JNTU MCA 2019 DATA WAREHOUSING AND DATA MINING Previous Question Paper (December - 2019 - R13)

Data preprocessing is a crucial data mining technique that involves transforming raw data into a clean, consistent, and usable format. Since real-world data is often incomplete, inconsistent, and lacking in certain behaviors or trends, preprocessing is necessary to improve the quality of the data and, consequently, the accuracy and efficiency of the data mining process.

The various forms of data preprocessing include:

• Data Cleaning: Filling in missing values, smoothing noisy data, identifying or removing outliers, and resolving inconsistencies.
• Data Integration: Merging data from multiple heterogeneous sources (such as databases or files) into a coherent data store.
• Data Transformation: Normalizing, aggregating, or generalizing data to make it suitable for mining, such as scaling numerical data or constructing new attributes.
• Data Reduction: Reducing the volume of the data while maintaining the integrity of the original analytical results, using techniques like dimensionality reduction, numerosity reduction, or data compression.

A data warehouse architecture is typically organized into a three-tier structure that facilitates the flow of data from operational sources to end-user analysis:

1. Bottom Tier (Data Warehouse Server): This layer consists of the data warehouse database system. It includes the extraction, cleaning, transformation, and loading (ETL) tools that integrate data from multiple heterogeneous operational sources into a unified repository.
2. Middle Tier (OLAP Server): This tier acts as an abstraction layer between the database and the end-user. It implements an Online Analytical Processing (OLAP) model, which can be either ROLAP (Relational OLAP) or MOLAP (Multidimensional OLAP), to provide fast, complex analytical queries and multidimensional data views.
3. Top Tier (Front-end Client): This is the presentation layer where end-users interact with the system. It includes query and reporting tools, data analysis tools, and data mining software used for decision support and business intelligence.

The architecture also incorporates metadata repositories to manage data definitions and administrative tools to handle scheduling, security, and system monitoring.