Reliable Context

Tags

Browse articles by topic (134)

17a-4 audit trail alternative 17a-4 worm 21 cfr 11.10 21 cfr part 11 MaDI-Bench abstention abstention testing acceptance testing access control agent evaluation agent evidence verification agent long-term memory provenance agent memory agent memory admission control agent memory evaluation agent memory write validation agent trace logging ai act 2 august 2026 ai act annex iv technical documentation ai act application dates ai act article 19 logs retention ai act article 50 ai act automatic logs high-risk ai act chatbot disclosure ai act data governance ai act data quality ai act high-risk deadline ai act logging requirements ai act record keeping ai act training data requirements ai act transparency obligations ai act vs iso 42001 ai act watermarking ai agent audit log ai agent evaluation ai agent handoff testing ai agent logging best practices ai agent memory admission policy ai agent memory deletion verification ai agent memory update ai answer attribution ai citation laundering ai citation poisoning ai context lineage ai credibility framework fda ai factual accuracy ai generated content labeling ai generated evidence admissibility ai governance evidence ai governance frameworks comparison ai hallucinated case citations ai in pharmacovigilance regulation ai management system ai rmf crosswalk ai rmf data provenance ai rmf govern map measure manage ai tool output verification alce alcoa alcoa data integrity examples alcoa plus alcoa principles alcoa+ data integrity algorithmic fairness anchor citation to text passage answer leakage answer provenance answer reuse validity answer traceability append only log integrity as-of queries as-of reconstruction attribution evaluation audit trail audit trail annex 11 audit trail for data values audit trail review audit trail review procedure audit trails auditable ai answers authenticating electronic records authoritative definitions automated fact checking average proximity b-cubed basel risk data aggregation bcbs 239 bcbs 239 compliance data lineage bcbs 239 principles belief revision benchmark audit benchmark contamination benchmark datasets benchmark design benchmark evaluation benchmark metrics beneficial ownership data best table format for llm prompts bitemporal data bitemporal sql blocking audit blocking recall broker dealer electronic records c-rag c2pa c2pa manifest c2pa specification cache invalidation for ai answers cached answer validity test calibration candidate omission canonical record catd cdd rule beneficial owner certificate transparency merkle tree chain of custody chain of custody digital evidence chain of custody form change data capture vs event sourcing chart claim refutation chatgpt gdpr accuracy chunk overlap citation accuracy citation completeness citation correctness citation evaluation citation evidence citation faithfulness citation grounding citation verification claim specificity claim verification claim-level reliability claim-level validation claimdb closed world assumption cluster evaluation cluster repair cluster validation codex competency questions ontology completeness statements comptruthhyp computer software assurance computer system validation concept drift concrete cross-lingual retrieval confabulation ai risk confidence calibration confidence thresholds confidence-aware truth discovery conflicting data sources conflicting numerical claims conflicting sources fact checking conformal language modeling conformal prediction conformal rag guarantees conformal risk control connected components clustering constraint preservation constraint-based ai evaluation constraint-based reasoning content authenticity initiative content credentials content drift context engineering context engineering for ai agents context engineering vs prompt engineering context of use ai model fda context versioning context window management agents contextual retrieval controlled vocabularies conversational query rewriting copy-aware truth discovery correlated sources counterfactual evaluation crh cross bundle provenance cross-lingual claim verification cross-lingual fact checking cross-modal claim verification crowdsourcing crypto shredding gdpr csv vs csa csv vs json token cost dart data contract example data contract ontology data contract vs schema data contracts data extraction error analysis data fusion data fusion benchmark data fusion error propagation benchmark data governance data integration data integrity data integrity guidance fda vs mhra data lineage data lineage requirements data lineage standard data lineage vs data provenance data lineage vs provenance data normalization data provenance data provenance and lineage data provenance definition data provenance privacy data provenance requirements data provenance standard data provenance standards data provenance tracking data provenance validation json data provenance vs data lineage data quality data quality assessment data source evaluation data source reliability data traceability data verification vs validation database claim verification database records as evidence dbt model contracts dcat deduplication record matching explained delta lake time travel delta lake vacuum retention denial constraints depen dependency tracking deterministic transformation baseline difference between audit trail and data lineage digital evidence chain of custody digital forensics preservation digital omnibus ai act delay digital signatures document order dqv dscsa traceability dynamic truth discovery dynatd e-sens edpb chatgpt taskforce accuracy eiopa ai opinion eiopa artificial intelligence electronic audit trail fda ema reflection paper ai empirical bayes enterprise knowledge graph model enterprise rag evaluation entity linking entity matching entity resolution entity resolution acceptance testing entity resolution benchmark entity resolution blocking entity resolution blocking recall entity resolution bridge records entity resolution candidate generation entity resolution evaluation metrics entity resolution false merges entity resolution human review entity resolution pipeline entity resolution techniques epcis events what when where why epistemic status error analysis error propagation in data pipelines eu ai act article 10 eu ai act article 12 eu ai act timeline eu sanctions list eudr due diligence statement evaluation metrics event sourcing event sourcing audit trail event sourcing explained event sourcing vs audit log event time evidence conflict fact checking evidence disclosure evidence grounding evidence handoff evidence loss evidence provenance evidence retrieval evidence review evidence sets evidence span citation evidence sufficiency evidence traceability evidence validity evidence-conditioned evaluation exactmatch exception-aware evaluation excluir explainable fact checking explicit negation extraction errors fact adjudication fact checking fact verification fact verification audit fact verification benchmark fact verification dataset factkg fair aggregation faircrowd fairness tests for truth discovery aggregation fairtd faitcrowd false memory resistance false merges false reuse rate false splits fatf beneficial ownership guidance fault injection fda ai guidance drug development fda csa guidance fda electronic records electronic signatures federated query completeness fellegi sunter model fellegi-sunter fitness for purpose forecasting fre 803(6) business records gamp 5 gcp electronic source data gdpr accuracy principle gdpr append only database gdpr article 16 rectification gdpr article 5(1)(d)gdpr event sourcing gdpr right to erasure immutable genai risk management profile generalized merge distance generative ai model risk banks gleif golden copy golden record golden record survivorship good clinical practice data integrity gql iso 39075 graph evidence graph provenance graphrag provenance grounded ai answers groundedness groundedness calibration gs1 epcis gxp audit trail requirements gxp data integrity regulations compared hallucinated edges hallucination in large language models halumem handoff reliability hierarchical data fusion hierarchical truth discovery historical graph replay historical source verification how provenance how to evaluate llm citations how to reduce llm hallucinations html vs markdown table http resolution human evaluation human in the loop hyde iceberg snapshots time travel ich e6 r3 ich e6 r3 changes ich e6(r3) data governance identifier governance identity decisions incremental reasoning independent holdouts independent recovery indirect prompt injection inferwiki information fusion information integrity genai nist information produced by the entity insurance ai governance insurers ai record keeping integrated bayesian model inverse relation leakage ipe audit evidence iso 27037 iso 42001 iso 42001 certification iso 42001 requirements iso 42001 vs nist ai rmf iso/iec 42001 json evidence binding test json pointer rfc 6901 json-ld json-ld context changes json-ld validation judge reliability knowledge base construction knowledge engineering knowledge fusion knowledge governance knowledge graph knowledge graph access control knowledge graph assertion retraction knowledge graph completeness knowledge graph completion knowledge graph completion evaluation protocol knowledge graph edge citation verification knowledge graph embedding predictions knowledge graph fact checking knowledge graph fact verification knowledge graph governance knowledge graph identifier versioning knowledge graph provenance knowledge graph provenance wikidata knowledge graph quality knowledge graph quality assessment knowledge graph rdf or lpg knowledge graph reasoning knowledge graph reproducibility knowledge graph truth maintenance knowledge graphs knowledge provenance knowledge representation knowledge update knowledge-based trust late chunking latent truth inference latent truth model lawyers chatgpt fake cases legal ai hallucination rate legal entity identifier legal entity identifier parent legal research ai reliability lei level 2 data lei parent relationship link prediction link prediction validation llm abstention llm agent memory poisoning llm agent observability llm as a judge llm citation evaluation llm conflicting sources llm context compression llm context provenance llm data provenance llm evaluation llm evidence position bias llm hallucination causes llm hallucination detection llm hallucination gdpr personal data llm json record misplacement llm judge groundedness evaluation llm knowledge graph extraction llm source conflict resolution llm table understanding benchmark local differential privacy long context evaluation long context llm robustness long-tail evidence longmemeval lost in the middle m and u probabilities machine unlearning madi-bench mapping quality mariadb system versioned tables markdown vs json vs csv master data master data best value master data management master data management examples mata v avianca mcp authorization oauth mcp hosts clients servers mcp prompt injection mcp security mcp security best practices mcp server mcp tools resources prompts mdm for compliance mdm survivorship rules examples mdm vs data governance mean reciprocal rank measurement uncertainty memory consistency merge or relate organization records merkle tree audit log mhra gxp data integrity guidance migration notice minimum data provenance metadata missed matches missing evidence detection model context protocol explained model context protocol vulnerabilities model risk model risk management llm multi agent evaluation multi truth discovery multi-hop rag citation completeness multi-hop reasoning multi-source data fusion multilingual evidence retrieval multilingual fact checking multimodal claim verification multiple sources of truth named graphs negation in information retrieval negative constraints negative knowledge graph negative sampling nevir nist ai 600-1 nist ai risk management framework nist ai rmf nist ai rmf playbook nist generative ai profile no-truth questions numerical data fusion obo foundry observation time odcs 3.2 odrl ofac 50 percent rule omission test ontology ontology alignment ontology change impact ontology crosswalk acceptance test ontology dependency management ontology engineering ontology governance ontology mapping validation ontology modeling ontology provenance ontology release beyond shacl conformance ontology requirements ontology term deprecation ontology term replacement ontology testing ontology versioning open data contract standard open source data lineage open world assumption open-world evaluation openlineage openlineage facets openlineage tutorial opentelemetry generative ai oracle substitution organization identity resolution original evidence osfi e-23 owl owl 2 owl deprecated owl import closure owl imports owl reasoning owl vs shacl paired bootstrap paired evaluation pairs completeness pairwise f1 parent child consistency part 11 audit trail requirements part 11 compliance partial pooling pcaob as 1105 permission-aware rag evaluation permission-aware retrieval persistent identifiers persistent uris pic/s data integrity point in time evaluation point in time verification preserve evidence when a cited source changes privacy preserving truth discovery privacy utility tradeoffs probabilistic knowledge bases probabilistic record linkage prov bundle prov data model explained prov-links prov-o prov-o ontology provenance provenance access control provenance bundle provenance conformance provenance evaluation provenance metadata model provenance redaction provenance semirings publication time quality scorecard query completeness certificate query decomposition query drift query provenance query rewriting in rag query transformation rag answer consistency rag chunk boundary effects rag chunking evaluation rag chunking sensitivity rag citation provenance rag data poisoning rag debugging rag evaluation rag evaluation dataset leakage audit rag failure modes rag fault isolation rag groundedness rag grounding verification rag query rewriting ragas ragchecker rdf rdf 1.2 rdf 1.2 provenance rdf canonicalization rdf dataset verification rdf graph signature rdf provenance rdf reasoning rdf reification rdf vs neo4j rdf vs property graph rdf-star record linkage record linkage blocking record linkage steps reduction ratio reifiers relation extraction evaluation release gates remote contexts reproduce past table state reproducibility reproducible llm answers reproducible reasoning requirements traceability reranking retained utility retrieval augmented generation retrieval evaluation retrieval metrics retrieval sensitivity reuse decision ledger risk data aggregation and reporting robust truth discovery row-level provenance rule 902(13) 902(14)safer sanctions entity resolution sanctions name matching sanctions screening false positives sciclaimeval scientific claim verification scientific figure claim verification scifact sciver sec 17a-4(f) requirements sec rule 17a-4 secure aggregation selective prediction semantic cache correctness semantic caching semantic change semantic compatibility semantic crosswalk semantic data contract semantic disclosure semantic drift semantic entropy semantic entropy hallucination detection semantic entropy limitations semantic entropy probes semantic interoperability semantic search sensitive data provenance sentence level citation shacl shacl validation signed rdf export silent tool failures skos skos concept vs owl class solvency ii ai data quality source admission source agreement source attribution source attribution for data values source credibility source dependence source errors source independence source independence provenance source integrity source provenance source reliability source reliability by topic source reliability estimation source selection source selection in data fusion source span source versioning sox 404 data controls sox it general controls reports sparql sparql 1.2 sparql federated query sparql regression testing sparse records entity matching splink fellegi sunter sql 2011 temporal tables sql witness sets sr 11-7 sr 26-2 model risk sssom stale data risk streaming truth discovery streamtf structured data verification structured output evaluation supply chain traceability data survivorship rules sybil attacks sybil attacks in truth discovery synthetic benchmarks system of record system of record for ai agent memory system versioned temporal tables system versioning table formatting for rag table serialization llm tabverse tamper evident logging tdssa tem temporal database temporal databases temporal fact checking temporal fact verification temporal knowledge graph benchmark data leakage temporal knowledge graph evaluation temporal knowledge graphs temporal leakage temporal link prediction temporal memory temporal provenance temporal tables sql server text fragment selector text-to-sql evaluation time travel vs audit trail tool calling evaluation topic-specific source reliability train test split transaction time transitive closure entity resolution translation errors in fact checking transparency log triple terms truth discovery truth discovery accuracy truth discovery algorithm comparison truth discovery algorithms truth discovery benchmark truth discovery benchmarks truth discovery evaluation truth discovery from conflicting sources truth discovery with hierarchical claims truth discovery without ground truth truth existence truth maintenance systems truthfinder ubo register entity resolution ultimate beneficial owner identification uncertainty quantification llm uncertainty-aware data integration unknown triples unknown vs false unsupported claims detection upstream data errors valid time valid time vs transaction time value path binding value-level provenance variation of information verifiable computation version iri versioned records w3c prov w3c prov for llm context provenance wasderivedfrom wasgeneratedby web annotation textquoteselector web archiving web-scale knowledge extraction what is context engineering what is data provenance what is mcp what is mdm where provenance who data integrity guideline who owns whom lei why do llms hallucinate why provenance wikibase data model wikidata qualifiers wikidata ranks preferred deprecated wikidata statements references x-fact benchmark xfever