DABCC flashcards

32 cards. Read the term and the definition here. The deck at the top of this page is optional.

Data Governance

The framework of decision rights, policies, and accountability that determines who may make decisions about data assets and how those decisions are enforced across the organization.

Data Steward

A role that operationalizes governance within a domain by maintaining definitions, quality rules, and issue logs, and by escalating unresolved problems to the data owner.

Data Owner

The accountable role for a data domain who approves access, sets quality targets, and accepts risk; accountability cannot be delegated to the steward.

Critical Data Element

A data element whose quality materially affects regulatory compliance, financial reporting, or strategy, and which therefore receives the strictest controls and monitoring thresholds.

RACI Matrix

A responsibility assignment grid mapping activities to Responsible, Accountable, Consulted, and Informed parties, with exactly one Accountable per activity.

Conceptual Data Model

The highest-level model capturing business entities and their relationships in business vocabulary, independent of attributes, keys, or any database technology.

Third Normal Form

A table state in which every non-key attribute depends on the whole key and on nothing but the key, eliminating transitive dependencies and update anomalies.

Star Schema

A dimensional design with a central fact table joined to denormalized dimension tables, optimizing query simplicity and read performance over storage efficiency.

Fact Table Grain

The declared meaning of a single fact row, such as one order line per day, which determines which measures are additive and which queries are valid.

Slowly Changing Dimension Type 2

A dimension design that preserves history by inserting a new row with a new surrogate key, version number, and effective dates whenever tracked attributes change.

Surrogate Key

A system-generated identifier assigned to a dimension or entity row, decoupled from business keys so history remains stable when source keys are reused.

Change Data Capture

A technique that identifies rows changed in a source since the last load, commonly by reading transaction logs, triggers, or timestamp columns.

Idempotent Load

A load process that produces the same target state whether run once or many times, typically implemented with upserts on natural keys or partition replacement.

ELT Pattern

An integration approach that loads raw source data into the target platform first and performs transformations there, exploiting warehouse compute and retaining raw history.

Late-Arriving Dimension

A dimension record referenced by a fact before the dimension row exists, handled by inserting a placeholder row later updated or by back-dating a Type 2 version.

Load Reconciliation

A post-load control comparing source and target totals such as row counts, summed amounts, or hash totals to confirm the load is complete and correct.

Data Completeness

The proportion of records with a populated value for a required attribute, calculated as non-null records divided by total records for that column.

Data Validity

The degree to which values conform to defined formats, ranges, or allowed value sets, independent of whether the values are factually correct.

Data Accuracy

The degree to which a stored value matches the verified real-world or authoritative reference value, measured by comparison against a trusted source.

Data Profiling

An analysis producing null counts, distinct counts, value ranges, frequency distributions, and pattern discovery to reveal the actual condition of a dataset.

Fuzzy Matching

A duplicate-detection technique that scores string similarity using edit distance, phonetic codes, or token comparison rather than requiring exact equality.

Data Classification

The assignment of data to sensitivity tiers such as public, internal, confidential, and restricted, which determines the encryption, access, and retention controls applied.

Tokenization

A protection method replacing sensitive values with non-reversible tokens stored alongside a secure vault mapping, keeping the original value out of ordinary systems.

Dynamic Data Masking

A control that obscures sensitive values at query time based on the requesting user's role, leaving the stored data unchanged for authorized users.

k-Anonymity

A privacy property requiring each record to be indistinguishable from at least k minus one other records on quasi-identifiers, reducing re-identification risk.

Attribute-Based Access Control

An authorization model granting access by evaluating attributes of the user, the resource, and the environment against policy rules, enabling context-aware decisions.

Semi-Additive Measure

A fact measure that can be summed across some dimensions but not all, such as an account balance that sums across accounts but not across time.

Drill-Down

An OLAP operation that navigates from a coarser aggregate to a finer level of detail, for example from annual revenue to monthly revenue within a region.

As-Of Join

A join that matches each fact row to the dimension version whose effective date range covers the event date, required for accurate Type 2 history reporting.

Golden Record

The single trusted, consolidated version of a master entity, assembled from source systems by applying survivorship rules that decide which source wins per attribute.

Business Glossary

A governed catalog of approved business terms, definitions, owners, and synonyms that establishes shared vocabulary and links to physical data assets.

Data Lineage

A record of how data moves from source systems through transformations to reports, used for impact analysis, root cause tracing, and regulatory evidence.