- Year not given · 2+3+5 marks
- Year not given
DMQL; transactional vs warehouse database
Short answer
DMQL is a SQL-based data mining query language for specifying mining tasks and defining warehouse schemas. It applies to ad hoc mining (characterisation, association, classification, clustering) and to defining star/snowflake/constellation schemas. Transactional databases store current detailed normalised operational data; data warehouses store historical summarised dimensional data for analysis.
What is DMQL? Where is it applicable? Differentiate between Transactional databases with Data warehouse databases.
DMQL
DMQL (Data Mining Query Language) is a query language proposed by Han et al. (in the DBMiner system) for specifying data mining tasks and for defining data warehouses. It is based on SQL. Its main parts are:
- the task-relevant data (
use database,from,where), - the kind of knowledge to mine (
mine characteristics,mine associations,mine classification), - background knowledge (concept hierarchies),
- interestingness thresholds (
with support threshold,with confidence threshold), - presentation of the discovered patterns.
It also has commands to define warehouse schemas: define cube and define dimension.
Example:
use database sales_db
mine associations as buying_habits
matching P(X: customer, W) ^ Q(X, Y) => buys(X, Z)
from sales
with support threshold = 2%
with confidence threshold = 60%
Where it is applicable
- Specifying and running ad hoc mining tasks (characterisation, association, classification, clustering).
- Defining data warehouse schemas (star, snowflake, constellation) and cubes.
- As a standard interface between mining front ends and mining engines; it lets the user interact with the system and refine the task.
Transactional database vs data warehouse
| Point | Transactional database | Data warehouse |
|---|---|---|
| Purpose | Daily operations (OLTP) | Analysis and decision support |
| Data | Current, detailed | Historical, summarised |
| Design | Normalised ER | Dimensional (star/snowflake) |
| Updates | Frequent | Periodic load, read-mostly |
| Queries | Short, simple | Complex, aggregate |
| Users | Clerks, thousands | Managers, analysts |
| Size | MB to GB | GB to TB |
✓ Answer checked by Claude Opus 5.5 (high effort)
Related questions
- Data warehouse vs database · Data Mining, chapter 1
- Where data mining is applicable · Data Mining, chapter 1
- Data warehouse, data mart and need · Data Mining, chapter 1
- OLAP vs OLTP; schema modeling; star DMQL · Data Mining, chapter 2
- DW characteristics vs operational database · Data Mining, chapter 1
- RDBMS vs DW database; DW terms · Data Mining, chapter 1
- Data warehouse and data mining · Information System, chapter 4 · asked 2×
- Column vs row-oriented NoSQL databases · Big Data Technologies, chapter 4
- Hybrid data in big data warehouse · Big Data Technologies, chapter 1
- SQL: create and use bank database · Database Management System, chapter 3