Conducting data mining and analytics on high-volume transactional database systems is recommended because such systems have the most up-to-date data.

• A. True
• B. False
• C. Depends On Workload
• D. Never

Answer: (B)

Data mining works best on data that is prepared for analysis—historical, integrated, and optimized for complex queries. High-volume transactional systems (OLTP) are designed for fast, concurrent updates and keeping data normalized for operational processes. They’re great for current transactions, but their structure and workload aren’t ideal for analytics: heavy joins, frequent updates, and locking can slow down analyses, and the data may be scattered across many normalized tables rather than consolidated into a single, stable dataset.

The usual approach is to extract and transform data from these transactional systems into an analytical store (a data warehouse or data mart) where data is cleaned, consolidated, and often denormalized. This setup supports efficient scans, aggregations, and historical analysis, which are essential for meaningful data mining. The freshness of data in the warehouse may involve some latency due to ETL/ELT processes, but it provides the consistency and scale needed for mining. Therefore, the statement is not accurate: you don’t rely on the live high-volume transactional systems for mining simply because they have the most up-to-date data.