Database Management Systems

Enterprise Information Management (EIM) is a broad discipline encompassing data governance, and data governance is indeed a key focus area within EIM frameworks. EIM provides the organizational structure for managing information as an asset.

Aggregation in OLAP pre-calculates and stores summarized measures (sums, averages, etc.) at various dimensional levels. This optimization dramatically improves query performance by avoiding runtime calculations on large datasets.

Rotating (or 'pivot') in OLAP changes the dimensional orientation of the report - essentially rotating the cube to view the same data from a different perspective, swapping rows and columns to highlight different dimensions.

Report bursting minimizes database queries by generating separate report outputs for different recipients or groups from a single execution pass, rather than running the entire report multiple times. This reduces database load and improves performance. Option B is incorrect because report bursting executes once, not multiple times. Option C is incorrect because it actually reduces network traffic by sending targeted data.

The staging area is the correct interim location for storing data extracted from source systems before it's transformed and loaded into the data warehouse. It provides a buffer where data can be validated, cleansed, and transformed without impacting source systems. A data mart is a subset of the data warehouse, ODS (Operational Data Store) stores current operational data, and staging is the correct answer.

Data warehouses are organized around important subject areas like customers, products, or sales, making them different from operational systems that are organized around business processes. They contain historical data (not just current), have consistent naming conventions, and are not updated by end users. Option C is correct because subject-oriented organization is a key characteristic of data warehouses.

Data profiling is the correct answer - it's the practice of examining data to understand its content, quality, and structure. It includes analyzing data patterns, identifying anomalies, and assessing data quality metrics. Data enrichment adds information, data caving is not a standard term, and data refinement typically refers to improving data quality after profiling has identified issues.

Data cleansing is the correct process for removing or correcting erroneous, incomplete, or old data from databases. It involves identifying and fixing data quality issues like errors, inconsistencies, and outdated records. Data eradication is not a standard term, data unloading refers to extracting data, and data purging typically refers to deleting old data regardless of quality.

Master data management (MDM) is the correct answer - it identifies and maintains a single, unified view of reference data (like customer, product, or supplier data) across the enterprise. MDM ensures consistency and accuracy of core business entities. Data Quality Management focuses on data quality, Metadata Management manages data about data, and Data Security Management protects data from unauthorized access.

ETL (Extract, Transform, Load) is recommended for large data volumes, low data quality, and frequent loads because transformations happen before loading, allowing for complex data cleaning and validation. ETL can handle poor data quality better than ELT (Extract, Load, Transform), which transforms after loading. ELTL adds logging, and EII (Enterprise Information Integration) provides virtual views.

Metadata provides context to data by describing its meaning, structure, and relationships, acting as the semantic layer between IT systems and business users. Metadata answers questions like what the data means, where it came from, and how it relates to other data. Master data refers to core business entities, not contextual information.

When an enterprise utilizes multiple heterogeneous ETL and reporting tools, a distributed metadata management architecture is recommended to allow individual tools to maintain their metadata while enabling sharing.

Metadata Management is applicable across all layers of the data architecture - the source layer (describing source systems), the data warehouse (describing warehouse structures and mappings), and the report layer (describing report definitions and business terms). It provides context and documentation throughout the data flow, making all layers correct.