Skip to main content

Short note: Noise in data warehouse

Short answer

Noise in data warehouses is random errors (e.g., age=250) from faulty instruments, human entry, transmission, or inconsistent sources. It causes wrong aggregates and patterns. Handled via binning, regression, clustering, human inspection, and prevention through ETL validation and quality checks.

Write a short note on Noise in Data Warehouse.

Noise in a data warehouse is random error, variance or meaningless/incorrect data in the stored values, for example age = 250, a negative salary, a misspelt city name, or a duplicate record.

Causes

  • Faulty data collection instruments and sensors.
  • Human or data-entry errors.
  • Transmission errors and technology limits.
  • Inconsistent naming and units across the source systems merged by ETL.

Effects: wrong aggregates and reports, wrong patterns, poor decisions, and lower trust in the warehouse.

Handling noisy data

  • Binning: sort values, partition into bins, smooth by bin means, medians or boundaries.
  • Regression: smooth values using a fitted function.
  • Clustering: detect and remove outliers that fall outside clusters.
  • Computer and human inspection: flag suspicious values for manual check.
  • Prevention: validation rules at the source and data quality checks in the ETL (cleaning) stage.

Cleaning noise is a main part of the ETL process, done before data are loaded.

✓ Answer checked by Claude Opus 5.5 (high effort)

Related questions

All Data Mining chapter 2 questions →