A Multiview Visualization System for Car Accident Data
This project presents an interactive Information Visualization System for analyzing U.S. traffic accident data (2016–2023). The system provides multiview visualization techniques to explore accident trends across different geographical regions, weather conditions, and temporal factors.
The visualizations aim to:
- Identify accident hotspots.
- Analyze correlations between weather and accident severity.
- Investigate temporal patterns to detect peak accident hours.
- Provide interactive, user-friendly data exploration.
- Source: Kaggle - US Accidents Dataset
- Attributes Analyzed:
- Geographical Factors: County, city, and state.
- Temporal Factors: Time of the day, year, and seasonality.
- Weather Conditions: Rain, fog, snow, and temperature.
- Road Conditions: Intersection, highway type, and urban vs. rural areas.
- Accident Severity: Categorical classification from minor to major accidents.
The project implements three distinct visualization systems to explore the dataset:
- Focus: Accident frequency over time.
- Key Features:
- Interactive zooming and filtering based on time of the day.
- Strip plots and jitter effects to represent severity.
- A slider to adjust temporal resolution dynamically.
- Focus: Spatial distribution of accidents.
- Key Features:
- Interactive map visualization with color-coded accident markers.
- Dropdown filters for road conditions, weather conditions, and severity.
- Brushing interactions to refine spatial queries.
- Focus: Correlation between weather conditions and accident severity.
- Key Features:
- Heatmap representation of accident density in major cities.
- Scatter Plot Matrix (SPLOM) to explore weather-accident relationships.
- Interactive bar charts for road condition-based filtering.
- Multi-view Visualization: Combines spatial, temporal, and weather-based analyses.
- Interactivity: Users can filter, zoom, and brush through different datasets.
- Comparative Analysis: Systems allow direct comparison between different views of the dataset.
- Data Filtering: Users can explore data subsets based on accident severity, weather, and location.
- Clone this repository:
git clone https://github.com/Mejorarsim/Information-Visualization.git cd Information-Visualization - Install dependencies:
pip install pandas numpy vega_datasets altair
- Run the visualization tool:
python main.py
- Simran Garg
- Anjali Gedam
- Shirisha Mahesh
- Zhihao Yan
- Yumeng Zhang