introduction to data mining

Prof. S. K. Pandey, I.T.S, Ghaziabad

Data Warehousing & Mining

UNIT – I

Prof. S.K. Pandey, I.T.S, Ghaziabad 2

Syllabus of Unit - ISyllabus of Unit - I

DSS-Uses, definition, Operational Database. Introduction to DATA Warehousing. Data-Mart, Concept of Data-Warehousing, Multi Dimensional Database Structures. Client/Server Computing Model & Data

Warehousing Parallel Processors & Cluster Systems. Distributed

DBMS implementations.

Introduction – Introduction – Decision Support Decision Support System (DSS)System (DSS)

A Decision Support System (DSS) is an interactive computer-based system or subsystem intended to help decision makers use communications technologies, data, documents, knowledge and/or models to identify and solve problems, complete decision process tasks, and make decisions.

It is clear that DSS belong to an environment with multidisciplinary foundations, including (but not exclusively):– Database research, – Artificial intelligence, – Human-computer interaction, – Simulation methods, – Software engineering, and – Telecommunications.



DSSDSS

• A Decision Support System (DSS) is a computer-based information system that supports business or organizational decision-making activities.

• DSSs serve the management, operations, and planning levels of an organization (usually mid and higher management) and help to make decisions, which may be rapidly changing and not easily specified in advance (Unstructured and Semi-Structured decision problems).

• Decision support systems can be either fully computerized, human or a combination of both.

Historical Evolution of DSSHistorical Evolution of DSS



Typical DSS ArchitectureTypical DSS Architecture

TPS: transaction processing system

MODEL: representation of a problem

OLAP: on-line analytical processing

USER INTERFACE: how user enters problem & receives answers

DSS DATABASE: current data from applications or groups

DATA MINING: technology for finding relationships in large data bases for prediction

TPSEXTERNAL

DATADSS DATA

BASE

DSS SOFTWARE SYSTEMMODELS

OLAP TOOLS

DATA MINING TOOLS

USERINTERFACE

USER

Why DSS?Why DSS?

Increasing complexity of decisions– Technology– Information:

“Data, data everywhere, and not the time to think!”– Number and complexity of options– Pace of change

Increasing availability of computerized support– Inexpensive high-powered computing– Better software– More efficient software development process

Increasing usability of computers



Operational DatabasesOperational Databases Operational database management systems (also referred to as OLTP

databases), are used to manage dynamic data in real-time. These types of databases allow you to do more than simply view archived

data. Operational databases allows to modify that data (add, change or delete data), doing it in real-time.

Since the early 90's, the operational database software market has been largely taken over by SQL engines.

Today, the operational DBMS market (formerly OLTP) is evolving dramatically, with new, innovative entrants and incumbents supporting the growing use of unstructured data and NoSQL DBMS engines, as well as XML databases and NewSQL databases.

Operational databases are increasingly supporting distributed database architecture that provides high availability and fault tolerance through replication and scale out ability.


FEATURES DATABASE DATA WAREHOUSECharacteristic It is based on Operational Processing. It is based on Informational Processing.

Data It mainly stores the Current data which always guaranteed to be up-to-date.

It usually stores the Historical data whose accuracy is maintained over time.

Function It is used for day-to-day operations. It is used for long-term informational requirements and decision support.

User The common users are clerk, DBA, database professional.

The common users are knowledge worker (e.g., manager, executive, analyst)

Unit of work Its work consists of short and simple transaction.

The operations on it consists of complex queries..

Focus The focus is on “Data IN” The focus is on “Information OUT”

Orientation The orientation is on Transaction. The orientation is on Analysis.

DB design The designing of database is ER based and application-oriented.

The designing is done using star/snowflake schema and its subject-oriented.

Summarization The data is primitive and highly detailed.

The data is summarized and in consolidated form.

View The view of the data is flat relational. The view of the data is multidimensional.

Differences between the Databases and Data Warehouses


FEATURES DATABASE DATA WAREHOUSEFunction It is used for day-to-day operations. It is used for long-term informational

requirements and decision support.

User The common users are clerk, DBA, database professional.

The common users are knowledge worker (e.g., manager, executive, analyst)

Access The most frequent type of access type is read/write.

It mostly use the read access for the stored data.

Operations The main operation is index/hash on primary key.

For any operation it needs a lot of scans.

Number of records accessed

A few tens of records. A bunch of millions of records.

Number of users In order of thousands. In the order of hundreds only.

DB size 100 MB to GB. 100 GB to TB.

Priority High performance, high availability High flexibility, end-user autonomy

Metric To measure the efficiency, transaction throughput is measured.

To measure the efficiency, query throughput and response time is

measured.


Concept ofConcept ofData WarehousingData Warehousing


Why Separate Data Warehouse?Why Separate Data Warehouse?

High performance for both systems– DBMS— tuned for OLTP: access methods, indexing,

concurrency control, recovery– Warehouse—tuned for OLAP: complex OLAP queries,

multidimensional view, consolidation. Different functions and different data:

– missing data: Decision support requires historical data which operational DBs do not typically maintain

– data consolidation: DS requires consolidation (aggregation, summarization) of data from heterogeneous sources

– data quality: different sources typically use inconsistent data representations, codes and formats which have to be reconciled


DATA Warehousing - IntroductionDATA Warehousing - Introduction

A data warehouse is a subject-oriented,

integrated, nonvolatile, time-variant collection

of data in support of management's decisions.

- WH Inmon


Data Warehouse UsageData Warehouse Usage Three kinds of data warehouse applications

– Information processing supports querying, basic statistical analysis, and reporting using

crosstabs, tables, charts and graphs

– Analytical processing multidimensional analysis of data warehouse data supports basic OLAP operations, slice-dice, drilling, pivoting

– Data mining knowledge discovery from hidden patterns supports associations, constructing analytical models, performing

classification and prediction, and presenting the mining results using visualization tools.

Differences among the three tasks


Data Warehouse: Subject-OrientedData Warehouse: Subject-Oriented

Organized around major subjects, such as customer, product,

sales.

Focusing on the modeling and analysis of data for decision

makers, not on daily operations or transaction processing.

Provide a simple and concise view around particular

subject issues by excluding data that are not useful in the

decision support process.


Subject-OrientedSubject-Oriented

Quotes Orders

ProspectsLeads

Operational Data Warehouse

Customers Products

Regions Time

Focus is on Subject Areas rather than ApplicationsFocus is on Subject Areas rather than Applications


Data Warehouse—IntegratedData Warehouse—Integrated

Constructed by integrating multiple, heterogeneous data sources– relational databases, flat files, on-line transaction records

Data cleaning and data integration techniques are applied.– Ensure consistency in naming conventions, encoding

structures, attribute measures, etc. among different data sources

E.g., Hotel price: currency, tax, breakfast covered, etc.

– When data is moved to the warehouse, it is converted.


Data Warehouse—Time VariantData Warehouse—Time Variant

The time horizon for the data warehouse is significantly longer

than that of operational systems.

– Operational database: current value data.

– Data warehouse data: provide information from a historical

perspective (e.g., past 5-10 years)

Every key structure in the data warehouse

– Contains an element of time, explicitly or implicitly

– But the key of operational data may or may not contain

“time element”.


Time VariantTime Variant


Current Value data• time horizon : 60-90 days

Snapshot data• time horizon : 5-10 years•data warehouse stores historical data

Data Warehouse Typically Spans Across TimeData Warehouse Typically Spans Across Time


Data Warehouse—Non-VolatileData Warehouse—Non-Volatile

A physically separate store of data transformed from the

operational environment.

Operational update of data does not occur in the data

warehouse environment.

– Does not require transaction processing, recovery, and

concurrency control mechanisms

– Requires only two operations in data accessing:

initial loading of data and access of data.


Non-volatileNon-volatile


replacechange

insert

changeinsert

delete load

read only access

Data Warehouse Is Relatively Static In NatureData Warehouse Is Relatively Static In Nature


Data Warehouse vs. Heterogeneous Data Warehouse vs. Heterogeneous DBMSDBMS

Traditional heterogeneous DB integration: – Build wrappers/mediators on top of heterogeneous databases

– Query driven approach When a query is posed to a client site, a meta-dictionary is used to

translate the query into queries appropriate for individual heterogeneous sites involved, and the results are integrated into a global answer set

Complex information filtering, compete for resources

Data warehouse: update-driven, high performance– Information from heterogeneous sources is integrated in advance and

stored in warehouses for direct query and analysis


Data Warehouse vs. Operational DBMSData Warehouse vs. Operational DBMS

OLTP (on-line transaction processing)

– Major task of traditional relational DBMS

– Day-to-day operations: purchasing, inventory, banking, manufacturing, payroll, registration, accounting, etc.

OLAP (on-line analytical processing)

– Major task of data warehouse system

– Data analysis and decision making Distinct features (OLTP vs. OLAP):

– User and system orientation: customer vs. market

– Data contents: current, detailed vs. historical, consolidated

– Database design: ER + application vs. star + subject

– View: current, local vs. evolutionary, integrated

– Access patterns: update vs. read-only but complex queries


OLTP vs. OLAPOLTP vs. OLAP OLTP OLAP

users clerk, IT professional knowledge worker

function day to day operations decision support

DB design application-oriented subject-oriented

data current, up-to-date detailed, flat relational isolated

historical, summarized, multidimensional integrated, consolidated

usage repetitive ad-hoc

access read/write index/hash on prim. key

lots of scans

unit of work short, simple transaction complex query

# records accessed tens millions

#users thousands hundreds

DB size 100MB-GB 100GB-TB

metric transaction throughput query throughput, response

Prof. S.K. Pandey, I.T.S, Ghaziabad 27Slide 29- 27

Characteristics of Data WarehousesCharacteristics of Data Warehouses

Multidimensional conceptual view Generic dimensionality Unlimited dimensions and aggregation levels Unrestricted cross-dimensional operations Dynamic sparse matrix handling Client-server architecture Multi-user support Accessibility Transparency Intuitive data manipulation Consistent reporting performance Flexible reporting

Prof. S.K. Pandey, I.T.S, Ghaziabad28

Multi-Tiered ArchitectureMulti-Tiered ArchitectureComponents & Framework

Data Integration Stage


Data MartData Mart

The data mart is a subset of the data warehouse that is usually oriented to a specific business line or team. Data marts are small slices of the data warehouse.

Whereas data warehouses have an enterprise-wide depth, the information in data marts pertains to a single department.

Data marts improve end-user response time by allowing users to have access to the specific type of data they need to view most often by providing the data in a way that supports the collective view of a group of users.

Contd………….Contd…………. A data mart is basically a condensed and more focused version

of a data warehouse that reflects the regulations and process specifications of each business unit within an organization.

Each data mart is dedicated to a specific business function or region.

This subset of data may span across many or all of an enterprise’s functional subject areas.

It is common for multiple data marts to be used in order to serve the needs of each individual business unit (different data marts can be used to obtain specific information for various enterprise departments, such as accounting, marketing, sales, etc.).


Reasons for creating a data martReasons for creating a data mart

Easy access to frequently needed data Creates collective view by a group of users Improves end-user response time Ease of creation Lower cost than implementing a full data warehouse Potential users are more clearly defined than in a full

data warehouse Contains only business essential data and is less

cluttered.


Types of Data MartsTypes of Data Marts Dependent Data Mart: A dependent data mart is one

whose source is another data warehouse, and all dependent data marts within an organization are typically fed by the same source — the enterprise data warehouse.


Contd…Contd…

Independent Data Mart: An independent data mart is one whose source is directly from transactional systems, legacy applications, or external data feeds.



Data warehouse:

i. Holds multiple subject areasii. Holds very detailed informationiii. Works to integrate all data sourcesiv. Does not necessarily use a dimensional model but feeds dimensional models.

Data mart:

i. Often holds only one subject area- for example, Finance, or Sales ii. May hold more summarized data (although many hold full detail)iii. Concentrates on integrating information from a given subject area or set of source systemsiv. Is built focused on a dimensional model using a star schema.

Data mart vs data warehouse

Multi-Dimensional Database Multi-Dimensional Database StructureStructure



Multi Dimensional Database Multi Dimensional Database StructuresStructures

Sales volume as a function of product, month, and region

Pro

duct

Regio

n

Month

Dimensions: Product, Location, TimeHierarchical summarization paths

Industry Region Year

Category Country Quarter

Product City Month Week

Office Day

Prof. S.K. Pandey, I.T.S, Ghaziabad 37Slide 29- 37

Data Modeling for Data WarehousesData Modeling for Data Warehouses

Example of Two- Dimensional vs. Multi-Dimensional

REGION

REG1 REG2 REG3

P123

P124

P125

P126::

PRODUCT

Two Dimensional Model

::

Three dimensional data cube

Product

Reg 1P123

P124

P125

P126

Reg 2 Reg 3

Region


From Tables and Spreadsheets From Tables and Spreadsheets to Data Cubesto Data Cubes

A data warehouse is based on a multidimensional data model

which views data in the form of a data cube

A data cube, such as sales, allows data to be modeled and viewed in

multiple dimensions

– Dimension tables, such as item (item_name, brand, type), or

time(day, week, month, quarter, year)

– Fact table contains measures (such as dollars_sold) and keys to

each of the related dimension tables

In data warehousing literature, an n-D base cube is called a base

cuboid. The top most 0-D cuboid, which holds the highest-level of

summarization, is called the apex cuboid. The lattice of cuboids

forms a data cube.


Cube: A Lattice of CuboidsCube: A Lattice of Cuboids

all

time item location supplier

time,item time,location

time,supplier

item,location

item,supplier

location,supplier

time,item,location

time,item,supplier

time,location,supplier

item,location,supplier

time, item, location, supplier

0-D cuboid

1-D cuboids

2-D cuboids

3-D cuboids

4-D(base) cuboid


Warehouse Database SchemasWarehouse Database Schemas

Star SchemaSnow-flake SchemaFact constellation (Gathering/ Togetherness)

schema


Conceptual Modeling of Data Conceptual Modeling of Data WarehousesWarehouses

Modeling data warehouses: dimensions & measures

– Star schema: A fact table in the middle connected to a set of

dimension tables

– Snowflake schema: A refinement of star schema where some

dimensional hierarchy is normalized into a set of smaller

dimension tables, forming a shape similar to snowflake

– Fact constellations: Multiple fact tables share dimension

tables, viewed as a collection of stars, therefore called galaxy

schema or fact constellation


Example of Star SchemaExample of Star Schema

time_keydayday_of_the_weekmonthquarteryear

time

location_keystreetcityprovince_or_streetcountry

location

Sales Fact Table

time_key

item_key

branch_key

location_key

units_sold

dollars_sold

avg_sales

Measures

item_keyitem_namebrandtypesupplier_type

item

branch_keybranch_namebranch_type

branch


Example of Snowflake SchemaExample of Snowflake Schema


time

location_keystreetcity_key

location

Sales Fact Table

time_key

item_key

branch_key

location_key

units_sold

dollars_sold

avg_sales

Measures

item_keyitem_namebrandtypesupplier_key

item


branch

supplier_keysupplier_type

supplier

city_keycityprovince_or_streetcountry

city


Example of Fact ConstellationExample of Fact Constellation


time

location_keystreetcityprovince_or_streetcountry

location

Sales Fact Table

time_key

item_key

branch_key

location_key

units_sold

dollars_sold

avg_sales

Measures

item_keyitem_namebrandtypesupplier_type

item


branch

Shipping Fact Table

time_key

item_key

shipper_key

from_location

to_location

dollars_cost

units_shipped

shipper_keyshipper_namelocation_keyshipper_type

shipper


Client/Server Computing Model & Client/Server Computing Model & Data WarehousingData Warehousing

The fundamental characteristic of client/server computing is distribution of computing resources (e.g. data, compute power) across different computers.

The idea is to divide applications into logical segments (tasks) so that they are then performed on platforms most appropriate.

A client/server database system increases processing power by separating the database management system from the application; the client as the front-end system handling the user interface and the server as the back-end system accessing the database, which cooperate to run an application.

Contd….Contd….

Data Warehousing is a continual process which enables a corporation to assemble operational and other data from a variety of internal and external sources, and transform that data into consistent, high-quality, business information, distribute that information to the points of maximum value within the organizations, and provide easy, flexible and fast access for busy non-technical users.


Reasons for using client/serverReasons for using client/server

Exploitation of centralized computing power /data capacity

Scalability Performance Flexibility (in order to adjust to changing demands) GUI on desktop Protection of investment, strategic software,

strategic data Client/server provides an integrated solution.



Parallel Processors & Cluster Parallel Processors & Cluster SystemsSystems


Loosely Coupled - ClustersLoosely Coupled - Clusters Collection of independent whole uni-processors or SMPs

– Usually called nodes

Interconnected to form a cluster Working together as unified resource

– Illusion of being one machine

Communication via fixed path or network connections

Cluster BenefitsCluster Benefits Absolute scalability Incremental scalability High availability Superior price/performance


Distributed DBMS implementationsWhat Is A Distributed DBMS?What Is A Distributed DBMS?

Decentralization of business operations and globalization of businesses created a demand for distributing the data and processes across multiple locations.

Distributed database management systems (DDBMS) are designed to meet the information requirements of such multi-location organizations.

A DDBMS manages the storage and processing of logically related data over interconnected computer systems in which both data and processing functions are distributed among several sites.

Distributed processing shares the database’s logical processing among two or more physically independent sites that are connected through a network.

DDBMS AdvantagesDDBMS Advantages

Data located near site with greatest demand Faster data access Faster data processing Growth facilitation Improved communications Reduced operating costs User-friendly interface Less danger of single-point failure Processor independence



Distributed ProcessingDistributed ProcessingShares database’s logical processing among physically, networked independent sites


DDBMS ComponentsDDBMS Components Computer workstations that form the network

system. Network hardware and software components that

reside in each workstation. Communications media that carry the data from one

workstation to another. Transaction processor (TP) receives and processes

the application’s data requests. Data processor (DP) stores and retrieves data

located at the site. Also known as data manager (DM).


Distributed DB TransparencyDistributed DB Transparency

A DDBMS ensures that the database operations are transparent to the end user.

Different types of transparencies are:– Distribution transparency– Transaction transparency– Failure transparency– Performance transparency– Heterogeneity transparency

55

Distributed Database DesignDistributed Database Design

All design principles and concepts discussed in the context of a centralized database also apply to a distributed database.

Three additional issues are relevant to the design of a distributed database:– data fragmentation– data replication– data allocation

Prof. S.K. Pandey, I.T.S, Ghaziabad

56

Data FragmentationData Fragmentation

Data fragmentation allows us to break a single object (a database or a table) into two or more fragments.

Three type of fragmentation strategies are available to distribute a table: - Horizontal, Vertical, Mixed.

Horizontal fragmentation divides a table into fragments consisting of sets of tuples:– Each fragment has unique rows and is stored at a different

node– Example: A bank may distribute its customer table by

location


57

Contd……Contd……

Vertical fragmentation divides a table into fragments consisting of sets of columns– Each fragment is located at a different node and

consists of unique columns - with the exception of the primary key column, which is common to all fragments

– Example: The Customer table may be divided into two fragments, one fragment consisting of Cust ID, name, and address may be located in the Service building and the other fragment with Cust ID, credit limit, balance, dues may be located in the Collection building.


58

Data FragmentationData Fragmentation

Mixed fragmentation combines the horizontal and vertical strategies.

A fragment may consist of a subset of rows and a subset of columns of the original table.

Example: Customer table may be divided by state and grouped by columns. The service building in Texas will store Customer service related information for customers from Texas.


59

Data ReplicationData Replication

Data replication involves storing multiple copies of a fragment in different locations. For example, a copy may be stored in New Delhi and another in Mumbai.

It improves response time and data availability. Data replication requires the DDBMS to maintain data

consistency among the replicas. A fully replicated database stores multiple copies of each

database fragment. A partially replicated database stores multiple copies of

some database fragments at multiple sites.


60

Data AllocationData Allocation Data allocation decision involves determining the location of

the fragments so as to achieve the design goals of cost, response time and availability.

Three data allocation strategies are: centralized, partitioned and replicated.

A centralized allocation strategy stores the entire database in a single location.

A partitioned strategy divides the database into disjointed parts (fragments) and allocates the fragments to different locations.

In a replicated strategy copies of one or more database fragments are stored at several sites.


introduction to data mining

Documents