(SEM VII) THEORY EXAMINATION 2023-24 DATA WAREHOUSING AND DATA MINING

SECTION A – Very Short Answer Type

(2 × 10 = 20 Marks)

a) Key steps of Data Mining

The key steps are:                                               Data cleaning

Data integration                                                  Data selection

Data transformation                                            Data mining

Pattern evaluation                                               Knowledge presentation

These steps convert raw data into useful knowledge.

b) Support and Confidence

Support: Frequency of an itemset appearing in the database.

Confidence: Probability that item Y is purchased when item X is purchased.

c) Data Warehouse Process

It involves data extraction from multiple sources, transformation to a consistent format, and loading into a central repository for analysis.

d) Warehousing Strategy

Warehousing strategy defines how data is collected, stored, and accessed, such as enterprise warehouse, data mart, or virtual warehouse.

e) Statement of Apriori Algorithm

“All non-empty subsets of a frequent itemset must also be frequent.”
This property reduces the search space in association rule mining.

f) Drawbacks of K-Means Algorithm

Requires predefined number of clusters                       Sensitive to initial centroids

Fails with non-spherical clusters                                    Affected by noise and outliers

g) Classification

Classification assigns data objects to predefined classes based on labeled training data, using algorithms like decision trees and Naive Bayes.

h) Clustering

Clustering groups similar data objects without predefined labels, aiming to maximize similarity within clusters and minimize similarity between clusters.

i) Need for Data Mining

Data mining helps discover hidden patterns, trends, and relationships in large datasets to support decision-making.

j) Binning

Binning is a data smoothing technique that reduces noise by grouping values into intervals or bins.

SECTION B – Long Answer Type

(Attempt any three – 10 Marks each)

2(a) Knowledge Extraction Process

Knowledge extraction is the process of discovering meaningful patterns from large datasets.

Steps:

Data Selection – Choose relevant data

Data Preprocessing – Clean noisy and missing data

Transformation – Normalize and aggregate data

Data Mining – Apply algorithms (classification, clustering, association)

Pattern Evaluation – Identify interesting patterns

Knowledge Presentation – Visualize results using charts and reports

This process converts raw data into actionable knowledge.

2(b) OLAP Functions, Tools, and Servers

OLAP Functions:                                                        Roll-up

Drill-down                                                                 Slice

Dice                                                                           Pivot

OLAP Tools:                                                               MOLAP (Multidimensional OLAP)

ROLAP (Relational OLAP)                                          HOLAP (Hybrid OLAP)

OLAP Servers:

OLAP servers store and process multidimensional data, enabling fast analytical queries.

2(c) Database Schemas

Types:

Star Schema – Central fact table connected to dimension tables

Snowflake Schema – Normalized version of star schema

Fact Constellation Schema – Multiple fact tables sharing dimensions

Example:
Sales fact table linked to time, product, and location dimensions.

2(d) Statistical Measures in Classification

Key measures include:                                               Mean

Median                                                                      Variance

Standard Deviation                                                   Correlation

These measures summarize data distribution and improve classification accuracy in large databases.

2(e) Building a Data Warehouse

Steps include:                                                             Business analysis

Data source identification                                          ETL design

Data modeling                                                           Data loading

Testing and deployment

A data warehouse supports long-term decision making.

SECTION C – Descriptive Answer Type

3(a) Mapping Data Warehouse to Multiprocessor Architecture

Steps:

Partition data across processors                                  Assign fact and dimension tables

Enable parallel query processing                                 Synchronize data access

Optimize workload distribution                                   This improves performance and scalability.

3(b) Data Cubes with Example

A data cube represents data in multiple dimensions.

Example:
Sales analyzed by time, location, and product.
Each cell stores aggregated values like total sales.

4(a) Concept Hierarchy

Concept hierarchy organizes data from general to specific levels.

Example:
Location → Country → State → City                             It supports roll-up and drill-down operations.

4(b) Warehouse Management and Support Process

Includes:                                                                       Data refresh

Indexing                                                                       Backup and recovery

Query management                                                     Performance tuning

Ensures smooth warehouse operation.

5(a) Data Mining System Integration Approaches

Semi-tight coupling provides better performance.

5(b) Data Consolidation Statement Justification

Yes, data consolidation is a data modeling activity because it integrates data from multiple sources into a unified schema for analysis.

6(a) Measures of Central Tendency

Mean – Average value                                       Median – Middle value

Mode – Most frequent value                             These summarize dataset characteristics.

6(b) Quartiles and Histograms

Quartiles divide data into four equal parts       Histograms graphically represent data distribution

Both help in understanding data spread.

7(a) Distance-Based vs Decision Tree Algorithms

7(b) Web Mining

Types:                                                                         Web Content Mining – Extracts text and media

Web Structure Mining – Analyzes link structure     Web Usage Mining – Studies user behavior

Used in recommendation systems and search engines.