核心采集门禁当前没有阻断项。
采集监控
系统是否在产出可读情报
监控页优先回答三件事:来源是否稳定、队列是否在推进、最新文档是否已经具备中文阅读价值。底部诊断抽屉保留完整队列、适配器、去重和合并细节。
展示来源32
S 级来源14
原始来源14
监控样本32
预览可导入3
跳过重复0
采集适配器4
去重簇3
合并建议3
生产源池100
生产任务100
生产就绪production_scheduler
运行任务5077
异常来源12
可重试0
持久化状态Dry
落库文档32
落库建议3
今日监控重点
先看阻断,再看吞吐
默认视图只放需要判断的信号;完整运维明细在下方诊断抽屉中展开。
Google Cloud Blog:检查 endpoint、限频或授权状态。
这里统计当前监控样本,不代表全库翻译率;用于判断最新队列是否适合直接阅读。
生产采集系统
registry100
queues8
leased0
live fetchers100
stable100
watch0
normalized0
dead letters0
metadata only11
POST /v1/ingestion/production/run-cycle 为手动动作,监控页不会自动租约、抓取或写入。后台 worker 负责持续采集;该页只读取队列和健康状态。
生产门禁
pass100 个专业来源注册100 sources registered.
pass生产调度计划100 jobs planned across 8 queues.
passworker 执行闭环25 succeeded, 0 dead letters.
pass标准化、去重、候选事件25 normalized docs, 25 candidates.
pass授权和内容边界11 metadata-only sources, 0 violations.
生产调度策略
scheduler优先调度 S 级、监管、临床、公司官方和高频媒体来源。
scheduler每个任务带 lease_ttl、idempotency_key、queue_name 和 rate_limit_policy。
scheduler授权源进入 metadata-only 队列,授权确认前不抓取付费全文。
scheduler失败任务按指数退避重试,超过最大次数进入 dead-letter。
调度预览
计划任务4
活跃来源4
预计文档7
预计去重簇3
预计合并建议3
阻断来源0
先运行官方 API 和授权 metadata 适配器。完成文档标准化后进入去重簇。去重簇生成事件合并建议,再进入审核与版本发布。
展开持久化运行明细
持久化运行
adapters4
jobs10147
health12
documents32
clusters3
merge3
dry-runPOST /v1/ingestion/persist-run 为手动动作,监控页只展示当前状态。
展开来源库与文档队列样本
来源库
A
AAAI Digital Libraryofficial_dataset · en
参考来源AACL Anthologyofficial_dataset · en
参考来源AACM CHIresearch_literature · en
参考来源AACM CSCW / PACMHCIresearch_literature · en
参考来源AACM Digital Libraryresearch_database · en
参考来源AACM Digital Libraryresearch_literature · en
参考来源AACM ICMIresearch_literature · en
参考来源AACM/IEEE HRIresearch_literature · en
参考来源AACM UISTresearch_literature · en
参考来源SAMD Newsroomprofessional_media · en
原始来源AAmerican Journal of Psychiatryresearch_literature · en
参考来源SAnthropic Newsprofessional_media · en
原始来源AAPA PsycNetofficial_dataset · en
参考来源SArm Newsroomprofessional_media · en
原始来源AarXivprofessional_source · en
参考来源SASML Newsprofessional_media · en
原始来源AAssociation for Psychological Scienceprofessional_source · en
参考来源SAWS Machine Learning Blogofficial_company_source · en
原始来源文档队列
medical literature metadata
Ovariocytotoxic serum enhances humoral innate immunity and reproductive performance in Kazakh Merino ewes: A controlled field study.PubMed Central · 待翻译 · en
3cc6f62500medical literature metadataTRAQ in clinical practice: A narrative review for health care transition readiness.PubMed Central · 待翻译 · en
1d563fe598official company sourceAmazon SageMaker Inference: 2026 year-to-date launches in reviewAWS Machine Learning Blog · 待翻译 · en
299ae5aa46medical literature metadataBreed- and age-associated variation in hematological parameters and Toll-like receptor 2 gene expression in peripheral blood mononuclear cells of commercial and purebred laying hens during the laying cycle.PubMed Central · 待翻译 · en
37dd012511official company sourceIntroducing Kimi K3 on Amazon BedrockAWS Machine Learning Blog · 待翻译 · en
25f2596ef4medical literature metadataLabel-free ultraviolet photoacoustic microscopy for multiscale structural phenotyping of epithelial morphogenesis in organoid-derived intestinal epithelial wrinkle tissue.PubMed Central · 待翻译 · en
cc7b280c22official company sourceHow BMW Group detects cost anomalies across 14,000 cloud accountsAWS Machine Learning Blog · 待翻译 · en
f8a011d069medical literature metadataArtificial intelligence-guided nanozyme engineering for chronic wound healing: from rational design to precision therapeutics.PubMed · 待翻译 · en
126631f56amedical literature metadataOrganoid Intelligent Morphomics: Decoding the organoid morphome through artificial intelligence from phenotypic quantification to mechanistic insight.PubMed · 待翻译 · en
31fcae4c94medical literature metadataBeyond myocardial infarction: Point-of-care cardiac troponin biosensors for heart failure prognosis and remote monitoring.PubMed · 待翻译 · en
36eb19d495medical literature metadataMachine learning prediction for epilepsy treatment selection and prognosis: achievements and challenges.PubMed · 待翻译 · en
a87418783aresearch literatureSex-specific biological aging clocks across organs and omicsNature Medicine · 待翻译 · en
ede22c9797medical literature metadataSelective Kv1.3 inhibition promotes M2 microglial polarization, accelerates hematoma resolution, and improves neurological recovery following collagenase-induced spontaneous intracerebral hemorrhage in mice.PubMed Central · 待翻译 · en
8b62ea1db6medical literature metadataEffectiveness of underwater treadmill rehabilitation following femoral head and neck ostectomy in dogs: A prospective controlled pilot study.PubMed Central · 待翻译 · en
1b50a8d4e4official company sourceFrom Wafer-Out to First Token: Codifying Supply Chain Expertise with Nemotron and Palantir FoundryNVIDIA Developer Blog · 待翻译 · en
258039c8a1research literatureHow genes shape personality and experienceNature Medicine · 待翻译 · en
d3db97ffdaresearch literatureLevetiracetam therapeutically targets GABAergic synapses in diffuse midline gliomaNature Medicine · 待翻译 · en
b6673cb3f5official company sourceReducing medical claims review time with AI on AWS: The EXL Medical IDP solutionAWS Machine Learning Blog · 待翻译 · en
c2785ddb08展开任务队列与来源健康
任务队列
queueddelta_syncattempt 1/3
ACM Digital Librarynext 2026/6/13 04:01:50succeededdelta_syncattempt 1/3
OpenAI Blognext 2026/6/6 05:01:50queueddelta_syncattempt 1/3
ClinicalTrials.govnext 2026/6/7 04:01:50runningdelta_syncattempt 1/3
SEC EDGARnext 2026/6/7 04:01:50queueddelta_syncattempt 1/3
ACM Digital Librarynext 2026/6/13 01:26:01succeededdelta_syncattempt 1/3
OpenAI Blognext 2026/6/6 02:26:01queueddelta_syncattempt 1/3
ClinicalTrials.govnext 2026/6/7 01:26:01runningdelta_syncattempt 1/3
SEC EDGARnext 2026/6/7 01:26:01queueddelta_syncattempt 1/3
ACM Digital Librarynext 2026/6/13 01:25:13succeededdelta_syncattempt 1/3
OpenAI Blognext 2026/6/6 02:25:13来源同步健康
healthylag 0mfail 0
SEC EDGARcadence_based_delta_sync · 无需人工处理。healthylag 0mfail 0
SEC Press Releasescadence_based_delta_sync · 无需人工处理。healthylag 0mfail 0
NVIDIA Newsroomcadence_based_delta_sync · 无需人工处理。healthylag 0mfail 0
OpenAI Newscadence_based_delta_sync · 无需人工处理。healthylag 0mfail 0
Microsoft Research Blogcadence_based_delta_sync · 无需人工处理。healthylag 0mfail 0
AWS News Blogcadence_based_delta_sync · 无需人工处理。healthylag 0mfail 0
AWS Machine Learning Blogcadence_based_delta_sync · 无需人工处理。healthylag 0mfail 0
Microsoft Azure Blogcadence_based_delta_sync · 无需人工处理。healthylag 0mfail 0
Google DeepMind Blogcadence_based_delta_sync · 无需人工处理。degradedlag 60mfail 1
Google Cloud Blogexponential_backoff_until_retry · 检查 endpoint、限频或授权状态。展开真实来源适配器
真实来源适配器
Sofficial_apibusiness_day_polling
SEC EDGARrespect_sec_fair_access_policy_and_user_agent · max_10_requests_per_second_with_backofftechnology_financelife_sciences
Sofficial_apidaily_polling
ClinicalTrials.govuse_public_api_and_record_query_params · polite_daily_delta_synclife_sciences
Srss_or_sitemaphourly_polling
OpenAI Blogrespect_robots_and_use_public_pages_only · low_frequency_public_page_checktechnology_financehuman_intelligence
Ametadata_searchweekly_metadata_review
ACM Digital Librarymetadata_only_unless_authorized · manual_or_authorized_low_frequency_synchuman_intelligence
展开去重与事件合并队列
去重预览
doc-cluster-1 · 3 文档 · 2 重复云厂商上调 AI 基础设施资本开支指引SEC EDGAR · create_or_update_event_candidate
doc-cluster-2 · 2 文档 · 1 重复AI 制药公司报告肿瘤 II 期中期数据ClinicalTrials.gov · create_or_update_event_candidate
doc-cluster-3 · 2 文档 · 1 重复AI 陪伴应用新增长期记忆和语音情绪识别ACM Digital Library · create_or_update_event_candidate
事件合并队列
merge_into_existing_event · 3 文档AI 基础设施资本开支出现新变化AI 基础设施资本开支出现新变化 · 候选事件与现有事件高度相似,建议合并为同一事件。
merge_into_existing_event · 2 文档AI 制药临床管线出现新进展AI 制药临床管线出现新进展 · 候选事件与现有事件高度相似,建议合并为同一事件。
merge_into_existing_event · 2 文档AI 陪伴产品能力和风险出现新变化AI 陪伴产品能力和风险出现新变化 · 候选事件与现有事件高度相似,建议合并为同一事件。
展开失败重试策略
失败重试预览
retry_policyPOST /v1/ingestion/jobs/retry-failed 为手动动作,监控页不会自动触发重试。
retry_policy失败任务只在人工确认来源、频率限制和错误类型后重试。