FDST-BDS key terms

A person accountable for a defined subject area or data set, who resolves quality issues, approves definitions, and enforces policy within business processes rather than owning the technical storage.

The assignment of ultimate accountability for a data domain to a named role that approves access, defines acceptable use, and answers for the data's fitness to support decisions.

A cross-functional decision body that sets enterprise data policy, arbitrates conflicts between domains, prioritizes remediation funding, and reviews compliance with agreed standards on a recurring cadence.

A written directive stating mandatory rules for how data is created, accessed, shared, retained, and retired, together with the roles empowered to grant exceptions.

The practice of labeling data by sensitivity and business criticality, such as public, internal, confidential, or restricted, so that handling controls can be applied consistently.

A recorded trace of where data originated, how it was transformed at each step, and which reports or models consume it, used for impact analysis and audit evidence.

A curated list of business terms with approved definitions, synonyms, and accountable owners, giving analysts and systems one shared vocabulary for reporting and requirements.

A store holding technical, business, and operational metadata about data assets, such as schemas, owners, definitions, and refresh schedules, queryable by tools and people.

A searchable inventory of data assets that exposes descriptions, owners, classifications, quality indicators, and usage guidance so consumers can find and trust suitable data.

The core reference entities shared across processes, typically customers, products, suppliers, and locations, whose consistent identity and attributes other systems depend on.

A controlled set of permissible values or codes, such as country codes or status flags, used to classify and constrain transactions consistently across applications.

The single trusted version of a master entity assembled by merging and survivorship rules from multiple sources, published for downstream consumers to reference.

The process of deciding which records from different sources refer to the same real-world thing, using matching rules, blocking keys, and similarity scoring before merging.

A precedence rule that decides which source value wins when merged records disagree, based on recency, trust ranking, completeness, or source system priority.

A technology-neutral diagram of the main business entities and their relationships, used to agree scope and meaning before physical design begins.

A detailed model of entities, attributes, keys, and relationships that is independent of any specific database product but precise enough to guide physical implementation.

The implementation-specific design of tables, columns, data types, indexes, partitions, and constraints targeting a particular database engine and workload pattern.

Structuring tables so each fact is stored once by removing repeating groups and transitive dependencies, reducing update anomalies at the cost of more joins.

Deliberately duplicating or pre-joining data in a physical design to cut query joins and improve read performance, accepting extra storage and refresh effort.

A dimensional design with a central fact table holding measures and foreign keys, surrounded by denormalized dimension tables that supply descriptive context for filtering and grouping.

A table whose rows record measurable events or snapshots at a declared grain, holding numeric measures plus foreign keys to the dimensions that describe them.

An explicit statement of what one row in a fact table represents, such as one order line per day, which fixes the level of detail available to every query.

A dimension design pattern for tracking attribute changes over time, either by overwriting values or by adding versioned rows with effective and expiry dates.

A named, measurable aspect of data fitness, such as completeness, accuracy, consistency, timeliness, validity, or uniqueness, used to frame quality requirements and tests.

An executable check that compares data against an expected condition, such as a null threshold or allowed value list, and returns pass or fail with counts.

Statistical inspection of a data set to reveal value distributions, null rates, distinct counts, patterns, and key candidates before rules or models are designed.

Correcting, standardizing, or removing defective records so they meet agreed rules, with each change logged to preserve traceability of what was altered.

A recurring report that scores defined data sets against quality dimensions and thresholds, showing trends and owners so remediation can be prioritized.

A scheduled sequence that extracts data from sources, transforms it to conform to target rules, and loads it into a warehouse or other store, with error handling and restart logic.

An integration approach that loads raw source data into a target platform first and applies transformations there, exploiting the target's compute and keeping raw history.

A technique that identifies and propagates only rows changed since the last run, using logs, timestamps, or triggers, to reduce load on sources and targets.

A property of a load or job whereby rerunning it with the same input produces the same end state, making retries after failure safe.

An agreed specification of endpoints, request and response schemas, status codes, and versioning rules that lets producers evolve services without breaking consumers.

A service that stores versioned message schemas and enforces compatibility rules so producers and consumers of a streaming or messaging platform can evolve independently.

Granting permissions to job roles rather than individuals, so users inherit the access their duties require and changes are managed by adjusting role membership.

The principle of granting only the minimum access needed for a task and only for as long as needed, reducing the blast radius of misuse or compromise.

Replacing or obscuring sensitive values with realistic substitutes so non-production users or analysts can work without seeing the original confidential content.

Replacing direct identifiers with tokens that can be re-linked to the original only through separately secured mapping, lowering exposure while preserving analytical joins.

A rule set specifying how long each data category is kept, where it resides, and how it is destroyed, balancing legal obligations against storage and privacy risk.

Tracing dependencies to determine which reports, models, or processes would break or change if a source, definition, or transformation were altered.

A governed model of metrics, dimensions, and joins defined once and reused by reporting tools, so business calculations stay consistent across dashboards.

A quantified measure tied to a strategic objective, with an agreed definition, target, and owner, used to judge whether performance is on track.

A reference describing each field in a data set, including its name, meaning, data type, allowed values, nullability, and source, aimed at implementers rather than executives.