Rapid Recall Key 39–76
- Logical taxonomy/model: common enterprise semantic frame helps recognize equivalent concepts across differently named/structured sources.
- Remediation vs transformation: fix defect/root cause vs intentionally convert/derive/restructure valid data for target/business rule.
- Optimistic vs pessimistic: provisional member + later reconciliation vs recycle/reject until dependency valid + reload.
- Population factors: source capability/volume, history, CDC, ordering, DQ/exceptions, latency, target design, performance windows, audit/restart/recovery.
- CDC contribution: identifies inserts/updates/deletes efficiently so recurring loads need not rebuild all data.
- Group users: executives, staff, analysts/power users need different interaction, latency, exploration and support; one tool does not fit all.
- Tactical vs strategic BI: short-term operational/management decisions vs long-term trends/goals/planning/performance.
- Self-service vs unmanaged: authorized user exploration on governed trustworthy data/semantics/Metadata/security/support vs uncontrolled local copies/definitions/access.
- Operational analytics: analytical information applied to current/near-current operational decisions, often lower latency.
- OLAP: interactive multidimensional slicing/dicing/drilling/comparison of measures across dimensions.
- Release plan: coordinates incremental value, production deployment, dependencies/readiness/maintenance and data+BI changes.
- Release vs iteration: coordinated business-facing deployment/package vs development cycle/increment.
- Pilot vs production: exploratory/limited environment vs supported live environment requiring business+IT readiness, governance/security, monitoring and quality.
- Load monitoring: start/end/duration, completeness/counts, rejects/recycles/errors, dependencies, restart, freshness, DQ, service-window result.
- Usage-driven tuning: optimize high-frequency/high-value objects; retire unused ones rather than tune everything equally.
- Transparent status: consumers/support can verify freshness, failures, delays and completion instead of guessing.
- Dictionary/glossary: business meaning plus relevant structure/types/definitions/security/context.
- Lineage vs impact: origin/movement/transformation path vs what a proposed change would affect.
- Model synchronization: prevents implemented structures from drifting from governed meaning and improves Metadata/lineage/maintenance/impact.
- Integration tools beyond transform: audit/control, orchestration/scheduling, restart/recovery, dependencies, monitoring, rejects/errors, Metadata/lineage capture, operational management.
- BI families: operational reporting, query/reporting, tactical/strategic BI, dashboards/visualization, self-service, operational analytics, BPM, OLAP/advanced analytics.
- Prototype: let users react to representative data, clarify uncertain requirements and expose source/DQ/feasibility problems early.
- Queryable audit: verify arrival/load/reject/transformation/freshness/quality evidence and troubleshoot at useful grain.
- Readiness: sponsorship, source/DQ, skills/staffing, architecture/capacity, security/privacy/legal, Governance, Metadata/lineage, support, performance/availability, service expectations.
- Release roadmap: connects incremental deliveries to enterprise target and avoids big-bang paralysis or disconnected local solutions.
- Configuration management: version/trace/transport models, mappings, code, semantic definitions, config, tests and artifacts consistently across environments.
- Five readiness/change factors: business sponsor/SMEs; source/DQ readiness; skills/capacity; architecture/platform/production support; security/privacy/legal/governance. Other source-aligned factors also count.
- DW governance: decision rights/controls for sources, boundaries, DQ, Metadata/lineage, access/security/privacy, retention, releases, self-service/discovery, risk, service expectations, exceptions—without arbitrary blockage.
- Business acceptance: understandable data + verifiable DQ + demonstrable lineage + business validation/sign-off.
- UAT: business sign-off and structured BI-vs-source comparisons across initial load and several update cycles, including quality/lineage expectations.
- Trust-supporting elements: Conceptual Data Model/meaning, DQ feedback/remediation loop, end-to-end Metadata/context, traceability/lineage plus business validation; core test = understandability + quality + lineage.
- SLAs: agreed service expectations such as availability, delivery/load timing, freshness/latency, response/support, recovery/retention as appropriate.
- Reporting strategy: users/access/security, report/analysis types, frequency, distribution, retention/storage, visualization, tool fit, support, timeliness/performance trade-offs.
- CoE: training, reusable patterns/templates, source/tool guidance, expertise/support, governance/standards communication, coordination.
- Usage metrics: actual connected/concurrent users, query/report activity/frequency/trends—not merely licenses.
- Coverage metrics: breadth/penetration of subjects/sources/departments represented/consuming the analytical environment.
- Response/performance: load duration/success, refresh timeliness, query response, extracts completed, service-window/SLA behavior.
- Final model: business need → history/grain/latency → source/DQ reality → architecture → integrate/store/present → BI delivery → Metadata/lineage/Governance/Security → product operations/releases → usage/coverage/performance/satisfaction → improvement.
Source range: pp. 375–393.