WEB AND SOCIAL MEDIA ANALYTICS

UNIT-1

An Overview of Business Intelligence, Analytics, and Decision Support

Analytics to Manage a Vaccine Supply Chain Effectively and Safely, Changing Business Environments and Computerized
Decision Support, Information Systems Support for Decision Making, The Concept of Decision Support Systems (DSS),
Business Analytics Overview, Brief Introduction to Big Data Analytics

1.1 An Overview of Business Intelligence, Analytics, and Decision Support

What is Business Intelligence?

Business Intelligence (BI) is a technology-driven process that involves the collection, integration, analysis, and
presentation of business data to help organizations make informed decisions.
It turns raw data into meaningful and actionable insights through reporting, dashboards, querying, and data
visualization.
Processes Involved in Business Intelligence

Data Collection

Data Integration & Cleaning

Data Warehousing (Storage)

Data Analysis & Querying

Reporting & Visualization

Decision-Making

MIETW WEB AND SOCIAL MEDIA ANALYTICS FACULTY NAME: MANJU

1. Data Collection
 Gathering data from various internal (ERP, CRM, databases) and external (social media,
surveys, APIs) sources.
 Structured (tables, records) or unstructured (emails, reviews).

2. Data Integration and Cleaning

 Combine data from different systems into a unified format.
 Remove duplicates, correct errors, handle missing values.

3. Data Storage (Data Warehousing)

 Data is stored in a data warehouse, which acts as a central repository.
 Historical data is maintained for analysis.

4. Data Analysis and Querying

 Use SQL or analytical tools to extract and analyze patterns, trends, and insights from the data.
 May use OLAP for multidimensional analysis

(OLAP Online Analytical Processing is a technology used in Business Intelligence (BI)

that allows users to analyze multidimensional data interactively from multiple
perspectives. It enables fast querying, data summarization, and drill-down into large
datasets for better decision-making.)

5. Reporting and Visualization

 Create reports, charts, dashboards, and KPIs using visualization tools.
 Helps managers and stakeholders understand key metrics at a glance.

6. Decision-Making

 Business users use insights from reports and dashboards to make informed, data-driven
decisions.

Key Function Purpose Example

Data Collection & Combine data from multiple
Sales + social media + CRM data
Integration sources
Amazon Redshift stores 5 years of
Data Warehousing Store historical, integrated data
customer data
Retrieve specific data and generate
Querying & Reporting Weekly sales report by product category
summaries
Dashboards &
Present key metrics visually Real-time dashboard for marketing KPIs
Visualization
Enable fast, multidimensional data Analyzing sales by product, region, and
OLAP
analysis time simultaneously

MIETW WEB AND SOCIAL MEDIA ANALYTICS FACULTY NAME: MANJU

Example:

A retail company uses BI to:

1. Collect data from sales, inventory, and customer feedback.

2. Store the data in a cloud warehouse (e.g., Snowflake).
3. Clean and integrate the data using Talend.
4. Analyze trends (e.g., best-selling products, sales by region).
5. Build dashboards in Power BI.
6. Make strategic decisions (e.g., where to open a new store).

Summary:

Aspect Details

Definition BI converts raw data into actionable business insights

Goal Better, data-driven decision-making

Processes Data collection → Integration → Warehousing → Analysis → Reporting

Tools Power BI, Tableau, SQL, Talend, Redshift, OLAP tools

Tools used in every function:

Function Tools

Data Integration (ETL) Talend, SSIS, Apache NiFi, Informatica

Data Warehousing Amazon Redshift, Snowflake, Google BigQuery

Reporting & Visualization Power BI, Tableau, QlikView, Looker

Querying & Analysis SQL, Excel, R, Python

MIETW WEB AND SOCIAL MEDIA ANALYTICS FACULTY NAME: MANJU

What is Business Analytics?

Business Analytics (BA) is the practice of using data analysis, statistical models, and predictive
techniques to explore historical data, discover patterns, and make better business decisions. Unlike
Business Intelligence (BI), which focuses more on reporting and monitoring, Business Analytics focuses
on analysis, prediction, and optimization.

Business Analytics Process Flow:

Data Collection
↓
Data Cleaning & Preparation
↓
Data Exploration & Visualization
↓
Statistical Analysis / Modeling
↓
Interpretation of Results
↓
Decision-Making & Optimization

1. Data Collection

 Gather data from various sources like sales logs, CRM, websites, sensors, etc.
 Types of data: Structured (tables), Semi-structured (JSON), Unstructured (text, images)

2. Data Cleaning & Preparation

 Fix missing values, remove duplicates, standardize formats

 Feature selection and transformation (for ML)

3. Data Exploration & Visualization

 Use graphs and charts to understand trends and patterns

 Techniques: Histograms, scatter plots, box plots

MIETW WEB AND SOCIAL MEDIA ANALYTICS FACULTY NAME: MANJU

4. Statistical Analysis / Modeling

 Use mathematics and algorithms to discover relationships

 Apply machine learning models for prediction and classification

5. Interpretation of Results

 Convert technical output into business insights (valuable understandings or conclusions)

 Evaluate accuracy, confidence, and practical value

6. Decision-Making & Optimization

 Use the insights to improve processes, reduce cost, increase ROI

 Tools like optimization models (e.g., linear programming) help choose the best action

Tools Used in Business Analytics:

Stage Tools

Data Collection SQL, Excel, APIs, Web scraping tools

Data Cleaning & Prep Python (Pandas), R, Power Query, Alteryx

Visualization Tableau, Power BI, Matplotlib, Seaborn

Statistical Analysis R, Python (NumPy, SciPy, Scikit-learn), SAS, Excel

Predictive Modeling Python (Scikit-learn, XGBoost), RapidMiner, IBM SPSS

Optimization Excel Solver, Python (SciPy), Gurobi

Types of Business Analytics:

TYPE DESCRIPTIVE EXAMPLE

Descriptive What happened? Monthly sales report

Diagnostic Why did it happen? Drop in sales due to pricing changes

Predictive What will happen? Forecasting future demand

Prescriptive What should be done? Optimizing delivery routes

MIETW WEB AND SOCIAL MEDIA ANALYTICS FACULTY NAME: MANJU

example:

A food delivery company uses Business Analytics to:

 Analyze past delivery times (Descriptive)

 Identify causes of late deliveries (Diagnostic)
 Predict high-demand areas (Predictive)
 Optimize driver routes (Prescriptive)

Summary:

Aspect Details

Definition Use of data, statistics, and models for decisions

Goal Discover insights and predict outcomes

Process Collect → Clean → Analyze → Interpret → Decide

Tools Python, R, Power BI, Excel, Tableau

Applications Marketing, Finance, Operations, HR, Healthcare

What is Business Decision Support?

It focuses on enhancing the quality, speed, and effectiveness of decision-making by
providing relevant information, models, and simulations.
Business Decision Support Process:

Problem Identification
↓
Data Collection
↓
Model Building & Analysis
↓
Evaluation of Alternatives
↓
Decision Implementation
↓
Feedback & Monitoring

MIETW WEB AND SOCIAL MEDIA ANALYTICS FACULTY NAME: MANJU

1. Problem Identification

 Understand and define the business issue that needs a decision (e.g., declining sales,
poor resource allocation)

2. Data Collection

 Gather relevant internal and external data (sales records, customer feedback,
competitor info)

3. Model Building & Analysis

 Use quantitative models such as forecasting, simulation, or optimization

 Analyze scenarios and predict outcomes

4. Evaluation of Alternatives

 Compare different decision options using what-if analysis, cost-benefit analysis, etc.

5. Decision Implementation

 Choose the most effective solution and apply it in the business process

6. Feedback & Monitoring

 Monitor the outcomes of the decision to assess its effectiveness

MIETW WEB AND SOCIAL MEDIA ANALYTICS FACULTY NAME: MANJU

Tools Used in Decision Support:

Function Tools / Examples

Data Access & Integration SQL, Excel, Power BI, Tableau

Analytical Modeling Python, R, SAS, IBM SPSS, MATLAB

Optimization Excel Solver, Gurobi, LINGO

Simulation @RISK, AnyLogic, Arena Simulation

Decision Trees / Rules Scikit-learn, DecisionTools Suite, Oracle DSS tools

Dashboards Power BI, QlikView, Tableau

Summary:

Aspect Details

Definition Systems and tools that help support business decisions

Purpose Improve quality, speed, and effectiveness of decisions

Process Identify → Collect Data → Model → Evaluate → Act → Monitor

Tools Excel Solver, Power BI, R, Python, DSS software

Applications Finance, logistics, HR planning, marketing strategies

MIETW WEB AND SOCIAL MEDIA ANALYTICS FACULTY NAME: MANJU

1.2 Analytics to Manage a Vaccine Supply Chain Effectively and Safely

Managing a vaccine supply chain is a complex task due to the sensitive nature of vaccines,
the need for strict temperature control, inventory accuracy, and timely delivery. Analytics
plays a crucial role in optimizing and securing this supply chain from manufacturing to
vaccination.

Key Objectives of Using Analytics in Vaccine Supply Chains:

 Ensure availability of vaccines when and where needed

 Maintain cold chain integrity to preserve vaccine efficacy
 Minimize wastage and stockouts
 Detect and prevent fraud or counterfeiting
 Support data-driven policy and logistics planning

Analytics Process in Vaccine Supply Chain:

Data Collection
↓
Data Integration (from warehouses, GPS, IoT, clinics)
↓
Analysis & Modeling
↓
Forecasting & Optimization
↓
Visualization & Decision Support
↓
Action & Monitoring

MIETW WEB AND SOCIAL MEDIA ANALYTICS FACULTY NAME: MANJU

How Analytics Helps at Each Stage:

Stage Analytics Role

Demand Forecasting Predict future vaccine needs using historical and real-time data

Inventory Track inventory levels and optimize stock to avoid overstock or

Management shortages

Cold Chain
Use IoT sensors and analytics to ensure temperature compliance
Monitoring

Use GIS and real-time traffic data to reduce delays and fuel
Route Optimization
consumption

Fraud Detection Use pattern analysis to identify suspicious distribution or usage

Monitor key performance indicators (KPIs) like delivery time,

Performance Tracking
wastage, coverage

Types of Analytics Used:

Type Application

Descriptive Analytics Monitor supply chain health, stock levels, delivery timelines

Predictive Analytics Forecast future vaccine demand, risk of spoilage or stockouts

Prescriptive Analytics Recommend optimal routes, reorder levels, buffer stocks

Tools & Technologies:

 Data Collection: IoT devices, barcodes, RFID

 Data Processing: Python, R, SQL, Apache Spark
 Visualization: Tableau, Power BI, dashboards
 Forecasting & Optimization: Machine Learning models, Excel Solver, GIS tools
 Security: Blockchain for traceability and authentication

MIETW WEB AND SOCIAL MEDIA ANALYTICS FACULTY NAME: MANJU

Real-World Example:

During COVID-19, many governments used analytics to:

 Predict vaccine demand based on population and infection rates

 Track and optimize last-mile delivery using GPS data
 Monitor cold chain compliance via IoT-enabled sensors
 Detect anomalies in distribution to prevent fraud

Benefits of Analytics in Vaccine Supply Chain:

 Enhanced transparency and traceability

 Improved vaccine coverage and accessibility
 Reduced wastage and losses
 Quicker and smarter decision-making
 Greater public trust through data-backed assurance

MIETW WEB AND SOCIAL MEDIA ANALYTICS FACULTY NAME: MANJU

1.3 Changing Business Environments and Computerized Decision Support

Changing Business Environments

Modern businesses operate in rapidly changing, complex, and competitive environments.

These changes affect how organizations make decisions and require more real-time, data-
driven support.

Key Factors Driving Change:

Factor Description

Businesses operate across borders; decisions involve global data and

Globalization
competition.

Technological
AI, IoT, cloud computing increase data volume and complexity.
Advances

Market Volatility Demand, prices, and competition can change suddenly.

Customer
Personalization, speed, and service matter more than ever.
Expectations

Businesses must adapt quickly to new laws, privacy, and compliance

Regulatory Changes
standards.

Need for Computerized Decision Support

Why Manual Decisions Are Not Enough:

 Too much data to process manually

 Real-time decisions are needed
 Risks and uncertainty are high
 Decisions need to be consistent and data-driven

MIETW WEB AND SOCIAL MEDIA ANALYTICS FACULTY NAME: MANJU

What is Computerized Decision Support?

Computerized Decision Support refers to the use of software systems, data, models, and
tools to assist managers and employees in making better, faster, and more informed
[Link] are often implemented as Decision Support Systems (DSS) or Business
Intelligence platforms.

Components of Computerized Decision Support Systems:

Component Function

Database Stores relevant internal and external data

Model Base Contains analytical, statistical, and simulation models

User Interface Allows users to interact with the system

Knowledge Base (Optional) Stores business rules, best practices

Benefits of Computerized Decision Support:

 Faster and more accurate decision-making

 Improved efficiency and productivity
 Better handling of complex and uncertain situations
 Enhanced competitive advantage
 Supports data-driven culture

Tools & Technologies Used:

Tool Type Examples

BI Tools Tableau, Power BI, Qlik

DSS Software IBM Cognos, SAP BusinessObjects, Oracle DSS

Data Analysis Excel, R, Python

Optimization Excel Solver, Gurobi

Simulation Arena, AnyLogic, @RISK

MIETW WEB AND SOCIAL MEDIA ANALYTICS FACULTY NAME: MANJU

1.4 Information Systems Support for Decision Making

Information Systems (IS) that support decision making are computer-based systems
designed to help individuals and organizations make better, faster, and data-driven
decisions. They collect, process, store, and analyze data to provide relevant information,
insights, and recommendations.

These systems support decisions that may be structured, semi-structured, or unstructured.

Types of Decisions:

1. Structured Decisions (Programmed Decisions)

These are routine, repetitive, and well-defined decisions made using established rules,
procedures, or algorithms.

Features:

 Clear problem definition

 Standard procedure or formula
 Often automated or rule-based
 Minimal human judgment required

Examples:

 Reordering inventory when stock is low

 Payroll processing
 Approving a loan under preset criteria

2. Semi-Structured Decisions

These decisions are partially routine but still require human judgment and interpretation
alongside data and models.

MIETW WEB AND SOCIAL MEDIA ANALYTICS FACULTY NAME: MANJU

Features:

 Some known rules, some uncertainty

 Supported by Decision Support Systems (DSS)
 Combination of logic + intuition

Examples:

 Marketing budget allocation

 Setting product pricing
 Forecasting sales

3. Unstructured Decisions (Non-programmed Decisions)

These are complex, unique, and non-routine decisions with no clear procedures. They
require creativity, judgment, and strategic thinking.

Features:

 High uncertainty and risk

 No predefined process
 Typically made by senior management
 Supported by Executive Information Systems (EIS) or Expert Systems

Examples:

 Launching a new product line

 Entering a new international market
 Crisis management (e.g., after a data breach)

MIETW WEB AND SOCIAL MEDIA ANALYTICS FACULTY NAME: MANJU

Types of Information Systems That Support Decision Making:

System Type Function Example

Transaction Processing Record day-to-day business

Transactions
Systems (TPS) operations

Management Information Summarize and report structured

Monthly sales report
Systems (MIS) data

Decision Support Systems Help with semi-structured decisions

Product pricing tool
(DSS) using data + models

Executive Information Dashboard-style systems for top

Strategic KPI dashboard
Systems (EIS) executives

Use rules and knowledge base to

Expert Systems Medical diagnosis system
offer recommendations

Analyze data and trends, support Power BI dashboards,

Business Intelligence (BI)
forecasting Tableau reports

Use machine learning to recommend Chat bots,

AI-based Systems
or automate decisions recommendation engines

Decision-Making Process Supported by IS:

Problem Identification
↓
Data Collection
↓
Data Analysis and Modeling
↓
Generation of Alternatives
↓
Evaluation and Choice
↓
Implementation and Monitoring

MIETW WEB AND SOCIAL MEDIA ANALYTICS FACULTY NAME: MANJU

Benefits of Information Systems in Decision Making:

 Improved data accuracy and access

 Faster decision-making through automation
 Better quality decisions based on real-time information
 Collaboration support for group decision-making
 Supports strategic, tactical, and operational planning

Real-World Example:

A retail company uses:

 TPS to record sales
 MIS to summarize weekly performance
 DSS to forecast inventory needs
 BI tools to monitor trends
 AI to personalize product recommendations

All of these systems together help inform both day-to-day operations and long-term
strategies.

Summary:

Aspect Details

Definition IS helps gather, analyze, and present information for decisions

Supported Decision Types Structured, Semi-structured, Unstructured

Key Systems TPS, MIS, DSS, EIS, Expert Systems, BI, AI

Benefits Accuracy, speed, insight, consistency, collaboration

Application Areas Finance, marketing, operations, HR, strategy

MIETW WEB AND SOCIAL MEDIA ANALYTICS FACULTY NAME: MANJU

1.5 The Concept of Decision Support Systems (DSS)

A Decision Support System (DSS) is a computer-based system that helps in making semi-
structured or unstructured decisions by combining:

 Data
 Analytical models
 User-friendly software tools
DSS supports decision-makers by providing relevant information, analysis, and
recommendations, not just raw data.
Objectives of DSS:
 Improve decision quality
 Provide interactive tools for analysis
 Handle complex, uncertain problems
 Support non-routine, strategic decisions
 Facilitate “what-if” and scenario analysis

Key Components of a DSS:

Component Function

Database Stores relevant internal/external data

Contains analytical models (e.g., simulation,
Model Base
optimization, forecasting)
User Interface Allows users to interact with the system easily
Knowledge Base (optional) Stores rules, procedures, and best practices

DSS Process Flow:

Problem Identification
↓
Data Collection and Access
↓
Use of Models/Analysis Tools
↓
Evaluate Alternatives / What-if Scenarios
↓
Decision Recommendation

MIETW WEB AND SOCIAL MEDIA ANALYTICS FACULTY NAME: MANJU

Characteristics of DSS:

 Supports semi-structured and unstructured decisions

 Provides interactive, real-time decision-making support
 Allows users to explore multiple scenarios
 Designed to be flexible and user-friendly
 Often used by middle and upper management

Examples of DSS Tools and Technologies:

Function Examples

Data Management SQL, Excel, Microsoft Access

Modeling/Analysis Python, R, MATLAB, Excel Solver

Visualization Tableau, Power BI

Integrated DSS Tools IBM Cognos, SAP BusinessObjects, Oracle DSS

Types of DSS:

Type Description Example

Focuses on access to and manipulation Sales analysis, KPI

Data-Driven DSS
of large data sets dashboards

Uses mathematical and statistical Financial forecasting,

Model-Driven DSS
models simulations

Knowledge-Driven DSS Provides expert recommendations Medical diagnosis system

Analyzes and retrieves documents for

Document-Driven DSS Legal research system
decision support

Communication- Supports collaboration and group Groupware tools, video

Driven DSS decisions conferencing

MIETW WEB AND SOCIAL MEDIA ANALYTICS FACULTY NAME: MANJU

Example Use Case:

A supply chain manager uses a DSS to:

 Analyze transportation costs

 Simulate delivery routes
 Compare cost vs. time trade-offs
 Choose the optimal distribution plan

Summary

Aspect Details

Definition Computer-based system to assist complex decision-making

Purpose To improve decision accuracy, speed, and consistency

Users Managers and analysts (middle/top management)

Core Components Database, Model base, User interface, Knowledge base

Application Areas Supply chain, finance, marketing, HR, healthcare, etc.

MIETW WEB AND SOCIAL MEDIA ANALYTICS FACULTY NAME: MANJU

1.6 Business Analytics Overview

Business Analytics (BA) is the scientific process of analyzing data to discover useful
information, uncover patterns, and support decision-making. It involves using statistical,
mathematical, and computational techniques to convert raw data into actionable insights.

Goals of Business Analytics:

 Understand business performance

 Predict future trends
 Optimize decision-making
 Drive strategic and operational improvements
 Support data-driven culture in organization

Business Analytics Process:

Data Collection
↓
Data Cleaning and Preparation
↓
Data Exploration and Visualization
↓
Modeling and Analysis
↓
Interpretation of Results
↓
Decision-Making and Implementation

Types of Business Analytics:

Type Purpose Example

Descriptive Analytics What happened? Monthly sales reports

Diagnostic Analytics Why did it happen? Identifying causes of customer churn

MIETW WEB AND SOCIAL MEDIA ANALYTICS FACULTY NAME: MANJU

 Type Purpose Example

Predictive Analytics What is likely to happen? Sales forecasting for next quarter

Prescriptive Analytics What should be done? Suggesting best pricing strategies

Benefits of Business Analytics:

 Data-driven decision-making
 Improved operational efficiency
 Enhanced customer experience
 Better risk management
 Identification of new business opportunities

Common Tools Used in Business Analytics:

Category Tools / Technologies

Data Analysis Excel, R, Python (Pandas, NumPy), SAS

Data Visualization Tableau, Power BI, Google Data Studio

Predictive Modeling Python (scikit-learn), R, IBM SPSS, RapidMiner

Data Management SQL, Microsoft Access, Google BigQuery

Machine Learning Python, TensorFlow, AutoML platforms

Example Use Case:

A telecom company uses Business Analytics to:

 Analyze call and usage patterns (Descriptive)

 Identify causes of customer churn (Diagnostic)
 Predict customers likely to leave (Predictive)
 Suggest loyalty offers to retain them (Prescriptive)

Summary

Aspect Details

Definition Data analysis for business insight and decisions

MIETW WEB AND SOCIAL MEDIA ANALYTICS FACULTY NAME: MANJU

 Aspect Details

Core Process Collect → Clean → Analyze → Interpret → Act

Key Types Descriptive, Diagnostic, Predictive, Prescriptive

Tools Used Excel, Python, R, Tableau, Power BI, SQL

Application Areas Marketing, finance, HR, supply chain, operations

1.7 Brief Introduction to Big Data Analytics

Big Data Analytics is the process of examining large, diverse, and complex datasets—called
Big Data—to uncover hidden patterns, unknown correlations, market trends, customer
preferences, and other valuable insights. It helps businesses make better, faster, and more
informed decisions using data that is too large or complex for traditional tools.

What is Big Data?

Big Data refers to datasets that are too large (volume), too fast (velocity), and too varied
(variety) for traditional data processing systems.

The 5 V’s of Big Data:

1. Volume – Massive amounts of data (TBs, PBs)

2. Velocity – Data generated in real-time or near real-time
3. Variety – Structured, semi-structured, and unstructured data (text, images, videos,
logs)
4. Veracity – Uncertainty or quality of data
5. Value – Extracting useful insights and business value

Goals of Big Data Analytics:

 Improve decision-making
 Predict future outcomes
 Personalize customer experiences
 Detect fraud and anomalies
 Optimize operations

MIETW WEB AND SOCIAL MEDIA ANALYTICS FACULTY NAME: MANJU

Big Data Analytics Process:

Data Collection (from sensors, logs, web, social media)

Data Storage (using Hadoop, cloud, data lakes)

Data Processing (batch or real-time processing)

Data Analysis (using analytics models, ML, AI)

Insight Generation and Visualization

Popular Big Data Analytics Tools:

Purpose Tools

Data Storage Hadoop HDFS, Amazon S3, Azure Data Lake

Data Processing Apache Spark, Apache Storm, Flink

Querying/Analysis Hive, Pig, Presto, SQL on Hadoop

Machine Learning Apache Mahout, MLlib, TensorFlow

Visualization Tableau, Power BI, Qlik, Kibana

MIETW WEB AND SOCIAL MEDIA ANALYTICS FACULTY NAME: MANJU

.

Example

An e-commerce company uses Big Data Analytics to:

 Track millions of customer interactions

 Recommend personalized products
 Predict inventory demand
 Analyze customer sentiment from reviews

Summary
Aspect Details

Definition Analyzing massive, complex datasets for insights

Key Features High volume, velocity, variety, veracity, and value

Tech Stack Hadoop, Spark, Hive, NoSQL, ML, Cloud

Applications Retail, healthcare, banking, IoT, telecom, social media

MIETW WEB AND SOCIAL MEDIA ANALYTICS FACULTY NAME: MANJU