n- **RL\u002F**: 强化学习在推荐中的应用。\n- **EE\u002F**: 探索与利用（Exploration & Exploitation）策略。\n\n### 2. 快速查找示例\n假设你想研究阿里巴巴提出的 **DIN (Deep Interest Network)** 模型：\n\n1. 打开本地 `DeepRec\u002FCTR\u002F` 文件夹。\n2. 找到文件：`[DIN][KDD 18][Alibaba] Deep Interest Network for Click-Through Rate Prediction.pdf`。\n3. 使用 PDF 阅读器打开，即可阅读原始论文。\n\n假设你想学习多任务学习中的 **MMoE** 模型：\n\n1. 打开本地 `DeepRec\u002FMTL\u002F` 文件夹。\n2. 找到文件：`[MMoE][KDD 18][Google] Modeling Task Relationships in Multi-task Learning with Multi-gate Mixture-of-Experts.pdf`。\n\n### 3. 进阶学习路径建议\n对于中国开发者，建议按照以下脉络进行学习（参考项目提供的脉络图）：\n\n1. **入门基础**: 从 `CTR` 目录下的 `Wide & Deep (WDL)`, `DeepFM` 开始，理解特征交叉的基本原理。\n2. **序列建模**: 阅读 `DIN`, `DIEN`, `DSIN`，掌握用户行为序列的建模方法。\n3. **工业实战**: 重点研读标注为 `[Alibaba]`, `[Google]`, `[Huawei]` 的论文，这些通常包含更多工程落地细节。\n4. **前沿探索**: 查看 `Match` 目录中的 `MIND` (多兴趣提取) 和 `Ranking` 目录中的 `BERT4Rec`，了解 Transformer 在推荐中的应用。\n\n### 4. 获取额外资源\n项目中还引用了王喆老师整理的优质 Paper 列表，可作为补充阅读：\n- 推荐系统论文列表：`https:\u002F\u002Fgithub.com\u002Fwzhe06\u002FReco-papers`\n- 广告系统论文列表：`https:\u002F\u002Fgithub.com\u002Fwzhe06\u002FAd-papers`\n\n---\n\n*提示：本项目资料均整理自互联网，仅供学习交流。如需深入交流，可参考原仓库中的联系方式加入相关技术社群。*","某电商平台的算法团队正致力于优化首页商品流的点击率（CTR）预估模型，以应对大促期间流量激增带来的转化压力。\n\n### 没有 DeepRec 时\n- **资料搜集耗时巨大**：工程师需手动在 arXiv、Google Scholar 及各大会官网大海捞针，难以系统性地获取如 DIN、DIEN 等阿里系经典论文及最新工业界成果。\n- **技术选型缺乏脉络**：面对碎片化的文献，团队难以理清从 FNN、DeepFM 到 xDeepFM、AutoInt 的技术演进逻辑，导致模型迭代方向盲目，容易重复造轮子。\n- **复现门槛高企**：缺乏经过筛选的权威资料库，初级成员在理解特征交叉（如 FiBiNET）或序列兴趣演化等复杂机制时，常因找不到核心源码或解读资料而陷入停滞。\n\n### 使用 DeepRec 后\n- **一站式资源聚合**：团队直接利用 DeepRec 中动态更新的论文集合，瞬间获取包含华为 FGCNN、谷歌 DCN 等大厂前沿成果，将数周的调研工作压缩至几天。\n- **清晰的技术路线图**：借助工具提供的论文脉络图，团队迅速定位到适合当前业务阶段的模型架构，明确了从显式到隐式特征交互的优化路径。\n- **高效落地与复现**：成员按图索骥下载带标注的经典论文（如 DSIN、PNN），快速吃透工业界实战细节，显著缩短了从理论验证到线上 A\u002FB 测试的周期。\n\nDeepRec 通过构建结构化的工业界知识图谱，将推荐算法团队的研发效率从“盲人摸象”提升为“按图索骥”，加速了高精度 CTR 模型的落地进程。","https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fimsheridan_DeepRec_00fa11d9.png","imsheridan","deepdeliever","https:\u002F\u002Foss.gittoolsai.com\u002Favatars\u002Fimsheridan_3ec82034.jpg",null,"Beijing","https:\u002F\u002Fweibo.com\u002Fimsheridan","https:\u002F\u002Fgithub.com\u002Fimsheridan",1020,218,"2026-03-15T14:22:49","MIT",5,"","未说明",{"notes":87,"python":85,"dependencies":88},"该仓库主要是一个推荐系统与 CTR 预估领域的论文、业界分享集合（包含 PDF 文档链接），并非一个可直接运行的软件工具或代码库，因此 README 中未提供具体的运行环境、依赖库或硬件需求信息。用户需根据仓库中列出的具体论文（如 DIN, DeepFM, BERT4Rec 等）去查找对应的原始开源代码实现以获取环境配置。",[],[14],[91,92,93,94,95,96],"deep-learning","recommendation-system","recommendation","reinforcement-learning","exploration-exploitation","computational-advertising","2026-03-27T02:49:30.150509","2026-04-11T18:31:18.072437",[],[]]