DATA VISUALIZATION
UNIT-2
R23 3-1 AIDS
Creating Visual Representations
Creating visual representations of data involves transforming raw data into
graphical formats that allow users to understand, explore, and interpret
information effectively. This process is central to data visualization and requires
careful attention to the type of data, target audience, and visualization goals.
🔷 1. Definition
A visual representation is the graphical depiction of data, ideas, or concepts
using visual elements like charts, graphs, maps, icons, shapes, or images. The
aim is to convert complex or abstract information into a clear, meaningful, and
interpretable format.
🔷 2. Steps in Creating Visual Representations
🔹 1. Understand the Data
Identify the type (quantitative, categorical, time-series, etc.)
Clean and preprocess the data (remove errors, fill missing values)
🔹 2. Define the Objective
What insights are you trying to deliver?
Is it for comparison, distribution, trend analysis, or correlation?
🔹 3. Choose the Right Chart or Graph
Chart Type Use Case
Bar Chart Compare categories
Line Chart Show trends over time
Scatter Plot Show relationships/correlations
Pie Chart Show part-to-whole relationships
Tree Map Represent hierarchical data
Heatmap Show intensity across a matrix
🔹 4. Map Data to Visual Elements
Variables are mapped to visual properties like:
o Position (x, y coordinates)
o Color
o Shape
o Size
o Length
o Orientation
🔹 5. Design for Readability
Use appropriate labels, titles, legends
Avoid clutter, 3D effects, and overuse of colors
Ensure accessibility (e.g., colorblind-friendly palettes)
🔷 3. Principles of Good Visual Representation
Principle Description
Clarity Viewers should quickly grasp the message
Accuracy No distortion or misleading elements
Principle Description
Consistency Use same colors/scales for similar data
Simplicity Focus on essential elements
Visual Hierarchy Guide the viewer’s attention to key points
🔷 4. Tools for Creating Visual Representations
Excel, Power BI, Tableau – Business dashboards
Python (Matplotlib, Seaborn, Plotly) – Programmatic visualizations
R (ggplot2) – Statistical charts
[Link] – Custom web-based interactive visuals
🔷 5. Real-world Examples
Domain Visual Used
Finance Line charts for stock trends
Healthcare Heatmaps of disease spread
Marketing Funnel or bar charts for campaign analysis
Geography Maps to show demographic distributions
Visualization Reference Model
The Visualization Reference Model is a conceptual framework that describes the
flow of data through various stages of transformation before it is presented
visually to a user. It serves as a guide for designing, analyzing, and
understanding the process of data visualization—from raw data acquisition to
the final visual output and user interaction.
🔷 1. Definition
A Visualization Reference Model outlines the structured
steps and components involved in converting raw data
into an effective and interactive visual representation. It
is used by researchers, analysts, and developers to design
systematic visual systems.
🔷 2. Purpose of the Model
To standardize the visualization process
To support the design of interactive visualization
systems
To help in identifying the roles of each stage (data,
processing, visual encoding, etc.)
To separate concerns between data management and
visual representation
🔷 3. Key Components of the Visualization Reference Model
The model typically includes four main stages:
🔹 A. Raw Data Stage
Input: Unprocessed, real-world data (e.g., from
sensors, databases, logs)
Goal: Gather all necessary information
Examples: CSV files, IoT sensor readings, survey
results
🔹 B. Data Transformation Stage
Processes: Cleaning, filtering, aggregating,
normalizing
Tools: SQL, Python (Pandas), Excel
Purpose: Prepare data for mapping into visual form
Example: Aggregating sales by month, removing
missing values
🔹 C. Visual Mapping (Encoding) Stage
Definition: Data values are mapped to visual
attributes
Visual Channels Used:
o Position – X and Y coordinates
o Color – Categorical or continuous values
o Shape – To distinguish different data types
o Size – To show magnitude or weight
Example: Sales amount represented by the size of a
circle
🔹 D. View and Interaction Stage
Visual Output: The final rendered chart, graph, or
dashboard
User Interaction Features:
o Zoom, filter, sort, tooltip
o Drill-down, pan, animation
Goal: Let users explore and extract insights
interactively
🔷 4. Extended Models (Like Card et al., 1999)
Some advanced reference models break the process
further into:
1. Data Transformation
2. Visual Mapping
3. View Transformation
4. User Interaction
This highlights that the visualization is not static—users
play an active role in modifying views or exploring
different parts of the data.
🔷 5. Applications of the Model
Area Use Case
Designing dashboards, BI tools, or
System Design
analytics software
Academic Evaluating or proposing new
Research visualization techniques
Teaching how data is transformed into
Education
visuals
Tool Guiding the architecture of data
Development visualization libraries
🔷 6. Benefits of Using the Visualization Reference Model
Clarifies the data-to-visual flow
Separates data logic from presentation logic
Enables modularity and scalability in visualization
systems
Supports user-centered design by integrating
interaction
Helps in identifying performance bottlenecks
Visual Mapping
Visual Mapping is a core concept in data visualization that
refers to the process of translating data values into visual
elements such as position, size, color, shape, and
orientation. It plays a key role in making data
understandable by leveraging human visual perception.
🔷 1. Definition
Visual mapping is the process of encoding data attributes
into graphical or visual features, allowing users to
interpret and analyze the data visually. It determines how
the data is presented on screen or paper.
🔷 2. Purpose of Visual Mapping
To visually encode data in an intuitive and perceptible
way
To support pattern recognition, comparison, and
trend analysis
To highlight relationships, differences, or anomalies in
the data
To enhance the clarity and usability of visualizations
🔷 3. Common Visual Channels for Mapping
Visual
Used For
Channel
Most accurate for comparing numerical
Position
values
Representing categories or gradients
Color
(e.g., heatmaps)
Size Showing magnitude (e.g., bubble charts)
Shape Differentiating types or groups
Length/
Used in bar charts or histograms
Width
Indicating direction or trend (e.g.,
Orientation
arrows)
Texture/ Less common; used for accessibility or
Pattern emphasis
🔷 4. Examples of Visual Mapping
Mapping sales figures to bar height
Mapping regions to colors on a map
Mapping profit to the size of circles in a bubble chart
Mapping temperature over time to line position in a
graph
🔷 5. Good Practices in Visual Mapping
Use consistent scales and ranges
Avoid using too many colors or shapes (overload)
Choose mappings that match human perceptual
strengths
Ensure accessibility (e.g., colorblind-friendly palettes)
Visual Analytics
Visual Analytics is an advanced field that combines data analysis, visualization,
and human interaction to support complex decision-making. It helps users
explore large and complex datasets visually, enabling them to identify patterns,
trends, correlations, and anomalies effectively.
🔷 1. Definition
Visual Analytics is the science of analytical reasoning supported by interactive
visual interfaces. It integrates automated data analysis techniques (such as
statistics, machine learning, and AI) with interactive visual representations that
enable users to make sense of massive, dynamic, and diverse datasets.
📌 Coined by Jim Thomas and Kristin Cook in 2005 under the U.S. DHS’s initiative.
🔷 2. Key Features
Feature Description
Interactive
Users can zoom, filter, sort, and manipulate views
Visualizations
Human judgment and reasoning are supported by visual
Human-Centric
tools
Scalable to Big Data Handles large, heterogeneous, and real-time datasets
Combines algorithms (e.g., clustering, predictive models)
Integrates Analytics
with visual output
🔷 3. Core Components
1. Data Management – Cleaning, integrating, and preparing datasets
2. Analytical Computation – Applying statistical, AI, or machine learning
models
3. Visualization Interface – Presenting results visually (graphs, maps,
timelines)
4. User Interaction – Allowing dynamic exploration, hypothesis testing, and
discovery
🔷 4. Applications of Visual Analytics
Domain Use Case
Cybersecurity Detecting anomalies in network traffic
Finance Risk analysis, fraud detection, and market forecasting
Healthcare Patient monitoring, outbreak analysis
Intelligence Crime pattern detection, surveillance
Business Intelligence Sales trend exploration, customer segmentation
Scientific Research Simulation analysis, data exploration in large experiments
Domain Use Case
🔷 5. Benefits
Enables quick, data-driven decision-making
Improves accuracy and confidence in analysis
Allows for real-time monitoring of systems
Bridges the gap between data science and human intuition
🔷 6. Popular Tools and Technologies
Tableau, Power BI – Dashboards with analytics capabilities
Qlik Sense – Associative data modeling with rich visuals
[Link], Plotly – Custom interactive web-based visual analytics
R + ggplot2 / Python + Seaborn/Plotly – Data science-focused visual
exploration
[Link], Kibana, Grafana – Geospatial and real-time analytics
Design of Visualization Applications
The design of visualization applications involves creating
software or tools that enable users to interact with,
explore, and interpret data visually. These applications
serve as the bridge between raw data and human
understanding, enabling effective decision-making
through intuitive visual interfaces.
🔷 1. Definition
A visualization application is a software system designed
to represent data graphically, allowing users to analyze,
filter, and interact with the data in meaningful ways. The
design process involves understanding user needs, data
structure, visualization goals, and technological
constraints.
🔷 2. Goals of Visualization Application Design
Present complex data simply and clearly
Enable user-driven exploration and filtering
Support decision-making through visual insights
Make the system scalable, interactive, and responsive
🔷 3. Key Design Considerations
Aspect Description
🎯 User-Centric Understand the target audience’s
Design skills, goals, and cognitive load
📊 Data Consider volume, type (categorical,
Characteristics numerical), and structure of the data
🧩 Visualization Choose the right type (bar, line,
Techniques scatter, heatmap, 3D, etc.)
Include features like zoom, filter,
⚙️Interactivity
drag, tooltips, animation
🎨 Aesthetics & Use effective colors, typography, and
Clarity layout to enhance readability
🧠 Cognitive Load Avoid clutter, emphasize key patterns,
Management support human perception
🔐 Performance & Handle large or real-time datasets
Scalability without lag
🔷 4. Design Process
1. Requirement Analysis
o Identify users, goals, and context
2. Data Understanding
o Analyze data quality, format, and complexity
3. Wireframing & Prototyping
o Sketch interface and layout before coding
4. Visualization & Interaction Design
o Map data to visuals and define user interactions
5. Implementation
o Use frontend/backend tools (e.g., [Link], Python,
Power BI)
6. Testing & Evaluation
o Check usability, performance, and visual accuracy
🔷 5. Tools & Frameworks Used
Tool Use
Tableau, Power Rapid dashboard and report
BI creation
[Link], [Link] Custom web-based visualizations
Python (Plotly, Data science-focused interactive
Dash) apps
Statistical data visualization and
R (Shiny)
interaction
[Link], WebGL 3D and immersive visualizations
🔷 6. Real-World Examples
Business Intelligence Dashboards (e.g., sales
tracking, KPI monitoring)
Scientific Visualization Tools (e.g., gene expression,
climate models)
Geospatial Mapping Apps (e.g., crime maps, delivery
tracking)
Healthcare Systems (e.g., patient vitals monitoring,
EHR dashboards)
🔷 7. Challenges in Designing Visualization Apps
Balancing complexity vs. simplicity
Ensuring real-time performance with big data
Designing for accessibility (colorblind-friendly,
mobile-responsive)
Maintaining data privacy and security