SRI ESHWAR COLLEGE OF ENGINEERING

Kinathukadavu , Coimbatore-641202

Department of Computer Science Engineering

(ARTIFICIAL INTELLIGENCE AND MACHINE LEARNING)

ACADEMIC YEAR 2024-25 (ODD SEMESTER)

DATA EXPLORATION AND VISUALIZATION

ASSIGNMENT NO :02

PRAVEEN S
22AM033
CSE(AIML)

1) You are tasked with analyzing the impact of a recent earthquake using geospatial data, including satellite
images, building damage reports with coordinates, and road network data. Your goal is to create a detailed
map visualizing the extent of the damage, identify the most affected areas to prioritize rescue operations,
and assess the connectivity of the road network to determine which regions are inaccessible. Using this
data, propose evacuation routes for future disaster preparedness, ensuring they are optimized for safety and
efficiency. How would you integrate geospatial analysis and visualization tools to achieve these objectives
effectively?

Data Collection and Preprocessing

Satellite Images: Use high-resolution satellite imagery (pre- and post-earthquake) to detect changes in
infrastructure and land. Images can be sourced from platforms like Google Earth Engine, Sentinel-2, or
commercial providers (e.g., Maxar, Planet).
Building Damage Reports: Gather geo-referenced building damage reports containing details like severity,
structural type, and location (latitude and longitude).
Road Network Data: Collect road network data, including information on road types, conditions, and
closures post-earthquake. Sources like OpenStreetMap (OSM) and local GIS databases can be valuable.

Geospatial Analysis Tools

GIS Software: Use Geographic Information System (GIS) software like QGIS or ArcGIS to handle,
visualize, and analyze spatial data.
Remote Sensing Tools: Utilize tools like Google Earth Engine for processing satellite images.
This includes change detection to identify damage extent and classify affected areas.
Python Libraries: Incorporate Python libraries like GeoPandas for spatial data manipulation, Fiona and
Shapely for handling spatial files, and Folium or Plotly for web-based interactive mapping.
Network Analysis: Use NetworkX in Python or ArcGIS Network Analyst for analyzing road connectivity
and determining the accessibility of affected regions.

Visualization and Damage Assessment

Satellite Image Analysis: Perform a comparison between pre- and post-earthquake satellite images to
identify damage. Use supervised or unsupervised classification techniques (like CNN or random forests) to
detect damaged buildings and areas.
Building Damage Visualization: Plot damaged buildings on a GIS platform. Categorize them by severity
(e.g., light, moderate, severe) using a color-coded system (e.g., green for minor, yellow for moderate, red
for severe).
Heatmaps: Create heatmaps to visualize the density of damaged buildings, highlighting the most affected
areas. This can be done with tools like QGIS or Python’s Folium.
Elevation and Hazard Layers: Overlay elevation data (from DEMs) and known hazard zones (like
landslide-prone areas) to better understand the geographical impact. Road Network Analysis
Network Accessibility: Use road network data to assess which areas have become inaccessible. Perform a
network analysis to determine disruptions (e.g., blocked roads, damaged bridges) and identify accessible
and non-accessible regions.
Shortest Path Analysis: Apply shortest path algorithms (e.g., Dijkstra’s) to find optimal routes for rescue
operations. Consider constraints like blocked roads and prioritize routes that remain operational. Road
Condition Visualization: Generate a visual map showing the condition of roads (passable, blocked,
partially accessible) using a distinct color code (e.g., green for accessible, red for blocked).
Time-Distance Analysis: Estimate the time required to access different parts of the affected area using
current road conditions to prioritize rescue efforts.
Evacuation Route Planning
Route Optimization: Utilize optimization algorithms (like A* or Dijkstra) to propose evacuation routes.
These should factor in:
o Safety: Avoid hazardous areas like landslide-prone or severely damaged zones. o Accessibility:
Ensure routes pass through areas with the least damage.
o Efficiency: Minimize travel time and distance.
Simulation Tools: Use software like ArcGIS’s Evacuation Modeling tool or Python’s pgRouting (a PostGIS
extension) for route simulations.
Multiple Scenarios: Simulate different scenarios (e.g., nighttime evacuation, bad weather) to ensure
robustness in evacuation plans.
Integration and Visualization for Decision-Making
Web-Based Visualization: Develop a web-based dashboard using Dash, Streamlit, or
Leaflet.js for interactive mapping and visualization. This dashboard should allow stakeholders to explore: o
Real-time damage assessments.
o The status of road networks. o Proposed evacuation routes.
3D Visualization: Use 3D GIS tools like CesiumJS or ArcGIS Pro to visualize the affected terrain and
infrastructure in three dimensions for a better perspective on elevation and the impact of topography. Story
Map: Create an interactive story map using ArcGIS StoryMaps or QGIS Plugin (e.g., qgis2web) to
communicate the analysis results and evacuation strategy clearly to stakeholders.

Future Preparedness and Recommendations

Evacuation Plan Refinement: Use the collected data and analysis to propose optimized and safer evacuation
routes, integrating them into the local disaster response plans.
Infrastructure Weakness Identification: Identify regions where infrastructure was most vulnerable to the
earthquake, providing data-driven insights to strengthen buildings and road networks.
Regular Monitoring: Recommend ongoing monitoring of key infrastructure using satellite data and
realtime road sensors to ensure quick updates in case of future disasters. Technological Workflow
Data Ingestion: Load satellite images, building damage reports, and road network data into a
GIScompatible database.
Data Processing: Pre-process satellite imagery (e.g., correcting for distortions, filtering) and perform
change detection.
Spatial Analysis: Conduct spatial analyses like clustering of damage points, connectivity assessment of the
road network, and route optimization.
Visualization: Implement visualizations in GIS software, web-based platforms, and 3D tools for an
interactive exploration of the results.
Automated Reporting: Generate automated reports using tools like Python's ReportLab or ArcGIS’s
reporting features to provide a summary of findings.
By leveraging these tools and techniques, the integration of geospatial analysis would ensure effective
disaster response, help prioritize resources efficiently, and contribute to safer evacuation planning for future
emergencies.
2) You are working as a data analyst for a retail company and have been tasked with presenting insights using
Tableau and Power BI. The dataset includes sales figures, customer demographics, product categories, and
regional performance data. Your objective is to visualize sales trends, identify high-performing regions, and
analyze customer behavior across different demographics. Using Tableau, how would you create an
interactive dashboard to drill down into regional sales performance over time? For Power BI, explain how
you would set up dynamic filters and slicers to allow stakeholders to explore sales by product category and
customer age group. How would you use both tools to create a compelling story with actionable insights
for the leadership team?

Tableau - Interactive Dashboard for Regional Sales

Data Import: Load the dataset into Tableau and connect to the source.
Dashboard Creation:
o Sales Trends Over Time: Create a line chart showing sales by month/quarter/year. o Regional
Performance: Use a map visualization with color-coded regions based on sales volume. Add a filter
for regions.
o Product & Demographics Analysis: Add bar charts for product categories and customer
demographics.
o Drill Down: Enable interactive drill-down by allowing users to click on regions to see detailed
trends over time or by product category.
o Filters: Include filters for date range, region, product category, and customer demographics for
deeper insights.

Power BI - Dynamic Filters & Slicers

Data Import: Load the dataset into Power BI and set relationships.
Dashboard Setup:
o Sales by Product Category: Create a bar chart with product categories, enabling a drill- down into
subcategories.
o Customer Age Group Analysis: Create a column chart or tree map segmented by age group.
o Dynamic Slicers: Add slicers for product category, customer age group, region, and time period,
allowing stakeholders to filter data dynamically.
o Visual Interactivity: Use sync slicers and cross-highlighting to explore related data in other visuals.
Storytelling with Tableau & Power BI
Tableau: Use the interactive dashboard to showcase sales trends, highlight top regions, and identify
seasonal patterns. Focus on visual storytelling with intuitive visuals and filters.
Power BI: Highlight customer behavior insights by using slicers to segment data. Showcase top product
categories by age group and identify buying trends.
Actionable Insights: Combine the findings from both tools to present targeted strategies for increasing
sales, focusing on high-performing regions, and tailoring marketing to key demographics. Use clear visuals
and concise storytelling for a compelling presentation to the leadership team.