Compare commits

..

4 Commits

Author SHA1 Message Date
codex-bot 2f445eddc8 Record CIKM programme and publisher registry reconciliation
Metadata integrity checks / metadata (push) Successful in 7m6s
2026-11-18 15:28:08 +08:00
Zayden-Jung f982851560 Merge pull request '记录 MM 出版核对与 UAI 原始决定来源边界' (#67) from codex/mm-publication into main
Metadata integrity checks / metadata (push) Successful in 9m35s
Reviewed-on: #67
2026-11-17 18:44:00 +00:00
Zayden-Jung 88aeb4d2eb Merge pull request '封装 MM 2026 全量出版记录与独立节目证据' (#68) from codex/mm-publications into main
Metadata integrity checks / metadata (push) Successful in 10m5s
Reviewed-on: #68
2026-11-17 18:43:55 +00:00
codex-bot 7e63035570 Document MM publication anchors and UAI raw decision-source limits
Metadata integrity checks / metadata (push) Successful in 4m8s
2026-11-17 23:43:21 +08:00
6 changed files with 853 additions and 3 deletions
+2 -2
View File
@@ -1,6 +1,6 @@
# 会议覆盖与公开决定核查
最新生产进展和剩余工作见 [2026-11-17 补漏状态](backfill-status-2026-11-17.md):固定 278 个会议年份槽位中,270 个有记录、8 个仍为空;这不代表完整覆盖,第 14 步仍未完成。
最新生产验收与剩余工作见 [2026-11-17 补漏状态](backfill-status-2026-11-17.md):当日 `14:06:28 UTC` 固定范围快照中,278 个会议年份槽位有 272 个含记录、6 个仍为空;这不代表完整覆盖,第 14 步仍未完成。后续 MM 规范化、只读候选审计与出版核查不等于生产导入。
**当前工作范围:2020 年起主动补漏。** 2000 年起已发现资料保留,恢复工作从 [补漏范围与恢复清单(2026-11-10)](backfill-handoff-2026-11-10.md) 开始;下文旧日期快照不改写,较早年份排除不算补齐。
@@ -72,7 +72,7 @@ PGSERVICEFILE=/etc/paper-atlas/pg_service.conf paper-atlas coverage \
## 优先缺口
- **UAI 2024–2026**:官方均允许作者选择公开拒稿,当前主库拒稿仍为零。原 notes 接口要求 Challenge 验证,公开镜像尚无已核验来源,必须保留这个真实缺口;不能根据 PMLR 录用名单推断拒稿。参见 [2025 投稿指南](https://www.auai.org/uai2025/submission_instructions)、JSON 内其他年份官方证据。
- **UAI 2024–2026**:官方均允许作者选择公开拒稿,首份生产快照拒稿为零。原 notes 接口要求 Challenge 验证;新核验镜像的 2022–2025 年 902 个论坛来自录用子集,不能填补公开拒稿缺口。2024 年的元评审也不能直接转成最终决定,见 [2026-11-17 原始来源与回复结构核查](uai-public-negative-source-status-2026-11-17.md)。不能根据 PMLR 或镜像录用名单反推拒稿;政策参见 [2025 投稿指南](https://www.auai.org/uai2025/submission_instructions)、JSON 内其他年份官方证据。
- **EMNLP 2023**:官方允许公开 opt-in 拒稿,目前只导入 9 条。需继续核验实际公开全集。ARR 评审、conference commitment 与最终会议决定分开;[ARR 作者说明](https://aclrollingreview.org/authors) 明确其自愿捐赠数据与 EMNLP 2023 直接投稿发布的范围不同。
- **ICLR**:固定快照已观察 2017 与 2020–2026;快照之后,2013/2018/2019 的 1,654 条来源记录已完成独立 PostgreSQL 验收(报告时间 `2026-11-07T19:52:28.378Z`),不回写旧快照计数,也不据此宣称逐届全集已齐。快照中的 10,208 次 unknown 不等于 rejected,近期补采与冲突核对仍在进行;评审推荐不能当最终决定。
- **MM / MLSys**:MM 11,200 条书目记录保留 `proceedings_unspecified`;MLSys 的 448 次官方投稿与 313 条 DBLP 记录重叠。2026 MLSys 的研究/工业轨道在 [官方 CFP](https://mlsys.org/Conferences/2026/CallForResearchPapers) 中分开,仍须核对逐篇轨道证据,不能仅凭会议名全部归 main。
+3 -1
View File
@@ -33,7 +33,7 @@ SDM 新增事件使 SeBA 和 AdamG 同时保有较早 ICLR 拒稿与后来 SDM
| ECCV | 2,834 个不同初步节目 ID 及公开摘要已独立采集;1,997 个只读候选历史证据包已保存,尚未入库 | 初步日程不等于 Springer 最终论文集;继续作品/版本和出版核查。见 [来源状态](../eccv2026-source-status.md) |
| ICDM | 324 项线下节目观察,明确研究241/应用83;演讲长168/短156,摘要均缺失;规范化和80个候选历史包已核验 | 展示长短不能当成论文类型;需审查作品与正式出版版本,尚未导入主库。见 [采集器说明](../icdm2026-program.md) |
| ISMIR | 140 项节目和正式名单均已核验,140份摘要,101组自愿公开评审;两处标题、三处作者名单差异保留,78个候选历史包已核验 | 数字 miniconf ID 不是原始 OpenReview ID;评审意见不转成最终拒稿。继续作品身份/历史关联,尚未入库。采集器见 [PR #63](https://gitea.entroia.com/Zayden-Jung/paper-atlas/pulls/63) |
| MM | 官方10.8MB日程JSON已冻结,4,326次观察对应2,143个节目ID、28个来源track;重复展示核心字段一致,216个ID缺摘要,其中两项主轨同时缺作者 | 不是2,143篇已核验科研论文。先区分各轨贡献与活动、保留全部上下文并单列缺失;尚未规范化/入库。见 [来源与访问证据](../mm2026-source-status.md) |
| MM | 全28轨2143个节目ID/4326展示位置已规范化;814个只读候选历史证据包已核验,24篇候选涉及24次负面投稿/37条来源观察;67项原始arXiv DOI中66项注册元数据一致、1项404暂缓 | 不是2143篇已核验科研论文。继续逐轨贡献/出版依据与作品审查,0身份批准、0入库。ISBN查询与父XML均只返回父记录,不作逐篇清单。见 [来源、审计与恢复证据](../mm2026-source-status.md) |
| CIKM | 官方录用/日程入口与五个分轨页面已保存;五个分轨链接实际是CFP/政策,不是逐篇录用名单 | 查找实际节目或出版社完整目录;TBD和没有静态表不能当作没有论文 |
| IROS | 官方论文索引入口的原始HTTP403已保存 | 保留访问限制,不重试绕过;寻找可合法读取、能独立核验的出版元数据 |
@@ -43,6 +43,8 @@ SDM 新增事件使 SeBA 和 AdamG 同时保有较早 ICLR 拒稿与后来 SDM
MM 官方群组元数据可匿名读取,但根据其给出的 submission invitation 和 rejected/withdrawn/desk-rejected venue ID,四次限1条的匿名 notes 核查均为403。查询、时间、原文和哈希已保留,没有认证重试或绕过权限。访问限制不等于零条负面记录;详见 [MM来源证据](../mm2026-source-status.md)。ISMIR公开的101组评审来自录用节目,不能记成101条拒稿。
UAI新增固定镜像已核对2022–2025的902个原始论坛及3053个回复的元数据/字段结构。整理版均标录用,2022保存230条明确Accept决定;2024仅有评审、作者回复、元评审,2023/2025未保存回复,不能据此补拒稿。来源限制和恢复见 [逐年原始核查](uai-public-negative-source-status-2026-11-17.md)。这不是UAI没有公开拒稿的结论,缺口保持未补齐。
第14步仍需逐会议、年份、track完成公开拒稿/desk-reject/withdrawn来源核查和可得数据采集;继续核查UAI近年自愿公开范围、EMNLP实际覆盖范围及其余已列明缺口。不能从未见录用、评分、评审推荐、缺少摘要或接口不可访问制造负面决定。
之后仍需完成跨会作品/历次投稿与类型标签关联(第15步),所有导入结束后的独立全库去重、标识冲突、近似标题作者、摘要/约束/查询及覆盖审计(第16步),以及长期运行与恢复文档最终整理(第17步)。阶段性来源入库、测试通过、某年度有记录均不能替代这些验收。
@@ -0,0 +1,759 @@
{
"query_records": 1081,
"official_volume": "10.1145/3799682",
"registry_records": 1081,
"programme_records": 1008,
"shared": 1008,
"shared_by_track": {
"SHORT": 234,
"APPLIED": 86,
"RESOURCE": 46,
"DEMO": 46,
"FULL": 596
},
"programme_only": [],
"registry_only": [
{
"doi": "10.1145/3799682.3839246",
"title": "Files Search Relevance using Wide and Deep Networks",
"page": "9464-9465"
},
{
"doi": "10.1145/3799682.3839247",
"title": "Capacity-Aligned Graph Learning for AML: From Top-K Alert Ranking to Candidate Fraud-Ring Discovery at PayPay",
"page": "9487-9489"
},
{
"doi": "10.1145/3799682.3839248",
"title": "Audit-Ready RAG Retrieval in Industrial Evidence Generation: Positional HNSW and Human Review Signals",
"page": "9449-9451"
},
{
"doi": "10.1145/3799682.3839250",
"title": "Adapting Code LLMs to Domain-specific Languages via Synthethic Instruction Generation and Reinforcement Learning with AI Feedback in Low-resource Settings",
"page": "9481-9483"
},
{
"doi": "10.1145/3799682.3839251",
"title": "Aligning Real-World Document Schemas to Fine-Grained Concept Vocabularies",
"page": "9476-9478"
},
{
"doi": "10.1145/3799682.3839252",
"title": "From Demo to Production: Generative Search Lessons from Alexa+",
"page": "9473-9475"
},
{
"doi": "10.1145/3799682.3839253",
"title": "A Phased Workflow for Operating LLM-Based Coding Agents",
"page": "9468-9469"
},
{
"doi": "10.1145/3799682.3839254",
"title": "Edge-Native Agentic AI for Scalable Knowledge Systems in Industrial Applications",
"page": "9496-9498"
},
{
"doi": "10.1145/3799682.3839255",
"title": "CERTBOT: An Agentic System for Cybersecurity",
"page": "9455-9457"
},
{
"doi": "10.1145/3799682.3839256",
"title": "Optimize the Harness, Not the Model: Fitting a Minimal Agent to a Self-Hostable LLM for Investigative Analysis",
"page": "9484-9486"
},
{
"doi": "10.1145/3799682.3839257",
"title": "Vector Index-Native Relevance Feedback in Production Search",
"page": "9490-9492"
},
{
"doi": "10.1145/3799682.3839258",
"title": "Supporting Human Memory with Audiobook Recaps: From Quality Discovery to Automatic Prompt Optimization",
"page": "9458-9460"
},
{
"doi": "10.1145/3799682.3839259",
"title": "Reliable Evaluation of AI Assisted Peer Review",
"page": "9452-9454"
},
{
"doi": "10.1145/3799682.3839260",
"title": "Temporal Moment Generation: A New Recommendation Task and Cost-Efficient Distillation for Outbound Marketing at Scale",
"page": "9493-9495"
},
{
"doi": "10.1145/3799682.3839261",
"title": "The Lift Is Usually Wrong: A Field Guide to Event Studies on Cross-Surface Behavioral Logs",
"page": "9466-9467"
},
{
"doi": "10.1145/3799682.3839262",
"title": "Contextual Ranker in Production: Session-Aware Similar Item Ranking for Lower-Funnel E-Commerce",
"page": "9461-9463"
},
{
"doi": "10.1145/3799682.3839263",
"title": "Preserving Utility When Watermarking Fine-Tuned Vision Classifiers",
"page": "9479-9480"
},
{
"doi": "10.1145/3799682.3839264",
"title": "To Re-embedd or Not to Re-embedd",
"page": "9499-9500"
},
{
"doi": "10.1145/3799682.3839265",
"title": "Rebuilding the Content Ranking Stack for Prime Video: Lessons Learned Building from Scratch",
"page": "9501-9503"
},
{
"doi": "10.1145/3799682.3839266",
"title": "Amplified and Emergent Safety Risks in Multi-Agent Recommendation and Retrieval",
"page": "9398-9401"
},
{
"doi": "10.1145/3799682.3839267",
"title": "A Tutorial on Retrieval-Augmented Generation for LLM-based Recommender Systems",
"page": "9415-9418"
},
{
"doi": "10.1145/3799682.3839268",
"title": "Knowledge Graphs Data Quality at Scale: Challenges, Solutions, and Applications",
"page": "9424-9428"
},
{
"doi": "10.1145/3799682.3839269",
"title": "Neural Networks meet Explicit Knowledge Representation: Towards Mechanistic Interpretability and Neuro-symbolic Modeling by-design",
"page": "9445-9448"
},
{
"doi": "10.1145/3799682.3839270",
"title": "Resource-Efficient Generative AI: Rethinking Knowledge Management and Delivery",
"page": "9402-9406"
},
{
"doi": "10.1145/3799682.3839271",
"title": "Beyond Benchmarks: Robustness for Real-World Information Systems",
"page": "9411-9414"
},
{
"doi": "10.1145/3799682.3839272",
"title": "Graph Correspondence Algorithms: A Comprehensive Tutorial on Graph Alignment and Subgraph Matching",
"page": "9433-9436"
},
{
"doi": "10.1145/3799682.3839273",
"title": "'Did something just break-' -- A Stream-Learning View on Monitoring: the Geometry and Causes of Change",
"page": "9407-9410"
},
{
"doi": "10.1145/3799682.3839274",
"title": "IDs That Mean Something: A Hands-On Tutorial on Semantic IDs for Recommender Systems",
"page": "9419-9423"
},
{
"doi": "10.1145/3799682.3839275",
"title": "Building Production LLM Agents: An Evaluation-Driven Playbook",
"page": "9437-9440"
},
{
"doi": "10.1145/3799682.3839276",
"title": "Path-based Reasoning on Knowledge Graphs for Explainable Recommender Systems",
"page": "9394-9397"
},
{
"doi": "10.1145/3799682.3839277",
"title": "Agentic Search: Foundations, Optimization, and Long-Horizons",
"page": "9429-9432"
},
{
"doi": "10.1145/3799682.3839278",
"title": "Query Expansion Revisited: From Classical Retrieval to Agentic Search",
"page": "9388-9393"
},
{
"doi": "10.1145/3799682.3839279",
"title": "Neural Lexical Search with Learned Sparse Retrieval",
"page": "9441-9444"
},
{
"doi": "10.1145/3799682.3839280",
"title": "Artificial Intelligence for Data Engineering and Management Of Networked Systems (AI4DEMONS)",
"page": "9553-9556"
},
{
"doi": "10.1145/3799682.3839281",
"title": "EARL: The 3rd Workshop on Evaluating and Applying Recommender Systems with Large Language Models",
"page": "9542-9545"
},
{
"doi": "10.1145/3799682.3839282",
"title": "Second International Workshop on Hybrid AI for Human-Centric Personalization and Recommendation (HyPeR 2026)",
"page": "9530-9533"
},
{
"doi": "10.1145/3799682.3839283",
"title": "GREEN-AI: Grounding Research on Energy- and Resource-Efficient Networks for Sustainable AI",
"page": "9508-9511"
},
{
"doi": "10.1145/3799682.3839284",
"title": "STMA4KD: Spatio-Temporal Data Management and Analysis for Knowledge Discovery Workshop",
"page": "9534-9537"
},
{
"doi": "10.1145/3799682.3839285",
"title": "LLM4XAI: Generative Models for Explainable AI Narratives",
"page": "9550-9552"
},
{
"doi": "10.1145/3799682.3839286",
"title": "VERIMEDIA: The second International Workshop on Media Verification and Integrity",
"page": "9504-9507"
},
{
"doi": "10.1145/3799682.3839287",
"title": "GRAIL 2026: Workshop on Generative, Retrieval-augmented, and Agentic Intelligence for Personalization",
"page": "9526-9529"
},
{
"doi": "10.1145/3799682.3839288",
"title": "KEEPER: Knowledge Elicitation, Externalization, and Preservation for Expert Repositories",
"page": "9538-9541"
},
{
"doi": "10.1145/3799682.3839289",
"title": "Clinical-MIRF: Multimodal Information Retrieval and Fusion in Clinical Workflows",
"page": "9519-9522"
},
{
"doi": "10.1145/3799682.3839290",
"title": "IWILDS'26: The 6th International Workshop on Investigating Learning during Web Search and AI-Mediated Information Access",
"page": "9523-9525"
},
{
"doi": "10.1145/3799682.3839291",
"title": "KEIR @ CIKM 2026: The Third Workshop on Knowledge-Enhanced Information Retrieval",
"page": "9515-9518"
},
{
"doi": "10.1145/3799682.3839292",
"title": "The 2nd Workshop on LLM Agents for Social Simulation",
"page": "9560-9563"
},
{
"doi": "10.1145/3799682.3839293",
"title": "REACH-AI 2026: Community-Informed Evaluation and Benchmarking of Culturally Competent Generative AI",
"page": "9546-9549"
},
{
"doi": "10.1145/3799682.3839294",
"title": "GenAI4SG: Generative AI for Social Good and Vulnerable Populations",
"page": "9512-9514"
},
{
"doi": "10.1145/3799682.3839295",
"title": "Graph-Augmented LLMs (GaLM): Bridging Language and Structured Knowledge",
"page": "9557-9559"
},
{
"doi": "10.1145/3799682.3840216",
"title": "From Interaction Traces to Reusable Experiences for LLM Agents",
"page": "9319-9323"
},
{
"doi": "10.1145/3799682.3840217",
"title": "Towards Controllable, Stable, and Trustworthy Audio-Driven Talking-Head Generation",
"page": "9383-9387"
},
{
"doi": "10.1145/3799682.3840218",
"title": "Towards Faithful and Actionable Explanations for Historical Handwriting Identification",
"page": "9300-9304"
},
{
"doi": "10.1145/3799682.3840219",
"title": "Towards Semantic Integration of Heterogeneous Energy Data",
"page": "9349-9353"
},
{
"doi": "10.1145/3799682.3840220",
"title": "Towards Privacy-Preserving and Adaptive Distributed Learning for Cybersecurity Applications",
"page": "9339-9343"
},
{
"doi": "10.1145/3799682.3840221",
"title": "Towards Multi-Agent Conversational Proactive Information Retrieval via Reinforcement Learning",
"page": "9374-9378"
},
{
"doi": "10.1145/3799682.3840222",
"title": "Graph Mining and Network Reconstruction from Incomplete Information: From Missing Statistics to Fragmentary Evidence",
"page": "9314-9318"
},
{
"doi": "10.1145/3799682.3840223",
"title": "Studying and Improving Explanation Preservation in Distilled Recommender Systems",
"page": "9309-9313"
},
{
"doi": "10.1145/3799682.3840224",
"title": "Performance Prediction for LLM-Era Information Access",
"page": "9362-9364"
},
{
"doi": "10.1145/3799682.3840226",
"title": "Pride in Precision, Prejudice in Prediction: Bias and Stereotyping in Recommendation Systems",
"page": "9344-9348"
},
{
"doi": "10.1145/3799682.3840227",
"title": "From Unstructured Text to Executable Knowledge: Automated Transformation of High-Stakes Domain Documents",
"page": "9354-9357"
},
{
"doi": "10.1145/3799682.3840228",
"title": "ML Inference Acceleration via Scalable In-network Computing",
"page": "9324-9328"
},
{
"doi": "10.1145/3799682.3840229",
"title": "Neuro-symbolic Recommender Systems",
"page": "9334-9338"
},
{
"doi": "10.1145/3799682.3840230",
"title": "Sensing How Users Adapt to AI Decision Support: A Multimodal Eye and Mouse Tracking Study",
"page": "9379-9382"
},
{
"doi": "10.1145/3799682.3840231",
"title": "Unsupervised Clustering of Longitudinal Data to Identify Disease Progression Subtypes in Heterogeneous Diseases",
"page": "9305-9308"
},
{
"doi": "10.1145/3799682.3840232",
"title": "Knowledge-Driven Recommendations in Decentralized Systems: Integrating Federated Personal Data and Public Contexts",
"page": "9358-9361"
},
{
"doi": "10.1145/3799682.3840233",
"title": "Disentangling Capabilities and Boundaries in Tokenizer-Free Language Models",
"page": "9329-9333"
},
{
"doi": "10.1145/3799682.3840234",
"title": "Multi-Mode Interest Diffusion for Online Activity Tensor Forecasting",
"page": "9365-9369"
},
{
"doi": "10.1145/3799682.3840235",
"title": "A Knowledge Graph-Based Framework for Interoperable Supply-Chain Data Ecosystems",
"page": "9370-9373"
},
{
"doi": "10.1145/3799682.3840304",
"title": "Learn to Compress or Compress to Learn? Shannon and Turing meet again in the LLM era",
"page": "2-2"
},
{
"doi": "10.1145/3799682.3842658",
"title": "A Shared Task on Agentic AI in University Administration",
"page": "9564-9569"
},
{
"doi": "10.1145/3799682.3846821",
"title": "Query Explorer: Domain-Aware Hierarchical Query Suggestions for Exploratory Search",
"page": "9470-9472"
},
{
"doi": "10.1145/3799682.3846822",
"title": "AI and Feminist Worldmaking in Technofascist Times",
"page": "1-1"
},
{
"doi": "10.1145/3799682.3848139",
"title": "Reliable by Construction: Query Processing over Unstructured Data with Guarantees",
"page": "3-3"
}
],
"outside_volume": [],
"normalized_title_differences": [
{
"doi": "10.1145/3799682.3840095",
"id": "cap0528",
"registered": "OneMall: One Architecture, More Scenarios — End-to-End Generative Recommender Family for E-commerce",
"programme": "OneMall: One Architecture, More Scenarios — End-to-End Generative Recommender Family for E-commerce at Kuaishou"
},
{
"doi": "10.1145/3799682.3840614",
"id": "cfp0317",
"registered": "A Theoretical Framework and Efficient Algorithm for Computing Condensed Multi-Level Bisimulation Graph Summaries",
"programme": "Condensed Multi-Level Bisimulation Summaries"
},
{
"doi": "10.1145/3799682.3840724",
"id": "cfp0892",
"registered": "SemCoT: A Semantic Step-level Reasoning Optimization Approach for Collaborative Chain-of-Thoughts",
"programme": "SemCoT: A Semantic Step-level Reasoning Optimization Approach for Collaborative Chain-of-Thought"
},
{
"doi": "10.1145/3799682.3840773",
"id": "cfp1077",
"registered": "SAERec: Constructing Fine-grained Interpretable Intent Priors via Sparse Autoencoders for Recommendation",
"programme": "SAERec: Constructing Fine-grained Interpretable Intents Priors via Sparse Autoencoders for Recommendation"
}
],
"ordered_author_differences": [
{
"doi": "10.1145/3799682.3840031",
"id": "csp4144",
"registered": [
"Amin Bigdeli",
"Radin Hamidi Rad",
"Negar Arabzadeh",
"Sajad Ebrahimi",
"Hai Son Le",
"Charles L. A. Clarke",
"Ebrahim Bagheri"
],
"programme": [
"Amin Bigdeli",
"Radin Hamidi Rad",
"Negar Arabzadeh",
"Sajad Ebrahimi",
"Hai Son Le",
"Charles Clarke",
"Ebrahim Bagheri"
]
},
{
"doi": "10.1145/3799682.3840143",
"id": "cap2170",
"registered": [
"Yice Luo",
"Yun Zhu",
"Xi Chen",
"Yongchao Liu",
"Xintan Zeng",
"Chengying Huan",
"Kai Zhang",
"Jinrui Zhang",
"Juelu Zhang",
"Jiajun Zheng"
],
"programme": [
"Yice Luo",
"Yun Zhu",
"Chen Xi",
"Yongchao Liu",
"Xintan Zeng",
"Chengying Huan",
"Kai Zhang",
"Jinrui Zhang",
"Juelu Zhang",
"Jiajun Zheng"
]
},
{
"doi": "10.1145/3799682.3840158",
"id": "cap2931",
"registered": [
"Shwetha Salimath",
"Francesca Bugiotti",
"Sylvain Wlodarczyk",
"Sohaib Ouzineb"
],
"programme": [
"Shwetha Nagaraj Salimath",
"Francesca Bugiotti",
"Sylvain Wlodarczyk",
"Sohaib Ouzineb"
]
},
{
"doi": "10.1145/3799682.3840212",
"id": "crp4242",
"registered": [
"Hu YiRan",
"Rongxin Cheng",
"Haitao Li",
"Wenbo Lv",
"Xin Meng",
"Qingjing Chen",
"Huanghai Liu",
"Jing Xu",
"Fuheng Jiang",
"Xuanhang Wu",
"Xinyu Wang",
"Chong Wang",
"Weihang Su",
"Zibing Que",
"Yun Liu",
"Weixing Shen",
"Qingyao Ai",
"Yiqun Liu",
"Charles L.A. Clarke"
],
"programme": [
"Hu YiRan",
"Rongxin Cheng",
"Haitao Li",
"Wenbo Lv",
"Xin Meng",
"Qingjing Chen",
"Huanghai Liu",
"Jing Xu",
"Fuheng Jiang",
"Xuanhang Wu",
"Xinyu Wang",
"Chong Wang",
"Weihang Su",
"Zibing Que",
"Yun Liu",
"Weixing Shen",
"Qingyao Ai",
"Yiqun Liu",
"Charles Clarke"
]
},
{
"doi": "10.1145/3799682.3840737",
"id": "cfp0236",
"registered": [
"Yitong Zhou",
"Yucong Luo",
"Mingyue Cheng",
"Qi Liu",
"Jiahao Wang",
"Daoyu Wang",
"Enhong Chen"
],
"programme": [
"Yitong Zhou",
"Yucong Luo",
"Cheng Mingyue",
"Qi Liu",
"Jiahao Wang",
"Daoyu Wang",
"Enhong Chen"
]
},
{
"doi": "10.1145/3799682.3840789",
"id": "cfp1122",
"registered": [
"Koki Kawabata",
"Pedro Fidalgo",
"Mirela Cazzolato",
"Saranya Vijayakumar",
"Yasuko Matsubara",
"Yasushi Sakurai",
"Christos Faloutsos"
],
"programme": [
"Koki Kawabata",
"Pedro Fidalgo",
"Mirela Teixeira Cazzolato",
"Saranya Vijayakumar",
"Yasuko Matsubara",
"Yasushi Sakurai",
"Christos Faloutsos"
]
},
{
"doi": "10.1145/3799682.3840812",
"id": "cfp1288",
"registered": [
"Jian Wang",
"Wendong Zhang",
"Peng Lan",
"Zhifang Liao"
],
"programme": [
"Jian Wang",
"Wendong Zhang",
"Zhifang Liao",
"Peng Lan"
]
},
{
"doi": "10.1145/3799682.3840855",
"id": "cfp1488",
"registered": [
"Shiqi Fan",
"Shi Li",
"Xiangzhi Mao",
"Zeqing Zang",
"Hongyi Nie",
"Tong Li",
"Wen Hua"
],
"programme": [
"Shiqi Fan",
"Shi Li",
"Xiangzhi Mao",
"Zeqing Zhang",
"Hongyi Nie",
"Tong Li",
"Wen Hua"
]
},
{
"doi": "10.1145/3799682.3840858",
"id": "cfp1513",
"registered": [
"Rongxiao Huang",
"Hao Zhou",
"Jiaqi Zheng",
"Jiecheng Guo",
"Guobin Wu",
"Qun Li",
"Guihai Chen"
],
"programme": [
"Rongxiao Huang",
"Zhou Hao",
"Jiaqi Zheng",
"Jiecheng Guo",
"Guobin Wu",
"Qun Li",
"Guihai Chen"
]
},
{
"doi": "10.1145/3799682.3840978",
"id": "cfp2149",
"registered": [
"Henry Mukalazi Serugunda",
"Ouyang Jianquan",
"Jacob Katende"
],
"programme": [
"Serugunda Henry Mukalazi",
"Ouyang Jianquan",
"Jacob Katende"
]
},
{
"doi": "10.1145/3799682.3841124",
"id": "cfp2986",
"registered": [
"Qian Zhang",
"Lech Szymanski",
"Jeremiah D. Deng",
"Haibo Zhang"
],
"programme": [
"Qian Zhang",
"Lech Szymanski",
"Jeremiah Deng",
"Haibo Zhang"
]
}
],
"registry_with_abstract": 0,
"page_evidence": [
{
"file": "crossref-page-1.json",
"sha256": "16bed29f8ee82d668b124c369b632c5c1de0753bdbec7eb72ecb22ffdda5c710",
"records": 250,
"at": "2026-11-18T07:24:02.465647+00:00"
},
{
"file": "crossref-page-2.json",
"sha256": "54424ca31d99462394c9fde5fecfd7dd16337351d21ede3169f19af54d4f49c5",
"records": 250,
"at": "2026-11-18T07:24:38.062056+00:00"
},
{
"file": "crossref-page-3.json",
"sha256": "2dc31d89f41966614ed66a11d1ebaa4b2a6904dcf92945b7185e72e7e65382dd",
"records": 250,
"at": "2026-11-18T07:25:12.169451+00:00"
},
{
"file": "crossref-page-4.json",
"sha256": "fbbec2b52650dc99a0251c743f035a6fd95be263b04c55c12134cf8473bdde76",
"records": 250,
"at": "2026-11-18T07:25:46.159940+00:00"
},
{
"file": "crossref-page-5.json",
"sha256": "06151c5f770c1e37e8ea9dabe5210be9c969a14fa3bf9925a2d7730ced47d6e5",
"records": 81,
"at": "2026-11-18T07:25:53.938425+00:00"
}
],
"database_written": false,
"programme": {
"receipt": {
"url": "https://cikm2026.diag.uniroma1.it/cikm-program/program.html",
"final_url": "https://cikm2026.diag.uniroma1.it/cikm-program/program.html",
"bytes": 2689414,
"http_status": 200,
"sha256": "3eaa668b6aaf09a671d8a3e19a2889b9ab21afaa7ded013794afa94bf35281d6",
"at": "2026-11-18T07:16:19.758324+00:00"
},
"observations": 1008,
"program_ids": 1008,
"observations_by_tab": {
"tab-oral": 412,
"tab-poster": 550,
"tab-demo": 46
},
"ids_by_track": {
"FULL": 596,
"APPLIED": 86,
"RESOURCE": 46,
"SHORT": 234,
"DEMO": 46
},
"ids_with_abstract": 1008,
"ids_with_doi_link": 1008,
"different_title_ids": [],
"different_abstract_ids": [],
"different_link_ids": [],
"database_written": false
},
"second_programme": {
"second": {
"url": "https://cikm2026.diag.uniroma1.it/cikm-program/CIKM2026_program.html",
"final_url": "https://cikm2026.diag.uniroma1.it/cikm-program/CIKM2026_program.html",
"bytes": 2808252,
"http_status": 200,
"sha256": "e3630bd1247df713b8b618f1d3cd9709e7020b2524bb677f82a10f83badd1732",
"at": "2026-11-18T07:19:27.528387+00:00"
},
"shared_ids": 1005,
"id_difference": [
[
"cde3519",
"cde4012",
"crp3459"
],
[]
]
},
"crossref_parent": {
"doi": "10.1145/3799682",
"isbn": "9798400725395",
"published": "2026-11-06"
},
"pg_readonly": {
"snapshot": {
"at": "2026-11-18 07:25:36.282346+00",
"snapshot": "1103274:1103274:"
},
"existing_editions": [],
"sha256": "f64cc2307b9229e4962ee741f9b0a58ca16daeabc1dc97923c0c7efcc3d2f439",
"doi_bindings": 97,
"title_hits": 304,
"doi_and_title_agree": 95,
"doi_bound_without_title_hits": 2,
"imports": 0
},
"source_scope": "Crossref query pins publication date and validates DOI prefix and exact container. Counts are evidence for this snapshot. Registry-only records need type and scope review. No public negative-decision export was verified."
}
@@ -0,0 +1,9 @@
# CIKM 2026 节目与出版登记核对
官方节目列有 1,008 项带 DOI 的条目,逐篇均有摘要。五个标签的数量分别为 Full Research 596、Short Research 234、Applied Research 86、Resources 46、Demo 46。另一份官方节目快照少三个节目编号,其余 1,005 个编号重合。两份原始快照和响应哈希保存在本机的 `cikm2026-official-detailed-program-v1` 工作目录。[带 DOI 的节目](https://cikm2026.diag.uniroma1.it/cikm-program/program.html) · [另一份官方节目](https://cikm2026.diag.uniroma1.it/cikm-program/CIKM2026_program.html)
正式卷 DOI 为 `10.1145/3799682`,ISBN 为 `9798400725395`。ISBN 查询只返回卷级登记。按卷名及出版日期分页取得 1,081 条登记,逐条核对卷 DOI 前缀、完整容器标题及来源类型。1,008 个节目 DOI 全部匹配,另有 73 条逐篇登记只见于 Crossref。差集包含产业报告等条目,需按出版目录逐条确认类型。该次 Crossref 响应中摘要字段均缺失。[ACM 出版条目](https://doi.org/10.1145/3799682.3840615) · [Crossref 卷登记](https://api.crossref.org/works/10.1145/3799682)
按已知排版标记、Unicode 和标点变化比较,同 DOI 的条目有 4 处标题差异、11 处有序作者列表差异。差异条目保留两份原文,等后续核对后再决定如何关联。当前 PostgreSQL 快照中,97 个节目 DOI 已绑定论文,304 个节目标题有候选,其中 95 个 DOI 绑定与标题候选重合。仍需核查已有作品与历次投稿历史,当前尚未写入。
[逐条差集、字段差异及证据哈希](cikm2026-source-reconciliation-2026-11-18.json)。先将出版登记与节目摘要作为各自来源采集,再核查两者的作品关系和轨道。公开拒稿、撤回及桌拒的可得性继续单独核查。
@@ -0,0 +1,51 @@
# UAI 公开负面决定来源:2026-11-17 核查
**公开拒稿缺口仍未补齐,总体仍为 13/17。** 本页保存新来源的实际范围与恢复线索,避免下次把录用子集误当作全部投稿。主动补漏范围仍从2020年开始;本次镜像只含2022–2025,不能替代其他年份或其他会议的核查。
## 政策与访问边界
已有逐年核查记录保存在 [覆盖清单](inventory-2026-11-08.json) 的 UAI `policy_checks`:2023–2026 官方说明允许决定后公开录用及作者选择公开的拒稿,默认私密拒稿不属于公开可采集范围。政策并不证明任何镜像完整。
- [2023 投稿说明](https://www.auai.org/uai2023/submission_instructions)
- [2024 投稿说明](https://auai.org/uai2024/submission_instructions)
- [2025 投稿说明](https://www.auai.org/uai2025/submission_instructions)
- [2026 投稿说明](https://www.auai.org/uai2026/submission_instructions)
先前匿名 notes 请求遇到 ChallengeRequiredError/403,仍作为访问限制保留。本轮没有认证、重试或绕过这些入口;读取的是整理者公开发布的固定版本镜像。复制的 `readers`、`signatures` 和邀请字段仍只是镜像内的来源断言,不是对当前原始接口可见性或签名真实性的验证。
## 实际镜像范围
来源:[latkes/ai-research-index-iclr-openreview](https://huggingface.co/datasets/latkes/ai-research-index-iclr-openreview/tree/4c9e09931a7814d7500630ffd064b74f1c01bcf3),固定提交 `4c9e09931a7814d7500630ffd064b74f1c01bcf3`。目录中存在 `papers/uai-YYYY/papers.jsonl` 与四个 `raw/uaiYYYY.jsonl`,不能仅凭数据集说明宣称拒稿覆盖。
| 年份 | 整理版及原始论坛数 | 原始回复对象 | 实际回复角色 | 标准结构化决定字段 |
| --- | ---: | ---: | --- | --- |
| 2022 | 230 | 1059 | Official_Review 829,Decision 230 | Accept (Poster) 194,Accept (Oral) 36 |
| 2023 | 243 | 0 | 没有保存回复 | 没有保存决定记录 |
| 2024 | 200 | 1994 | Official_Review 905,Rebuttal 889,Meta_Review 200 | 没有单独 Decision 记录或结构化 decision 字段 |
| 2025 | 229 | 0 | 没有保存回复 | 没有保存决定记录 |
整理版902条均被整理者标为 `accepted=true`;原始根记录的 `venue` 为相应会议名称或 oral/poster/spotlight 文本,不能替代逐次决定证据,更不能算作902条已认证的原始录用决定。这不是四届全部投稿或全部录用的完整性证明。2024年还有 `Edit` 邀请,属于同一批回复上的另一个邀请,不能再累加成1994个新回复。
为排除字段漏读,对含回复的2022、2024快照另做了全部3053个回复的字段名、包装结构、邀请角色和短决定类字段核查;重新读取的正文哈希与首次捕获完全一致。2022的230条明确字段是录用标签。2024仅保存评审、作者回复和元评审,没有可据此新增的结构化拒稿/撤稿决定。没有从元评审文本、评分、推荐或缺少录用推断负面决定,也没有保存评审正文、作者回复正文或论文全文。
**这只描述上述固定镜像子集,不证明UAI没有公开拒稿。** 本批没有生成负面决定导入队列,也未写数据库。2020、2021、2026及其他可用来源仍需分别核查;某文件未出现在这个镜像目录中不代表全网不存在。
## 产物与恢复
服务器目录均相对 `/data/paper-index/collected/`:
- `uai2022-2025-mirror-metadata-v1`:902条整理版元数据投影、字段盘点与标签计数。
- `uai2022-2025-raw-decisions-v1`:902个原始论坛的元数据、明确决定字段及来源正文/逐行哈希。
- `uai2022-2024-reply-schema-v2`:3053个回复的元数据、字段名与短标签;评审正文不保存。角色核查包含直接记录和 `note` 包装,未发生静默遗漏。
源响应正文仅在内存中读取、校验,保存元数据投影与固定来源URL/字节哈希;不是保留了含评审文本的完整原始快照。复核原始源字节时应使用固定提交,并区分投影文件哈希与源正文哈希。恢复包已在本机独立核验全部902个论坛、3053个回复的元数据结构、逐文件哈希及来源关联。
| 恢复证据 | SHA256 |
| --- | --- |
| 整理版元数据恢复包 | 55889fb2b9005260cef1001b2cf36e203f3b7e1deeec0291f058b52eeae880e7 |
| 原始元数据/回复结构恢复包 | caade7df899dd3e65572d06e2bf2326fad692cc5e1c1e4ab002fdc6d7823743f |
| 原始结构独立核验报告 | 5f0ef2948d0db68fe0cd89eb4c8b57a1088b94b53bed98f180d2b96629746e72 |
服务器恢复包分别为 `/data/paper-index/validation/uai2022-2025-mirror-metadata-recovery-v1.tar.gz` 与 `uai2022-2025-raw-reply-recovery-v1.tar.gz`。本地同名备份位于 `/Users/zzyang/Documents/Codex/2026-11-06/paper-index/`。先检查实际单元、完整manifest和哈希,再恢复;不要因进度文件旧状态或日志安静重跑。
后续继续寻找能明确覆盖作者公开拒稿选择的来源,保留每次投稿及决定原文标签、原始ID与镜像关系,核对跨来源重叠和后续录用。UAI缺口不关闭,全部43个目标的逐年公开负面决定核查也不缩减;最终全库去重仍为独立必做阶段。
+29
View File
@@ -108,3 +108,32 @@ paper-atlas audit source-identities \
候选审计中的 `status=unknown` 仅为内部查询占位符,不代表原始提交的最终状态。节目ID以 `mm_program` 来源键查询,不转成全局 DOI、arXiv 或 OpenReview 标识;保留原始 `listed_in_official_program` 观察与全部展示位置。所有贡献类型、出版身份、最终决定核验标志均为false;缺作者的两项也保留并标记。
查询使用 PostgreSQL 只读一致性快照,候选作品的原始来源、标识、历次投稿和逐次决定压缩保存并记录哈希;截断会显式标记。标题/作者候选匹配不是作品身份批准,没有匹配也不证明是新论文。既往拒稿和后续录用不能互相覆盖,评分或评审推荐不能充当最终决定。输出目录必须不存在;遇到已有输出或静默运行,先检查具体单元和报告,不重启重放。
## 已完成的规范化、只读审计与出版核对
采集器提交 `c8db611` 和直接节目审计提交 `09a2d4e` 已通过PR #65、#66由用户合并。授权远程runner的535/543项测试、wheel构建安装和安装后全量命令检查通过;没有本机测试或构建。服务器规范化和只读候选报告也已独立验收,并在本机备份、复核全部2143项与814个候选证据包。
| 只读候选分类 | 来源节目项数 |
| --- | ---: |
| 标题与完整作者多重集合一致,仍需作品核查 | 660 |
| 仅部分作者重叠 | 153 |
| 无精确标题候选,不证明是新作品 | 1329 |
| 有标题候选但无作者一致性 | 1 |
全部4326个展示上下文一致,814个证据包均无截断。24篇候选作品涉及24次已有负面投稿、37条负面来源观察;37不能直接算作37次拒稿。所有候选均尚未批准身份或入库,旧负面决定必须在之后关联时保留。
对已有作品的原始arXiv记录、标题/完整作者多重集合和MM2026语境进一步筛查,得到67个 Main Track 出版DOI核对候选。独立查询其明确给出的DOI及父论文集,共68次单次请求,67次HTTP200、1次404,没有认证或重试。父注册记录为[第34届ACM Multimedia论文集](https://doi.org/10.1145/3767308),ISBN `9798400722134`,出版者ACM,出版年2026。
其中66项的原始arXiv引用DOI与注册DOI一致,按既有审计规则规范化的标题、完整作者多重集合、出版年、出版者、明确会议名和论文集标题也一致。注册文章未提供ISBN,不能伪称逐篇ISBN已核验;归属依据是文章明确给出的会议和论文集字段,以及独立的原始arXiv语境和DOI。仍需逐项作品/版本审查及新鲜数据库保护检查,**不是66项已入库或最终身份批准**。
`mfp7067` 对应既有候选作品2749947,原始arXiv提供的 `10.1145/3767308.3836169` 返回404,暂缓。404不是撤稿/拒稿证据,也不据此删除或修正既有标识;需独立核对后再处理。
完整会议出版名单仍需继续核对。实际ISBN查询只返回父论文集一个条目,父DOI的公开Unixref XML也只含该父DOI、0个 `conference_paper` 节点;都不能代表逐篇名单。按会议名检索得到的 `total-results` 是出版者全年范围内的宽泛检索量,不能当作MM篇数;必须逐项核对明确会议/论文集字段、节目与分轨归属,并保留未匹配项。全部28个来源track继续保留,不能把本批Main Track候选当成会议完整覆盖。
| 已独立验证的恢复证据 | SHA256 |
| --- | --- |
| 全轨规范化恢复包 | b119081618b8ce98037dc139be159512f9fdb86a0ccfb127c8b8a1c5722f3d87 |
| 只读候选/814证据包恢复包 | 42d0d0404bba04e64e638344b050ad4e7287c3e4c75a777b7069ecc7d7a6ed60 |
| 67项DOI及父论文集原始注册回执恢复包 | dd7a0ff76f12070f403cc7352ac54269e454758d807896a086c58c1c1ac7b798 |
服务器恢复包位于 `/data/paper-index/validation/`,分别名为 `mm2026-native-recovery-v1.tar.gz`、`mm2026-preflight-recovery-v1.tar.gz`、`mm2026-publication-metadata-recovery-v1.tar.gz`;本地同名备份在 `/Users/zzyang/Documents/Codex/2026-11-06/paper-index/`。来源目录为 `collected/mm2026-native-v1`、`mm2026-preflight-v1`、`mm2026-publication-metadata-v1`。先核对完整manifest及实际单元结果,不能用旧进度文件的running字样重启。其他会议已消费的导入队列仍禁止重放;本轮均为来源核查、只读候选审计,没有生产写入。